(The Oily Press Lipid Library) Frederic Destaillats, Jean-Louis Sebedio, Fabiola Dionisi, Jean-Michel Chardigny - Trans Fatty Acids in Human Nutrition-Woodhead Publishing (2009)

Trans fatty acids in
human nutrition
Also in the Oily Press Lipid Library:
Volume 22. Phospholipid Technology and Applications
Edited by Frank D. Gunstone
Volume 21. Long-Chain Omega-3 Specialty Oils
Edited by Harald Breivik
Volume 20. Antioxidants in Food and Biology: Facts and Fiction
Written by Edwin N. Frankel
Volume 19. Lipids: Structure, Physical Properties and Functionality
Written by Kåre Larsson, Peter Quinn, Kiyotaka Sato and Fredrik Tiberg
Volume 18. Lipid Oxidation (second edition)
Written by Edwin N. Frankel
Volume 17. Bioactive Lipids
Edited by Anna Nicolaou and George Kokotos
Volume 16. Advances in Lipid Methodology – Five
Edited by Richard O. Adlof
Volume 15. Lipid Analysis (third edition)
Written by William W. Christie
Volume 14. Confectionery Fats Handbook
Written by Ralph E. Timms
Volume 13. Lipids for Functional Foods and Nutraceuticals
Edited by Frank D. Gunstone
Volume 12. Lipid Glossary 2
Written by Frank D. Gunstone and Bengt G. Herslöf
Volume 11. Lipids in Nutrition and Health: A Reappraisal
Written by Michael I. Gurr
Volume 9. Trans Fatty Acids in Human Nutrition (first edition)
Edited by Jean Louis Sébédio and William W. Christie
Volume 8. Advances in Lipid Methodology – Four
Edited by William W. Christie
Volume 7. Advances in Lipid Methodology – Three
Edited by William W. Christie
Volumes 1– 6 and 10. Out of print
Woodhead Publishing in Food Science, Technology and Nutrition
Trans fatty acids in

human nutrition
Second edition
Edited by
FRÉDÉRIC DESTAILLATS
Nestlé Research Center, Lausanne, Switzerland
JEAN-LOUIS SÉBÉDIO
UMR 1019, Plateforme d’exploration du métabolisme, INRA centre
de Theix, St Genès Champanelle, France
FABIOLA DIONISI
JEAN-MICHEL CHARDIGNY
UMR 1019 INRA Université Clermont I,
Clermont-Ferrand, France
Oxford Cambridge Philadelphia New Delhi

Published by Woodhead Publishing Limited,
80 High Street, Sawston, Cambridge CB22 3HJ, UK
www.woodheadpublishing.com
www.woodheadpublishingonline.com
Woodhead Publishing, 1518 Walnut Street, Suite 1100, Philadelphia,

PA 19102-3406, USA
Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road,
Daryaganj, New Delhi – 110002, India
www.woodheadpublishingindia.com
First published by The Oily Press, 2009

Reprinted by Woodhead Publishing Limited, 2012
© PJ Barnes & Associates, 2009; © Woodhead Publishing Limited, 2012

The authors have asserted their moral rights
This book contains information obtained from authentic and highly regarded sources.
Reprinted material is quoted with permission, and sources are indicated. Reasonable
efforts have been made to publish reliable data and information, but the authors and the
publisher cannot assume responsibility for the validity of all materials. Neither the authors
nor the publisher, nor anyone else associated with this publication, shall be liable for any
loss, damage or liability directly or indirectly caused or alleged to be caused by this book.
Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming and recording, or
by any information storage or retrieval system, without permission in writing from
Woodhead Publishing Limited.
The consent of Woodhead Publishing Limited does not extend to copying for general
distribution, for promotion, for creating new works, or for resale. Specific permission
must be obtained in writing from Woodhead Publishing Limited for such copying.
Trademark notice: Product or corporate names may be trademarks or registered trade-

marks, and are used only for identification and explanation, without intent to infringe.
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
ISBN 978-0-9552512-3-8 (print)

ISBN 978-0-85709-787-3 (online)
This book is Volume 23 in The Oily Press Lipid Library
Typeset by Ann Buchan (Typesetters), Middlesex, UK

Printed by Lightning Source
Preface
The chemistry of fats and oils has enjoyed a long and successful history. The
first evidence of the occurrence of trans fatty acids (TFA) in edible fats was
demonstrated by direct chemical analysis more than 80 years ago by Bertram in
1928 (Biochem. Z., 197, 433–441). In studying ruminant fats, Bertram discov-
ered the trans-11 18:1 acid and named it vaccenic acid. It was shown later that
vaccenic acid is not the only TFA found in ruminant fats and more recent
research revealed that vaccenic acid is further metabolized in ruminants as well
as in other animals.
Over the last thirty years numerous studies have been carried out in a number
of fields including analytical chemistry, food science, nutrition and epidemiol-
ogy to understand the composition, physical properties and health implications
of TFA found in partially hydrogenated vegetable oils. The basic chemical
information gained was that partial hydrogenation of vegetable oils generates
a very complex and diverse profile of TFA isomers. These TFA were found to
be conspicuously more stable toward chemical oxidation reactions compared
to their polyunsaturated precursors and to exhibit distinct physical properties.
However, the most recent research over the past two decades has documented
various detrimental effects of consumption of TFA on risk factors of vascular
health.
Since the 1990s there has been increasing regulatory concern about the
health effects of the trans mono-ethylenic acid isomers formed during partial
hydrogenation of vegetable oils. Consequently public health policies have
been implemented in various countries including Denmark, USA and Canada,
to ban the use or limit the consumption of TFA from industrial origin. However,
debate still rages around the world as to agreeing the most appropriate policies,
determining which specific chemicals are deleterious and by what mechanisms
and in what quantities, and deciding how regulatory agencies should guide the
public to appropriate food choices based on their TFA contents. To frame this
debate, scientific knowledge must take a central role. Therefore the editors
undertook to produce a state-of-the-art book that assembles the scientific
knowledge of trans fats – what is known and what needs to be determined.
An earlier book carried out a very similar task for the state of our knowledge
in the late 1990s; this was Trans Fatty Acids in Human Nutrition co-edited by
one of us ( Jean-Louis Sébédio) together with William W. Christie who also
contributes to the present book. Also like the present book, the earlier volume
was published by The Oily Press. Therefore it was decided that the new book
v
vi PREFACE
should become the Second Edition of Trans Fatty Acids in Human Nutrition
even though the rapid expansion and progress in the subject meant that it would
be completely re-written and be expanded from the original 9 to the present 15
chapters.
In this book, authors who are recognized international authorities in their
field have addressed various domains of TFA research such as consumption,
analysis, biochemistry, synthesis and natural TFA biosynthesis, health effects,
food formulation, and also regulation and consumer perception.
Each chapter contains the latest references and major advances and break-
throughs in the different areas of scientific research. Furthermore, the book also
includes a discussion of a major question on the health effects of the ‘natural
trans isomers’, comparing their effects to those observed for the industrially
produced TFA. We hope that the availability of so much information in a single
volume will help to clarify the major effects of TFA in human nutrition
discovered over the last two decades and guide the next generation of scientists
to the important opportunities for making further progress in this challenging
field of research.
Frédéric Destaillats (Lausanne, Switzerland)

Jean-Louis Sébédio (St Genès Champanelle, France)
Fabiola Dionisi (Lausanne, Switzerland)
Jean-Michel Chardigny (Clermont-Ferrand, France)
Contents
Preface v
List of Contributors xv
Prologue xix
1 Biosynthesis of trans fatty acids in ruminants 1

FRANCIS ENJALBERT AND ANNABELLE TROEGELER-MEYNADIER
A. Introduction 1
B. Physiology of rumen biohydrogenation 2
1. Lipolysis of dietary lipids: a prerequisite for biohydrogenation
2. Biohydrogenation of mono and polyunsaturated fatty acids
3. Extent of lipolysis and biohydrogenation
4. Microorganisms involved in lipolysis and biohydrogenation
5. Enzymes involved in biohydrogenation and their mechanisms
6. Intestinal digestion
C. Factors affecting rumen outflow of trans fatty acids 15
1. Factors affecting extent of biohydrogenation
2. Factors affecting trans-18:1 and CLA isomeric profile
D. Use of vaccenic acid for mammary synthesis of rumenic acid: 30
a brief comment
E. Conclusion 30
References 31
2 Formation of trans fatty acids during catalytic hydrogenation 43

of edible oils
JEAN-BAPTISTE BEZELGUES AND ALBERT J. DIJKSTRA
A. Mechanism of trans fatty acid (TFA) formation during 44
catalytic hydrogenation of edible oils and fats
1. Theories that turned out to be wrong
2. Currently accepted hydrogenation mechanism
B. Industrial hydrogenation of edible oils and fats 50
1. The hydrogenation process
2. Hydrogenation parameters
3. Reduction of formation of trans fatty acids (TFA) during partial
hydrogenation
D. Conclusion 59
References 60
vii
viii CONTENTS
3 Formation of trans fatty acids during deodorization of edible oils 65

JEAN-BAPTISTE BEZELGUES AND FRÉDÉRIC DESTAILLATS
A. Deodorization: a critical operation 65
B. Geometrical isomerization of linoleic acid and α-linolenic acid 66
C. Geometrical isomerization of long-chain polyunsaturated fatty 70
acids during marine oil deodorization
D. Technical solutions to prevent excessive isomerization of 72
polyunsaturated fatty acids during deodorization
E. Conclusion 74
References 74
4 Chemical synthesis of monounsaturated trans fatty acids 77

ZEPHIRIN MOULOUNGUI AND LAURE CANDY
A. Unsaturated fatty acids synthesis 77
1. The Wittig olefination reactions
2. The Horner-Emmons, Wadsworth-Emmons and Wittig-Horner
olefination reactions
3. The Wittig olefination under phase transfer catalysis conditions
4. Wittig type olefination applied to unsaturated fatty acid synthesis
B. Cis-trans isomerization 87
C. Fractionation techniques 94
1. Fractional crystallization in solvent
2. Dry fractionation
3. Fractionation in presence of a surfactant (‘Lanza’ process)
4. Comparison of the processes
D. Conclusions and outlook 99
References 101
5 Analysis of trans fatty acids of partially hydrogenated 105

vegetable oils and dairy products
W.M. NIMAL RATNAYAKE AND CRISTINA CRUZ-HERNANDEZ
A. Introduction 105
B. Analysis of partially hydrogenated oils 106
1. Methylation of oils, fats and lipids
2. Extraction and transmethylation of fat from food and tissue
samples
3. Cis and trans isomers of partially hydrogenated vegetable oils
CONTENTS ix
4. Analysis of cis and trans isomers of partially hydrogenated

vegetable oils
5. Analysis of trans isomers by capillary GC in combination with
silver-ion chromatography
6. GC-MS analysis of fatty acids
C. Analysis of ruminant fats 128
1. Ruminant Fats
2. Lipid extraction
3. Methylation
4. Capillary GC of FAME from dairy and meat fat
5. Analysis of monounsaturated FAME by combining
AgNO3-TLC and GC
6. Calculation of monounsaturated FAME
7. Separation and identification of CLA isomers using GC
and Ag+-HPLC
8. Identification of all CLA isomers
D. Summary 140
References 141
6 Replacement of partially hydrogenated oils in food products: 147

a technological challenge
GUILLERMO NAPOLITANO AND FRANCESCA GIUFFRIDA
A. Introduction 147
B. Modification of the hydrogenation process 147
C. Interesterified low-TFA fats 148
1. Chemical interesterification
2. Enzymatic interesterification
3. Applications
D. Modification of the food formulation 154
E. Oils with modified composition 156
1. High-oleic, non-genetically-modified sunflower oil
2. High-oleic, non-genetically-modified canola oil
3. Low-linolenic, high-stearic, non-genetically-modified soybean oil
4. High-oleic, genetically-modified soybean oil
5. Issues related to transgenic oilseed
F. Conclusions 158
G. Sources of further information and advice 159
References 159
x CONTENTS
7 Metabolism of trans fatty acid isomers 163

JEAN-LOUIS SÉBÉDIO AND WILLIAM W. CHRISTIE
A. Introduction 163
B. Metabolism of trans isomers in vivo 164
1. Absorption and incorporation into tissues
2. Desaturation and elongation
3. β-oxidation
4. Effects on the metabolism of other fatty acids
C. Metabolism of trans isomers in vitro 174
1. Incorporation into lipid classes
3. β-Oxidation
D. Effects on eicosanoid production 180
E. Human studies 184
References 188
8 Biosynthesis and biological activity of rumenic acid: 195

a natural CLA isomer
ADAM L. LOCK, JANA KRAFT, BETH H. RICE AND DALE E. BAUMAN
A. Introduction 195
B. The ruminant dimension 198
1. Origin of rumenic acid
2. The Δ9-desaturase enzyme system
3. Origin of other CLA isomers in milk fat
4. Endogenous synthesis of rumenic acid in humans and other
species
5. Factors affecting rumenic acid content in milk fat
6. Manufacturing and product quality considerations related to
rumenic acid enriched milk fat
C. Biological effects of rumenic acid: cancer 210
1. Cell culture studies
2. Animal studies
3. Human studies
4. Potential mechanisms of action
D. Biological effects of rumenic acid: atherosclerosis 216
1. Animal studies
2. Human studies
E. Summary 221
References 222
CONTENTS xi
9 Biosynthesis, synthesis and biological activity of 231

trans-10,cis-12 conjugated linoleic acid (CLA) isomer
DELPHINE TISSOT-FAVRE AND MARK WALDRON
A. Introduction 231
B. Biosynthesis 232
1. Biosynthesis in mammals
2. Production by bacteria, fungi and yeasts
3. Production of trans-10,cis-12 CLA isomer by genetically-modified
organisms and microorganisms
C. Chemical synthesis of trans-10,cis-12 CLA isomer 235
1. Alkaline isomerization of linoleic acid
2. Photocatalytic synthesis of CLA isomers
5. Purification of CLA isomers
D. Biological activity of trans-10,cis-12 CLA isomer 237
1. Effects of trans-10,cis-12 CLA isomer on body composition and
energy metabolism
2. Anti-obesity mechanisms of trans-10,cis-12 CLA isomer
3. Effects of trans-10,cis-12 CLA isomer on gene expression
4. Effect of trans-10,cis-12 CLA isomer in cancer
5. Effect of trans-10,cis-12 CLA isomer on immune function
6. Effects of trans-10,cis-12 CLA isomer on cardiovascular disease
7. Effect of trans-10,cis-12 CLA isomer on glucose tolerance and
insulin sensitivity
E. Conclusion 246
References 248
10 Observational epidemiological studies on intake of trans 255

fatty acids and risk of ischaemic heart disease
MARIANNE UHRE JAKOBSEN AND KIM OVERVAD
A. Background 255
1. Sources and intake
2. Biological effects of TFA and mechanisms
B. Methods 257
C. Results 258
1. Included studies
2. Ecologic and cross-sectional studies
3. Case-control studies
4. Follow-up studies
5. Summary of the study results
D. Discussion 297
xii CONTENTS
2. Quality of the studies

3. Specific TFA and risk of IHD
4. Conclusions and perspectives
Acknowledgements 302
References 302
11 Dietary Trans Fatty Acids and Cardiovascular Disease Risk 307

CORINNE MALPUECH-BRUGÈRE, BÉATRICE MORIO AND
RONALD P. MENSINK
A. Evidence from epidemiological studies 307
B. Evidence from metabolic, randomized studies 308
1. Effect of TFA on lipid and lipoprotein metabolism
2. Effect of TFA on haemostatic function
3. Effect of TFA on inflammatory markers and endothelial cell
function
4. Effect of TFA on endothelial cell function
5. Relation of cardiovascular disease risk to risk of type 2 diabetes
C. Conclusion 316
References 316
12 Dietary trans fatty acids: from the mother’s diet to the infant 319
JEAN-MICHEL CHARDIGNY AND NICOLE COMBE
A. Introduction 319
B. TFA in infant nutrition 320
1. TFA in human milk
2. TFA in infant formulas
C. TFA: from the mother to foetus and infant 322
1. Placental transfer
2. Incorporation into cord tissues
D. Effects of TFA on infant growth and development 323
E. TFA and essential fatty acid metabolism in newborns 324
F. Conjugated linoleic acid (CLA) isomers in infant nutrition 325
G. Conclusion 325
References 325
13 Evolution of worldwide consumption of trans fatty acids 329

MARGARET C. CRAIG-SCHMIDT AND YINGHUI RONG
A. Methods used to estimate trans fatty acid consumption 331
B. Estimates of worldwide consumption of trans fatty acids 335
CONTENTS xiii
1. North America
2. UK and Ireland
3. Continental Europe
4. Nordic Countries
5. Australia
6. Asian/Pacific Region
7. Central and South America
8. Iran
C. Consumption of trans fatty acids: future trends 372
Acknowledgement 374
References 374
14 Legislation relating to trans fatty acids 381

KOENRAAD DUHEM
A. Definition of trans fatty acids: a prerequisite to regulation 381
1. Australia and New Zealand
2. France
3. Definition excluding conjugated isomers of linoleic acids:
rationale
4. Denmark
5. Canada
6. USA
7. Codex Alimentarius definition
8. European Food Safety Authority definition
B. Regulatory decisions 383
1. CODEX/WHO/FAO
2. Denmark and Switzerland
3. UK
4. European Union
5. USA
6. Canada
7. Australia
8. Asia
C. Advantages and disadvantages of the different regulatory models 391
1. Advantages and disadvantages of banning TFA: lessons from
the Danish experience
2. Advantages and disadvantages of labelling TFA
D. Remarks and future perspective 392
References 393
xiv CONTENTS
15 Consumer concerns and risk perception related to 395

trans fatty acids
CLOTILDE AUBERTIN
A. Introduction and context around the consumer 395
1. Entities and interactions conditioning consumers’ concerns and
their risk perception
2. The Internet as an open window on ‘powerful consumer writing’
of health concerns and risk perception
B. Evidence of consumer’s attitudes and concerns about food and 398
health worldwide
1. Individual attributes and collective values: the construction of
health concerns
2. Consumer concerns related to trans fatty acids
C. The role of risk and benefit perceptions in the construction of 403
consumer attitude
1. Lessons from the past: food safety risk perception
2. Individual knowledge and consumer risks perception: overview
and case of trans fatty acids
3. Internet: agora or cacophony to clarify concerns about trans fatty
acids and to perceive risk
4. Influence and role of each entity concerning trans fatty acids
D. Conclusions and perspectives 410
References 411
Index 415
List of Contributors
Clotilde Aubertin, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

26, Switzerland, Switzerland
Dale E. Bauman, Department of Animal Science, Cornell University, Ithaca,

New York, NY 14853, USA
Jean-Baptiste Bezelgues, Nestlé Product Technology Center, 809 Collins

Avenue, Marysville, Ohio 43040, USA
Laure Candy, Laboratoire de Chimie AgroIndustrielle; ENSIACET, 4 Allées

Emile Monso, F-31029 Toulouse, France
Jean-Michel Chardigny , INRA, UMR1019, Clermont-Ferrand, 63000 France;

and CRNH Auvergne, 63000 Clermont-Ferrand, France
William W. Christie, Scottish Crop Research Institute and Mylnefield Lipid

Analysis, Invergowrie, Dundee DD2 5DA, Scotland
Nicole Combe, ITERG, Département de Nutrition, 33000 Bordeaux, France
Margaret C. Craig-Schmidt, Department of Nutrition and Food Science,

Auburn University, Auburn, AL 36849, USA
Cristina Cruz-Hernandez, Nutrient Bioavailability, Nestlé Research Centre,

PO Box 44, CH-1000 Lausanne 26, Switzerland
Frédéric Destaillats, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

26, Switzerland
Albert J. Dijkstra, Carbougnères, 47210 St Eutrope-de-Born, France
Fabiola Dionisi, Nestlé Research Center, PO Box 44, CH-1000 Lausanne 26,
Switzerland
Koenraad Duhem, Centre National Interprofessionnel de l’Économie Laitière

(CNIEL), 42 rue de Châteaudun, 75314 Paris CEDEX 09, France
xv
xvi CONTRIBUTORS
Francis Enjalbert, Université de Toulouse, INPT, ENVT, UMR 1289 Tan-

dem, Tissus Animaux, Nutrition, Digestion, Ecosystème et Métabolisme,
ENVT, 31076 Toulouse CEDEX 3, France; and INRA, UMR 1289 Tandem,
Tissus Animaux, Nutrition, Digestion, Ecosystème et Métabolisme, Chemin de
Borde-Rouge, Auzeville, 31326 Castanet-Tolosan, France
J. Bruce German, Department of Food Science and Technology, University of

California, One Shields Avenue, Davis, CA 95616-8598, USA
Francesca Giuffrida, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

26, Switzerland
Marianne Uhre Jakobsen, Department of Clinical Epidemiology, Aarhus

University Hospital, Sdr. Skovvej 15, Box 365, DK-9100 Aalborg, Denmark
Jana Kraft, Department of Animal Science, University of Vermont, 219

Terrill Hall, 570 Main Street, Burlington, VT 05405, USA
Adam L. Lock, Department of Animal Science, University of Vermont, 219

Corinne Malpuech-Brugère, Clermont Université, UFR Médecine, UMR

1019 Nutrition Humaine, F-63000, Clermont-Ferrand, France; and INRA,
UMR 1019 Nutrition Humaine, F-63122, Saint Genès Champanelle, France
Ronald P. Mensink, Maastricht University, Nutrition and Toxicology Re-

search Institute Maastricht, Department of Human Biology, Maastricht,
NL-6200 MD, The Netherlands
Béatrice Morio, Clermont Université, UFR Médecine, UMR 1019 Nutrition

Humaine, F-63000, Clermont-Ferrand, France; and INRA, UMR 1019 Nutri-
tion Humaine, F-63122, Saint Genès Champanelle, France
Zephirin Mouloungui, Laboratoire de Chimie AgroIndustrielle; ENSIACET,

4 Allées Emile Monso, F-31029 Toulouse, France
Guillermo Napolitano, Nestlé Product Technology Center, 809 Collins Av-

enue, Marysville, Ohio 43040, USA
Kim Overvad, Department of Clinical Epidemiology, Aarhus University

Hospital, Sdr. Skovvej 15, Box 365, DK-9100 Aalborg, Denmark
W.M. Nimal Ratnayake, Nutrition Research Division, Food Directorate,

Health Products and Food Branch, Health Canada, 251 Sir Frederick Banting
Driveway, Ottawa, Ontario K1A 0L2, Canada
CONTRIBUTORS xvii
Beth H. Rice, Department of Animal Science, University of Vermont, 219

Yinghui Rong, Department of Nutrition and Food Science, Auburn Univer-

sity, Auburn, AL 36849, USA
Jean-Louis Sébédio, UMR 1019, Unité de Nutrition Humaine, Plateforme

d’exploration du métabolisme, INRA centre de Theix, 63122 St Genes
Champanelle, France
Delphine Tissot-Favre, Nestlé Purina PetCare, Checkerboard Square - 1RS,

St Louis, MO 63164, USA
Annabelle Troegeler-Meynadier, Université de Toulouse, INPT, ENVT,

UMR 1289 Tandem, Tissus Animaux, Nutrition, Digestion, Ecosystème et
Métabolisme, ENVT, 31076 Toulouse CEDEX 3, France; and INRA, UMR
1289 Tandem, Tissus Animaux, Nutrition, Digestion, Ecosystème et
Métabolisme, Chemin de Borde-Rouge, Auzeville, 31326 Castanet-Tolosan,
France
Mark Waldron, Nestlé Research Center, PO Box 44, CH-1000 Lausanne 26,
Switzerland
Prologue
J. BRUCE GERMAN
University of California, Davis, California, USA, and Nestlé Research Center,

Lausanne, Switzerland
Few innovations in the brief history of industrialized food production have

been as interesting or eventful as the hydrogenation of edible fats. In many
ways this process serves as a model for how chemical innovation for one
purpose can set in motion a series of gradual events that produce long-term
consequences that impact on many aspects of agriculture, food and human
health. Hydrogenation of unstable, polyunsaturated commodity oils was one of
the first commercial successes of the industrialization of chemistry in the 19th
century. The scientific revolution of chemistry and the emergence of molecular
theory of matter were propelled in the laboratory by the ability to separate pure
elements. Separation of pure gases, including hydrogen, was in turn one of the
first of these chemical processes to be industrialized.
Once it was made available commercially, pure hydrogen, it could be said,
then went looking for real world applications. It found them. In a short time, the
ability of purified hydrogen to react spontaneously with unsaturated oils in the
presence of a simple catalyst and to alter their physical, chemical and biological
properties was one of the first technological successes of this new science:
applied chemistry. First developed as an industrialized process to stabilize
whale oil for lamp and candle making, the potential of this chemistry to create
unintended effects was evidenced by the increased viability of commercial
whaling eventually leading to the devastation of the world’s whale populations.
With commercial success, the process of chemical hydrogenation expanded.
Hydrogenation of plant oils to produce stable, functionally superior edible fats
was industrialized rapidly to produce margarines and shortenings as less
expensive substitutes for the more expensive animal fats – butter and lard.
Not surprisingly, the arrival of these substitutes was not met with enthusiasm
from all sectors. The first battles in what would become a century-long
political, regulatory and scientific war began as the producers of traditional
xix
xx PROLOGUE
edible fats (butter, lard, tallow) acted to have the cheaper and inferior hydro-
genated substitutes banned, regulated and labelled.
During the 20th century continuous improvements in the chemistry of hydro-
genation, from improved selectivity of catalysts and reactors to blending of
diverse feedstocks, provided considerable control over the overall hydrogena-
tion process and its products. The quality of hydrogenated oil products increased
dramatically and led to the development of highly tailored functional edible
fats in shortening, confectionery and baking applications. One by-product of
the reaction conditions designed to induce addition of hydrogen to fatty acids,
though largely unnoticed, was critical to their success. Within the reactor,
interaction of fatty acids with the metal catalyst surface led to the isomerization
of cis double bonds to trans configuration. Interestingly, the consequences for
the physical properties of the products were not undesired. As important
components of partially hydrogenated oils, the trans fatty acids in oleomarga-
rine and shortenings actually provided the industrial chemists with a valuable
functionality. The apparently simple structural differences between the cis and
trans isomer configuration of a double bond have profound physical conse-
quences for the triacylglycerol molecules containing them. For C18 fatty acids
in triacylglycerols, monounsaturated fatty acids with a cis double bond are
liquid at room temperature whereas monounsaturated fatty acids with a trans
double bond are solid fats. The trans fatty acids allowed the fats to match the
melting points and crystal habits specified for particular product applications.
Hence, the trans structure was part of the physical properties that propelled
these hydrogenated oils into the global food supply. Nonetheless, as well as
being an effective means of conferring the physical properties of traditional
fats, these hydrogenated fats provided the additional benefit of being an
inexpensive substitute for the more valuable animal products such as butter,
lard and tallow. This situation was changed by a new factor in the marketing of
edible fats: the risk of causing disease.
Research in the mid-1900s began to link diet to long-term health. In
particular, early studies that examined the newly developed biomarker of heart
disease risk, serum cholesterol, found that consuming very high levels of
saturated fatty acids and animal fats caused an elevation of plasma cholesterol.
Over time, even though the risk of heart disease was clearly a highly complex
and multi-faceted problem, saturated fats became viewed as the villain at the
‘heart’ of heart disease. In the media and the marketing of ingredients in the
food marketplace, worldwide attention turned to the composition of edible fats
and oils. For animal fats and their saturated fatty acids there was nowhere to
hide. The public, agriculture and the food industry were advised to lower the
abundance in foods, and the dietary intakes of, saturated fats. Into this oppor-
tunity, partially hydrogenated oils came running. Previously perceived as a
poor substitute for edible fats, the lower proportions of saturated fatty acids in
the hydrogenated vegetable fats was marketed as a nutritional advantage. The
PROLOGUE xxi
lack of routine chemical methods at the time to distinguish cis double bonds
from trans double bonds was not considered an issue: they were simply
unsaturated. Remarkably, little human clinical research was pursued to chal-
lenge the assumption that these hydrogenated alternatives to saturated fats,
with or without trans fatty acids, did not adversely affect blood lipids. Marga-
rines and shortenings were assumed to be healthier alternatives to animal fats
simply because they were not saturated. As a result of this perceived nutritional
advantage of hydrogenated vegetable oils, their health advantages were ag-
gressively marketed by producers and recommended by health agencies and
there was a rapid growth in the use of vegetable shortenings and margarines at
the expense of saturated fat commodities.
The scientific and public health perception that the presence of trans fatty
acids in the diet provided a beneficial, or at worst neutral, contribution to blood
lipoprotein cholesterol was changed dramatically by a single study published
by Mensink and Katan (1990). These investigators for the first time had access
to a source of trans fatty acids at relatively high concentration and could
examine trans fats as an independent experimental variable in a large human
clinical study. This first report of a well-designed and well-executed human
clinical study examining trans fatty acids as a single, independent variable
while measuring both low-density lipoprotein (LDL) and high-density lipopro-
tein (HDL) cholesterol in blood as discrete end points was striking. It became
one of the most public and widely read scientific papers on food ever published.
Results were contrary to the assumption that trans fats were harmless. In fact,
Mensink and Katan documented that the consumption of trans fatty acids by
normal humans raised blood levels of LDL and lowered HDL, both deleterious
actions in terms of the risk of heart disease. This landmark study was made
possible by the availability of experimental oils containing high concentrations
of trans fatty acids. An interesting historical note is that this experimental
material, provided by , was manufactured not by hydrogenation, but
by an isomerization reaction to ensure that the levels of trans isomers were
sufficiently high to test their biological effects.
The publication of a scientific study documenting the deleterious conse-
quences of trans fatty acids for risk factors of heart disease was met with some
debate from the various organizations associated with edible oils and fats. Not
surprisingly, nutrition and health scientists called for the removal of trans fatty
acids from foods. However, the relevance of the study to the normal human
population was questioned due to the high content of trans fatty acids in the
experimental diets in the Mensink and Katan study. Debate continued until the
1992 study by Zock and Katan that repeated the same basic clinical trial
protocols of the previous work, but included trans fatty acids in the diet at
approximately half the concentration of the previous study. This new study,
again using normal men and women, reported that the effects of a lower trans
fatty acid intake were consistent with a linear dose effect of trans fatty acids on
xxii PROLOGUE
raising LDL and lowering HDL cholesterol levels in blood. With these con-
vincing studies establishing that trans fatty acids were clearly not producing
metabolic effects similar to those of unsaturated fatty acids, considerable
scrutiny began to focus on whether they were truly deleterious to health as the
initial studies implied.
Various epidemiologic databases were rapidly polled to determine whether
or not the predictions that trans intakes would be deleterious to risk of heart
disease were consistent with human health. Indeed, trans intakes as an epide-
miological variable appeared to coincide not with a decrease in heart disease
risk but with an increase as predicted from the now known consequences to
lipoprotein concentrations. Unfortunately, to date no prospective clinical trial
has yet been conducted that could define unequivocally the effects of trans fats
alone. Because the use of these fats commercially is not simply to achieve a
health effect but also to obtain functionality in a wide variety of food products
from baked goods to fried foods, it remains practically impossible to distin-
guish trans fats in the diet as a truly independent variable. Those who have high
trans fat intakes also have distinct dietary habits in many ways. Nonetheless,
although definitive scientific evidence was lacking, and in fact may be impos-
sible to obtain in a normal human population, regulatory agencies and public
health agencies worldwide began to act.
The tale of trans fats is not simply related to chemical hydrogenation.
Scientists examining the determinants of fat compositions in animals had
discovered that one of the interesting microbial effects of rumen digestion and
fermentation in ruminant animals was the biological hydrogenation of polyun-
saturated fatty acids. Grazing ruminants consume a low fat intake but the fatty
acids in plants, especially leafy matter, are highly unsaturated. Why then are the
fat depots of these animals not polyunsaturated? The polyunsaturated fatty
acids are hydrogenated in the rumen prior to digestion and absorption by the
animal, hence the animal’s diet is effectively low in polyunsaturated fats. Even
more strikingly, a small proportion of the polyunsaturated fatty acids are
converted by microbial fermentation into the trans isomer of oleic acid,
vaccenic acid (trans 18:1n–11). This observation would have remained an
obscure fact of interest only to ruminant lipid biochemists except for one
completely unexpected finding. In studies attempting to identify potential pro-
carcinogenic substances in processed foods, Michael Pariza and colleagues
(Ha et al., 1989) made an astonishing discovery — an isomer of linoleic acid
(9c,12c, 18:2n–6) containing two double bonds but in the cis,trans conjugated
conformation, a substance found in foods from ruminants, was apparently anti-
carcinogenic in animals. In now hundreds of cellular and animal studies, the
molecule identified as conjugated linoleic acid CLA has shown significant
anti-carcinogenic properties. CLA is in fact a misnomer as the fatty acid is not
linoleic acid. As Dale Bauman and colleagues at Cornell University have
shown (Griinari et al., 2000), CLA is instead the biological product formed
PROLOGUE xxiii
when the rumen trans fatty acid vaccenic acid is converted to a polyunsaturated
acid 18:2 9c,11t, octadecadienoic acid, by the action of the Δ9 desaturase in
animals.
With these results emerging, a major dilemma began to shape up for
regulatory agencies. Not all trans fats are equal. The chemistry and chemical
composition of trans fats from industrial hydrogenation differ from those of
trans fats produced by biological hydrogenation in ruminants. Even more
perplexing, their health effects appear to be quite different as well. Discourag-
ingly for regulatory agencies, the complexity and small magnitude of effects on
health, combined with the invariably confounding co-occurrence of the differ-
ent trans fats with different dietary choices and even lifestyles, makes it
practically impossible to develop convincing scientific evidence of the inde-
pendent effects of the different trans isomers on actual human health outcomes.
The agencies must decide how to act in the absence of definitive scientific
evidence.
From the mid-1990s forward, trans fats became the subject of increasing
calls from scientists and public health officials for the regulation of their
content in foods and their inclusion on product labels. In 2003, the food
regulatory agency of Denmark banned the use of all hydrogenated fats from
food products, but at the same time made an explicit exception allowing the use
of animal fats containing natural trans fatty acids as these were viewed as
chemically different. In 2006, the USA’s mandatory labelling of all trans fats
irrespective of source came into effect. While it is too early to conclude
unequivocally, trans fats obtained by chemical hydrogenation are rapidly
disappearing from the industrial food supply. As for trans fats from ruminants,
the future is much more complex. It is not yet known whether the properties of
the different trans isomers of ruminants will have an important different effect
on the incidence of heart disease relative to chemically hydrogenated fats. Nor
is it yet known whether the remarkable anti-cancer properties of the trans
isomers in ruminant fats that have been observed in animals will translate into
similar benefits in lowering the risk of human cancers. Scientific research in
these areas will be important, and prior to its completion it is impossible to
know – time alone will tell.
The issue of trans fats has provided a controversy encompassing all of the
sciences related to diet and health. There are villains, heroes, wars and even
life and death. This book provides a unique opportunity to gain a detailed
and relatively comprehensive overview of the full spectrum of the issues of
trans fats from many of the scientists who have made the key discoveries
and who have brought the science to the level where it is today. The book is
very enjoyable reading for all those who appreciate science, mystery and
drama.
xxiv PROLOGUE
References
Mensink, RP, and Katan, MB (1990) Effect of dietary trans fatty acids on high-density and
low-density lipoprotein cholesterol levels in healthy subjects. New Eng. J. Med., 323, 439–
445.
Zock, P, and Katan, MB (1992) Hydrogenation alternatives: effects of trans fatty acids and
stearic acid versus linoleic acid on serum lipids and lipoproteins in humans. J. Lipid Res.,
33, 399–410.
Griinari, JM, Corl, BA, Lacy, SH, Chouinard, PY, Nurmela, KV and Bauman, DE (2000)
Conjugated linoleic acid is synthesized endogenously in lactating dairy cows by Delta(9)-
desaturase. J. Nutr., 130, 2285-2291.
Ha, YL, Grimm, NK and Pariza, MW (1989) Newly recognized anticarcinogenic fatty acids:
identification and quantification in natural and processed cheeses. J. Agric. Food Chem.,
37, 75–81.
CHAPTER 1
Biosynthesis of trans fatty acids in ruminants
FRANCIS ENJALBERT1,2 AND ANNABELLE TROEGELER-

MEYNADIER1,2
1
Université de Toulouse, INPT, ENVT, UMR 1289 Tandem, Tissus Animaux,
Nutrition, Digestion, Ecosystème et Métabolisme, ENVT, Toulouse, France
2
INRA, UMR 1289 Tandem, Tissus Animaux, Nutrition, Digestion, Ecosystème
et Métabolisme, Auzeville, Castanet-Tolosan, France
A. Introduction
Meat and milk from ruminants constitute an important natural source of trans
fatty acids (TFA) for humans. These TFA are formed from the lipids contained
in ruminant diets. Common forages and concentrates used to feed ruminants
typically contain a moderate amount of fat (i.e. under 5% of dry matter). The
main fatty acids (FA) found in ruminant diets are palmitic (16:0), oleic
(c9-18:1), linoleic (c9,c12-18:2) and α-linolenic (c9,c12,c15-18:3) acids. The
main lipids found in forages are pigments, and FA, the latter being normally
present in galactolipids and phospholipids and representing about 50% of total
fat. In grains and seeds, lipids are present mainly as triacylglycerols. In silages
(Lee et al., 2006a) and crushed or extruded feedstuffs stored for a long time
(Dierick & Decuypere, 2002), the main lipids are free FA formed by hydrolysis
of acyl ester linkages. In grains and peas, the level of fat is lower than 5%
whereas in oilseeds it ranges from 20% (dry matter basis) for soybean to 45%
for canola (rapeseed) and sunflower. In forages, c9,c12,c15-18:3 acid is the
most important FA except in corn silage where c9,c12-18:2 acid is the most
prevalent FA. In oilseeds, the main FA found depends on species. Indeed, in
canola and some cultivars of sunflower the main FA is c9-18:1 acid and in other
seeds (e.g. sunflower, soybean), c9,c12-18:2 is the main FA. In linseed
(flaxseed), the main FA found is c9,c12,c15-18:3 acid. Edible oils such as palm
oil (rich in 16:0 and c9-18:1), fats from terrestrial animals (rich in 16:0, 18:0
and c9-18:1), or marine oils rich in eicosapentaenoic (c5,c8,c11,c14,c17-20:5)
and docosahexadecenoic (c4,c7,c10,c13,c16,c19-22:6) acids can be used as
pure fat-based feedstuffs.
In ruminants, dietary FA undergo rumen microbial digestion. The two main
steps are lipolysis and biohydrogenation of unsaturated FA (UFA). During
1
2 TRANS FATTY ACIDS IN HUMAN NUTRITION
lipolysis, free FA are released from acyl lipids and during biohydrogenation,
UFA are sequentially reduced to saturated FA, and the final end product is
stearic (18:0) acid. The number of biohydrogenation steps depends on the
initial structure of the UFA and some intermediate TFA are formed. Due to the
continuous rumen outflow, some biohydrogenation intermediates can be
absorbed in the small intestine, metabolized and deposited in tissues or
excreted into milk fat.
B. Physiology of rumen biohydrogenation
1. Lipolysis of dietary lipids: a prerequisite for biohydrogenation

Ruminal lipolysis is responsible for the formation of free FA from galactolipids,
phospholipids and triacylglycerols. It is the first step of ruminal lipid metabo-
lism except for free FA or supplements such as FA calcium salts (see below).
Lipolysis is catalysed by bacterial exoenzymes attached to rumen particles
(Prins et al., 1975). Lipolysis of dietary triacylglycerols is almost complete
since low concentrations of partial acylglycerols are found in ruminal fluid
(Noble et al., 1974), which suggest that lipolysis of di- and monoacylglycerols
is much more rapid than lipolysis of triacylglycerols.
Free FA resulting from lipolysis remain mainly adsorbed onto feed particles.
It has been reported by Bauchart and co-workers (1990b), that some FA can be
incorporated into solid adherent bacteria (about 10 to 20% of bacterial FA) and
can be therefore protected against biohydrogenation. However, in other studies
(Kim et al., 2005; Or-Rashid et al., 2007; Kucuk et al., 2008) lower levels (3 to
6%) of PUFA in solid adherent or mixed rumen bacteria are reported. Rumen
bacteria can resynthesize esterified lipids taken up from free FA, as demon-
strated for vaccenic (t11-18:1) acid with Bifidobacterium adolescentis (Fukuda
et al., 2006c). This explains how esterified trans-18:1 acid isomers can be found
in very low concentrations in the duodenal flow (Atkinson et al., 2006). Results
obtained by Bauchart and co-workers (1990b) and Demeyer and co-workers
(1978) suggest that most of these bacterial lipids are phospholipids. The
phenomenon of bacterial re-esterification can bias the measurement of lipolysis
somewhat.
2. Biohydrogenation of mono and polyunsaturated fatty acids

Biohydrogenation is an extracellular process (Kim et al., 2005), which needs free
FA resulting from the lipolysis of triacylglycerols (Hawke & Silcock, 1969) or
galactolipids (Dawson 1974), or from the dissociation of FA calcium salts.
Ruminal biohydrogenation of UFA is mainly associated with the food-particle
fraction of the rumen (Harfoot et al., 1973a), so that removal of particulate
material reduces the biohydrogenation extent in vitro (Hawke & Silcock, 1970).
BIOSYNTHESIS OF TRANS FATTY ACIDS IN RUMINANTS 3
c9,c12,c15-18:3
c9,t11,c15-18:3
c9,c12-18:2
t10 ,c12-18 :2 c9,t11-18:2 t11,c13; t11,t13-18:2

t11,c15-18:2
t10,t12-; t9,t11-; t8-t10-; t7,t9-; t8,c10-18:2
c13-18:1
c12-18:1
c11-18:1
c15-18:1
t10-18:1 interconversions t11-18:1
t4-; t5-; t6+7+8-; t12-; t1 3+14-; t15- ; t16- 18:1
c9-18:1 t9-18:1
18:0
Figure 1. Major biohydrogenation pathways of oleic (c9 18:1), linoleic (c9,c12 18:2) and α-linolenic
(c9,c12,c15-18:3) acids in ruminants.
a. Biohydrogenation of oleic acid

Biohydrogenation of c9-18:1 has long been considered as direct (Figure 1),
leading without intermediate to 18:0 (Morris, 1970). However, feeding c9-18:1
to dairy cows increases trans-18:1 acid isomers in milk fat (Sellner & Schultz,
1980), and the amount of these TFA in the duodenal flow is similar when high
oleic or high linoleic sunflower oils are fed to dairy cows (Kalscheur et al.,
1997b). The isomerization of c9-18:1 in the rumen, recently demonstrated,
results in trans isomers from t6- to t16-18:1 acid isomers, elaidic (t9-18:1) acid
being the most abundant (Mosley et al., 2002; AbuGhazaleh et al., 2005). In
vitro, t9-18:1 acid can be hydrogenated to 18:0 or converted into t6- to t16-18:1
acid isomers, while 18% is isomerized from trans to cis, resulting in c11-18:1
and c9-18:1 acid isomers (Proell et al., 2002). It is unknown if formation of the
different trans-18:1 acid isomers from c9-18:1 acid is direct or comprises first
a geometrical isomerization from cis to trans followed by a positional isomeri-
zation (Proell et al., 2002).
These observations are consistent with the in vivo data obtained by Loor and
co-workers (2002a) showing that rumen concentrations of t6- to t9-18:1 acid
isomers are higher when cows are fed canola oil, which contains mainly
c9-18:1 acid, than soybean oil, which contains mainly PUFA. A strong rela-
tionship has been observed between c9-18:1 supply and ruminal t10-18:1
formation (Loor et al., 2002a). These biohydrogenation pathways are also
consistent with the positive correlation between c9-18:1 intake and concentra-
tion of t7,c9-18:2 conjugated acid isomer in milk fat (Collomb et al., 2004).
The t7,c9-18:2 acid almost exclusively originates from mammary desaturation
of ruminally derived t7-18:1 acid (Piperova et al., 2002, Corl et al., 2002).
It has been shown that the c9-18:1 biohydrogenation isomeric profile is
affected by the mass of the 13C labelled c9-18:1 acid (Mosley et al., 2006a).
However differences were not highly significant and therefore do not make
previous results (Mosley et al., 2002; Proell et al.; 2002, and AbuGhazaleh
et al., 2005) questionable.
Despite the formation of biohydrogenation intermediates, the concentrations
of 18:0 are much higher than those of trans-18:1 when c9-18:1 is incubated
with ruminal microorganisms (Mosley et al., 2002). These observations sug-
gest either a rapid reduction of biohydrogenation trans intermediates, or that
reduction of c9-18:1 acid is mainly direct. The hypothesis of a mainly direct
reduction is probable since the ruminal biohydrogenation of cis isomers is more
rapid than those of trans isomers (Kemp et al., 1984).
In the rumen, c9-18:1 can also be hydrated to 10-hydroxystearic acid
(Hudson et al., 1995). It has been shown that hydroxystearic acid is almost
exclusively formed from c9-18:1 acid and it is further converted to 10-
ketostearic acid but not to TFA (Jenkins et al., 2006).
b. Biohydrogenation of linoleic acid

The ruminal biohydrogenation of c9,c12-18:2 acid comprises three successive
steps: an isomerization and two reductions as described in Figure 1. Positional
and geometrical isomerization seems to be necessary before reduction of
c9,c12 double bonds (Kemp & Lander, 1984). Positional and geometrical
isomerization results in the formation of a conjugated linoleic acid (CLA)
intermediate with a trans double bond. The most abundant CLA intermediate
formed during biohydrogenation of c9,c12-18:2 is rumenic (c9,t11-18:2) acid
(Kepler et al., 1966; Kramer et al., 1998). The isomerization of the Δ9 double
bond can also occur, resulting in the formation of t10,c12-18:2 CLA isomer as
hypothesized by Griinari and Bauman (1999), and demonstrated by Kim and
co-workers (2002). Addition of c9,c12-18:2 acid to in vitro cultures also results
in formation of t8,t10-, t9,t11-, c9,t12-, t9,c12- and t9,t12-18:2 (Jouany et al.,
2007). When c9,c12-18:2 is added to cow’s diet, t10,t12-, t9,t11-, t8,t10-, t7,t9,
t8,c10-, t7,c9-18:2 CLA isomers are found in milk fat (Collomb et al., 2004;
Roy et al., 2006). These various isomers have been identified in the digestive
tract of ruminants as intermediates of rumen biohydrogenation (Shingfield
et al. 2003; Piperova et al., 2002). The precise pathways resulting in their
appearance remain unclear.
The second step is a reduction of the conjugated intermediates which mainly

affects the cis double bond and results in the formation of t11-18:1 and t10-18:1
acid isomers (Figure 1). Partial hydrogenation of t10,c12-18:2 to c12-18:1 acid
and subsequent isomerization to t12-18:1 acid seems to be responsible for the
occurrence of these isomers in digestive contents and milk fat (Loor & Herbein,
2003). The formation of t7-, t8- and t9-18:1 acid isomers could result from the
partial reduction of the t9,t11-, t8,t10-, t7,t9-, t8,c10-, t7,c9- and t8,c10-18:2
CLA isomers (Figure 1). The last step is the reduction of the remaining double
bond resulting in the formation of 18:0 acid. This last reduction is slower than
the isomerization and the first reduction, therefore the concentration of trans-
18:1 in the rumen or the digesta entering the duodenum is much higher than that
of CLA.
Hydration of linoleic acid has been reported by Morvan and Joblin (2001)
and Hudson and co-workers (1998) who identified the formation of 10-
hydroxy,12-18:1 and 13-hydroxy,9-18:1 acids.
c. Biohydrogenation of α-linolenic acid

The ruminal biohydrogenation of c9,c12,c15-18:3 also begins by an isomeriza-
tion, with the formation of a conjugated linolenic acid c9,t11,c15-18:3 isomer
(Kemp & Dawson, 1968). Alternatively, isomerization of c9,c12,c15-18:3 to
t10,c12,c15-18:3 (Griinari & Bauman, 1999) and c9,t13,c15-18:3 (Destaillats
et al., 2005) have been suggested, and are supported by identification of these
isomers in ewe’s cheese (Winkler & Steinhart, 2001) and in cow’s milk fat
(Destaillats et al., 2005), respectively. However, it can be hypothesized that
these isomers are formed by microbial isomerization in cheese or, in the case of
c9,t13,c15-18:3 acid, from a tissue Δ9-desaturation of t13,c15-18:2 CLA
isomer. Loor and co-workers (2005c) reported that the levels of t10-18:1 acid
in the duodenal FA is much lower with dietary linseed oil (source of c9,c12,c15-
18:3 acid) than sunflower oil (source of c9,c12-18:2 acid), suggesting a lower
importance of the t10 pathway with c9,c12,c15-18:3 than with c9,c12-18:2.
This isomerization is followed by three successive reduction steps (Fig-
ure 1). The first one results mainly in the formation of t11,c15-18:2 acid, which
represents a large proportion of duodenal FA in ruminants receiving diets
containing c9,c12,c15-18:3 acid (Loor et al., 2004; Loor et al., 2005c). A 10-
fold higher level of t11,c15-18:2 acid than c9,t11,c15-18:3 has been reported in
duodenal flow material of cows receiving linseed supplement (Enjalbert et al.,
2006). Rumenic (c9,t11-18:2) acid, which might be formed by reduction of
c9,t11,c15-18:3 acid, seems not to be found as an intermediate of c9,c12,c15-
18:3 acid biohydrogenation (Harfoot & Hazlewood, 1997). This hypothesis is
further supported by recent observations showing that increasing c9,c12,c15-
18:3 acid intake does not increase the duodenal flow of c9,t11-18:2 CLA
isomer (Lock & Garsnworthy, 2002), and that supplementation with linseed
results in an increase of duodenal t11,c15-18:2 acid 40 times greater than the
increase of c9,t11-18:2 acid (Loor et al., 2004). Other isomers have been
shown to be related to c9,c12,c15-18:3 acid biohydrogenation in vitro (Jouany
et al., 2007) and in milk (Kraft et al., 2003; Collomb et al., 2004). In the study
by Collomb and co-workers (2004), the most abundant intermediates found
were 12,14-18:2 and t11,c13-18:2 CLA isomers and they hypothesized that
these conjugated isomers are formed by isomerization of t11,c15-18:2 acid.
This isomerization has been recently demonstrated by Hino & Fukuda (2006).
In addition, other geometrical isomers such as c11,c13-, c11,t13-and t11,t13-
18:2 acids, have been identified in the duodenal or omasal content of cows or
in in vitro rumen cultures (Duckett et al., 2002; Piperova et al., 2002; Sackman
et al., 2003; Shingfield et al., 2003; Loor et al. 2004, 2005c; Jouany et al.,
2007). The t11,t13-18:2 CLA isomer is more abundant when the diet contains
linseed oil than sunflower oil (Loor et al, 2005c).
The second reduction step results in the formation of t11-18:1 and c15-18:1
acid isomers from t11,c15-18:2 acid (Figure 1). Vaccenic acid (t11-18:1) is also
formed during biohydrogenation of c9,c12-18:2 acid, and is usually the most
important trans-18:1 isomer produced in the rumen, independently of the fat
source. The c15-18:1 acid and its geometrical isomer t15-18:1 have been
reported in higher concentrations in the digesta of cows receiving linseed,
compared to cows receiving diets without fat (Loor et al., 2004) or diets with
sunflower oil or fish oil (Loor et al., 2005c). These observations are consistent
with the study performed by White and co-workers (1970) who showed in vitro
that incubation of c9,c12,c15-18:3 acid with rumen bacteria results in the
formation of large amounts of c15-18:1 acid, but also t15-, t13- and t14-18:1 acid
isomers. In lactating cows, high linseed diets also result in high proportions of
t12-, t16- and t13+t14-18:1 acid isomers, the latter being the most abundant
trans-18:1 following vaccenic acid (Loor et al., 2004; Loor et al., 2005c; Akraim
et al. 2006a). The separation of the t13- from t14-18:1 by gas-liquid chromatog-
raphy is very critical and it is difficult to identify further if there is a difference
between the production of t13 and t14-18:1 acids. It has been reported that 18:2
isomers having a t13 double bond are much more abundant in the rumen or the
duodenum than 18:2 acid isomers with a t14 double bond (Palmquist et al.,
2005), therefore it is probable that t13-18:1 acid isomer is produced at a higher
level than t14-18:1 acid isomer. However, White and co-workers (1970), using
labelled FA, observed that in vitro incubations of c9,c12,c15-18:3 produced four
times more t14-18:1 than t13-18:1 acid isomers.
Based on the data reported by Kemp and co-workers (1975), in the
biohydrogenation pathway proposed by Harfoot and Hazlewood (1997), c15-
18:1 and t15-18:1 acid are not further hydrogenated, but only limited data are
available to ascertain this fact. The kinetics in vitro study by Akraim and co-
workers (2006b) showed that all trans-18:1 isomers, including t15-18:1,
reached a maximal proportion of FA at 8 hours of incubation, and decreased
thereafter, which could be due to either a hydrogenation or an isomerization.
d. Biohydrogenation of other polyunsaturated fatty acids

The biohydrogenation pathway of γ-linolenic (c6,c9,c12-18:3) acid (Kemp &
Lander, 1983; Harfoot & Hazlewood, 1997) also involves an isomerization and
three reduction steps. However, this FA is found in minute amount in ruminant
diets, and therefore both biological significance and data are limited. The
c6,t11-18:2 acid, a supposed intermediate of c6,c9,c12-18:3 biohydrogenation,
has not been identified in the digestive contents of ruminants.
The biohydrogenation pathways of long-chain PUFA, mainly found in
marine algae or fish oil, have not been investigated. Ashes and co-workers
(1992) found that c5,c8,c11,c14,c17-20:5 and c4,c7,c10,c13,c16,c19-22:6
acids were not biohydrogenated by rumen microorganisms. However, it is now
well established that these long-chain PUFA are metabolized by ruminal
microorganisms (Gulati et al., 1999; AbuGhazaleh & Jenkins, 2004a). In a
study performed by Loor and co-workers (2005c) on cows supplemented with
fish oil, it has been observed that the duodenal flow of trans C18 FA isomers was
much greater than the amount of dietary C18 UFA that disappeared in the rumen.
In this study, the duodenal flow of C20 to C24 FA was much lower than the
intake, which could suggest a possible conversion of C20 to C18 FA (Loor et al.,
2005c). Shingfield and co-workers (2003) also observed a much lower duode-
nal flow of C20 to C24 FA than the dietary intake in cows receiving fish oil, but
did not observe an increased flow of C18 FA.
In conclusion, the best known pathways of rumen biohydrogenation of UFA
in the rumen result in the production of t11 intermediates. However, a very
large number of other positional and geometrical isomers can be formed. Some
of these 18:1 isomers were identified many years ago in ruminant digesta in
vitro (Ward et al., 1964), in vivo (Bickerstaffe et al., 1972), and in milk fat
(Parodi, 1976). Recent data provided further insight on the distribution of
trans-18:2 isomers and they help better to understand the biohydrogenation
pathways.
3. Extent of lipolysis and biohydrogenation

The extent of biohydrogenation is evaluated by measuring the disappearance of
either UFA or double bonds in the rumen or in in vitro cultures (Fievez et al.,
2007). As outlined previously, lipolysis is a prerequisite step for isomerization
of PUFA or direct reduction of monounsaturated FA.
Methodology can affect the measurement of extent of biohydrogenation. In
vivo, the determination is based on the comparison of FA profile between
dietary intake and duodenal profiles or flows. This latter method needs the
measurement of duodenal flows using digestive markers, and provides esti-
mates of biohydrogenation extent different to those obtained by comparison of
profiles. This is mainly due to duodenal flows that are higher than intake with
low-fat diets but lower with high-fat diets (Doreau & Chilliard, 1997; Schmidely
et al., 2008). Analytical problems such as coelution of dietary UFA with their
isomers can also result in biased estimations. In vitro measurements usually
result in slower biohydrogenation rates and lower biohydrogenation extents
than in vivo studies, with high variability among studies (Moate et al., 2004).
Lipolysis is rapid and in earlier in vitro studies lipolysis was considered to
happen completely (Garton et al., 1958). Later it was shown that between 50
and 70% of dietary galactolipids are hydrolysed within 1 hour after a meal in
sheep (Dawson et al., 1974), and then that about 5% of lipids entering the
duodenum of goats are mono-, di- or triaclylglycerols or phospholipids
(Bickerstaffe et al., 1972). Similarly, Bauchart and co-workers (1990a) showed
that more than 90% of dietary triacylglycerols disappear before the duodenum.
Overall, using results from published in vivo trials, Moate et al. (2004) esti-
mated that more than 80% of lipids are hydrolysed in the rumen.
In vitro data using incubation of triacylglycerols do not suggest that the rate
of lipolysis can limit the biohydrogenation rate (Hawke & Silcock, 1970; Beam
et al., 2000). Similarly, based on eight experiments, Moate and co-workers
(2004) demonstrated that lipolysis rates were in most cases higher than
Table 1. Extent of biohydrogenation of unsaturated C18 acids and relative importance of

trans intermediates and 18:0 measured by comparison between intake and duodenal or
abomasal flow, reported by Duckett et al. (2002), Piperova et al. (2002), Sackmann et al.
(2003), Shingfied et al. (2003), Loor et al. (2004), Lundy et al. (2004), Lee et al. (2005),
Loor et al. (2005c), Akraim et al. (2006a)* and Lee et al. (2006a).
Main added fat Percentage of

concentrate
No Source of Source of Fish oil** ≤ 50 > 50
added fat c9,c12-18:2 c9,c12,
c15-18:3
Observations 13 12 6 4 20 15
Biohydrogenation extent, %
c9-18:1*** 56.9 73.1 80.9 69.8 67.7 69.4
c9,c12-18:2 83.5 90.2 86.4 88.8 89.2 84.7
c9,c12,c15-18:3 89.2 90.6 92.0 92.0 92.2 89.1
Biohydrogenation completion, %
trans-18:2/ 1.27 0.87 8.91 8.16 2.51 4.59
disappeared PUFA
trans-18:1/ 15.9 17.3 31.7 55.5 22.2 27.7
disappeared UFA
appeared 18:0/ 86.5 55.4 46.6 49.2 68.9 64.8
disappeared UFA
*Biohydrogenation measured using fatty acid profiles. **Either sole fat source or associated with
another fat source. ***Values from Lee et al. (2006a), which were unexpectedly low due to a diet with
a low concentration of c9-18:1, were not taken into account. PUFA, polyunsaturated fatty acids; UFA,
unsaturated fatty acids.
biohydrogenation rates in vivo. Recently, Atkinson and co-workers (2006)

showed that the total duodenal flow of c9,c12-18:2 acid does not significantly
differ compared to esterified c9,c12-18:2 acid in sheep receiving 0 to 9%
safflower oil. This observation suggests an incomplete lipolysis, but more than
94% of dietary c9,c12-18:2 acid disappeared before the duodenum, whatever
the amount of added safflower. Therefore, large amounts of esterified c9,c12-
18:2 in the duodenum are not necessarily a result of incomplete lipolysis,
because bacterial uptake and re-esterification of free FA can occur.
Typical reported values for the extent of biohydrogenation are >70% for c9-
18:1 acid, >80% for c9,c12-18:2 acid and about 90% for c9,c12,c15-18:3 acids
(Doreau & Ferlay, 1994; Enjalbert, 1995; Jenkins & Bridges, 2007). In Table 1
results are reported from recent experiments in which duodenal or abomasal FA
flows were measured and FA analysed with chromatographic methods suitable
for separation of the main cis and trans isomers. The biohydrogenation rates
determined by Moate and co-workers (2004) from in vivo experiments were
27.4, 87.6 and 243.9%/h for c9 18:1, c9,c12 18:2 and c9,c12,c15 18:3 acids,
respectively. In this model, the estimated biohydrogenation rate of trans-18:1
acid isomers was 22.8%/h. Biohydrogenation kinetic data obtained in vivo
have not been published for the other TFA. However, since trans-18:3 or trans-
18:2 intermediates are found in very small amounts in the rumen or the
duodenal flow, compared to their precursors and trans-18:1 acid, their
biohydrogenation rates can be assumed to be very high. Some differences
between isomers have been observed: results from Gulati and co-workers
(2000) and Loor and Herbein (2003) suggested that t10,c12-18:2 acid is more
resistant to biohydrogenation than c9,t11-18:2 acid. It seems that t11,c15-18:2
acid is rather resistant to biohydrogenation since this acid is found in higher
proportion than any trans-18:2 isomers in duodenal fluids when cow’s diet is
supplemented with c9,c12,c15-18:3 acid (see Table 2).
The extent of biohydrogenation of c5,c8,c11,c14,c17-20:5 and
c4,c7,c10,c13,c16,c19-22:6 acids measured in vitro seems to be lower than for
C18 UFA (Chow et al., 2004; Sinclair et al., 2005b). However, when the extent
of biohydrogenation of c5,c8,c11,c14,c17-20:5 and c4,c7,c10,c13,c16,c19-
22:6 acids were measured in vivo, levels around 90% were reported (Shingfield
et al., 2003; Sinclair et al., 2005a; Loor et al., 2005c).
4. Microorganisms involved in lipolysis and biohydrogenation

The best known bacterium responsible for ruminal lipolysis is Anaerovibrio
lipolytica, which produces a cell-bound esterase and an extracellular lipase
(Harfoot, 1978). Growth of Anaerovibrio lipolytica is decreased at pH 5.7 and
inhibited at pH 5.3 (Hobson, 1965); the optimal pH for this bacterium is 7.4
(Henderson, 1971). Its lipase hydrolyses triacylglycerols but not phospholipids
and galactolipids, which can be attacked by Butyrivibrio strains (Harfoot &
10
Table 2. Average percentages of main trans fatty acids (g/100 g of total fatty acids (number of observations)) in the duodenal or abomasal
flow, reported by Duckett et al. (2002), Piperova et al. (2002), Sackmann et al. (2003), Shingfied et al. (2003), Loor et al. (2004), Lundy et
al. (2004), Lee et al. (2005), Loor et al. (2005c), Akraim et al. (2006a) and Lee et al. (2006a).
Main added fat Percentage of concentrate

No added fat Source of c9,c12-18:2 Source of c9,c12,c15-18:3 Fish oil* ≤ 50 >50
t6+t7+t8-18:1 0.35(12) 1.16 (2) 0.87 (6) 1.16 (4) 0.57 (17) 0.97 (7)
t9-18:1 0.25 (13) 0.64 (12) 0.55 (6) 0.80 (4) 0.39 (20) 0.64 (15)
t10-18:1 0.86 (13) 8.35 (12) 2.89 (3) 4.98 (4) 1.53 (17) 7.60 (15)
t11-18:1 3.90 (13) 5.22 (12) 10.44 (3) 16.85 (4) 7.96 (17) 5.11 (15)
t12-18:1 0.50 (13) 1.04 (13) 1.01 (6) 1.50 (4) 0.89 (20) 0.88 (15)
t13+t14-18:1 1.16 (12) 2.41 (11) 3.76 (6) 2.49 (2) 1.76 (14) 2.76 (7)
t15-18:1 0.71 (12) 0.94 (2) 1.82 (6) 1.09 (4) 1.02 (17) 1.19 (7)
c9,t11-18:2 0.112 (9) 0.093 (9) 0.154 (6) 0.244 (4) 0.180 (15) 0.080 (13)
t9,t11-18:2 0.034 (10) 0.034 (6) 0.122 (3) 0.068 (12) 0.013 (7)
t10,c12-18:2 0.061 (13) 0.114 (9) 0.016 (3) 0.059 (4) 0.072 (14) 0.073 (15)
c11,t13-18:2 0.009 (7) 0.017 (8) 0.011 (3) 0.050 (1) 0.013 (4) 0.016 (15)
t11,t13-18:2 0.038 (7) 0.021 (6) 0.160 (3) 0.059 (2) 0.063 (6) 0.051 (12)
t11,c15-18:2 0.41 (3) 0.86 (1) 2.97 (6) 3.30 (2) 1.68 (7) 2.95 (5)
c9,t11,c15-18:3 0.163 (3) 0.0163 (3)
TRANS FATTY ACIDS IN HUMAN NUTRITION
* Either sole fat source or associated with another fat source.

Hazlewood, 1997). Butyrivibrio bacteria can also hydrolyse triacylglycerols

(Latham et al., 1972; Paillard et al., 2007a). The involvement of rumen proto-
zoa in the lipolysis remains unclear.
Similarly, protozoa must play a minor direct role in the isomerization of
c9,c12-18:2 acid since this FA does not disappear when incubated with
protozoa alone (Girard & Hawke, 1978; Devillard et al., 2006), and removal of
protozoa from the rumen (defaunation) does not modify the metabolism of
c9,c12-18:2 acid (Dawson & Kemp, 1969; Devillard et al., 2006). However,
c9,c12-18:2 acid mainly results in trans-18:1 when incubated with bacteria
only, as opposed to 18:0 when incubated with both bacteria and holotrichs
(Girard & Hawke, 1978). It has been shown that supplementation with dietary
algae simultaneously increases biohydrogenation intermediates and disappear-
ance of ciliates (Boeckaert et al., 2007a) suggesting either an interaction of
protozoa with the action of bacteria, or a possible implication of protozoa in the
last biohydrogenation reduction step.
Despite the lack of biohydrogenation capacity, protozoa were shown to
contain more CLA and t11-18:1 acid than do bacteria (Devillard et al., 2006;
Or-Rashid et al., 2007). These FA are probably incorporated preferentially in
protozoa after bacterial formation (Devillard et al., 2006). Jenkins and co-
workers (2008) hypothesized that during digestion of bacteria inside protozoa,
the first steps of biohydrogenation could continue for a longer time than the last
step, resulting in a high t11-18:1 acid concentration. Ciliate protozoa are
retained selectively within the rumen, so that their high content of trans
intermediates could result in lower proportions of trans-18:2 and trans-18:1
FA isomers in the duodenal flow than in the rumen (Devillard et al., 2006). In
the experiments where both rumen and duodenum were sampled for FA
analysis (Akraim et al., 2006a; Loor et al., 2005b and 2005c), the proportion of
CLA among FA was from 2 to 12 times lower in the duodenum than in the
rumen, whereas duodenal and ruminal proportions of trans-18:1 acid where
similar. In spite of this selective retention, Yanez-Ruiz and co-workers (2006)
demonstrated that protozoa can account for 40% of duodenal t11-18:1 and
c9,t11-18:2 acids contents.
Rumen fungi have a limited ability to biohydrogenate c9,c12-18:2 acid
(Nam & Garsworthy, 2007; Maia et al., 2007).
The involvement of bacteria in the biohydrogenation process was studied
extensively during the 1960s and 1970s, and resulted in the identification of
rumen bacterial strains that can hydrogenate UFA. Very recent data, based on
molecular methods analysing the sequence of ribosomal RNA, provide further
insights into the phylogenetic relationships between biohydrogenating bacteria.
Only a limited number of strains can hydrogenate FA. Kemp and co-workers
(1975) found that among 200 bacterial strains, 30 showed biohydrogenation
activity and only 5 of these 30 strains had significant capacity. Unfortunately,
many of the original well-characterized biohydrogenation strains have been
lost (Van de Vossenberg & Joblin, 2003) and further studies of their character-
istics are not possible. These results have been extensively reviewed by
Harfoot and Hazlewood (1997). During the last 10 years, some other strains
have been isolated (Kim et al., 2002: Megasphaera elsdenii YJ-4; Van de
Vossenberg & Joblin, 2003: Butyrivibrio hungatei Su6; Fukuda et al., 2005:
B. fibrisolvens TH1; Fukuda et al., 2006a: B. fibrisolvens MDT-5; Fukuda
et al., 2006b: B. fibrisolvens MDT-10; Maia et al., 2007: Clostridium
proteoclasticum B316 and P-18; Wallace et al., 2007: Propionibacterium
acnes), and Paillard et al. (2007a) recently demonstrated the biohydrogenation
capacity of many Butyrivibrio-like bacteria.
Butyrivibrio fibrisolvens is the most extensively investigated bacterium
(Polan et al., 1964; Kepler et al., 1966). Some strains are also cellulolytic, but
their capacity to digest cellulose is limited (Halliwell & Bryant, 1963). Polan
and co-workers (1964) showed that Butyrivibrio fibrisolvens is able to hydro-
genate c9,c12-18:2 to 18:1 acid isomers but not to 18:0 acid, concluding that
another biohydrogenation system was necessary for complete reduction. Moreo-
ver, they demonstrated an enhanced biohydrogenation activity when
Butyrivibrio fibrisolvens was incubated with two species of rumen bacteria:
Megaspahaera elsdenii (old name Peptostreptococcus elsdenii) and
Selenomonas strain 233 (Polan et al., 1964). Butyrivibrio fibrisolvens cannot
metabolize c5,c8,c11,c14,c17-20:5 or c4,c7,c10,c13,c16,c19-22:6 acids
(Wasowska et al., 2006; Maia et al., 2007).
Kemp and Lander (1984) proposed the classification of hydrogenating
bacteria into two groups: group A bacteria, which hydrogenate c9,c12-18:2
and c9,c12,c15-18:3 mostly to t11-18:1 or t11,c15-18:2, but not 18:0, and
group B bacteria which hydrogenate c9,c12,c15-18:3 to t11,c15-18:2, t15- or
c15-18:1, and hydrogenate c9,c12-18:2, c9-18:1 and t11-18:1 to 18:0. Earlier
studies isolated only a few group B bacteria, including two bacteria of the
genus Fusocillus (Harfoot, 1978). The Butyrivibrio hungatei Su6 strain is able
to complete the biohydrogenation of both c9-,c12-18:2 and c9,c12,c15-18:3 to
18:0, and therefore does not completely fit with this classification of bacteria
into A and B groups (Van de Vossenberg & Joblin, 2003). Fusocillus spp. and
Butyrivibrio hungatei are phenotypically similar. Phylogenetic analysis based
on 16S rRNA indicates that Butyrivibrio hungatei clusters with Clostridium
proteoclasticum, a specific branch of the Butyrivibrio tree (Wallace et al.,
2006; Paillard et al., 2007a; Jenkins et al., 2008; Wallace, 2008). Clostridium
proteoclasticum population represents between 2 and 9% of the rumen
eubacterial community; variations can be due to diets and to a higher extent to
inter-individual variations (Paillard et al., 2007b).
Hydration of UFA in the rumen is due to facultative anaerobic bacteria, the
main type of which is Sreptococcus bovis, but several strains of Streptococcus,
Staphylococcus, Lactobacillus, Enterococcus, and Pediococcus can also cata-
lyse this reaction (Hudson et al., 2000).
5. Enzymes involved in biohydrogenation and their mechanisms
a. Enzymes involved in PUFA isomerization

Kepler and Tove (1967) isolated the c12 to t11 18:2 isomerase from Butyrivibrio
fibrisolvens, and the amino acid sequence was later determined by Park and co-
workers (1996). The CLA production by this enzyme has been later investigated
by Kim and co-workers (2000). The enzyme is bound to the cellular membrane,
and can isomerize c9,c12-18:2 at pH ranging from 5.5 to 8.5, the activity being
maximal at pH 7.0 to 7.2 and the activity at pH 6.0 being about half that at pH
7.0 to 7.2. Because the biohydrogenation of c9,c12,c15-18:3 also begins by a
c12 to t11 isomerization, Kepler and Tove (1967) assumed that it was due to the
same isomerase. This enzyme has an absolute requirement for a c9,c12 diene
system, so that c9,t12-18:2, t9,c12-18:2 and t9,t12-18:2, are not attacked by
Butyrivibrio fibrisolvens (Kepler et al., 1966). UFA without this diene system
exert a competitive inhibition on the enzyme, which works only on free FA
(Kepler et al., 1970).
When c9,c12-18:2 acid is incubated with Butyrivibrio fibrisolvens in a
medium containing deuterated water, a deuterium enrichment of c9,t11-18:2
acid is observed (Wallace et al., 2007). These authors proposed a mechanism
of isomerization starting with a hydrogen abstraction from the carbon 11 of
c9,c12-18:2, followed by shift of the double bond between carbon 12 and 13 for
thermodynamic stability, and finally by abstraction of a hydrogen atom from
water. Other 9,11-18:2 geometric isomers could result from the same pattern,
but simply be less abundant because they are less favourable energetically
(Wallace et al., 2007).
Some strains of Lactobacillus can synthesize CLA from 10-hydroxy,12-18:1
and 13-hydroxy,9-18:1 acids (Ogawa et al., 2005). However, hydrated 18:2
acid is not likely to be an intermediate of the isomerization of c9,c12-18:2 acid
to CLA in the rumen (Wallace et al., 2007).
As outlined above, ruminal digestion of UFA not only produces t11 isomers,
but also a large range of trans-18:2 or trans-18:1 acid isomers, and with 18:2
acid isomers with two trans double bonds also being formed. These isomers are
not explained by the c12 to t11 isomerase described by Kepler and Tove (1967).
Palmquist and co-workers (2005) hypothesized that these isomers could result
either from a low specificity of the described isomerase, or from the existence
of different isomerases that remain unidentified. Recently, Kim and co-work-
ers (2002) demonstrated that a strain of Megaspahaera elsdenii produced
mainly the t10,c12-18:2 CLA isomer. However, Wallace and co-workers
(2007) were not able to obtain t10,c12-18:2 with Megaspahaera elsdenii, but
did obtain it with Propionibacterium acnes. The isomerase of this bacterium
has been isolated (Deng et al., 2007) and its mode of action described by
Liavonchanka and co-workers (2006). The mechanism comprises a hydride
abstraction from carbon 11, followed by a migration of the double bond and a
reintroduction of a hydride on carbon 9. This reaction needs a bound flavin

adenine dinucleotide (FAD), and does not involve an exchange of hydrogen
with water, which is consistent with the low incorporation of deuterium in
t10,c12-18:2 when c9,c12-18:2 is incubated in vitro with deutered water
(Wallace et al., 2007). These results obtained with pure strains of bacteria show
that the mechanisms of isomerization depend on the bacterial species and
suggest that several isomerases are involved in biohydrogenation. Indeed,
when incubating c9-18:1, only t6- to t10-18:1 acid isomers are formed at pH 5.5
contrasting with t6- to t16-18:1 acid isomers at pH 6.5, indicating that more
than one isomerase is involved, some of them being inhibited by a low pH
(AbuGhazaleh et al., 2005).
b. Enzymes involved in double bond reduction

The second PUFA biohydrogenation step is a reduction. The reductase isolated
from Butyrivibrio fibrisolvens is bound to the membrane, its activity depends
on iron and reduced α-tocopherolquinol, and the optimum pH for activity is
between 7.2 and 8.2 (Hughes et al., 1982). C18 UFA increase its expression at
transcriptional level (Fukuda et al., 2006). Butyrivibrio fibrisolvens can also
reduce t10,c12-18:2 acid isomer, suggesting that its reductase is not very
substrate-specific. The same reductase can also hydrogenate c9,t11,c15-18:3
to t11,c15-18:2 acid, and hydrogenate t11,c15-18:2 or t11,c13-18:2 to t11-18:1
acid (Hino & Fukuda, 2006). Most microorganisms that are able to hydrogen-
ate these 18:2 acid isomers to t11-18:1 also exhibit capacity to hydrolyse
phospholipids (Hazlewood et al., 1976).
The reductase(s) that hydrogenates 18:1 acid isomers to 18:0 acid has not
been studied. Only a few species are responsible for this reduction, which could
explain why the level of trans-18:1 acid isomers in the rumen is much higher
than the concentration of CLA isomers. The efficiency of hydrogenation of
18:1 double bonds depends both on their position and their geometric configu-
ration: hydrogenation is very efficient for c4- to c11-18:1 acid isomers, but the
efficiency is much lower for c12-18:1 isomer and negligible for c13-18:1 acid
isomer (Kemp et al., 1984). The hydrogenation of trans isomers is less efficient
than for cis isomers (Kemp et al., 1984).
Moreover, hydrogenation of 18:1 acid isomers, including trans isomers, to
18:0 is not necessarily direct (Proell et al., 2002). By analogy with double bond
migration occurring at high temperature and with a catalyst, Mosley and co-
workers (2002) suggested that the variety of positional and geometrical 18:1
acid isomers formed in the rumen is due to a multitude of cis/trans isomerases.
6. Intestinal digestion
FA that leave the rumen can be absorbed in the small intestine. Some data have
been published on the intestinal digestibility of total 18:1 TFA, with a range
c9,c12-18:2
Promoted by
corn silage
HC and/or low pH Decreased by
HC x high LA hay vs grass or grass silage a
HC x high LA x adaptation fat protection a
monensin HC and/or low pH b
high LAc (in vitro)
t10,c12-18:2 c9,t11-18:2
Decreased by
high LA substratec (in vitro)
extrusion
t10-18:1 t11-18:1
Decreased by
HC and/or low pH b
high LA substratec (in vitro)
extrusion
fish oil
a: property of fat source

b: inhibition of the reaction
18:0 c: saturation of the reaction
Figure 2. Major effects of diet on the three steps of rumen biohydrogenation of linoleic (c9,c12 18:2,
LA) acid and the ratio of t11- to t10 18:1 acid isomers (adapted from Troegeler-Meynadier et al., 2006a).
HC, high concentrate diet.
from 82 to 96%, both when measured in the small intestine (Enjalbert et al.,
1997) or in the small intestine plus the hindgut (Romo et al., 2000; Loor et al.,
2004; Loor et al., 2005c). This is similar or slightly greater than values
observed for c9-18:1 acid. Loor and co-workers (2004; 2005c) reported similar
digestibility levels for various 18:1 and 18:2 TFA isomers ranging from 32 to
100%, with the lowest values for c9,t11-18:2 acid. In a recent review, Glasser
and co-workers (2008) reported apparent digestibility levels of 82.4% (45
observations) and 44.8% (17 observations) for t11-18:1 and c9,t11-18:2 acids,
respectively.
It has been reported that desaturation of 18:0 to c9-18:1 acid can happen in the
intestinal mucosa (Bickerstaffe et al., 1972), but Mosley and co-workers (2006b)
failed to demonstrate intestinal conversion of t11-18:1 to c9,t11-18:2 acid.
C. Factors affecting rumen outflow of trans fatty acids

Because trans FA are intermediates of UFA biohydrogenation, the amount
flowing out of the rumen depends on the efficiency of all the biohydrogenation
steps. The outflow of TFA increases with the efficiencies of lipolysis and
isomerization of dietary UFA, and decreases when the efficiency of the last
reduction step increases. Moreover, dietary factors can modulate the isomeric
profile of TFA. These major effects of diet on rumen biohydrogenation are
summarized in Figure 2.
1. Factors affecting extent of biohydrogenation

Factors that affect the extent of biohydrogenation are the diet composition, the
amount and treatment of fat source, and the use of additives.
a. Effect of diet composition

The effects of diet composition are related to forage source, proportion of
concentrates and nutrient composition.
In terms of forage source, it has been demonstrated with silage in dairy cows
(Dewhurst et al., 2003), with silage in steers (Lee et al. 2003 and 2006a) and
with fresh forage in vitro (Loor et al., 2003) that red clover lowers by 7 to 8%
the biohydrogenation extent of c9,c12,c15-18:3 acid compared to grass. In the
experiment by Dewhurst and co-workers (2003), white clover and alfalfa
silages resulted in intermediate values. The rate of transfer of dry matter from
the rumen was higher with white clover and alfalfa silages than for grass and
red clover silages (Dewhurst et al., 2003). This could explain the lower
biohydrogenation extent with white clover and alfalfa than with grass silage,
but not the difference between red clover and grass silages. The occurrence of
polyphenol oxidase in red clover could explain these differences since it has
been shown that the extent of lipolysis decreases in vitro when switching from
low to high polyphenol oxidase red clover lines (Lee et al., 2004 and 2007b).
FA in hay are less biohydrogenated than in fresh grass (Ribeiro et al., 2005
and 2007) or in both fresh grass and silage (Doreau et al., 2005). Hay usually
contains lower amounts of FA than fresh or ensiled grass, therefore the
duodenal flows of PUFA differ resulting in much lower TFA flows. Prelimi-
nary results suggest that FA oxidation products that are formed during early
wilting of grass increase PUFA biohydrogenation (Lee et al., 2007a).
The proportion of concentrates in the diet has an effect on biohydrogenation.
High concentrate diets lower ruminal pH and therefore affect the
biohydrogenation extent. In vitro studies demonstrated that at low pH, the
biohydrogenation of c9,c12-18:2 and c9,c12,c15-18:3 acids is less (Martin &
Jenkins, 2002; Troegeler-Meynadier et al., 2003; Ribeiro et al., 2007). This is
consistent with previous observations indicating that the optimal activity of the
Butyrivibrio fibrisolvens isomerase is reached at pH 7.0. Troegeler-Meynadier
and co-workers (2006a) showed that the inhibition begins at pH over 6.1. In
these experiments, donor animals received high forage diets. Comparing in
vitro cultures using donor cows receiving successively a low and a high
concentrate diet, Latham and co-workers (1972) and Gerson and co-workers
(1985) showed that the lipolysis and biohydrogenation of PUFA are slower
when the donor cows receive a high concentrate diet, and explained this effect
by the decreased number of Butyrivibrio spp.. The inhibition of lipolysis at low
pH increases with oil concentration (Van Nevel & Demeyer, 1996b). Overall,
these results indicate that both ruminal bacteria adapted to high concentrate
diets and ruminal bacteria adapted to a ruminal pH over 6.0 or lower have a low
ability to isomerize PUFA. However, Choi and co-workers (2005) showed only
minor effects of culture pH on biohydrogenation, but found that
biohydrogenation was more active with rumen bacteria from cows receiving a
high concentrate diet and having a 5.6 ruminal pH than with rumen bacteria
from cows receiving a low concentrate diet and having a 6.8 ruminal pH.
In vivo, the biohydrogenation extent of PUFA decreases with increasing
proportion of concentrate (Table 1). Sackmann and co-workers (2003) ob-
served, within a narrow range of concentrate (64 to 88%), a decrease of
biohydrogenation for c9,c12-18:2 but not for c9,c12,c15-18:3 acids. Similarly,
Lee and co-workers (2006b) observed when increasing concentrate level from
20 to 40% a decrease of c9,c12-18:2 biohydrogenation but not c9,c12,c15-
18:3; further increase of concentrate did not affect the extent of
biohydrogenation. This effect of high concentrate levels is prevented by
addition of sodium bicarbonate, which limits the drop of pH due to high
concentrate diets (Kalscheur et al., 1997a). However, sodium bicarbonate is
also known to increase ruminal outflow rate (Hart & Polan, 1984), and with
continuous in vitro cultures, a high outflow rate could abolish the effects of a
low pH on the biohydrogenation extent (Martin & Jenkins, 2002). The effect of
high concentrate diets could also result from a lack of large particles in the
rumen. Indeed, when increasing the size of particles in vitro from 0.1–0.4 mm
to 1–2 mm the rate of lipolysis increases by 25% and that of c9,c12-18:2
biohydrogenation by 60% (Gerson et al., 1988).
As opposed to PUFA, in most experiments, the biohydrogenation extent of
c9-18:1 was not significantly affected by high concentrate diets (Kalsheur
et al., 1997a; Loor et al., 2004) except when a high level of concentrates was
investigated (Kucuk et al., 2001). This difference between monounsaturated
FA and PUFA could be due to a difference in the effect of low pH on the
isomerase needed for the disappearance of PUFA and on the reductase which
accounts for most of c9-18:1 acid disappearance. Moreover, most hydrating
bacteria are lactic acid bacteria, which develop at low rumen pH, so that
hydration could be a major fate of UFA during ruminal acidosis (Hudson et al.,
2000). To our knowledge, this hypothesis has not yet been investigated, and
whether such a shift of FA metabolism toward hydration would affect c9-18:1
rather than c9,c12-18:2 acid is unknown.
The third effect of the diet is through its composition of nutrients. The
carbohydrate content of silages (Lee et al., 2006a) or grass (Scollan et al.,
2003) has been shown not to affect the extent of biohydrogenation. However,
in these experiments, the diets contained only forage so that increasing sugars
could have failed to modify the pH or the equilibrium between ruminal
bacterial species. On the contrary, addition of sucrose to continuous ruminal
cultures at constant pH linearly decreases the biohydrogenation extent of FA
from alfalfa hay (Ribeiro et al., 2005). Sucrose addition also decreased fibre
digestibility and, therefore, the authors attributed this effect on biohydrogenation
to a decrease in cellulolytic microorganisms which have biohydrogenation
activities. Gerson and co-workers (1985) observed an increased c9,c12-18:2
biohydrogenation rate when sucrose was added to the culture medium. The
effect of soluble carbohydrates supplements may depend on adaptation of
bacteria or on the time pattern of addition because in vitro an effect of sucrose
on biohydrogenation rates was not observed (Ribeiro et al., 2007).
Gerson and co-workers (1983) demonstrated that both lipolysis and
biohydrogenation rates increase in vitro when the nitrogen content increases
from 0.72 to 2.5% of dry matter. Since, the latter concentration is in the range
of nitrogen concentrations in most ruminant diets, nitrogen is not likely to be a
limiting factor in practice.
b. Effect of amount and treatment of fat source

The extent of biohydrogenation can be affected by the source and amount of fat,
and by the technological treatment of the fat source.
Lipolysis can be affected by the composition of the dietary fat source.
Indeed, it has been shown that ruminal hydrolysis of tristearin is under 50% in
sheep (Sklan et al., 1985), and the rate of lipolysis of tallow is lower than that
of soybean oil in vitro (Beam et al., 2000). These observations suggest a slower
lipolysis of fat containing no or little UFA. Interactions between different
esterified FA can affect the extent of lipolysis (Boeckaert et al., 2007b). For
instance, it has been shown that c4,c7,c10,c13,c16,c19-22:6 acid decreases the
lipolysis of c9,c12-18:2 and c9,c12,c15-18:3 acids (Boeckaert et al., 2007b).
The rate of lipolysis is only partially modified by the initial concentration of
triacylglycerols: it decreases from 44 to less than 30%/h when the in vitro
concentration of soybean oil increases from 2 to 10% (Beam et al., 2000).
However, the apparent preduodenal disappearance of triacylglycerols in-
creases when fat is added to the diet (Bauchart et al., 1990b). Moate and
co-workers (2004) concluded from a data set of in vivo studies, that the adverse
effect of fat concentration on lipolysis only applies to tallow.
Beam and co-workers (2000) measured a negative effect of c9,c12-18:2 acid
on its own isomerization. The c9,c12-18:2 acid disappearance in vitro de-
creased by 1.2% for each percentage of c9,c12-18:2 acid added (Beam et al.,
2000) which is consistent with the inhibition of the c9,c12-18:2 isomerase by
c9,c12-18:2 acid observed earlier (Kepler & Tove, 1967). However, when
increasing c9,c12-18:2 acid concentration, its biohydrogenation rate actually
decreases, but the amount of c9,c12-18:2 acid that disappears increases,

suggesting a limit on the capacity of isomerization rather than an inhibition of
the enzyme (Troegeler-Meynadier et al., 2003 and 2006a). The lack of recy-
cling of c9,c12-18:2 isomerase proposed by Kim and co-workers (2000) could
explain why its capacity is easily saturated. The same enzyme also catalyses
c9,c12,c15-18:3 acid biohydrogenation (Kepler & Tove, 1967), therefore a
high concentration of c9,c12,c15-18:3 acid in culture medium decreases the
disappearance of c9,c12-18:2 acid (Troegeler-Meynadier et al., 2003) due to
competitive inhibition (Troegeler-Meynadier et al., 2006a). The effects of
c9,c12-18:2 acid concentration on the c9,c12,c15-18:3 acid biohydrogenation
extent have not been studied. The disappearance of c5,c8,c11,c14,c17-20:5
and c4,c7,c10,c13,c16,c19-22:6 acids decreases in vitro when the level of fish
oil is increased (Dohme et al., 2003; AbuGhazaleh & Jenkins, 2004a; Chow
et al., 2004).
Contrary to these in vitro results, comparison of diets without and with added
sources of unprotected UFA demonstrate an increased biohydrogenation ex-
tent after fat addition as shown in Table 1 and in other experiments (Wu et al.,
1991; Ferlay et al., 1993; Lock & Garnsworthy, 2002; Shingfield et al., 2008).
Moreover, when experimenting with graded amounts of added fat, increasing
dietary fat does not affect (Wu et al., 1991; Sackmann et al., 2003; Jenkins &
Bridges, 2007) or slightly increases (Shingfield et al., 2008) the
biohydrogenation extent of C18 UFA and increases the biohydrogenation extent
of c5,c8,c11,c14,c17-20:5 and c4,c7,c10,c13,c16,c19-22:6 (Lee et al., 2005).
In a recent meta-analysis, Schmidely and co-workers (2008) calculated that the
increase of biohydrogenation extent of total UFA was 3.8% for each 1%
increase of dietary fat (dry matter basis). Moate and co-workers (2004), based
on published in vivo experiments, indicated that increasing the overall concen-
tration of rumen free long-chain FA from 0 to 5% of dry matter does not affect
the biohydrogenation rate of c9-18:1 and c9,c12-18:2 acids but decreases the
biohydrogenation rate of c9,c12,c15-18:3 acid. However, none of the fat
sources in the data set of these authors contained high proportions of this latter
FA, and on the contrary Loor and co-workers (2005c) found that the
biohydrogenation extent of c9,c12,c15-18:3 acid is increased when the diet of
cows is supplemented with linseed oil compared to a diet supplemented with
sunflower oil.
Most published data show that fish oil or marine algae, which contain large
proportions of c5,c8,c11,c14,c17-20:5 and c4,c7,c10,c13,c16,c19-22:6 acids,
decrease the biohydrogenation extent of C18 UFA, particularly c9,c12-18:2
acid, in vitro (AbuGhazaleh & Jenkins, 2004a; Wasowska et al., 2006; Boeckaert
et al., 2007b) and in vivo (Shingfield et al., 2003; Loor et al., 2005c). However,
some authors reported no effect in vitro (Chow et al., 2004) or in vivo (Lee
et al., 2005, Kim et al., 2008, for c9-18:1 and c9,c12,c15-18:3 acids), or an
increased biohydrogenation (Kim et al., 2008 for c9,c12-18:2 acid), so that
comparisons across experiments do not show any clear trend (Table 1). The
relative concentrations of c9,c12-18:2, c5,c8,c11,c14,c17-20:5 and
c4,c7,c10,c13,c16,c19-22:6 acids could account for these discrepancies, due to
complex interactions between c9,c12-18:2 acid and fish oil fatty acids
(Wasowska et al., 2006).
Technological treatment of the fat source also affects biohydrogenation. The
fat sources are strongly modified in the rumen, but rumen microorganisms also
can be affected by fat addition, as early in vitro and in vivo studies by Brooks
et al. (1954) demonstrated. Due to these reciprocal actions of microorganisms
and fat, specific treatments have been proposed to limit the inhibitory effect of
added fat on microorganisms, and the biohydrogenation of their FA. Other
processes applied to fat sources can incidentally affect rumen biohydrogenation.
Encapsulation of fat sources in a formaldehyde treated protein matrix was
first investigated in Australia (Scott et al., 1971). Formaldehyde-treated pro-
teins are disrupted only in the abomasum, and giving sheep a
formaldehyde-treated mixture of soybean oil and casein increases the duodenal
flow of c9,c12-18:2 + c9,c12,c15-18:3 acids from 2.5 to 13.0 g/d (Clapperton,
1978). Similarly, when applied to a synthetic mixture of c9,t11-18:2 and
t10,c12-18:2 acids, this method of protection strongly increases their abomasal
flows (Gulati et al., 2000). The efficiency of the protection decreases with the
oil/casein ratio (Clapperton, 1980), and is low when formaldehyde treatment is
applied without casein to oilseeds either directly (Bitman et al., 1973) or after
a formic acid treatment (Sinclair et al., 2005b). However, due to the cost of
casein and/or regulatory limitations on the use of formaldehyde in cattle
feeding in some countries, this method is not commercially used.
Recently, whey-protein complexes have proved to be efficient to protect
soybean oil, thus increasing the proportions of PUFA and decreasing the
proportions of TFA biohydrogenation intermediates in milk fat and plasma
triacylglycerols (Carroll et al., 2006; Heguy et al., 2006).
The use of FA calcium salts (also referred to as calcium soaps) was
developed two decades ago (Jenkins & Palmquist, 1982, 1984) based on
previous work performed by Davison & Woods (1963). The main purpose of
using FA Ca salts was to prevent the negative effect of FA on rumen microor-
ganisms. The protection of FA by Ca salts against biohydrogenation is moderate
(Wu et al., 1991; Wu & Palmquist, 1991; Ferlay et al., 1992; Enjalbert et al.,
1994, 1997) or absent (Fotouhi & Jenkins, 1992; Ferlay et al., 1993; Harvatine
& Allen, 2006). The use of CLA Ca salts instead of formaldehyde-protected
CLA results in much lower increase of milk CLA (De Veth et al., 2005). FA Ca
salts need to be dissociated prior to biohydrogenation. Sukhija and Palmquist
(1990) showed in vitro that dissociation of FA Ca salts can occur, with an extent
that increases when pH decreases and when the unsaturation level increases. At
pH 6.0, the percentage dissociation was around 50% for soybean oil FA and
20% for palm oil FA (Sukhija & Palmquist, 1990). Van Nevel and Demeyer
(1996a) showed that PUFA Ca salts partly escape rumen biohydrogenation at

pH 6.3 to 6.9, but are not protected at pH < 6.3.
The use of FA amide derivatives to protect FA against ruminal
biohydrogenation has also been investigated. Ruminal disappearance of
linoleoyl methionine is 70% compared to 93% for free c9,c12-18:2 acids in
vivo (Fotouhi & Jenkins, 1992). Hydroxyethylsoyamide was found to be more
efficient than butylsoyamide in increasing c9,c12-18:2 acid level in plasma
triacylglycerols (Jenkins & Thies, 1997), but had negative effects on intake
(Jenkins, 1997). Further studies on sheep treated with linoleamide showed a
38% relative increase of c9,c12-18:2 in the duodenum compared to free c9,c12-
18:2 (Jenkins & Adams, 2002). No effect on duodenal recovery of dietary C18
UFA in lactating dairy cows treated with canolamide compared to canola oil
was observed (Loor et al., 2002b). Dairy cows receiving soybean amides
showed no change in the rumen biohydrogenation of c9,c12-18:2 acid com-
pared with soybean oil (Lundy et al., 2004). Similarly, no significant
modification of c9-18:1 in the rumen of non-lactating cows were observed
when oleamide was used and compared to free c9-18:1 acid (Jenkins et al.,
2000). Oleamide resulted in greater proportion of c9-18:1 acid in milk fat
compared to canola oil (Jenkins, 1998), but soybean amides had no effect on
milk c9,c12-18:2 acid level (Lundy et al., 2004), suggesting that PUFA amide
derivatives are less efficient than their monounsaturated analogues.
Heat treatments of oilseeds may affect the biohydrogenation extent. It has
been shown that extrusion of oilseeds improves free FA release in the rumen
compared to raw or roasted material in vitro (Reddy et al., 1994). However in
this experiment, the fat concentration was high, and free FA release from
oilseeds both depended on fat release from feedstuffs and on the subsequent
microbial lipolysis (Reddy et al., 1994). In the same study, the authors reported
a slight reduction of biohydrogenation due to extrusion but a greater effect of
roasting, with a tendency toward a stronger effect when roasting temperature
increased (Reddy et al., 1994). Similarly, both roasting and extrusion were
shown to inhibit in situ the disappearance of c9,c12-18:2 acid (Troegeler-
Meynadier et al., 2006b). A similar effect of roasting has been reported with
cottonseed in vivo (Pires et al., 1997), but Tice and co-workers (1994) did not
observe any effect of roasting on biohydrogenation of FA from whole soybeans.
The effects of seed oil extrusion on the biohydrogenation extent are not quite
clear: extrusion of soybeans slowed down the disappearance of c9,c12-18:2
acid in situ but in the same experiment decreases of c9,c12-18:2 acid levels in
plasma and milk FA were observed (Chouinard et al., 1997b). Moreover, it has
been shown that extrusion of canola seeds hastens disappearance of c9,c12-
18:2 and c9,c12,c15-18:3 acids but not c9-18:1 acid in vitro (Enjalbert et al.,
2003). Extrusion of linseed seems not to affect disappearance of c9,c12-18:2
and c9,c12,c15-18:3 in vitro (Akraim et al., 2006b) and in vivo (Akraim et al.,
2006a), or slightly increases ruminal disappearance of c9,c12,c15-18:3 acid in
vivo (Gonthier et al., 2004). The reasons for the discrepancy between different
investigations remain unclear. The extrusion temperature has only minor
effects on the c9,c12-18:2 acid disappearance (Chouinard et al., 1997b), but
preconditioning and particle size seem to interact with the effects of extrusion
(Akraim et al., 2006a and 2006b). Particle size of fat-containing seeds has
minor effects on rumen biohydrogenation (Tice et al., 1994).
c. Use of additives
Some antimicrobials can inhibit lipolysis, specially ionophores and amoxicillin,
but have very limited effects on the disappearance of PUFA (Van Nevel &
Demeyer, 1995). Amoxicillin has a broad antibiotic spectrum, but ionophores
inhibit only gram-positive bacteria, therefore ionophores cannot inhibit lypolitic
activities of Anaerovibrio spp. (gram-negative bacteria). Recent data suggest
that in high-starch diets, monensin does not suppress classical gram-positive
bacteria but affects Megasphaera elsdenii and Butyrivibrio fibrisolvens (Weimer
et al., 2008). The effect of different ionophores could differ because monensin
results in a lower biohydrogenation rate than lasalocid (Martineau et al., 2008).
Treatment of linseed with quebracho condensed tannin reduced
biohydrogenation of c9,c12,c15-18:3 acid, but did not result in an improve-
ment of c9,c12,c15-18:3 transfer in the plasma lipids of steers (Kronberg et al.,
2007).
2. Factors affecting trans-18:1 and CLA isomeric profile

Due to rumen outflow, some biohydrogenation intermediates, which mainly
have trans double bonds, can leave the rumen before complete reduction to
18:0 acid. This outflow of TFA is often more important than the outflow of cis
isomers which mainly originate from dietary lipids. The ratios between poly-
and monounsaturated biohydrogenation intermediates and the biohydrogenation
end-product 18:0 acid depend on the relative activities of the successive
enzymes involved in the biohydrogenation process. On the contrary, the profile
of positional isomers depends on the profile of dietary FA and on the
isomerizations performed by bacteria.
a. Ratios between 18:0 acid, and monounsaturated and

polyunsaturated TFA
These ratios are representative of the biohydrogenation completion; an incom-
plete biohydrogenation results in more TFA flowing out of the rumen. Average
values across experiments are shown in Table 1. Duodenal flows often differ
from FA intake (Doreau & Chilliard, 1997), therefore the sum of TFA
biohydrogenation intermediates and 18:0 can differ from 100% of disappeared
UFA.
Factors affecting the ratios include inter-individual variation, high concen-
trate diets, amount of added fat, type of fat, technological treatment and quality
of the fat source, and additives.
An example of inter-individual variation is the large variations of milk CLA
content that have been observed between animals receiving the same diet
within the same herd. Peterson and co-workers (2002) showed that individual
animals were consistent over time when the diet remained unchanged, and that
the hierarchy of cows for milk c9,t11-18:2 production levels remained un-
changed when the diet was modified, suggesting that animals differ in both
their rumen outflow of TFA and their ability to produce CLA via mammary
desaturation.
High concentrate diets affect the ratios. In vitro, a low ruminal pH results in
an inhibition of the reduction of trans-18:1 to 18:0 acids, resulting in higher
concentrations of trans-8:1 isomers (Troegeler-Meynadier et al., 2006a). Simi-
larly, increasing in vivo the percentage of concentrate increases the
disappearance of PUFA in the rumen and the total 18:1 TFA isomers in the
duodenal flow (Piperova et al., 2002; Loor et al., 2004). Piperova and co-
workers (2002) showed that a high concentrate diet also increases CLA level in
the duodenum, suggesting an inhibition of the CLA reduction, which could not
be shown in vitro (Troegeler-Meynadier et al., 2006a). These effects of a high
concentrate diet can be prevented by dietary supply of sodium bicarbonate
(Piperova et al., 2002).
The amount of added fat influences ratios of the acids and, due to the interest
of CLA for nutritional applications, the formation of c9,c12-18:2 acid
biohydrogenation intermediates has been extensively investigated. In vitro,
increasing the initial concentration of c9,c12-18:2 increases the proportions of
conjugated 18:2 or trans-18:1 acid isomers but limits the formation of 18:0 acid
(Harfoot et al., 1973b; Troegeler-Meynadier et al., 2003). This modulation in
the formation of c9,c12-18:2 acid biohydrogenation intermediates is due to a
lowered reduction of conjugated 18:2 to trans-18:1 acid isomers and of trans
18:1 to 18:0. It has been shown that increased concentrations of CLA can
inhibit the growth of Butyrivibrio fibrisolvens cultures (Kim et al., 2000). An
effect on the enzyme activities catalysing the reduction of CLA to trans-18:1
isomers can explain the effect of initial c9,c12-18:2 acid concentration
(Troegeler-Meynadier et al., 2006a; Moate et al., 2008).
In vitro, high concentrations of c9,c12-18:2 acid inhibit trans-18:1 reduc-
tion. Polan and co-workers (1964) indicated that the reduction of trans-18:1
acid isomers only begins when their concentration is higher than c9,c12-18:2
acid concentration. However, Harfoot and co-workers (1973b) argued that the
inhibition due to a high initial concentration of c9,c12-18:2 is irreversible. The
inhibition threshold was determined to be about 1 mg of c9,c12-18:2/ml of
rumen contents (Harfoot et al., 1973b). More recently, Troegeler-Meynadier
and co-workers (2003), observed that a high initial concentration of c9,c12-
18:2 acid effectively decreases the 18:0/trans-18:1 ratio, but results in a linear
increase over time of 18:0 acid. The rate of 18:0 acid increase was found to be
higher with high initial c9,c12-18:2 acid concentration (1.87 mg/ml) than with
low concentration (0.62 mg/ml) (Troegeler-Meynadier et al., 2003). These
authors hypothesized that there is a maximal rate for the reduction of trans-18:1
acids, therefore a high production of trans-18:1 acids results in an accumula-
tion. (Troegeler-Meynadier et al., 2003). Using in vitro kinetics, Moate and
co-workers (2008) observed that t11-18:1 acid biohydrogenation is inhibited
when its concentration is over 0.5 mg/ml. A direct inhibition of t11-18:1 acid
on its own reduction can explain why the reduction of t11-18:1 acid is limited
even after disappearance of CLA (Moate et al., 2008), and explain the apparent
irreversibility of the inhibition stated by Harfoot (1973b). CLA could also
inhibit the reduction of trans-18:1 acids (Troegeler-Meynadier et al., 2006a).
In vivo, the effect of graded supplies of c9,c12-18:2 acid have not been
extensively studied: Sackmann and co-workers (2003) observed that increas-
ing sunflower oil supply to steers does not increase the duodenal proportion of
total trans-18:2 and trans-18:1 acid isomers despite an increased c9,c12-18:2
acid disappearance in the rumen. In dairy cows, Shingfield and co-workers
(2008) showed that increasing dietary sunflower oil increased the concentra-
tion of both trans-18:1 and CLA in omasal FA flow. Modelling experiments
obtained from several in vivo data showed that the biohydrogenation rate of
trans-18:1 isomers is negatively affected by the concentration of free FA in the
rumen (Moate et al., 2004).
Adaptation of cows to fat supplementation can modify the completion of
c9,c12-18:2 acid biohydrogenation. Indeed, the concentration of milk CLA
decreases over time when oil supplements are given (Bauman et al., 2000;
Chilliard & Ferlay, 2004). Palmquist and co-workers (2005) related this
decrease of milk CLA content to the positive effect of adaptation on the
tolerance of Butyrivibrio fibrisolvens to c9,c12-18:2 acid observed earlier
(Kim et al., 2000). However, Roy et al. (2006) showed that total milk trans-
18:1 acid isomers do not decrease over time when fat is added to the diet. This
suggests that the decrease of milk CLA is not due to an increased efficiency of
the reduction reactions but to the t11 to t10 shift (see later).
The effect of fat type is illustrated by the body of evidence showing that
dietary fish oil modifies ruminal biohydrogenation (see Tables 1 and 2). The
main reported effect is a large increase (about 4 fold) in the rumen outflow of
trans-18:1 acid isomers (Shingfield et al., 2003, Wonsil et al., 1994). A dietary
supplementation with 2.5% of fish oil results in an enhanced formation of
trans-18:1 acid isomers in the rumen as observed with 5% of sunflower oil
(Loor et al., 2005b and 2005c). In steers supplemented with a constant amount
of sunflower oil and gradual addition of fish oil, increases of duodenal flows of
trans-18:1 acid isomers were observed (Lee et al., 2005). The effects of fish oil
are isomer specific, t10-18:1 and t11-18:1 acids showing the greatest increases
(Shingfield et al., 2003; Lee et al., 2005). Fish oil addition to a diet without
addition of c9,c12-18:2 acid also increases the omasal flow of non-conjugated

trans-18:2 acids but tends to decrease the flow of total CLA (Shingfield et al.,
2003). It has been shown that a blend of fish oil and marine algae is as efficient
as the same amount of lipids from fish oil for increasing duodenal trans-18:1
acid isomer in sheep (Sinclair et al., 2005a). The effect of addition of marine
algae or fish oil on trans-18:1 acid isomers content in milk fat are equivalent
(Franklin et al., 1999).
Interestingly, fish oils contain low amounts of C18 UFA, the precursors of
trans-18:1 acid isomers, therefore fish oil does not increase the level of the
trans-18:1 acid isomers precursor pool. In vitro, AbuGazahleh and co-workers
(2004b) demonstrated that fish oil c4,c7,c10,c13,c16,c19-22:6 acid promotes
t11-18:1 acid accumulation due to an inhibition of t11-18:1 reduction
(AbuGhazaleh & Jenkins, 2004a). Similarly, it has been shown that
c4,c7,c10,c13,c16,c19-22:6 acid also inhibits the reduction of t11,c15-18:2
(Boeckaert et al., 2007b). This observation explains the increased duodenal
flow of non-conjugated trans-18:2 observed after fish oil addition (Shingfield
et al., 2003). It has been reported that c4,c7,c10,c13,c16,c19-22:6 is less
effective than t11-18:1 acid to inhibit the reduction of t11,c15-18:1 (Vlaeminck
et al., 2008). The effect of c4,c7,c10,c13,c16,c19-22:6 seems to be mediated
by a direct inhibition of bacteria (Boeckaert et al., 2007a).
Effect of technological treatment and quality of the fat source is shown by the
observation that extrusion of canola (Enjalbert et al., 2003), soybeans
(Troegeler-Meynadier et al., 2006b) and linseed (Akraim et al., 2006b) in-
creases the proportion of t10+t11-18:1 and c9,t11-18:2 acid isomers compared
to raw seeds (Figure 3). In vivo, contradictory effects of extruded linseed have
been shown: an increased proportion of trans-18:1 acid isomers (Gonthier
et al., 2004) or lack of effect (Akraim et al., 2006a). Differences could be
explained by the fact that cows were lactating in the experiment done by
Gonthier and co-workers (2004) while not in the experiment of Akraim and co-
workers (2006a). Extrusion of seeds to increase the proportions of trans-18:1
and/or CLA in the milk fat has been investigated (Chouinard et al. 1997a;
Bayourthe et al., 2000; Chouinard et al., 2001; Gonthier et al., 2005; Akraim
et al., 2007). To date, no explanation has been proposed for the effect of
extrusion. Moreover, among other heat treatments, micronization and roasting
have been shown to poorly modify the proportions of trans-18:1 in the
duodenal flow (Gonthier et al., 2004; Chouinard et al., 1997b; Gonthier et al.,
2005).
It has been shown that the level of fat oxidation negatively affects the
proportions of biohydrogenation intermediates and increases the proportion of
18:0 acid in vitro (Vázquez-Añón and Jenkins, 2007).
Additives can affect the ratios, as shown by the fact that copper depletion due
to molybdenum excess increases the proportion of biohydrogenation inter-
mediates in the plasma of cows (Morales et al., 2000). Copper supplementation
16
linseed
14
12
10
% of C18 FA
0
0 6 12 18 24
Incubation time, h
16
canola
14
12
10
% of C18 FA
0 6 12 18 24
Incubation time, h
Figure 3. In vitro synthesis of 9c,11t 18:2 () and t10+t11 18:1 (¸) during incubation of raw (——)
or extruded (- - - - -) linseed or canola (adapted from Enjalbert et al., 2003 and Akraim et al., 2006b).
above requirements decreases the formation of trans-18:1 acid isomers in the

ruminal fluid (Engle et al., 2000). In both reported experiments, no significant
effect on c9,c12-18:2 acid was observed, the effect was probably due to a
modification of the reduction steps of biohydrogenation. However, it seems
that the effects are of limited extent and depended on animal breeds and on the
addition of dietary fat. Therefore further research is needed to better understand
the effect of copper supplementation.
Recently, Fukuda and co-workers (2006b) showed in vitro that Butyrivibrio
fibrisolvens MDT-10 strain multiplies by four the concentration of t11-18:1
acid isomers when added to mixed ruminal cultures. Simultaneous addition of
Bifidobacterium adolescentis HF-11 strain, which has a high capacity to
incorporate t11-18:1 acid, further increases t11-18:1 acid level in the cultures
(Fukuda et al.,2006b). These authors concluded that these strains, used as
dietary additives, could possibly be used to increase the amount of t11-18:1
acid absorbed from the small intestine of ruminants (Fukuda et al.,2006b).
b. Ratio between positional trans isomers

The profile of positional trans isomers depends on the profile of dietary FA.
Compared to soybean oil, addition of canola oil (source of c9-18:1 acid) results
in more t4- to t10-18:1, and less t11- to t16-18:1 and c9,t11-18:2 (Loor et al.,
2002a) while addition of linseed oil (source of c9,c12,c15-18:3 acid) results in
less t10-18:1 and t10,c12-18:2, but more t13- to t16-18:1 acid isomers and
trans 18:2 with a t13 double bond (Loor et al., 2005a) (see Table 2 on page 10).
Beyond these effects of FA composition of dietary fat, other dietary modifi-
cation can affect the relative proportions of t10- and t11-18:1 acid isomers
(Table 2); in some experiments, t10- can overcome t11-18:1 acid isomers
(Figure 4).
Proportions of the isomers are related to forage source, diet composition, fat
addition, interaction between fat addition and diet composition, and additives.
The effect of forage source is seen by the use of corn silage which results in
higher proportions of t10-18:1 in milk fat with or without addition of dietary oil
(Shingfield et al., 2005; Kay et al., 2005; Roy et al., 2006). This effect could
partly be due to the concentrations of both starch and c9,c12-18:2 acid which
are higher in corn silage compared to grass as pasture, hay or silage.
In terms of diet composition, the first evidence that high concentrate diets
can result in a strong increase of t10-18:1 acid isomer was shown by Griinari
and co-workers (1998) by switching from 50 to 20% forage in a diet containing
unsaturated fat. The effect observed was a 7-fold increase of the t10-18:1 acid
proportion in milk fat (Griinari et al., 1998). Experiments measuring the FA
duodenal flow also indicated modification of isomeric profile (Table 2). In
cows receiving a diet without added fat, by modification of the percentage of
concentrate from 35–40 to 65–75, a 4-fold to 11-fold increase of t10-18:1 acid
isomer duodenal flow has been observed (Piperova et al., 2002; Loor et al.,
2004). It has been shown that the duodenal flow of t10-18:1 acid linearly
increases with the increase in concentrate percentage in the diet of steers
(Sackmann et al., 2003). The duodenal flow of t10,c12-18:2 acid increased in
the same way when increasing the concentrates level in diets (Piperova et al.,
30 30
25
25 t11-C18:1
t11-C18:1
t10-C18:1 t10-C18:1
20
20
% of FA
% of FA
15 15
10 10
5 5
0 0
C Cu4 0
40
88u2
4
C C i0 i3
C i0 3
6488
Li M0
M 25
L Li0
4 5
7564
S Li
Su Su
L Li
C C7
L L
0
C C i3
3
C
C2
S
C
M
C
35
35 35
65 65
35 65
2
M
C
C
C
65
C40 and C75: 40 and 75% of concentrate, duodenal flow (Piperova et al., 2002);
C64 and C88: 64 and 88% of concentrate, duodenal flow (Sackmann et al., 2003);
Su2 and Su4: 2 and 4% of added sunflower oil (Sackmann et al., 2003); C35, C65,
Li0 and Li3: 35 and 65% of concentrate, 0 and 3% of added linseed oil, duodenal
flow (Loor et al., 2004); M0 and M25: 0 or 25 mg of monensin /kg of dry matter,
in vitro continuous culture, (Jenkins et al., 2003).
Figure 4. Effects of the level of concentrate (% of dry matter), fat addition and monensin on t10- and
t11 18:1 acid levels (% of total FA) in the duodenal flow or in in vitro cultures. C40 and C75: 40 and 75%
of concentrate, duodenal flow, Piperova et al. (2002). C64 and C88: 64 and 88% of concentrate, duodenal
flow, Sackmann et al. (2003). Su2 and Su4: 2 and 4% of added sunflower oil, Sackmann et al. (2003).
C35, C65, Li0 and Li3: 35 and 65% of concentrate, 0 and 3% of added linseed oil, duodenal flow, Loor
et al. (2004). M0 and M25: 0 or 25 mg of monensin per kg of dry matter, in vitro continuous cultures,
Jenkins et al. (2003).
2002; Sackmann et al., 2003) but not in the experiment of Loor and co-workers
(2004). The same effect has been observed in ewes (Kucuk et al., 2001). The
effect of high concentrate diets on t10-18:1 and t10,c12-18:2 acids production
can be alleviated by the addition of sodium bicarbonate to the diet, suggesting
a pH dependent effect (Piperova et al., 2002).
It has been shown in vitro, using the same donor cow, that lowering the pH
of rumen fluid cultures only slightly modifies the t10/t11 ratio (Troegeler-
Meynadier et al., 2003), which contrasts with the effect observed in vivo after
adaptation of cows. However, it has been shown that whatever the pH, in vitro
cultures contained more t10,c12-18:2 when the donor cow received a high
concentrate diet (Choi et al., 2005). These results suggest that the effect of
concentrate is mediated by a modification of the ruminal ecosystem, but not by
a modification of enzyme activities. Dietary starch with a high degradation rate
results in a higher level of t10-18:1 acid. This has been demonstrated in dairy
cows switched from dry ground corn to high moisture corn (Bradford & Allen,
2004) or from potatoes to wheat (Jurjanz et al., 2004).
Dietary addition of linseed oil results in lower duodenal flows of t10-18:1
acid than addition of soybean oil (Loor et al., 2004) or sunflower oil (Loor et
al, 2005c), and graded increase of dietary sunflower oil increases t10 acid
isomers (Sackmann et al., 2003; Shingfield et al., 2008), confirming that
c9,c12-18:2 acid is the main precursor of the t10-18:1 and t10,c12-18:2 acid
isomers (Figure 1).
The ruminal production of t10 isomers is higher with c9,c12-18:2 acid in oils
than with c9,c12-18:2 acid seed sources (Duckett et al., 2002), which shows
that the availability of c9,c12-18:2 acid for bacteria can play an important role
in the t10 shift. Many experimental data have been obtained with oil supple-
mentation; however in practice oilseeds supplementation is more current than
oils supplementation. Limited data suggest that soybean amides increase
ruminal production of t10 isomers (Lundy et al., 2004).
The results related to the effect of fish oil on the production of t10 isomers are
conflicting. Shingfield and co-workers (2003) reported that fish oil addition
(1.6%) does not impact t10/t11 ratio. However, Loor and co-workers (2005c)
observed a t10/t11 ratio that was around 1 with either 2.5% of fish oil or 5% of
sunflower oil, compared to 0.3 with 5% of linseed oil suggesting that fish oil
can result in a high production of t10 isomers. These observations are consist-
ent with the higher t10/t11 ratio obtained with the addition of 3% fish oil
compared to 1% (Kim et al., 2008). In vitro, low pH or a low forage level result
in high concentrations of t10-18:1 acid when fish oil is added to the culture
(AbuGhazaleh & Jacobson, 2007a and 2007b).
It has been shown in lactating cows that the duration of c9,c12-18:2
supplementation affects the t10/t11 ratio. It has been found that t11-18:1 and
c9,t11-18:2 acids reach a maximal concentration 4–7 days after the beginning
of oil supplementation, whereas t10-18:1 reaches a plateau from 10 to 20 days
of supplementation (Bauman et al., 2000; Roy et al., 2006; Shingfield et al.,
2006). Such an effect on t11/t10 ratio was not observed when linseed oil was
added to a grass hay-based diet (Roy et al., 2006). However, it has been
observed when linseed oil was added to a corn silage based diet (Pottier et al.,
2006).
The interaction between fat addition and diet composition has been seen to
influence the isomeric profile. In the experiment of Griinari et al. (1998),
switching from 50 to 20% of forage in the diet of lactating cows increased t10-
18:1 acid by 27% in milk fat when the diet contained saturated fat, but a 314%
increase was observed when the diet contained corn oil.
Similarly, a high level of t10-18:1 acid isomer (> 10% of total FA) has been
observed in the duodenal flow when high concentrate diets are associated with
unsaturated fat addition (Sackmann et al., 2003; Loor et al., 2005c; Lundy
et al., 2004). However, experiments investigating the interaction of concen-
trate level and oil addition on TFA duodenal flow failed to demonstrate such
interaction (Sackmann et al., 2003, Loor et al., 2004).
Additives may affect isomeric profile and Fellner and co-workers (1997)
demonstrated that ionophores increased, in vitro, the production of trans-18:1
isomers, with or without addition of c9,c12-18:2 acid. It was shown later that
this modulation is mainly due to increase in the formation of t10-18:1 acid
isomer (Jenkins et al., 2003). In this study, complex interactions were reported:
monensin increased the effect of soybean oil on t10-18:1 acid production when
the starch source was barley but not corn (Jenkins et al., 2003). This observa-
tion suggests that ionophore can impact fermentation of starch (Jenkins et al.,
2003). These biohydrogenation changes affect milk fat composition, monensin
exacerbating the positive effects of sunflower oil on the proportions of t10-18:1
and t10,c12-18:2 acid (Bell et al., 2006; Cruz-Hernandez et al., 2006).
Pottier and co-workers (2006) showed that a supplementation of cows with
a high dosage of vitamin E prevents the shift toward the t10 pathway for at least
3 weeks but does not reverse the shift once it has occurred.
D. Use of vaccenic acid for mammary synthesis of rumenic acid: a

brief comment
Desaturation of saturated FA in mammary gland is possible due to Δ9-desaturase
enzyme activity (Annison et al., 1967). It has been also shown that vaccenic
(t11-18:1) acid formed by ruminal biohydrogenation is transported in the
arterial flow and taken up by the mammary gland and can be metabolized into
rumenic (c9,t11-18:2) acid. The efficiency of the conversion of vaccenic acid
has been estimated to be between 8 and 39% (Palmquist et al., 2005). This
metabolic pathway represents the main origin of c9,t11-18:2 acid in milk fat.
These metabolic aspects are detailed in Chapter 8.
E. Conclusion
Ruminal biohydrogenation of unsaturated FA is responsible for the synthesis of
TFA, mainly t11-18:1 and t10-18:1 acid isomers and various positional iso-
mers. Current knowledge allows to channel quantitatively and qualitatively the
rumen TFA outflow by modifying forage and concentrate sources, and propor-
tion of concentrate. The level and the profile of the different fat sources alone
or in combination, or via the use of feed additives also affect the quality of TFA
produced during biohydrogenation. The most specific way of control is via fat
supplementation: most experiments have used addition of oil. Current know-
ledge on the effects of lipids supplied not as oil but as processed oilseeds on the
duodenal flow of TFA is limited.
The formation of a limited number of TFA isomers (i.e. t11, t10-18:1,
t10,c12- and c9,t11-18:2 acids) during biohydrogenation has been studied for
their nutritional properties. The biological properties of the various conjugated
and non-conjugated 18:2 acid isomers remain unknown and their formation
during biohydrogenation unclear.
References
AbuGhazaleh, AA and Jacobson, BN (2007a) The effect of pH and polyunsaturated C18 fatty
acid source on the production of vaccenic acid and conjugated linoleic acids in ruminal
cultures incubated with docosahexaenoic acid. Anim. Feed Sci. Technol., 136, 11–22.
AbuGhazaleh, AA and Jacobson, BN (2007b) Production of trans C18:1 and conjugated
linoleic acid in continuous culture fermenters fed diets containing fish oil and sunflower
oil with decreasing levels of forage. Animal, 1, 660–665.
AbuGhazaleh, AA and Jenkins, TC (2004a) Disappearance of docosahexaenoic and
eicosapentaenoic acids from cultures of mixed ruminal microorganisms. J. Dairy Sci., 87,
645–651.
AbuGhazaleh, AA and Jenkins, TC (2004b) Short communication: docosahexaenoic acid
promotes vaccenic acid accumulation in mixed ruminal cultures when incubated with
linoleic acid. J. Dairy Sci., 87, 1047–1050.
AbuGhazaleh, AA, Riley, MB, Thies, EE and Jenkins, TC (2005) Dilution rate and pH effects
on the conversion of oleic acid to trans C18:1 positional isomers in continuous cultures.
J. Dairy Sci., 88, 4334–4341.
Akraim, F, Nicot, MC, Juaneda, P and Enjalbert, F (2007) Conjugated linolenic acid (CLnA),
conjugated linoleic acid (CLA), and other biohydrogenation intermediates in plasma and
milk fat of cows fed raw or extruded linseed. Animal, 1, 835–843.
Akraim, F, Nicot, MC, Weill, P and Enjalbert, F (2006a) Effects of preconditioning and
extrusion of linseed on the ruminal biohydrogenation of fatty acids. 1. In vivo studies.
Anim. Res., 55, 83–91.
Akraim, F, Nicot, MC, Weill, P and Enjalbert, F (2006b) Effects of preconditioning and
extrusion of linseed on the ruminal biohydrogenation of fatty acids. 2. In vitro and in situ
studies. Anim. Res., 55, 261–271.
Annison, EF, Linzell, JL, Fazakerley, S and Nichols, BW (1967) The oxidation and utilization
of palmitate, stearate, oleate and acetate by the mammary gland of the fed goat in relation
to their overall metabolism, and the role of plasma phospholipids and neutral lipids in
milk-fat synthesis. Biochem. J., 102, 637–647.
Ashes, JR, Siebert, BD, Gulati, SK, Cuthbertson, AZ and Scott, TW (1992) Incorporation of
n–3 fatty acids of fish oil into tissue and serum lipids or ruminants. Lipids, 27, 629–631.
Atkinson, RL, Scholljegerdes, EJ, Lake, SL, Nayigihugu, V, Hess, BW and Rule, DC (2006)
Site and extent of digestion, duodenal flow, and intestinal disappearance of total and
esterified fatty acids in sheep fed a high-concentrate diet supplemented with high-
linoleate safflower oil. J. Anim. Sci., 84, 387–396.
Bauchart, D, Legay-Carmier, F and Doreau, M (1990a) Ruminal hydrolysis of dietary
triglycerides in dairy cows fed lipid-supplemented diets. Reprod. Nutr. Dev., 30, suppl.,
187s.
Bauchart, D, Legay-Carmier, F, Doreau, M and Gaillard, B (1990b) Lipid metabolism of
liquid-associated and solid-adherent bacteria in rumen contents of dairy cows offered
lipid-supplemented diets. Br. J. Nut., 63, 563–578.
Bauman, DE, Barbano, DM, Dwyer, DA and Griinari, JM (2000) Technical note: production
of butter with enhanced conjugated linoleic acid for use in biomedical studies with animal
models. J. Dairy Sci., 83, 2422–2425.
Bayourthe, C, Enjalbert, F and Moncoulon, R (2000) Effects of different forms of canola oil
fatty acids plus canola meal on milk composition and physical properties of butter.
J. Dairy Sci., 83, 690–696.
Beam, TM, Jenkins, TC, Moate, PJ, Kohn, RA and Palmquist, DL (2000) Effects of amount
and source of fat on the rates of lipolysis and biohydrogenation of fatty acids in ruminal
contents. J. Dairy Sci., 83, 2564–2573.
Bell, JA, Griinari, JM and Kennelly, JJ (2006) Effect of safflower oil, flaxseed oil, monensin
and vitamin E on concentration of conjugated linoleic acid in bovine milk fat. J. Dairy
Sci., 89, 733–748.
Bickerstaffe, R, Noakes, DE and Annison, EF (1972) Quantitative aspects of fatty acids
biohydrogenation, absorption and transfer into milk in the lactating goat, with special
reference to the cis- and trans-isomers of octadecenoate and linoleate. Biochem. J., 130,
607–617.
Bitman, J, Dryden, LP, Goering, HK, Wrenn, TR, Yoncoskie, RA and Edmondson, LF (1973)
Efficiency of transfer of polyunsaturated fats into milk. J. Amer. Oil Chem. Soc., 50, 93–
98.
Boeckaert, C, Fievez, V, Van Hecke, D, Verstraete, W and Boon, N (2007a) Changes in
rumen biohydrogenation intermediates and ciliate protozoa diversity after algae supple-
mentation to dairy cattle. Eur. J. Lipid Sci. Technol., 109, 767–777.
Boeckaert, C, Vlaeminck, B, Mestdagh, J and Fievez, V (2007b) In vitro examination of
DHA-edible micro algae. 1. Effect on rumen lipolysis and biohydrogenation of linoleic
and linolenic acids. Anim. Feed Sci. Technol., 136, 63–79.
Bradford, BJ and Allen, MS (2004) Milk fat responses to a change in diet fermentability vary
by production level in dairy cattle. J. Dairy Sci., 87, 3800–3807.
Brooks, CC, Garner, GB, Gehrke, CW, Muhrer, ME and Pfander, WH (1954) The effect of
added fat on the digestion of cellulose and protein by ovine rumen micoorganisms. J.
Anim. Sci., 13, 758–764.
Carroll, SM, DePeters, EJ and Rosenberg, M (2006) Efficacy of a novel whey protein gel
complex to increase the unsaturated fatty acids composition of bovine milk fat. J. Dairy
Sci., 89, 640–650.
Chilliard, Y and Ferlay, A (2004) Dietary lipids and forages interactions on cow and goat milk
fatty acid composition and sensory properties. Reprod. Nutr. Dev., 44, 467–492.
Choi, NJ, Imm, JY, Oh, S, Kim, BC, Hwang, HJ and Kim, YJ (2005) Effect of pH and oxygen
on conjugated linoleic acid (CLA) production by mixed rumen bacteria from cows fed
high concentrate and high forage diets. Anim. Feed Sci. Technol., 123–124, 643–653.
Chouinard, PY, Corneau, L, Butler, RW, Chilliard, Y, Drackley, JK and Bauman, DE (2001)
Effect of dietary lipid source on conjugated linoleic acid concentration in milk fat. J.
Dairy Sci., 84, 680–690.
Chouinard, PY, Girard, V, Brisson and GJ (1997a) Performance and profiles of milk fatty
acids of cows fed full fat, heat-treated soybeans using various processing methods. J.
Dairy Sci., 80, 334–342.
Chouinard, PY, Levesque, J, Girard, V and Brisson, GJ (1997b) Dietary soybeans extruded
at different temperatures: milk composition and in situ reactions. J. Dairy Sci., 80, 2913–
2924.
Chow, TT, Fievez, V, Moloney, AP, Raes, K, Demeyer, D and De Smet, S (2004) Effect of
fish oil on in vitro lipolysis, apparent biohydrogenation of linoleic and linolenic acids, and
accumulation of biohydrogenation intermediates. Anim. Feed Sci. Technol., 117, 1–12.
Clapperton, JL (1978) Biohydrogenation of protected soya-bean oil in sheep fed a low-fibre
diet. Proc. Nutr. Soc., 37, 65A.
Clapperton, JL (1980) The extent of hydrogenation of two formaldehyde-treated spray-dried

mixtures of soya bean oil and casein fed to sheep. J. Sci. Food Agric., 31, 439–447.
Collomb, M, Sieber, R and Bütikofer, U (2004) CLA isomers in milk fat from dairy cows fed
diets with high levels of unsaturated fatty acids. Lipids, 39, 355–364.
Corl, BA, Baumgard, BH, Griinari, JM, Delmonte, P, Morehouse, KM, Yurawecz, MP and
Bauman, DE (2002) Trans-7,cis-9 CLA is synthesized endogenously by delta9 desaturase
in dairy cows. Lipids, 37, 681–688.
Cruz-Hernandez, C, Kramer, JKG, Kennelly, JJ, Okine, EK and Weselake, RJ (2006) Effect
of sunflower oil and monensin on the CLA and 18:1 isomer composition in milk fat of
dairy cattle. Fourth Euro Fed Lipid Congress, Madrid.
Davison, KL and Woods, W (1963) Effect of calcium and magnesium upon digestibility of
a ration containing corn oil by lambs, J. Anim. Sci., 22, 27–29.
Dawson, RMC, Hemington, N, Grime, D, Lander, D and Kemp, P (1974) Lipolysis and
hydrogenation of galactolipids and the accumulation of phytanic acid in the rumen.
Biochem. J., 144, 169–171.
Dawson, RMC and Kemp, P (1969) The effect of defaunation on the phospholipids and
on the hydrogenation of unsaturated fatty acids in the rumen. Biochem. J., 115, 351–
352.
Demeyer, DL, Henderson, C and Prins, RA (1978) Relative significance of exogenous and
de novo synthesized fatty acids in the formation of rumen microbial lipids in vitro. Appl.
Environment. Microbiol., 35, 24–31.
Deng, MD, Grund, AD, Schneider, KJ, Langley, KM, Wassink, SL, Peng, SS and Rosson,
RA (2007) Linoleic acid isomerase from Propionibacterium acnes: purification, charac-
terization, molecular cloning, and heterologous expression. Appl. Biochem. Biotechnol.,
143, 199–211.
Destaillats, F, Trottier, JP, Galvez, JMG and Angers, P (2005) Analysis of alpha-linolenic
acid biohydrogenation intermediates in milk fat with emphasis on conjugated linolenic
acids. J. Dairy Sci., 88, 3231–3239.
DeVeth, MJ, Gulati, SK, Luchini, ND and Bauman, DE (2005) Comparison of calcium-salts
and formaldehyde-protected conjugated linoleic acid in inducing milk fat depression. J.
Dairy Sci., 88,1685–1693.
Devillard, E, McIntosh, FM, Newbold, CJ and Wallace, RJ (2006) Rumen ciliate protozoa
contain high concentrations of conjugated linoleic acids and vaccenic acid, yet do not
hydrogenate linoleic acid or desaturate stearic acid. Br. J. Nutr., 96, 697–704.
Dewhurst, RJ, Evans, RT, Scollan, ND, Moorby, JM, Merry, RJ and Wilkins, RJ (2003)
Comparison of grass and legume silages for milk production. 2. In vivo and in sacco
evaluations of rumen function. J. Dairy Sci., 85, 2612–2621.
Dierick, NA and Decuypere, JA (2006) Endogenous lipolysis in feedstuffs and compound
feeds for pigs: effects of storage time and conditions and lipase and/or emulsifier addition.
Anim. Feed Sci. Technol., 102, 53–70.
Dohme, F, Fievez, V, Raes, K and Demeyer, DI (2003) Increasing levels of two different fish
oils lower ruminal biohydrogenation of eicosapentaenoic and docosahexaenoic acids in
vitro. Anim. Res., 52, 309–320.
Doreau, M and Chilliard, Y (1997) Digestion and metabolism of dietary fat in farm animals.
Br. J. Nutr., 78, (suppl. 1), S15–S35.
Doreau, M and Ferlay, A (1994) Digestion and utilisation of fatty acids by ruminants. Anim.
Feed Sci. Techn., 45, 379–396.
Doreau, M, Lee, MRF, Ueda, K and Scollan, ND (2005) Métabolisme ruminal et digestibilité
des acides gras des fourrages. Renc. Rech. Ruminants, 12, 101–104.
Duckett, SK, Andrae, JG and Owens, FN (2002) Effects of high oil corn or added corn oil on
ruminal biohydrogenation of fatty acids and conjugated linoleic acid formation in beef
steers fed finishing diets. J. Anim. Sci., 80, 3353–3360.
Engle, TE, Spears, JW, Fellner, V and Odle, J (2000) Effects of soybean oil and dietary copper
on ruminal and tissue lipid metabolism in finishing steers. J. Anim. Sci., 78, 2713–2721.
Enjalbert, F (1995) Les lipides dans l’alimentation des ruminants. 2. Particularités de
l’utilisation digestive. Rev. Méd. Vét., 146, 383–392.
Enjalbert, F, Akraim, F, Juaneda, P, Troegeler-Meynadier, A and Nicot, MC (2006) Cis-15
intermediates of biohydrogenation in the duodenal flow of cows receiving linseed. Fourth
Euro Fed Lipid Congress, Madrid.
Enjalbert, F, Eynard, P, Nicot, MC, Troegeler-Meynadier, A, Bayourthe, C and Moncoulon,
R (2003) In vitro versus in situ ruminal biohydrogenation of unsaturated fatty acids from
a raw or extruded mixture of ground canola seed/canola meal. J. Dairy Sci., 86, 351–359.
Enjalbert, F, Nicot, MC, Bayourthe, C, Vernay, M and Moncoulon, R (1997) Effects of
dietary calcium soaps of unsaturated fatty acids on digestion, milk composition and
physical properties of butter. J. Dairy Res., 64, 181–195.
Enjalbert, F, Nicot, MC, Vernay, M, Moncoulon, R and Griess, D (1994) Effect of different
forms of polyunsaturated fatty acids on duodenal and serum fatty acids profiles in sheep.
Can. J. Anim. Sci., 74, 595–600
Fellner, V, Sauer, FD and Kramer, JKG (1997) Effect of nigericin, monensin, and tetronasin
on biohydrogenation in continuous flow-through ruminal fermenters. J. Dairy Sci., 80,
921–928.
Ferlay, A, Chabrot, J, Elmeddah, Y and Doreau, M (1993) Ruminal lipid balance and
intestinal digestion by dairy cows fed calcium salts of rapeseed oil fatty acids or rapeseed
oil. J. Anim. Sci., 71, 2237–2245.
Ferlay, A, Chilliard Y and Doreau M (1992) Effects of calcium salts differing in fatty acid
composition on duodenal and milk fatty acid profiles in dairy cows. J. Sci. Food Agric.,
60, 31–37.
Fievez, V, Vlaeminck, B, Jenkins, T, Enjalbert, F and Doreau, M (2007) Assessing rumen
biohydrogenation and its manipulation in vivo, in vitro and in situ. Eur. J. Lipid Sci.
Technol., 109, 740–756.
Fotouhi, N and Jenkins, TC (1992) Ruminal biohydrogenation of linoleoyl methionine and
calcium linoleate in sheep. J. Anim. Sci., 70, 3607–3614.
Franklin, ST, Martin, KR, Baer, RJ, Shingoethe, DJ and Hippen, A (1999) Dietary marine
algae (Schizochytrium sp.) increases concentrations of conjugated linoleic,
docosahexaenoic and transvaccenic acids in milk of dairy cows. J. Nutr., 129, 2048–
2052.
Fukuda, S, Furuya, H, Suzuki, Y, Asanuma, N and Hino, T (2005) A new strain of
Butyrivibrio fibrisolvens that has high ability to isomerize linoleic acid to conjugated
linoleic acid. J. Gen. Appl. Microbiol., 51, 105–113.
Fukuda, S, Suzuki, Y, Komiri, T, Kawamura, K, Asanuma, N and Hino, T (2006a)
Purification and gene sequencing of conjugated linoleic acid reductase from a
gastrointestinal bacterium, Butyrivibrio fibrisolvens. J. Appl. Microbiol., 103, 365–371.
Fukuda, S, Suzuki, Y, Murai, M, Asanuma, N and Hino, T (2006b) Isolation of a novel strain
of Butyrivibrio fibrisolvens that isomerizes linoleic acid to conjugated linoleic acid
without hydrogenation, and its utilization as a probiotic for animals. J. Appl. Microbiol.,
100, 787–794.
Fukuda, S, Suzuki, Y, Murai, M, Asanuma, N and Hino, T (2006c) Augmentation of
vaccenate production and suppression of vaccenate biohydrogenation in cultures of
mixed ruminal microbes. J. Dairy Sci., 89, 1043–1051.
Garton, GA, Hobson PN and Lough AK (1958) Lipolysis in the rumen. Nature, 182, 1511–
1512.
Gerson, T, John, A and King, ASD (1985) The effects of dietary starch and fibre on the in vitro
rates of lipolysis and biohydrogenation by sheep ruminal digesta. J. Agric. Sci. (Camb.),
105, 27–30.
Gerson, T, John, A and Sinclair, BR (1983) The effect of dietary N on in vitro lipolysis and
fatty acid hydrogenation in rumen digesta from sheep fed diets high in starch. J. Agric.
Sci. (Camb.), 101, 97–101.
Gerson, T, King, ASD, Kelly, KE and Kelly, WJ (1988) Influence of particle size and surface
area on in vitro rates of gas production, lipolysis of triglycerol and hydrogenation of
linoleic acid by sheep rumen digesta or Ruminococcus flavefaciens. J. Agric. Sci.
(Camb.), 110, 31–37.
Girard, V and Hawke, JC (1978) The role of holotrichs in the metabolism of dietary linoleic
acid in the rumen. Biochim. Biophys. Acta, 528, 17–27.
Glasser, F, Schmidely, P, Sauvant, D and Doreau, M (2008) Digestion of fatty acids in
ruminants: a meta-analysis of flows and variation factors: 2. C18 fatty acids. Animal, 2,
691–704.
Gonthier, C, Mustafa, AF, Berthiaume, R, Petit, HV and Ouellet, DR (2004) Feeding
micronized and extruded flaxseed to dairy cows: effects on digestion and ruminal
biohydrogenation of long-chain fatty acids. Can. J. Anim. Sci., 84, 705–711.
Gonthier, C, Mustafa, AF, Ouellet, DR, Chouinard, PY, Bertiaume, R and Petit, HV (2005)
Feeding micronized and extruded flaxseed to dairy cows: effects on blood parameters and
milk fatty acid composition. J. Dairy Sci,. 88, 748–756.
Griinari, JM and Bauman, DE (1999) Biosynthesis of conjugated linoleic acid. In: Ad-
vances in conjugated linoleic acid research, Vol.1 (MP Yurawecz, MM Mossoba,
JKG Kramer, MW Pariza and GJ Nelson, eds), AOCS Press, Champaign, Illinois,
USA, pp.180–200.
Griinari, JM, Dwyer, DA, McGuire, MA, Bauman, DE, Palmquist, DL and Nurmela, KV
(1998) Trans-octadecenoic acids and milk fat depression in lactating dairy cows. J. Dairy
Sci., 81, 1251–1261.
Gulati, SK, Ashes, JR and Scott, TW (1999) Hydrogenation of eicosapentaenoic and
docosahexaenoic acids and their incorporation into milk fat. Anim. Feed Sci. Technol.,
79, 57–64.
Gulati, SK, Kitessa, SM, Ashes, JR, Fleck, E, Byers, EB, Byers, YG and Scott, TW (2000)
Protection of conjugated linoleic acids from ruminal hydrogenation and their incorpora-
tion into milk fat. Anim. Feed Sci. Technol., 86, 139–148.
Halliwell, G and Bryant MP (1963) The cellulolytic activity of pure stains of bacteria from
the rumen of cattle. J. Gen. Microbiol., 32, 441–448.
Harfoot, CG (1978) Lipid metabolism in the rumen. Prog. Lipid Res., 17, 21–54.
Harfoot, CG and Hazlewood, GP (1997) Lipid metabolism in the rumen. In: The rumen
microbial ecosystem (PN Hobson and DS Stewart), Chapman and Hall, London, UK,
pp.382–426.
Harfoot, CG, Noble, RC and Moore, JH (1973a) Food particles as a site for biohydrogenation
of unsaturated fatty acids in the rumen (short communication). Biochem. J., 132, 829–
832.
Harfoot, CG, Noble, RC and Moore, JH (1973b) Factors influencing the extent of
biohydrogenation of linoleic acid by rumen micro-organisms in vitro. J. Sci. Food. Agric.,
24, 961–970.
Hart, SP and Polan, CE (1984) Effect of sodium bicarbonate and disodium phosphate on
animal performance, ruminal metabolism, and rate of passage in ruminating calves. J.
Dairy Sci., 67, 2356–2368.
Harvatine, KJ and Allen, MS (2006) Fat supplements affect fractional rates of ruminal fatty
acid biohydrogenation and passage in dairy cows. J. Nutr., 136, 677–685.
Hawke, JC and Silcock, WR (1969) Lipolysis and hydrogenation in the rumen. Biochem. J.,
112, 131–132.
Hawke, JC and Silcock, WR (1970) The in vitro rate of lipolysis and biohydrogenation in
rumen contents. Bioch. Bioph. Acta, 218, 201–212.
Hazlewood, GP, Kemp, P, Lander, D and Dawson, RMC (1976) C18 unsaturated fatty acid
hydrogenation patterns of some rumen bacteria and their ability to hydrolyse exogenous
phospholipids. Br. J. Nutr., 35, 293–297.
Heguy, JM, Juchem, SO, DePeters, EJ, Rosenberg, M, Santos, JEP and Taylor, SJ (2006)
Whey protein gel composites of soybean and linseed oils as a dietary method to modify
the unsaturated fatty acid composition of milk lipids. Anim. Feed Sci. Technol., 131, 370–
388.
Henderson, C (1971) A study of the lipase produced by Anaerovibrio lipolytica, a rumen
bacterium. J. Gen. Microbiol., 65, 81–89.
Hino, T and Fukuda, S (2006) Biohydrogenation of linoleic and linolenic acids, and
production of their conjugated isomers by Butyrivibrio fibrisolvens. Fourth Euro Fed
Lipid Congress, Madrid.
Hobson, PN (1965) Continuous culture of some anaerobic and facultatively anaerobic rumen
bacteria. J. Gen. Microbiol., 38, 167–180.
Hudson, JA, Cai, Y, Corner, RJ, Morvan, B and Joblin, KN (2000) Identification and
enumeration of oleic acid and linoleic acid hydrating bacteria in the rumen of sheep and
cows. J. Appl. Microbiol., 88, 286–292.
Hudson, JA, MacKenzie, CAM and Joblin, KN (1995) Conversion of oleic acid to 10-
hydroxystearic acid by two species of ruminal bacteria. Appl. Microbiol. Biotechnol., 44,
1–6.
Hudson, JA, Morvan, B and Joblin, KN (1998) Hydration of linoleic acid by bacteria isolated
from ruminants. FEMS Microbiol. Letters, 169, 277–282.
Hughes, PE, Hunter, WJ and Toce, SB (1982) Biohydrogenation of unsaturated fatty acids.
Purification and properties of cis9,trans11 octadecadienoate reductase. J. Biol. Chem.,
257, 3643–3649.
Jenkins, TC (1997) Ruminal fermentation and nutrient digestion in sheep fed
hydroxyethylsoyamide. J. Anim. Sci., 75, 2277–2283.
Jenkins, TC (1998) Fatty acid composition of milk from Holstein cows fed oleamide or canola
oil. J. Dairy Sci., 81, 794–800.
Jenkins, TC, AbuGhazaleh, AA, Freeman, S and Thies, EJ (2006) The production of 10-
hydroxystearic and 10-ketostearic acids is an alternative route of oleic acid transformation
by the ruminal microbiota in cattle. J. Nutr., 136, 926–931.
Jenkins, TC and Adams, CS (2002) The biohydrogenation of linoleamide in vitro and its
effects on linoleic acid concentration in duodenal content of sheep. J. Anim. Sci., 80, 533–
540.
Jenkins, TC and Bridges, WC Jr (2007) Protection of fatty acids against ruminal
biohydrogenation in cattle. Eur. J. Lipid Sci. Technol., 109, 778–789.
Jenkins, TC, Fellner, V and McGuffey, RK (2003) Monensin by fat interaction on trans fatty
acids in cultures of mixed ruminal microorganisms grown in continuous fermentors fed
corn or barley. J. Dairy Sci., 86, 324–330.
Jenkins, TC and Palmquist, DL (1982) Effect of added fat and calcium on in vitro formation
of insoluble fatty acid soaps and cell wall digestibility. J. Anim. Sci., 55, 957–963.
Jenkins, TC and Palmquist, DL (1984) Effects of fatty acids or calcium soaps on rumen and
total nutrient digestibility of dairy rations. J. Dairy Sci., 67, 978–986.
Jenkins, TC and Thies, E (1997) Plasma fatty acids in sheep fed hydroxyethylsoyamide, a
fatty acid amide that resists biohydrogenation. Lipids, 32, 173–178.
Jenkins, TC, Thompson, CE and Bridges, WC Jr (2000) Site of administration and duration
of feeding oleamide to cattle on feed intake and ruminal fatty acid concentrations. J. Anim.
Sci., 78, 2745–2753.
Jenkins, TC, Wallace, RJ, Moate, PJ and Mosley, EE (2008) Board-invited review: Recent
advances in biohydrogenation of unsaturated fatty acids within the rumen microbial
ecosystem. J. Anim. Sci., 86, 397–412.
Jouany, JP, Lassalas, B, Doreau, M and Glasser, F (2007) Dynamic features of the rumen
metabolism of linoleic acid, linolenic acid and linseed oil measured in vitro. Lipids, 42,
351–360.
Jurjanz, S, Monteils, V, Juaneda, P and Laurent, F (2004) Variations of trans octadecenoic acid
in milk fat induced by feeding different starch based diets to cows. Lipids, 39, 19–24.
Kalscheur, KF, Teter, BB, Piperova, LS and Erdman, RA (1997a) Effect of dietary forage
concentration and buffer addition on duodenal flow of trans-C18:1 fatty acids and milk
fat production in dairy cows. J. Dairy Sci., 80, 2104–2114.
Kalscheur, KF, Teter, BB, Piperova, LS and Erdman, RA (1997b) Effect of fat source on
duodenal flow of trans-C18:1 fatty acids and milk fat production in dairy cows. J. Dairy
Sci., 80, 2115–2126.
Kay, JK, Roche, JR, Solver, ES, Thomson, NA and Baumgard, LH (2005) A comparison
between feeding systems (pasture and TMR) and the effect of vitamin E supplementation
on plasma and milk fatty acid profiles in dairy cows. J. Dairy Res., 72, 322–332.
Kemp, P, Dawson and RMC (1968) Isomerization of linolenic acid by rumen micro-
organisms. Biochem. J., 109, 477–478.
Kemp, P and Lander, DJ (1984) Hydrogenation in vitro of alpha-linolenic acid to stearic acid
by mixed cultures of pure strains of rumen bacteria. J. Gen. Microbiol., 130, 527–533.
Kemp, P, Lander, DJ and Gunstone, FD (1983) The biohydrogenation of gamma-linolenic
acid to stearic acid by pure cultures of two rumen bacteria. Biochem. J., 216, 519–522.
Kemp, P, Lander, DJ and Gunstone, FD (1984) The hydrogenation of some cis- and trans-
octadecenoic acids to stearic acid by a rumen Fusocillsu sp. Brit. J. Nutr., 52, 165–170.
Kemp, P, White, RR and Lander, DJ (1975) The hydrogenation of unsaturated fatty acids by
five bacterial isolates from the sheep rumen, including a new species. J. Gen. Microbiol.,
90, 100–114.
Kepler, CR, Hirons, KP, McNeill, JJ and Tove, SB (1966) Intermediates and products of the
biohydrogenation of linoleic acid by Butryrivibrio fibrisolvens. J. Biol. Chem., 241,
1350–1354.
Kepler, CR and Tove, SB (1967) Biohydrogenation of unsaturated fatty acids: purification
and properties of a linoleate Δ12 cis- Δ11 trans isomerase from Butyrivibrio fibrisolvens.
J. Biol. Chem., 242, 5686–5692.
Kepler, CR, Tucker, WP and Tove, SB (1970) Biohydrogenation of unsaturated fatty acids.
IV. Substrate specificity and inhibition of linoleate delta12-cis, delta11-trans isomerase
from Butyrivibrio fibrisolvens. J. Biol. Chem., 245, 3612–3620.
Kim, EJ, Huws, SA, Lee, MRF, Wood, JD, Muetzel, SM, Wallace, RJ and Scollan, ND (2008)
Fish oil increases the duodenal flow of long chain polyunsaturated fatty acids and trans-
11 18:1 and decreases 18:0 in steers via changes in the rumen bacterial community. J.
Nutr., 138, 889–896.
Kim, EJ, Sanderson, R, Dhanoa, MS and Dewhurst, RJ (2005) Fatty acids profiles associated
with microbial colonization of freshly ingested grass and rumen biohydrogenation. J.
Dairy Sci., 88, 3220–3230.
Kim, YJ, Liu, RH, Bond, DR and Russel, JB (2000) Effect of linoleic acid concentration on
conjugated linoleic acid production by Butyrivibrio fibrisolvens. Appl. Environ. Microbiol.,
66, 5226–5230.
Kim, YJ, Liu, RH, Rychlik, JL and Russel, JB (2002) The enrichment of a ruminal bacterium
(Megasphaera elsdenii YJ-4) that produces the trans-10, cis-12 isomer of conjugated
linoleic acid. J. Appl. Microbiol., 92, 976–982.
Kraft, J, Collomb, M, Mockel, P, Sieber and R,Jarheis, G (2003) Differences in CLA isomer
distribution of cow’s milk lipids. Lipids, 38, 657–664.
Kramer, JPG, Parodi, PW, Jensen, RG, Mossoba, MM, Yurawecz, MP and Adlof, RO (1998)
Rumenic acid: a proposed common name for the major conjugated linoleic acid isomer
found in natural products. Lipids, 33, 835.
Kronberg, SL, Schollegerdes, EJ, Barcelo-Coblijn, G and Murphy, EJ (2007) Flaxseed

treatments to reduce biohydrogenation of alpha-linolenic acid by rumen microbes in
cattle. Lipids, 42, 1105–1111.
Kucuk, O, Hess, BW, Ludden, PA and Rule, DC (2001) Effect of forage:concentrate ratio on
ruminal digestion and duodenal flow of fatty acids in ewes. J. Anim. Sci., 79, 2233–2240.
Kucuk, O, Hess, BW and Rule, DC (2008) Fatty acid compositions of mixed ruminal
microbes isolated from sheep supplemented with soybean oil. Res. Vet. Sci., 84, 215–224.
Latham, MJ, Storry, JE and Sharpe, E (1972) Effect of low-roughage diets on the microflora
and lipid metabolism in the rumen. Appl. Microbiol., 24, 871–877.
Lee, MRF, Connelly, PL, Tweed, JKS, Dewhurst, RJ, Merry, RJ and Scollan, ND (2006a)
Effects of high-sugar ryegrass silage and mixtures with red clover silage on ruminant
digestion. 2. Lipids. J. Anim. Sci., 84, 3061–3070.
Lee, MRF, Harris, LJ, Dewhurst, RJ, Merry, RJ and Scollan, ND (2003) The effect of clover
silages on long chain fatty acid rumen transformations and digestion in beef steers. Anim.
Sci., 76, 491–501.
Lee, MRF, Huws, SA, Scollan, ND and Dewhurst, RJ (2007a) Effects of fatty acids oxidation
products (green odor) on rumen bacterial populations and lipid metabolism in vitro. J.
Dairy Sci., 90, 3874–3882.
Lee, MRF, Parfitt, LJ, Scollan, ND and Minchin, FR (2007b) Lipolysis in red-clover with
different polyphenol oxidase activities in the presence and absence of rumen fluid. J. Sci.
Food Agr., 87, 1308–1314.
Lee, MRF, Tweed, JKS, Dewhurst, RJ and Scollan, ND (2006b) Effect of forage:concentrate
ratio on ruminal metabolism and duodenal flow of fatty acids in beef steers. Anim. Sci.,
82, 31–40.
Lee, MRF, Tweed, JKS, Moloney, AP and Scollan, ND (2005) The effects of fish oil
supplementation on rumen metabolism and the biohydrogenation of unsaturated fatty
acids in beef steers given diets containing sunflower oil. Anim. Sci., 80, 361–367.
Lee, MRF, Winters, AL, Scollan, ND, Dewhurst, RJ, Theodorou, MK and Minchin, FR
(2004) Plant-mediated lipolysis and proteolysis in red clover with different polyphenol
oxidase activities. J. Sci. Food Agr., 84, 1639–1645.
Liavonchanka, A, Hornung, E, Feussner, I and Rudolph MG (2006) Structure and mechanism
of the Propionibacterium acnes polyunsaturated fatty acid isomerase. Proc. Natl. Acad.
Sci., 103, 2576–2581.
Lock, AL and Garnsworthy, PC (2002) Independent effects of linoleic and linolenic fatty
acids on the conjugated linoleic acid content of cows’ milk. Anim. Sci., 74, 163–176.
Loor, JJ, Bandara, ABPA and Herbein JH (2002a) Characterization of 18:1 and 18:2 isomers
produced during microbial biohydrogenation of unsaturated fatty acids from canola and
soya bean oil in the rumen of lactating cows. J. Anim. Physiol. Anim. Nutr., 86, 422–432.
Loor, JJ, Ferlay, A, Ollier, A, Doreau, M and Chilliard, Y (2005a) Relationship among trans
and conjugated fatty acids and bovine milk fat yield due to dietary concentrate and linseed
oil. J. Dairy Sci., 88, 726–740.
Loor, JJ, Ferlay, A, Ollier, A, Ueda, K, Doreau, M and Chilliard, Y (2005b) High-concentrate
diets and polyunsaturated oils alter trans and conjugated isomers in bovine rumen, blood
and milk. J. Dairy Sci., 88, 3886–3899.
Loor, JJ and Herbein JH (2003) Dietary canola or soybean oil with two levels of conjugated
linoleic acid (CLA) alter profiles of 18:1 and 18:2 isomers in blood plasma and milk fat
from dairy cows. Anim. Feed Sci., Technol., 103, 63–83.
Loor, JJ, Herbein, JH and Jenkins, TC (2002b) Nutrient digestion, biohydrogenation, and
fatty acid profiles in blood plasma and milk fat from lactating Holstein cows fed canola
oil or canolamide. Anim. Feed Sci., Technol., 97, 65–82.
Loor, JJ, Hoover, WH, Miller-Webster, TK, Herbein, JH and Polan, CE (2003)
Biohydrogenation of unsaturated fatty acids in continuous culture fermenters during
digestion of orchardgrass or red clover with three levels of ground corn supplementation.
J. Anim. Sci., 81, 1611–1627.
Loor, JJ, Ueda, K, Ferlay, A, Chilliard, Y and Doreau, M (2004) Biohydrogenation, duodenal
flow, and intestinal digestibility of trans fatty acids and conjugated linoleic acids in
response to dietary forage:concentrate ratio and linseed oil in dairy cows. J. Dairy Sci.,
87, 2472–2485.
Loor, JJ, Ueda, K, Ferlay, A, Chilliard, Y and Doreau, M (2005c) Intestinal flow and
digestibility of trans fatty acids and conjugated linoleic acids (CLA) in dairy cows fed a
high-concentrate diet supplemented with fish oil, linseed oil, and sunflower oil. Anim.
Feed Sci. Technol., 119, 203–225.
Lundy, FP, Block, E, Bridges, WC Jr, Bertrand, JA and Jenkins, TC (2004) Ruminal
biohydrogenation in Holstein cows fed soybean fatty acids as amides or calcium salts. J.
Dairy Sci., 87, 1038–1046.
Maia, MRG, Chaudhary, LC, Figueres, L and Wallace, RJ (2007) Metabolism of polyunsatu-
rated fatty acids and their toxicity to the microflora of the rumen. Antonie Van
Leeuwenhoek, 91, 303–314.
Martin, SA and Jenkins, TC (2002) Factors affecting conjugated linoleic acid and trans-
C18:1 fatty acid production by mixed ruminal bacteria. J. Anim. Sci., 80, 3347–3352.
Martineau, R, Petit, HV, Benchaar, C, Lapierre, H, Ouellet, DR, Pellerin, D and Berthiaume,
R (2008) Effects of lasalocid or monensin on in situ biohydrogenation of flaxseed and
sunflower seed unsaturated fatty acids. Can. J. Anim. Sci., 88, 335–339.
Moate, PJ, Boston, RC, Jenkins, TC and Lean, IJ (2008) Kinetics of ruminal lipolysis of
triacylglycerol and biohydrogenation of long-chain fatty acids: new insights from old
data. J. Dairy Sci., 91, 732–742.
Moate, PJ, Chalupa, W, Jenkins, TC and Boston, RC (2004) A model to describe ruminal
metabolism and intestinal absorption of long-chain fatty acids. Anim. Feed. Sci. Technol.,
112, 79–105.
Morales, MS, Palmquist, DL and Weiss, WP (2000) Effects of fat source and copper on
unsaturatuon of blood and milk triacylglycerol fatty acids in Holstein and Jersey cows.
J. Dairy Sci., 83, 2105–2111.
Morris, LJ (1970) Mechanisms and stereochemistry in fatty acid metabolism. Biochem. J.,
118, 681–693.
Morvan, B and Joblin, KN (2001) Ruminal bacteria and hydration of long-chain unsaturated
fatty acids. Reprod. Nutr. Dev., 40, 201 (Abstr.).
Mosley, EE, Nuda, A, Corato, A, Rossi, E, Jenkins, TC and McGuire, MA (2006a) Diffrential
biohydrogenation and isomerization of [U-13C]oleic and [1-13C]oleic acids by mixed
ruminal microbes. Lipids, 41, 513–517.
Mosley, EE, Powell, GL, Riley, MB and Jenkins, TC (2002) Microbial biohydrogenation of
oleic acid to trans isomers in vitro. J. Lipid Res., 43, 290–296.
Mosley, EE, Shafi, B, Moate, PJ and McGuire, MA (2006b) Cis-9, trans-11 conjugated
linoleic acid is synthesized directly from vaccenic acid in lactating dairy cattle. J. Nutr.,
135, 571–575.
Nam, IS and Garnsworthy, PC (2007) Biohydrogenation of linoleic acid by rumen fungi
compared with rumen bacteria. J. Appl. Microbiol., 103, 551–556.
Noble, RC, Moore, JH and Harfoot, CG (1974) Observations on the pattern of biohydrogenation
of esterified and unesterified linoleic acid in the rumen. Br. J. Nutr., 31, 99–108.
Ogawa, J, Kishino, S, Ando, A, Sugimoto, S, Mihara, K and Shimizu, S (2005) Production
of conjugated fatty acids by lactic acid bacteria. J. Bioscience Bioengeneering, 100, 355–
364.
Or-Rashid, MM, Odongo, NE and McBride, BW (2007) Fatty acid composition of ruminal
bacteria and protozoa, with emphasis on conjugated linoleic acid, vaccenic acid, and odd-
chain and branched-chain fatty acids. J. Dairy Sci., 85, 1228–1234.
Paillard, D, McKain, N, Chaudhary, LCL, Walker, ND, Pizette, F, Koppova, I, McEwan, NR,
Kopecny, J, Vercoe, PE, Louis, P and Wallace, RJ (2007a) Relation between phylogenetic
position, lipid metabolism and butyrate production by different Butyrivibrio-like bacteria
from the rumen. Antonie Van Leeuwenhoek, 91, 417–422.
Paillard, D, McKain, N, Rincon, MT, Shingfield, KJ, Givens, DI and Wallace, RJ (2007b)
Quantification of ruminal Clostridium proteoclasticum by real-time PCR using a
molecular beacon approach. J. Applied Microbiol., 103, 1251–1261.
Palmquist, DL, Lock, AL, Shingfield, KJ and Bauman, DE (2005) Biosynthesis of conjugated
linoleic acid in ruminants and humans. In: Advances in food and nutrition research (S
Taylor, ed) Elsevier, London, UK, pp.179–218.
Park, SJ, Park, KA, Park, CW, Park, WS, Kim, JO and Ha, YL (1996) Purification and amino
acid sequence of the linoleate isomerase produced from Butyrivibrio fibrisolvens A38. J.
Food Sci. Nutr., 1, 244–251.
Parodi, PW (1976) Distribution of isomeric fatty acids in milk fat. J. Dairy Sci., 59, 1870–
1873.
Peterson, DG, Kelsey, JA and Bauman, DE (2002) Analysis of variations of cis-9, trans-11
conjugated linoleic acid (CLA) in milk fat of dairy cows. J. Dairy Sci., 85, 2164–2172.
Piperova, LS, Sampugna, J, Teter, BB, Kalscheur, KF, Yurawecz, MP, Ku, Y, Morehouse,
KM and Erdman, RA (2002) Duodenal and milk trans octadecenoic acid and conjugated
linoleic acid (CLA) isomers indicate that postabsorptive synthesis is the predominant
source of cis-9containing CLA in lactating dairy cows. J. Nutr., 132, 1235–1241.
Pires, AV, Eastridge, ML, Firkins, JL and Lin, YC (1997) Effects of heat treatment and
physical processing of cottonseed on nutrient digestibility and production performance
by lactating cows. J. Dairy Sci., 80, 1685–1694.
Polan, CE, McNeill, JJ and Tove, CB (1964) Biohydrogenation of unsaturated fatty acids by
rumen bacteria. J. Bacteriol., 88, 1056–1064.
Pottier, J, Focant, M, Debler, C, De Buysser, G, Goffe, C, Mignolet, E, Froidmond, E and
Larondelle, Y (2006) Effect of dietary vitamin E on rumen biohydrogenation pathways
and milk fat depression in dairy cows fed high fat diets. J. Dairy Sci., 89, 685–692.
Prins, RA, Lankhorst, A, Van Der Meer, P and Van Nevel, CJ (1975) Some characteristices of
Anaerovibrio lipolytica, a rumen lipolytic organism. Antonie Van Leeuwenhoek, 41,1–11.
Proell, JM, Mosley, EE, Powell, GL and Jenkins, TC (2002) Isomerization of stable
isotopically labeled elaidic acid to cis and trans monoenes by ruminal microbes. J. Lipid
Res., 43, 2072–2076.
Reddy, PV, Morril, JL, Nagaraja, TG (1994) Release of free fatty acids from raw or processed
soybeans and subsequent effects on fiber digestibilities. J. Dairy Sci., 77, 3410–3416.
Ribeiro, CVDM, Eastridge, ML, Firkins, JL, St-Pierre, NR and Palmquist, DL (2007)
Kinetics of fatty acid biohydrogenation in vitro. J. Dairy Sci., 90, 1405–1416.
Ribeiro, CVDM, Karnati, SKR and Eastridge, ML (2005). Biohydrogenation of fatty acids
and digestibility of fresh alfalfa or alfalfa hay plus sucrose in continuous cultures. J. Dairy
Sci., 88, 4007–4017.
Romo, GA, Erdman, RA, Teter, BB, Sampugna, J and Casper, DP (2000) Milk composition
and apparent digestibilities of dietary fatty acids in lactating dairy cows abomasally
infused with cis or trans fatty acids. J. Dairy Sci., 83, 2609–2619.
Roy, A, Ferlay, A, Shingfield, KJ and Chilliard, Y (2006) Examination of the persistency of
milk fatty acid composition responses to plant oils in cows fed different basal diets, with
particular emphasis on trans-C18:1 fatty acids and isomers of conjugated linoleic acid.
Anim. Sci., 82, 479–492.
Sackman, JR, Duckett, SK, Gillis, MH, Realini, CE, Parks, AH and Eggelston, RB (2003)
Effects of forage and sunflower oil levels on ruminal biohydrogenation of fatty acids and
conjugated linoleic acid formation in beef steers fed finishing diets. J. Anim. Sci., 81,
3174–3181.
Schmidely, P, Glasser, F, Doreau, M and Sauvant, D (2008) Digestion of fatty acids in

ruminants: a meta-analysis of flows and variation factors. 1. Total fatty acids. Animal, 2,
677–690.
Scollan, ND, Lee, MRF and Enser, M (2003) Biohydrogenation and digestion of long chain
fatty acids in steers fed on Lolium perenne bred for elevated levels of water-soluble
carbohydrate. Anim. Res., 52, 501–511.
Scott, TW, Cook, LJ and Mills, SC (1971) Protection of dietary polyunsaturated fatty acids
against microbial hydrogenation in ruminants. J. Am. Oil Chem. Soc., 48, 358–364.
Sellner, DR and Schultz, LH (1980) Effects of feeding oleic acid or hydrogenated vegetable
oils to lactating cows. J. Dairy Sci., 63, 1235–1241.
Shingfield, KJ, Ahvenjärvi, S, Toivonen, V, Ärölä, A, Nurmela, KVV, Huhtanen, P and
Griinari, JM (2003) Effect of dietary ûsh oil on biohydrogenation of fatty acids and milk
fatty acid content in cows. Anim. Sci., 77, 165–179.
Shingfield, KJ, Ahvenjärvi, S, Toivonen, V, Vanhatalo, A, Huhtanen P and Griinari, JM
(2008) Effect of incremental levels of sunflower-seed oil in the diet of ruminal lipid
metabolism in lactating cows. Br. J. Nutr., 99, 971–983.
Shingfield, KJ, Reynolds, CK, Hervas, G, Griinari, JM, Grandison and AS, Beever, DE
(2006) Examination of the persistency of milk fatty acid composition responses to fish
oil and sunflower oil in the diet of dairy cows. J. Dairy Sci., 89, 714–732.
Shingfield, KJ, Reynolds, CK, Lupoli, B, Toivonen, V, Yurawecz, MP, Delmonte, P,
Griinari, JM, Grandison, AS and Beever, DE (2005) Effect of forage type and proportion
of concentrate in the diet on milk fatty acid composition in cows given sunflower and fish
oil. Anim. Sci., 80, 225–238.
Sinclair, LA, Cooper, SL, Chikunya, S, Wilkinson, RG, Hallett, KG, Enser, M and Wood, JD
(2005a) Biohydrogenation of n–3 polyunsaturated fatty acids in the rumen and their
effects on microbial metabolism and plasma fatty acid concentration in sheep. Anim. Sci.,
81, 239–248.
Sinclair, LA, Cooper, SL, Huntington, JA, Wilkinson, RG, Hallett, KG, Enser, M and Wood,
JD (2005b) In vitro biohydrogenation of n–3 polyunsaturated fatty acids protected
against ruminal microbial metabolism. Anim. Feed Sci., Technol., 123–124, 579–593.
Sklan, D, Arieli, A, Chalupa, W and Kronfeld, DS (1985) Digestion and absorption of lipids
and bile acids in sheep fed stearic acid, oleic acid, or tristearin J. Dairy Sci., 68, 1667–
1675.
Sukhija, PS and Palmquist, DL (1990) Dissociation of calcium soaps of long-chain fatty acids
in rumen fluids. J. Dairy Sci., 73, 1784–1787.
Thompson, GE and Christie, WW (1991) Extraction of plasma triacylglycerols by the
mammary gland of the lactating cow. J. Dairy Res. 58, 251–255.
Tice, EM, Eastridge, ML and Firkins, JL (1994) Raw soybeans and roasted soybeans of
different particle sizes. 2. Fatty acid utilization by lactating cows. J. Dairy Sci., 77, 166–
180.
Troegeler-Meynadier, A, Bret-Bennis, L and Enjalbert, F (2006a) Rates and efficiencies of
reactions of ruminal biohydrogenation of linoleic acid according to pH and polyunsatu-
rated fatty acids concentrations. Reprod. Nutr. Dev., 46, 713–724.
Troegeler-Meynadier, A, Nicot, MC, Bayourthe, C, Moncoulon, R and and Enjalbert, F
(2003) Effects of pH and concentrations of linoleic and linolenic acids on extent and
intermediates of ruminal biohydrogenation in vitro. J. Dairy Sci., 86, 4054–4063.
Troegeler-Meynadier, A, Nicot, MC and Enjalbert, F (2006b) Effects of heating process of
soybeans on ruminal production of conjugated linoleic acids and trans-octadecenoic
acids in situ. Rev; Méd. Vét., 157, 509–514.
Van de Vossenberg, JLCM and Joblin, KN (2003) Biohydrogenation of C18 unsaturated fatty
acids to stearic acid by a strain of Butyrivibrio hungatei from the bovine rumen. Letters
Appl. Microbiol., 37, 424–428.
Van Nevel, CJ and Demeyer, DI (1995) Lipolysis and biohydrogenation of soybean oil in the
rumen in vitro: inhibition by antimicrobials. J. Dairy Sci., 78, 2797–2806.
Van Nevel, CJ and Demeyer, DI (1996a) Effect of pH on biohydrogenation of polyunsatu-
rated datty acids and their Ca-salts by rumen microorganisms in vitro. Arch. Anim. Nutr.,
49, 151–157.
Van Nevel, CJ and Demeyer, DI (1996b) Influence of pH on lipolysis and biohydrogenation
of soybean oil by rumen contents in vitro. Reprod. Nutr. Dev., 36, 53–63.
Vázquez-Añón, M and Jenkins, TC (2007) Effects of feeding oxidized fat with or without
dietary antioxidants on nutrient digestibility, microbial nitrogen, and fatty acid metabo-
lism. J. Dairy Sci., 90, 4361–4367.
Vlaeminck, B, Mengitsu, G, Fievez, V, de Jonge, L and Dijkstra, J (2008) Effect of in vitro
docosahexaenoic acid supplementation to marine algae-adapted and unadapted rumen
inoculum on the biohydrogenation of unsaturated fatty acids in freeze-dried grass. J.
Dairy Sci., 91, 1122–1232.
Wallace, RJ (2008) Gut microbiology – broad genetic diversity, yet specific metabolic niches.
Animal, 2, 661–668.
Wallace, RJ, Chaudhary, LC, McKain, N, McEwan, NR, Richardson, AJ, Vercoe, PE,
Walker, ND and Paillard, D (2006) Clostridium proteoclasticum: a ruminal bacteria that
forms stearic acid from linoleic acid. FEMS Microbiol. Lett., 265, 195–201.
Wallace, RJ, McKain, N, Shingfield, KJ and Devillard, E (2007) Isomers of conjugated
linoleic acids are synthesized via different mechanisms in ruminal digesta and bacteria.
J. Lipid Res., 48, 2247–2254.
Ward, PFV, Scott, TW and Dawson, RMC (1964) The hydrogenation of unsaturated fatty
acids in the ovine digestive tract. Biochem. J., 92, 60–68.
Wasowska, I, Maia, MRG, Niedzwiedzka, KM, Czauderna, M, Ramalho-Ribeiro, JMC,
Devillard, E, Shingfield, KJ and Wallace, RJ (2006) Influence of fish oil on ruminal
biohydrogenation of C18 unsaturated fatty acids. Br. J. Nutr., 95, 1199–1211.
Weimer, PJ, Stevenson, DM, Mertens, DR and Thomas, EE (2008) Effect of monensin
feeding and withdrawal on populations of individual bacterial species in the rumen of
lactating dairy cows fed high-starch rations. Applied Microbial Cell Physiol., 80, 135–
145.
White, RW, Kemp, P and Dawson, MC (1970) Isolation of a rumen bacterium that
hydrogenates oleic acid as well as linoleic acid and linolenic acid. Biochem. J., 116, 767–
768.
Winkler, K and Steinhart, H (2001) Identification of conjugated isomers of linolenic acid and
arachidonic acid in cheese. J. Separation Sci., 24, 663–668.
Wonsil, WB, Herbein, JH and Watkins, BA (1994) Dietary and ruminally derived trans-18:1
fatty acids alter bovine milk lipids. J. Nutr., 124, 556–565.
Wu, Z, Ohajuruka, A and Palmquist, DL (1991) Ruminal synthesis, biohydrogenation, and
digestibility of fatty acids by dairy cows. J. Dairy Sci., 74, 3025–3034.
Wu, Z and Palmquist, DL (1991) Synthesis and biohydrogenation of fatty acids by ruminal
microorganisms in vitro. J. Dairy Sci., 74, 3035–3046.
Yáñez-Ruiz, DR, Scollan, ND, Merry, RJ and Newbold, CJ (2006) Contribution of rumen
protozoa to duodenal flow of nitrogen, conjugated linoleic acid and vaccenic acid in steers
fed silages differing in their water-soluble carbohydrate content. Br. J. Nutr., 96, 861–
869.
CHAPTER 2
Formation of trans fatty acids during catalytic
hydrogenation of edible oils
JEAN-BAPTISTE BEZELGUES1 AND ALBERT J. DIJKSTRA2
1
Nestlé Product Technology Center, Marysville, Ohio, USA
2
Carbougnères, St Eutrope-de-Born, France
In the early 20th century, with the increasing world demand for solid, edible
fats, especially for bakery fats, margarine and shortenings, much attention was
given to the development of alternatives for lard and tallow that maintained a
good quality during prolonged storage. In 1897, catalytic hydrogen addition to
the double bond of organic material was demonstrated by Sabatier and, based
on Normann’s patent (1903), the hydrogenation of edible oils was quickly
developed on an industrial scale in the UK (Hastert, 1998). Progressively
during the 20th century, this technology became one of the most important
processes in fats and oils transformation, allowing available fats and oils to be
modified in accordance with demands for specific physical properties (such as
melting point and consistency), with improved oxidative and thermal stability
at an acceptable price.
In comparison with most industrial hydrogenation processes, the hydrogena-
tion of vegetable oils (also known as fat hardening) is exceptional. The former
aim at completing the reduction or saturation reactions, but in the case of
triacylglycerols, the reaction aims at partial conversion. Triacylglycerol spe-
cies in the starting material can have one, two or three monounsaturated or
polyunsaturated fatty acids. During hydrogenation, triacylglycerol ethylenic
double bonds progressively disappear by being saturated by hydrogen. Before
disappearing, their positions can shift along the fatty acid chain (positional
isomerization) and/or their geometry can change from cis to trans configura-
tion and back leading to various trans fatty acids (TFA) in the final product.
The TFA levels (up to 80% of the ethylenic double bonds) and the isomeric
profile can vary tremendously according to operational condition and the
starting material.
It is worth noting that high temperatures applied during refining of oils can
also cause geometrical isomerization of ethylenic double bonds. In particular
during the deodorization step, vegetable oils are exposed to temperatures
ranging from 180 to 260ºC depending on the type of oil and process conditions.
43
The sensitivity to heat-induced geometrical isomerization depends strongly on

the number of ethylenic double bonds in a fatty acid and their positions in the
fatty acid chain.
From a nutritional point of view, numerous epidemiological and clinical
studies have provided evidence for detrimental health effects of dietary TFA
(Willett & Ascherio, 1994; Mozaffarian, 2006). More particularly, trans
isomers of octadecenoic acid (trans-18:1), as present in partially hydrogenated
vegetable oils can modify plasma cholesterol levels in humans, raising the
concentration of low-density lipoprotein (LDL) and lowering that of high-
density lipoprotein (HDL), thus possibly increasing the risk of cardiovascular
disease (Mozaffarian, 2006). In this context, the health authorities have started
to set-up regulations (Eller et al., 2005) to limit the consumption of TFA by
labelling TFA content in food products. In Europe, a code of good practice was
introduced by the International Margarine Association (IMACE) recommend-
ing since 2002 a TFA level of less than 1% in margarine and spreads. In June
2003, Denmark enacted the most stringent rule with a maximum of 2% TFA of
industrial origin in all food products. Since January 2006 food labels in the
USA must mention TFA content as a separate line and a product containing less
than 0.5 g TFA per serving can be declared as “zero trans”.
Nevertheless, oils and fats manufacturers had already started more than
15 years ago to better control and reduce TFA formation during industrial
hardening. This chapter will describe some general aspects of the hydrogena-
tion reaction mechanism and the hydrogenation process, and will provide an
update on recent improvements in reduction of TFA levels in edible fats and
oils during hardening.
A. Mechanism of trans fatty acid (TFA) formation during catalytic

hydrogenation of edible oils and fats
The hydrogenation of a pure substance, for instance trilinoleate, will lead to the
formation of tens of different fatty acids and thousands of different
triacylglycerols. The analysis of such mixtures is a difficult, if not impossible
task. Therefore, more or less successful attempts have been made to describe in
simplified terms what happens during a hydrogenation reaction.
The progress of the actual hydrogenation reaction is generally monitored by
the decrease in the iodine value. Net geometrical isomerization of double bonds
can be described by the total TFA content of the oil, which can be determined
by Fourier transform near-infrared (FT-NIR) spectroscopy. Physical properties
of the hydrogenated product can be described by its melting point, its solid fat
index as determined by dilatometry or by its solid fat content as determined by
pulse nuclear magnetic resonance. With the advent of gas-liquid chromatogra-
phy, fatty acid composition could be determined much more accurately and
when capillary gas chromatography was introduced, TFA isomer distribution
FORMATION OF TFA DURING CATALYTIC HYDROGENATION 45
and content could be determined accurately. This greatly assisted the observa-
tion of what happens on a molecular scale during the hydrogenation reaction,
but did not reveal it in detail.
1. Theories that turned out to be wrong

These observations allowed various theories to be put forward concerning the
kinetics and mechanism of the hydrogenation reaction. As is only to be
expected, several of these theories turned out to be invalid or incomplete.
In 1946, before fatty acid compositions could be determined by gas-liquid
chromatography, Bailey and Fisher suggested the concept of a ‘common fatty
acid pool’, meaning that the rate of reaction of a given fatty acid does not
depend on the nature of the other fatty acid moieties present in the same
triacylglycerol molecule. This suggestion contravened the current thinking
since, although Bushell and Hilditch (1937) observed that “the tri-unsaturated
glycerides are attacked more rapidly than the di-oleo-glycerides, and the latter
somewhat more so than the mono-oleo-compounds.”, the differences were
much smaller than those corresponding to the oleic acid content.
When studying the hydrogenation of a mixture of trilinolenin and dipalmito-
monolinolenin, Schilling (1977; 1978) concluded that the trilinolenin does not
react three times as fast as the monolinolenin and thus provided firm evidence
that the ‘common fatty acid pool’ concept is not correct. Similar evidence was
provided by the hydrogenation of a mixture of sunflower seed oil, providing
trilinolein, and the randomization product of sunflower seed oil and a synthetic
triacylglycerol containing only caprylic (8:0) and capric (10:0) fatty acids,
providing monolinolein (Beyens & Dijkstra, 1983). Again, the linoleic acid
was observed to react faster when it was the only unsaturated moiety in a
triacylglycerol molecule than when there were other unsaturated moieties as
well. Accordingly, experimental evidence to the contrary necessitated that the
‘common pool concept’ be abandoned (Dijkstra, 1997). This also means that
nearly all kinetic studies of the hydrogenation of triacylglycerols as reported in
the literature have to be re-evaluated.
This also holds for fatty acid selectivities, like the linoleic acid selectivity,
which has been defined as the ratio of the hydrogenation rate constants of
linoleic acid and oleic acid. This definition assumes the common pool concept
to be valid. Moreover, it also assumes that the rates of reaction of the various
fatty acids depend in an identical manner on the hydrogen concentration. Being
a ratio of two constants, the linoleic acid selectivity should itself also be a
constant, but it is not (Dijkstra, 1997; 2002a): in the early stages of a hydro-
genation run, it is high and fairly constant, but when the linoleic acid content
has fallen to around 15%, the calculated selectivity decreases sharply. Accord-
ingly, the rates of reaction vary according to hydrogen concentration. This also
holds for the linolenic acid selectivity.
The literature tends to be rather vague about the maximum TFA content and
mentions, for example, a nickel sub sulphide catalyst (Baltes, 1970; 1972) that
causes close to 100% of the double bonds to have a trans configuration. This is
not correct. The cis/trans equilibrium is governed by the enthalpy difference of
ΔHiso = – 4 kJ /mol. Accordingly, its position is temperature dependent and
shows, for instance, 79% trans at 100ºC and 72% trans at 250ºC (Dijkstra,
2006).
The literature regularly mentions ‘shunt reactions’ taking place during the
hydrogenation of polyunsaturated triacylglycerols. The ‘oleate shunt’, being a
‘direct-through’ reaction of linolenic acid to oleic/elaidic acid, was suggested
by Bailey (1949). Subsequent authors (Mounts & Dutton, 1967) even sug-
gested stearate shunts in which linolenic acid and linoleic acid would react
straight through to stearic acid. Their conclusion that these reaction paths
existed was based on otherwise inexplicable deviations from kinetic models. In
retrospect, the validity of these models is doubtful, since they do not take into
account triacylglycerol selectivity or the order with respect to hydrogen of the
various fatty acid moieties. Besides, the only difference in reaction rate
between fatty acid isomers taken into account is between linoleic acid and
isolinoleic (cis-9,cis-15 18:2) acid. Accordingly, the models used are oversim-
plified and do not provide valid evidence for the existence of shunt reactions.
When an oil like sunflower oil is hydrogenated and elaidic acid is observed
in the reaction product, it is impossible to say whether this trans isomer of oleic
acid results from the isomerization of oleic acid or from the hydrogenation of
linoleic acid. On the other hand, when high-erucic-acid rapeseed (HEAR) oil is
hydrogenated, the isomerization of monounsaturated acid can be followed by
looking at just the erucic (22:1) acid since HEAR oil does not contain a large
amount of dodecadienoic acids. Coenen and Boerma (1968) studied the
formation of brassidic (trans-22:1) acid during hydrogenation of HEAR oil at
100ºC. They noted the formation of this trans isomer always coincided with the
formation of behenic (22:0) acid and concluded that monounsaturated fatty
acids cannot isomerize without some of them being reduced at the same time.
This turned out to be an unwarranted generalization. When the experiment with
HEAR oil was repeated at a much higher temperature and under selective
conditions, the erucic acid was observed to isomerize without behenic acid
being formed (W.L.J. Meeussen, personal communication).
As will be explained later on, the reason for the different outcome of the two
experiments lies in the hydrogen concentration, which was much higher in the
low temperature experiment than in the high temperature one. During most
hydrogenation experiments, the hydrogen concentration varies and this is
something most authors, with one exception (Coenen, 1978), do not take into
account or even deny (Jonker, 1999). In the beginning of an industrial hydro-
genation experiment or its laboratory mimic, the iodine value and the reactivity
of the reaction mixture will still be high. This high reactivity causes the
hydrogen concentration to be much lower than its solubility. In fact, the

hydrogen concentration will be governed by the demand for hydrogen as
determined by its concentration and the reactivity (iodine value) of the reaction
mixture, and its supply as determined by the gas-liquid volumetric gas transfer
coefficient kla and the difference between the actual hydrogen concentration
and its solubility. Since a decrease in hydrogen concentration causes the rate of
hydrogenation to decrease and the rate of hydrogen dissolution to increase, a
dynamic equilibrium between demand and supply will establish itself charac-
terized by an increase in hydrogen concentration during the course of the
experiment and in line with the decrease in reactivity of the reaction mixture
(Dijkstra, 1990; 1997).
2. Currently accepted hydrogenation mechanism

It is now commonly accepted that the nickel-catalysed hydrogenation of edible
oils follows a Horiuti-Polanyi mechanism (Horiuti & Polanyi, 1934). The
reactions involved in the hydrogenation of dienoic fatty acids according to this
mechanism have been depicted in Figure 1 (Jonker, 1999; Dijkstra, 2006). This
scheme is simplified since it does not show positional isomerization products;
it is limited to the geometrical cis/trans isomerization. The asterisk (*) indi-
cates that the compound concerned has been adsorbed onto the catalyst surface.
Accordingly, free molecular hydrogen is adsorbed from the bulk of the oil onto
the catalyst in reaction step (1) and, in reaction step (2), this adsorbed hydrogen
molecule dissociates to form two adsorbed hydrogen atoms; both steps are
reversible. In reaction step (3), all-cis linoleic acid (indicated as c,c-D for
‘diene’) is adsorbed onto the catalyst surface and in step (4), it reacts with an
adsorbed hydrogen atom to form a half-hydrogenated intermediate (c-DH*)
which still has one cis bond left. This intermediate c-DH* can now react in
three different ways. It can react with a further hydrogen atom in step (10) and
form a monoene (c-M) with the cis double bond that was still present in the
intermediate (c-DH*), or it can dissociate into a hydrogen atom and a diene,
whereby the double bond formed during this dissociation can have a cis or a
trans configuration. If the double bond has the cis configuration, the step is (4)
whereas step (5) leads to the trans configuration in c,t-D*. There is no
positional isomerization of the double bond if the dissociating hydrogen leaves
from the same carbon atom to which it was added in step (4). If a hydrogen atom
leaves from an adjacent carbon atom, a positional isomer results.
As shown in Figure 1, the exothermicity of the saturation of the double bond
causes the hydrogenated monoene to leave the catalyst’s surface. On the other
hand, if the half-hydrogenated intermediate (c-DH*) dissociates, the dissocia-
tion product (c,c-D*or c,t-D*) remains adsorbed. Which reaction will be
preferred in practice depends strongly upon the concentration of adsorbed
hydrogen atoms [H*]. If this is relatively high, hydrogen addition according to
1 2
H2 H2* 2 H*
c,c-D c,t-D t,t-D

3 6 9
4 5 7 8
c,c-D*+ H* c-DH* c,t-D*+ H* t-DH* t,t-D* + H*
+ +
H* H*
10 11
c-M t-M
13
12
14 15
c-M*+ H* MH* t-M* + H*
+
H*
16
S
Figure 1. Horiuti-Polanyi mechanism (see text for explanation).
step (10) will be favoured and if this is low, steps (4) and (5) will be favoured.
Accordingly, the hydrogen concentration determines how many double bonds
are isomerized per double bond being saturated. This ratio is commonly
referred to as the ‘isomerization index’.
The hydrogen concentration on the catalyst’s surface [H*] can be low for two
reasons. The catalyst can be poisoned (Baltes, 1972; Rijnten & Eikema, 1975)
so that a given concentration of hydrogen in the oil [H2] corresponds to a lower
concentration of atomic hydrogen on the catalyst’s surface [H*], or the
molecular hydrogen concentration [H2] in the bulk of the oil can be low. This
latter phenomenon can have a number of causes such as the high reactivity of
the hydrogenation substrate, a low pressure in the system, or a small gas-liquid
volumetric gas transfer coefficient kla due to, for instance, slow agitation.
Other reasons can be a high temperature, a relatively high amount of catalyst
and a relatively active catalyst, all of which increase the rate of the reaction and
thus lower the molecular hydrogen concentration [H2] in accordance with the
dynamic equilibrium mentioned above. According to the mechanism shown in
Figure 1, a monoene (c-M or t-M) can adsorb onto the catalyst surface via steps
(12) or (13) and react with an adsorbed hydrogen atom to form a half-
hydrogenated monoene (MH*). As before, this intermediate can react via step
(16) with a further hydrogen atom and become fully saturated, or dissociate.
This dissociation may lead to isomerization that can be geometrical, positional
or both.
In the partial hydrogenation of edible oils, it is often the intention to saturate
polyenes but to refrain from saturating monoenes and forming saturated fatty
acids. The reason for this intention is simply that saturated fatty acids lead to an
increase in triacylglycerol melting point and this can cause a sticky mouthfeel.
Accordingly, the ratio of the rates of reaction of dienes and monoenes (also
referred to as the ‘linoleic acid selectivity’) is an important process character-
istic that merits a detailed discussion. It has already been mentioned that this
ratio is not constant during the course of a hydrogenation process but decreases
when the reactivity of the substrate decreases. This has been tentatively
explained (Dijkstra, 1997; 2002a), by assuming that for diene hydrogenation,
the rate determining step is the formation of the half-hydrogenated intermedi-
ate (cDH*); accordingly this rate is proportional to the concentration of the
atomic hydrogen [H*]. For monoene hydrogenation on the other hand, the
saturation of the half-hydrogenated intermediate (MH*) determines the overall
rate of stearic acid formation. Since the equilibrium concentration of the half-
hydrogenated intermediate MH* is proportional to the concentration of adsorbed
atomic hydrogen [H*] and the rate of saturation of this intermediate is also
proportional to this adsorbed atomic hydrogen concentration [H*], the rate of
stearic acid formation is proportional to [H*]2 or [H2].
Accordingly, when the concentration of molecular hydrogen in the bulk of
the oil increases because the reactivity of the substrate decreases, the concen-
tration of the atomic hydrogen [H*] adsorbed increases. This favours monoene
saturation over diene saturation and causes the linoleic acid selectivity to
decrease. However, in addition to the hydrogen concentration, there is another
reason why monoenes may react differently from dienes. This reason has to do
with the catalyst. If the catalyst would only adsorb dienes and would show no
affinity for monoenes whatsoever, no monoenes would be hydrogenated and
no stearic acid would be formed. This would correspond to an infinite linoleic
acid selectivity. Much effort has gone into developing catalysts with this kind
of adsorption preference, but no nickel catalyst with an absolute linoleic acid
selectivity has as yet resulted.
On the other hand, copper catalysts that do not hydrogenate monoenes have
been developed and studied in depth. These catalysts do not cause monoenes to
be saturated since they only catalyse the hydrogenation of conjugated double
bonds. Accordingly, the hydrogenation process commences with this conjuga-
tion and it is likely that this conjugation is initiated by the abstraction of a
bis-allylic hydrogen atom rather than by the addition of a hydrogen atom to one
of the double bonds (Dijkstra, 2002b). As only to be expected, this hydrogen
abstraction and the subsequent addition of a hydrogen atom are reversible
reactions and the intermediates involved lose their original geometrical con
Table 1. Influence of lecithin content during hydrogenation of soybean oil.
Added lecithin (ppm phosphorus)

Fatty acids (g per 100 g of fatty acids) 0 4 8
18:0 15.7 18.0 33.0

18:1 69.8 65.2 41.8
18:2 3.6 5.9 13.8
Trans fatty acids 49 41 24
Calculated iodine value 64.7 64.7 58.3
Melting point (°C) 42.5 47.5 59.0
figuration. Consequently, this conjugation goes hand in hand with cis/trans

isomerization, which means that the use of copper catalysts does not avoid the
formation of TFA from polyunsaturated fatty acids.
Finally, catalyst structure and impurities in the oil being hydrogenated can
also affect the TFA content of the partially hydrogenated product, its fatty acid
composition and consequently, its physical properties. Catalysts with wide
pores permit the triacylglycerol molecules to travel quite freely to and from the
nickel surface, whereas a narrow pore catalyst tends to retain the substrate and
thus promote multiple hydrogenation of the same molecule (Coenen, 1978).
Phosphatides can have an even larger effect. As reported by Ariaansz (1996),
it has been postulated that these compounds attach themselves in slimy layers
to the catalyst surface and partially block the pore entry. They hinder diffusion
thus promoting multiple saturation of the same molecule (as illustrated in
Table 1).
B. Industrial hydrogenation of edible oils and fats

At the present time, industrial hydrogenation is still a classical way of convert-
ing liquid oils to solid or semi-solid plastic fats for bakery, margarine, shortening,
frying fats, confectionery etc. Since its development one hundred years ago,
catalytic hydrogenation of edible oils has remained relatively unchanged and
as described above, geometric cis/trans and positional isomerization occur
during the reaction leading to the formation of TFA. For the time being,
agricultural feedstocks such as soybean oil, palm oil, rapeseed oil, corn germ
oil and sunflower seed oil are usually hydrogenated or partially hydrogenated.
Even if fractionation and interesterification today constitute valuable technolo-
gies to produce ‘zero trans’ edible fats offering interesting physical
characteristics, hydrogenation still remains one of the most important indus-
trial processes for modification of oils. Its annual production, which is beginning
to decrease, is estimated at around 5 million tonnes (Farr, 2005), which
represents around 5% of world oil production.
As illustrated by Table 2, commercial partially hydrogenated vegetable oils
Table 2. Trans octadecenoic acid isomers level (g per 100 g of fatty acids) and profile
(% of total trans octadecenoic acid isomers) in commercial partially hydrogenated
vegetable oils (source: Nestlé, 2005)
Sample 1 2 3 4 5 6 7 8 9 10
Level of trans-18:1 acid isomers (g per 100 g of total fatty acids)

Total trans-18:1 60.7 50.8 51.4 48.8 51.1 12.7 28.5 35.6 42.0 52.3
trans-4 18:1 0.4 0.2 0.1 0.1 0.1 0.0 0.0 0.1 0.0 0.3
trans-5 18:1 1.1 0.3 0.3 0.3 0.2 0.0 0.1 0.4 0.1 0.9
trans-6/8 18:1 3.8 5.6 5.2 7.0 6.5 2.1 5.2 7.9 7.3 11.5
trans-9 18:1 6.9 8.2 10.4 7.0 9.0 2.3 10.6 6.1 15.4 7.0
trans-10 18:1 10.1 9.7 10.6 9.8 10.3 3.5 5.6 6.0 7.6 6.7
trans-11 18:1 7.4 9.4 11.2 7.9 8.6 2.3 2.6 3.4 3.3 5.6
trans-12 18:1 11.8 6.8 7.3 6.2 7.1 1.4 0.0 4.2 2.1 7.2
trans-13/14 18:1 16.8 5.6 5.2 6.6 6.0 0.0 2.4 4.6 3.4 8.7
trans-15 18:1 1.5 3.8 0.0 2.6 2.9 0.8 2.0 2.1 2.8 2.9
trans-16 18:1 1.0 1.2 1.1 1.1 0.5 0.3 0.1 0.9 0.1 1.5
Relative distribution of trans-18:1 acid isomers (%)
trans-4 18:1 0.7 0.3 0.3 0.3 0.2 0.0 0.0 0.4 0.0 0.6
trans-5 18:1 1.7 0.7 0.6 0.7 0.4 0.2 0.2 1.0 0.2 1.7
trans-6/8 18:1 6.3 11.0 10.2 14.4 12.7 16.5 18.4 22.2 17.3 21.9
trans-9 18:1 11.3 16.1 20.1 14.4 17.6 17.7 37.3 17.2 36.7 13.5
trans-10 18:1 16.7 19.0 20.6 20.0 20.2 27.6 19.6 16.7 18.0 12.8
trans-11 18:1 12.2 18.6 21.9 16.2 16.7 18.1 9.0 9.7 7.9 10.7
trans-12 18:1 19.5 13.4 14.1 12.8 13.8 10.9 0.0 11.7 4.9 13.7
trans-13/14 18:1 27.6 11.1 10.1 13.6 11.7 0.1 8.3 12.9 8.1 16.7
trans-15 18:1 2.4 7.5 0.0 5.3 5.6 6.5 6.9 5.8 6.7 5.6
trans-16 18:1 1.7 2.3 2.2 2.3 1.0 2.6 0.2 2.5 0.2 2.9
60
Soya bean oil
50 Fish oil
Palm oil
TFA content (%)
40
30
20
10
0
180 160 140 120 100 80 60 40 20 0
Iodine value
Figure 2. Trans isomer formation in three different oils under classical hydrogenation conditions
(adapted from Engelhard).
have TFA isomer contents up to 60%. Most of the TFA isomers found in
partially hydrogenated vegetable oils are monounsaturated fatty acids, which
are partly responsible for the physical properties (plasticity and melting behav-
iour). As depicted in Figure 2, TFA content varies with the degree of
unsaturation. The level depends strongly on the nature (iodine value) of the
starting material. The higher the starting iodine value, the higher will be the
level of TFA isomers at any point in the conventional hydrogenation process.
1. The hydrogenation process

Usually the reaction is carried out in a three-phase system consisting of gaseous
hydrogen, liquid oil and solid catalyst, generally nickel. Nowadays, hardening
is performed semi-continuously in vessels which have to withstand a tempera-
ture of up to 250ºC and usually a hydrogen pressure up to 5 bars (Bockisch,
1998; Beers, 2007).
The oil is pumped into the vessel and heated to the gassing temperature. A
mixture of catalyst with oil, prepared separately, is pumped into the reactor
under stirring. Hydrogen is then injected, the reaction starts and the mixture is
self heated to the targeted reaction temperature. Indeed, the reduction of
triacylglycerols containing unsaturated fatty acid by direct addition of hydro-
gen is highly exothermic with around 125 kJ being released per mole of double
bound, which is equivalent to a temperature increase of 1.6 to 1.7ºC per unit
lowering of iodine value (Faur, 1996). Therefore, at this stage strong cooling is
required to control the kinetics of the reaction. At the end of the reaction (as
determined by iodine value and/or refractive index) hydrogen is sucked from
the headspace. The mixture of oil and catalyst is then cooled to around 100ºC
and filtered to remove the catalyst. Hydrogenated oils must be post-treated with
bleaching earth and possibly filter aid to remove the catalyst completely;
finally they are deodorized.
Two principal types of batch process employing stirred autoclaves are
widely used: the ‘dead end’ process with internal hydrogen circulation and the
‘circulating’ process with external hydrogen circulation (Bockisch, 1998). The
first one, shown in Figure 3, needs intensive stirring to disperse the hydrogen
which is distributed near the bottom of the autoclave and which collects in the
headspace of the autoclave. Since volatile hydrogenation compounds are
formed and collect in the headspace their pressure may cause the pressure-
controlled hydrogen supply to be switched off. Therefore, by applying an
external re-circulation of hydrogen with compressors or jet pumps, the agitator
has only to keep the catalyst in suspension while volatile by-products are
continuously carried out and condensed outside the hydrogenator. The ther-
moregulation of the reaction is very important to control the quality of the
hydrogenated oil. Good isothermal conditions with temperature fluctuations of
only 1ºC can be achieved in a third type of reactor such as a Venturi loop-
Figure 3. Hydrogenation: an illustration of the dead end process.
Hydrogen
Venturi
mixer
Heat exchanger
Oil with
suspended
catalyst
Circulating pump
Figure 4. Venturi loop reactor (source: Buss AG, Basel, Switzerland).

reactor (Buss AG, Switzerland, see Figure 4) through which the reaction
mixture is continuously circulated at great velocities causing intense, intimate
mixing of the oil with the hydrogen being sucked into the venturi (Duveen &
Leuteritz, 1982).
An important aspect in hydrogenation is the design of the agitator, which has
to ensure a proper gas dispersion and good mass transfer during the reaction.
Radial flow flat-blade and axial flow flat-blade turbines are commonly used
providing high shear rates and vortex formation. However, it has been shown
that mixing intensities achievable with these rotary stirrers in industrial-scale
reactors are not high enough to control, significantly, the selectivity and
geometrical isomerization rate occurring during nickel-catalysed hydrogena-
tion of oils (Ackman & Mag, 1998).
Despite the interest in continuous hydrogenation, few continuous plants are
in operation.
2. Hydrogenation parameters
As described previously, the rate and selectivities of the hydrogenation reac-
tion are mainly the result of process parameters such as temperature, hydrogen
pressure, stirring conditions, catalyst concentration and catalyst type. By
varying these parameters, a great variety of finished products with different
physical properties can be obtained from the same starting material. To
illustrate this variety, various curves for solid fat content of partially hydrogen-
ated palm oil obtained using different hydrogenation conditions are given in
Figure 5. As shown in this Figure, the reaction time is a process parameter that
gives rise to different degrees of saturation as well as different TFA contents
and solid fat content profiles. The reaction time influences the final physical
properties, in this case the melting characteristics of the oils.
Referring back to the mechanism of hydrogenation, the availability of
hydrogen at the catalyst/reaction site is a crucial parameter in the cis/trans
isomerization process. Like nearly all chemical reactions, hydrogenation is
temperature dependent: the reaction rate constants increase with the tempera-
ture. Indeed, an increase in temperature directly induces higher solubility of
hydrogen and a decrease in viscosity which improves mass transfer during the
reaction. While increased temperature increases the hydrogenation reaction
rate it also tends to decrease the concentration of hydrogen at the surface of the
catalyst (Ariaansz, 2006). Under these hydrogen-starved conditions the level
of TFA will increase (Larsson, 1983). As depicted in Figure 6, during partial
hydrogenation of soybean oil with the same catalyst concentration and under
the same low pressure (3 bars), higher temperatures induce higher TFA
contents.
Hardening is usually carried out at pressures between 2 and 5 bars, but the
rate and selectivity of the reaction can be affected by modifying hydrogen
120
Hydrogenation time
100
80
Solid (%)
60
40
20
0
10 20 30 40 50 60
Temperature (°C)
Figure 5. Solids content of hardened palm oil dependent on reaction time. õ palm oil, IV 55, 0% trans;
palm oil, IV 48, 35% trans; ¸ palm oil, IV 42, 30% trans; ü palm oil, IV 1.5, 1.2% trans.
Selective conditions
40
180°C, 3 bar 35
150°C, 3 bars
30
120°C, 20 bars
120°C , 3 bars 25
% trans
20
15
Non-selective conditions
10
5
0
140 120 100 80 60
Iodine value
Figure 6. Trans fatty acid level in hardened soybean oil as a function of reaction temperature and
hydrogen pressure (nickel catalyst, 0.01%).
pressure. Indeed, a pressure increase leads to higher solubility of hydrogen

which is conducive to a better hydrogen supply at the surface of the catalyst. At
low pressure for example, hydrogen-starved conditions exist and the half-
hydrogenated intermediate has to wait longer to receive the second hydrogen
atom giving more probability of formation of trans isomers (Ariaansz, 2006).
The influence of pressure on the final TFA level is also illustrated in Figure 6.
Saturation at the same iodine value with nickel catalyst under high pressure and
low temperature is the best combination to limit TFA formation during
hydrogenation of soybean oil.
Unfortunately, the final TFA level remains quite important even under these
non-selective conditions.
Nickel catalysts are still the most commonly used catalyst for vegetable oil
hydrogenation (Farr, 2005). The amount of nickel used in conventional indus-
trial processes is between 0.005 and 0.01% (w/w % of oil) (Ackman & Mag,
1998). Even if these catalysts present several advantages including low cost,
high activity, potential tailored linolenic and linoleic selectivities, they also
promote the geometrical and positional isomerization of cis-ethylenic double
bonds during the hardening process. Increasing the catalyst concentration will
increase the hydrogen consumption that reinforces hydrogen-starved condi-
tions at the catalyst site leading to increased formation of TFA isomers. The
selectivity of the nickel catalysts is influenced by the particle size and porosity
(Beckmann, 1983). As previously mentioned, wide pore and small particle size
tend to favour higher selectivity (high linoleic and linolenic selectivities, SI,
SII respectively) because this allows shorter residence time at the catalyst site
(Beckmann, 1983). Under conventional process parameters (low pressure
3 bars, high temperature 180ºC) differences between catalyst structures have
only a small influence on TFA formation (Ariaansz, 2006, personal communi-
cation). However, hydrogenating soybean oil with the same iodine value using
a selective catalyst (wide pore) at low temperature and high pressure, results in
a significant reduction of 55% of TFA content compared to using non-selective
catalysts which leads to only 37% reduction (Ariaansz, 2006, personal commu-
nication).
Impurities in the feedstock can also have an influence on TFA formation, by
altering the activity of the catalyst. Sulphur compounds present in rapeseed oil
for instance tend to decrease the active sites in the catalyst surface thereby
increasing the likelihood of isomerization of cis-ethylenic double bonds
(Drozdowski & Szukalska, 2000). On the other hand, the presence of
phosphatides in the starting material tends to reduce the formation of TFA
(Beckmann, 1983). However, it appears that despite a decrease in the level of
TFA by optimizing temperature, pressure, and nickel catalyst structure, the
existing process does not allow operation under optimal conditions. For
example, to achieve a final TFA level below 10%, high hydrogen pressures
(above 50–60 bars) are required but this pressure cannot be achieved using the
existing hydrogenation plants. The overall influence of processing conditions
on hydrogenation characteristics is summarized in Table 3.
3. Reduction of formation of trans fatty acids (TFA) during partial

hydrogenation
In accordance with the hydrogenation mechanism described above, reducing
the TFA content of partially hydrogenated vegetable oils implies suppressing
the dissociation of half-hydrogenated intermediates by, for instance, promoting
Table 3. Effect of processing conditions on hydrogenation.
Increased Variables Trans fatty acids Selectivity Reaction rate
Temperature increasing increasing increasing

Pressure decreasing decreasing increasing
Agitation decreasing decreasing increasing
Catalyst concentration increasing increasing
their saturation. As previously mentioned, this can be done by increasing the

hydrogen concentration by operating the process at increased pressure and/or
reduced temperature. Reducing the catalyst concentration will also cause the
hydrogen concentration to increase but will also reduce the rate of reaction,
which may have already been reduced by the lowering of the temperature.
Accordingly, Hasman (1995) observes a reduction of the TFA content by 50%
and 70% for soybean and rapeseed oils, respectively, when decreasing the
temperature from 204ºC to 77ºC and increasing the pressure from 1 to 17 bars.
To prevent the reaction from slowing down too much, the catalyst concentra-
tion was increased from 0.02% to 0.5% (Hasman, 1995). Similarly, Van Toor
et al. (2005) proposed a hydrogenation temperature below 70ºC; in their study,
they use an even lower temperature of 40ºC to achieve a reduction of TFA
content. These conditions are more suitable for oils with a high oleic acid
content than for vegetable oils with high levels of linoleic acid (Dijkstra, 2006).
In this specific case, the level of stearic acid formation is increased but the
concentration of the catalyst should be increased to about 4%. In the corre-
sponding patent application an appropriate solution is provided to produce
suitable hardstock material (Dijkstra, 2006).
Another way of increasing the hydrogen concentration is to use a solvent
such as supercritical or near critical propane in which hydrogen is more soluble
than in oil. This type of solvent is compatible with both nickel and noble metal
catalysts (Härröd & Møller, 1996). Subsequently, it was proposed by Härröd
and co-workers, that under these operating conditions all the polyunsaturated
fatty acids are hydrogenated to cis-monounsaturated fatty acids; no TFA and no
saturated fatty acids are formed. The palladium catalyst used by Härröd and
co-workers had a low surface affinity for monoenoic acids. However, estab-
lishing adsorption equilibria that favour the adsorption of polyunsaturated fatty
acids over monounsaturated ones takes time. Accordingly, the productivity of
the process is lower than that of current nickel catalysed hydrogenations. High-
pressure equipment, the palladium catalyst and solvent recovery also add to the
cost and make the process expensive.
To reduce formation of TFA isomers during hardening, several studies on
hydrogenation with catalysts other than nickel have been carried out. They
were focused mainly on the use of heterogeneous catalysts made with palla-
dium or platinum (Beers, 2007; Savchenko & Makaryan, 1999) coupled with
solid supports (such as alumina and silica) and on mixtures of homogeneous

and heterogeneous catalysts (Dijkstra, 2006; Ackman & Mag, 1998). Attempts
have also been made to replace nickel catalyst with copper based catalysts
(Beers & Mangnus, 2004; Beers et al., 2008). With palladium catalysts, TFA
formation is reduced by 50% with good hydrogenation rates (Hsu et al., 1988;
Rylander, 1970), but industrial applications are limited by the operating
conditions, high hydrogen pressure, and associated cost of such a catalyst.
Therefore various patents suggest that nickel catalyst be conditioned with
agents such as ammonia, urea or aliphatic amines (Cahen, 1979; 1980), or
organic acid phosphates (Higgins, 2004). Platinum catalysts conditioned with
compounds such as tetraethyl ammonium hydroxide (Kuiper, 1980), gaseous
or liquid ammonia (Kuiper, 1981b) or amines (Kuiper, 1981a) can be made
more selective than other conditioned types of catalyst. Noble metal catalysts
such as platinum can also be modified to suppress the formation of both TFA
isomers and saturated fatty acids (Beers & Mangnus, 2004). They can be used
under supercritical conditions leading to a fairly low isomerization index
ranging from 0.2 to 0.4. These conditions are not used industrially.
The use of alternative homogeneous catalysts, for example ruthenium com-
plex, was investigated by Wright and co-workers (2003), who showed that
good hydrogenation selectivity could be achieved with low isomerization
index under high pressure (50 bars). However, industrial applications are
limited due to difficulties met during the removal of such catalyst and the high
hydrogen pressure required. The use of mixed catalyst systems, including
ruthenium complex paired with nickel, were studied by Wright et al. (2003a)
and their results showed that these systems could be suitable to selectively
hydrogenate cis-ethylenic double bonds in rapeseed oil. However, the benefit
of using such mixtures is mitigated again by the fact that ruthenium complexes
are oil-soluble and therefore difficult to remove at the end of the reaction.
Immobilization of the ruthenium onto an inert support should help to overcome
this limitation (Wright et al., 2003b).
Another way of changing the relative affinities of the unsaturated fatty acids
for the catalyst surface is by alloying nickel with other metals such as rare
earths. Schöön (1995) proposed that zeolite-based catalysts may be one possi-
ble solution for a zero TFA hydrogenation process. Indeed, zeolite that contains
platinum inside the pores can be used to reduce the fairly straight elaidic acid
chain that can enter, whereas the bent oleic acid chain cannot (Jacobs et al.,
2001). Schöön rated this type of catalyst higher than homogeneous catalysts
which are often poisonous, unstable and difficult to remove completely from
the hydrogenated oil (Schöön, 1995). For the time being the use on an industrial
scale of other metal catalysts or homogeneous catalytic complexes to replace
conventional nickel-based catalysts is technically and economically infeasible.
Another emerging technique suitable for limiting the production of TFA
during hydrogenation is the use of electrochemical hydrogenation (An et al.,
1998). Pintauro and co-workers described a low–temperature electro-catalytic

hydrogenation process for producing soybean oils with iodine value ranging
from 100 to 60. The partially hydrogenated vegetable oils thus obtained have
TFA levels of about 10 and 2% using a palladium catalyst cathode and a
platinum catalyst cathode, respectively (An et al., 1998; An et al., 1999;
Warner et al., 2000). This promising new approach has so far not been
implemented industrially.
Modification of nickel catalyst structural characteristics and process condi-
tions, as previously described, remains the easiest way to improve hydrogenation
selectivity for cis-ethylenic double bonds (Mangnus, 2004). Novel modified
nickel catalysts on an inert support (alumina or silica) and more resistance to
poisoning can be used to assure a good selectivity without excessive geometri-
cal isomerization. By combining this flexible catalyst with non-selective
process parameters (104ºC, 620 kPa), hydrogenation of rapeseed oil at iodine
value 60 can be performed with 50% lower TFA content.
Finally, full hydrogenation is another way to reduce TFA levels in edible fats
but it is not achievable in practice on an industrial scale. In fact, the residual
iodine value of a ‘fully’ hydrogenated oil is around 1–2 which corresponds to
residual trans isomers (Van Duijn, 2000). In practice, for almost fully hydro-
genated fat the thermodynamic equilibrium ratio between cis and trans is
reached meaning that a residual iodine value of 1 corresponds to about 0.85%
trans. To obtain a product with a TFA level below 1.25% (requirement for an
80% fat product having less than 1% TFA) the hydrogenation should be
conducted until the iodine value is below 1.5 (Van Duijn, 2005). These fully
hardened fats are characterized by high melting points, which can be detrimen-
tal for their direct use in food formulations. Generally, these oils are
interesterified with lauric oils or fully hydrogenated lauric oils and then the
interesterification products are blended with refined liquid vegetable oils to
achieve the required physical properties.
D. Conclusion
An understanding of the mechanism and key parameters involved in the
formation of TFA during catalytic hydrogenation of vegetable oils has been
acquired during recent decades. The performance of catalysts and plant designs
have been investigated in order to identify the necessary operating conditions
to limit the formation of TFA during partial hydrogenation of edible oils.
However, for the time being, the production of ‘zero trans’ hydrogenated fats
is elusive due, essentially, to cost limitations. Alternatives to partial hydro-
genation such as interesterification and fractionation become increasingly
popular. These processes allow the effective formulation of fats and oils having
physical properties that make them suitable to replace hydrogenated fats in
food products. The food industry has already started to reformulate food
products using these trans fat alternatives and recent dietary surveys indicate
that the TFA intake has decreased in many European Union countries (Morin,
2005). A new challenge for food manufacturers nowadays is to overcome the
increased content in food products of pro-atherogenic saturated fatty acids
introduced by these new fractionated or interesterified fats.
References
Ackman, RG and Mag, TK (1998) Trans fatty acid and the potential for less in technical
products. In: Trans Fatty Acids in Human Nutrition (J-L Sébédio and WW Christie, eds),
Oily Press, Bridgwater, UK, pp.35–36.
Alouche, A, Lambert, DC, Hubaut, R, Davies, P and Hertoghe, P (1994) Procédé
d’hydrogénation sélective d’une huile végétale avec rétention des doubles liaisons de
configuration cis et utilisation des huiles obtenues par ce procédé. FR Patent 2 694 015,
assigned to Société de la Raffinerie BP et ELF de Dunkerque.
An, W, Hong, JK, Pintauro, PN, Warner, K and Neff, W (1998) The electrochemical
hydrogenation of edible oils in a solid polymer electrolyte reactor. I. Reactor design and
operations. J. Am. Oil Chem. Soc., 75, 917–924.
An, W, Hong, JK, Pintauro, PN, Warner, K and Neff, W (1999) The electrochemical
hydrogenation of edible oils in a solid polymer electrolyte reactor. II. Hydrogenation
selectivity studies. J. Am. Oil Chem. Soc., 76, 215–222.
Ariaansz, RF (1996) Hydrogenation with minimum trans acids. In: Lipids & Nutrition:
Current Hot Topics (KG Berger, ed) PJ Barnes & Associates, Bridgwater, UK, pp.81–
103; proceedings of SCI conference held on 30 April 1996, London, UK.
Bailey, AE (1949) Theory and mechanics of the hydrogenation of edible fats. J. Am. Oil
Chem. Soc., 26, 596–601.
Bailey, AE and Fisher, GS (1946a) Modifications of vegetable oils. V. Relative reactivities
toward hydrogenation of the mono- di- and triethenoid acids in certain oils. Oil & Soap,
23, 14–18.
Baltes, J (1970) Selektive Härtung von Fetten und Fettsäuren mit Sulfiden der Übergangsmetalle
als Katalysatoren. Fette Seifen Anstrichm., 72, 425–432.
Baltes, J (1972) Process for the selective hydrogenation of fats and fatty acids. US Patent
3,687,989.
Beckmann, HJ (1983) Hydrogenation practice. J. Am. Oil Chem. Soc., 60, 282–290.
Beers, AEW, Berben, PH, Groen, C, Jagta, R, Lazar, G and Mangnus, G (2004) Lowering the
trans content in edible oils. Paper presented at the 3rd EuroFedLipid Congress,
Edinburgh, UK, page 62 of the abstracts.
Beers, AEW, Berben, PH and Okonek, DV (2005) Low trans hydrogenation of edible oils —
new developments by Engelhard. Paper presented at the 26th ISF World Congress and
Exhibition, Prague, page 78 in the abstracts.
Beers, AEW and Mangnus, G (2004) Hydrogenation of edible oils for reduced trans fatty acid
content. Inform, 15, 404–405.
Beers, AEW (2007) Low trans hydrogenation of edible oils. Lipid Technology, 19, 56–58.
Beers, AEW, Ariaansz, RF, Okonek, D (2008) Trans isomer control in hydrogenation of
edible oils. In: Trans Fatty Acids (AJ, Dijkstra, RJ Hamilton, W, Hamm, eds) Blackwell
Publishing Ltd, Oxford, UK, pp.147–180.
Beyens, Y and Dijkstra, AJ (1983) Positional and triglyceride selectivity of hydrogenation
of triglyceride oils. In: Fat Science 1983, Proceedings of 16th ISF congress, Budapest (J
Holló, ed), Akadémiai Kiadó, Budapest, Hungary, pp.425–432.
Bockisch, M (1998) Fats and Oils Handbook, AOCS Press, Champaign, Illinois, USA.
Boelhouwer, C, Snelderwaard, J, and Waterman, HI (1956) Problems of selectivity in the

hydrogenation of linoleic acid esters. J. Am. Oil Chem. Soc., 33, 143–146.
Bushell, WJ and Hilditch, TP (1937) The course of hydrogenation in mixtures of mixed
glycerides. J. Chem. Soc., 1767–1774.
Cahen, RM (1979) Hydrogenation process. US Patent 4,161,483, assigned to Labofina SA.
Cahen, RM (1980) Hydrogenation catalyst and hydrogenation process. US Patent 4,229,361,
assigned to Labofina SA.
Coenen, JWE (1978) The rate of change in the perspective of time. Chem. Ind., 709–722.
Coenen, JWE and Boerma, H (1968) Absorption der Reaktionspartner am Katalysator bei der
Fetthydrierung. Fette Seifen Anstrichm., 70, 8–14.
Dijkstra, AJ (1990) On the mechanism of the hydrogenation of edible oils. In: Beauty is our
Business. A Birthday Salute to Edsger W. Dijkstra (WHJ Feijen, AJM van Gasteren, D
Gries, and J Misra, eds), Springer Verlag, New York, USA, pp.102–111.
Dijkstra, AJ (1997) Hydrogenation revisited. Inform, 8, 1150–1158.
Dijkstra, AJ (2002a) Hydrogenation and fractionation. In: Fats in Food Technology (KK
Rajah, ed), Sheffield Academic Press, Sheffield, UK, pp.123–158.
Dijkstra, AJ (2002b) On the mechanism of the copper-catalysed hydrogenation; a reinterpre-
tation of published data. Eur. J. Lipid Sci. Technol., 104, 29–35.
Dijkstra, AJ (2006) Revisiting the formation of trans isomers during the partial hydrogena-
tion of triglyceride oils. Eur. J. Lipid Sci. Technol., 108, 249–264.
Drozdowski, B and Szukalska, E (2000) Effect of rapeseed oil hydrogenation conditions on
trans isomer formation. Eur. J. Lipid Sci. Technol., 102, 642–645.
Duijn, G van (2000) Technical aspects of trans reduction in margarines. Oléagineux Corps
Gras Lipides/OCL, 7, 95–98.
Duijn, G van. (2005), Technical aspects of trans reduction in modified fats, Oléagineux Corps
Gras Lipides/OCL, 12, 422–426.
Duveen, RF and Leuteritz, G (1982) Der BUSS-Schleifenreaktor in der Öl- und
Fetthärtungsindustrie. Fette Seifen Anstrichm., 84, 511–515.
Eller, FJ, List, GR, Teel, JA, Steidley, KR, and Adlof, RO (2005) Preparation of spread oils
meeting US Food and Drug Administration labeling requirements for trans fatty acids via
pressure controlled hydrogenation. J. of Agricultural and Food Chemistry, 53, 5982–
5984.
Farr, WE (2005) Hydrogenation: processing technologies. In: Bailey’s Industrial Oil & Fat
Products, sixth edition (F. Shahidi, ed), John Wiley Interscience, Hoboken, New Jersey,
USA, pp.385–396.
Faur, L (1996) Transformation of fat for use in food products: hydrogenation technology. In:
Oils and Fats Manual (A. Karleskind, ed), Lavoisier Tec et Doc, Paris, France, pp.897–
922.
Härröd, M, Hark, S van der, Holmqvist, A, and Møller, P (2004) Selective hydrogenation of
triglycerides at supercritical single-phase conditions. Paper presented at the 3rd
EuroFedLipid Congress, Edinburgh, page 85 in the abstracts.
Härröd, M, Holmqvist, A, and Hark, S van der (2005) Selective hydrogenation of functional
groups in substrates and partially hydrogenated fatty acids and fatty acid derivatives. PCT
Patent Application WO 2005/095306, assigned to Härröd Research AB.
Härröd, M and Møller, P (1996) Hydrogenation of substrate and products manufactured
according to the process. WO Patent Application 96/01304, assigned to Poul Møller
Ledelses- og Ingeniörrådgivning APS.
Hasman, JM (1995) Trans suppression in hydrogenated oils. Inform, 6, 1206–1210.
Hastert, RC (1981) Practical aspects of hydrogenation and soybean salad oil manufacture.
J. Am. Oil Chem. Soc., 58, 169–174.
Hastert, RC (1998) Past, present and future for the hydrogenation process. Lipid Technology,
10, 101–105.
Higgins, NW (2004) Low trans stereoisomer shortening systems. US Patent Application

Publication 2004/0146626 A1.
Horiuti, I and Polanyi, M (1934) Exchange reactions of hydrogen on metallic surfaces. Trans.
Faraday Soc., 30, 1164.
Hsu, N, Diosady, LL, and Rubin, LJ (1988) Catalytic behavior of palladium in the
hydrogenation of edible oils. J. Am. Oil Chem. Soc., 65, 349–356.
Jacobs, PA, Maes, PJ, Paulussen, SJ, Tielen, M, Van Steenkiste, DFE, and Van Looveren, LK
(2001) Elimination of trans unsaturated fatty acid compounds by selective adsorption
with zeolites. US Patent 6,229,032 B1, assigned to K.U.Leuven Research & Develop-
ment.
Jonker, GH (1999) Hydrogenation of edible oils and fats. PhD thesis, Rijksuniversiteit
Groningen, The Netherlands.
Koritala, S (1970) Selective hydrogenation of soybean oil. V. A novel copper catalyst with
excellent re-use properties. J. Am. Oil Chem. Soc., 47, 106–107.
Kuiper, J (1980) Selective hydrogenation. US Patent 4,228,088, assigned to .
Kuiper, J (1981a) Process for the selective hydrogenation of triglyceride oils with a metallic
catalyst in the presence of a diamine. US Patent 4,307,026, assigned to .
Kuiper, J (1981b) Process for the selective hydrogenation of triglyceride oils with a metallic
catalyst in the presence of ammonia. US Patent 4,278,609, assigned to .
Larsson, R (1983) Hydrogenation theory: some aspects. J. Am. Oil Chem. Soc., 60, 275–281.
Mangnus, G, (2004) Hydrogenation of oils at reduced TFA levels. Oils & Fats International,
7, 33–35
Morin, O (2005) Acides Gras Trans: récents développements. OCL, 12, 414–421.
˜Mounts, TL and Dutton, HJ (1967) Micro vapor-phase hydrogenation monitored with
tandem chromatography-radioactivity. II. Evaluation of catalyst selectivity for linolenate.
J. Am. Oil Chem. Soc., 44, 67–70.
Mozaffarian, D, Katan, MB, Ascherio, A, Stampfer, MJ, and Willett, WC (2006) Trans fatty
acids and cardiovascular diseases. New England J. Med., 354 (15), 1601–1613.
Normann, W (1903) Process for converting unsaturated fatty acids or their glycerides into
saturated compounds. British Patent 1 515, assigned to Herforder Maschinenfett-und
Ölfabrik Leprince und Siveke.
Okkerse, C, Jonge, A. de, Coenen, JWE, and Rozendaal, A (1967) Selective hydrogenation
of soybean oil in the presence of copper catalysts. J. Am. Oil Chem. Soc., 44, 152–156.
Rijnten, HT and Eikema, ETJ (1975) Process for the preparation of partially sulfided metallic
supported catalysts. US Patent 3,872,028, assigned to .
Rylander, PN (1970) Hydrogenation of natural oils with platinum metal group catalysts.
J. Am. Oil Chem. Soc., 47, 482–486.
Sanders, JH (1950) Partial hydrogenation of unsaturated glyceride oils in solvents. US Patent
2,520,440, assigned to The Procter & Gamble Company.
Savchenko, VI and Makaryan, IA (1999) Palladium catalyst for the production of pure
margarine. Platinum Metals Rev., 43, 74–82.
Schilling, K (1977) The mechanism of triglyceride hydrogenation. Paper presented at the 9th
Scandinavian Symposium on Lipids, F-116–F-119.
Schilling, K (1978) Der Reaktionsverlauf bei der Hydrierung von Triglyceriden.
Simultanhydrierung von Tri- und Monolinolenin. Fette Seifen Anstrichm., 80, 312–314.
Schöön, N-H (1995) Is a low trans content attainable by conventional hydrogenation of
vegetable oils? In: Oils-Fats-Lipids 1995 (WAM Castenmiller, ed), P.J. Barnes &
Associates, Bridgwater, UK, pp.155–158.
Tacke, T, Wieland, S, Panster, P, Bankmann, M, Brand, R, and Mägerlein, H (1998)
Hardening of unsaturated fats, fatty acids or fatty acid esters. US Patent 5,734,070,
assigned to Degussa AG.
Toor, H van, Rossum, GJ van, and Kruidenberg, M (2005) Low trans fatty acid compositions;
low-temperature hydrogenation, e.g. of edible oils. US Patent Application Publication
2005/0027136 A1, assigned to Cargill Incorporated.
Veldsink, JW, Bouma, MJ, Schöön, N-H, and Beenackers, AACM (1997) Heterogeneous
hydrogenation of vegetable oils: a literature review. Catal. Rev. -Sci. Eng., 39, 253–318.
Warner, K, Neff, WE, List, GR, and Pintauro, PN (2000) Electrochemical hydrogenation of
edible oils in a solid polymer electrolyte reactor. Sensory and compositional character-
istics of low trans soybean oils. J. Am. Oil Chem. Soc., 77, 1113–1117.
Willett, WC and Ascherio, A (1994) Trans fatty acids: are the effects only marginal? Am. J.
Public Health, 85, 722–744.
Wright, AJ, Mihele, AL, and Diosady, LL (2003a) Ni-catalyst promotion of a cis selective
Pd catalyst for canola oil. Food Res. Int., 36, 797–804.
Wright, AJ, Wong, A, and Diosady, LL (2003b) Cis selectivity of mixed catalysts systems
in canola oil hydrogenation. Food Res. Int., 36, 1069–1072.
CHAPTER 3
Formation of trans fatty acids during
deodorization of edible oils
JEAN-BAPTISTE BEZELGUES1 AND FRÉDÉRIC DESTAILLATS2
1
2
Consumers are increasingly more demanding and do not tolerate off-flavours

and after-tastes in finished food products. In that respect, to be suitable for
industrial food applications, crude fats and oils such as palm, soybean, sun-
flower, rapeseed and fish oils are usually refined using chemical or physical
processes. The refining process aims at producing vegetable oils with a
pleasant light colour, and a neutral flavour and odour. Typically, in the
chemical process oils are degummed, alkali neutralized, bleached and deodor-
ized to remove undesirable compounds including phosphatides, free fatty
acids, pigments, volatiles imparting unpleasant flavours and some contami-
nants such as heavy metals, polycyclic aromatic hydrocarbons, polychlorinated
biphenyls and pesticides (see Figure 1). Physical refining starts with a
degumming stage followed by a bleaching step and a deodorization conducted
at higher temperature than the chemical process.
A. Deodorization: a critical operation

Unrefined (crude) vegetable oils contain only a negligible level of TFA (< 0.3%
of total fatty acids); the most important process parameters to control to avoid
excessive formation of trans isomers are deodorization temperature and dura-
tion (Schwarz, 2000). Natural contaminants, such as phosphatides, heavy
metals or technological contaminants, like residual bleaching earth not com-
pletely separated during the previous refining steps, do not catalyse the
geometrical isomerization of polyunsaturated fatty acids (Pudel & Denecke,
1997). Formation of geometrical isomers from polyunsaturated fatty acids is
limited during the traditional alkali refining process because the deodorization
step is generally conducted below 240°C with short residence time (20–
60 min) (Greyt & Kellens, 2005). In deodorization columns used for physical
refining operating at higher temperature (230–260°C) and longer time, sub-
stantial formation of geometrical isomers of polyunsaturated fatty acids has
been reported (Greyt & Kellens, 2005). In the case of canola (rapeseed) oil
65
CRUDE OIL
C P
H H
E Y
M Degumming Degumming S
I I
C C
A Neutralisation FFA A
L Bleaching L
R Bleaching R
E FFA E
F Deodorisation Deodorisation F
I I
N volatiles contaminants N
I I
N N
G REFINED OIL G
Figure 1. Representation of chemical and physical oil refining processes.
under normal operating conditions at 250°C, the TFA content can reach 3%
(Pudel & Denecke, 1997; Wolff,1992).
Deodorization consists of injecting live steam into oil maintained under
vacuum (2 to 4 mbar) and at high temperature (220°C–260°C). It has been
shown by Ackman et al. (1974) in the early 1970s that small amounts of
geometrical (trans) isomers of the essential fatty acids can be formed. Geo-
metrical isomerization, also called stereomutation, refers to modification of the
configuration of the ethylenic double bond. Almost all the vegetable oils
contain linoleic (cis-9,cis-12 18:2) acid, and some of them such as soybean or
rapeseed oils also contain significant amount of α-linolenic (cis-9,cis-12,cis-
15 18:3) acid. The nutritional value of these vegetable oils is very important
since both linoleic and α-linolenic acids are essential fatty acids, required for
the endogenous synthesis of long-chain polyunsaturated fatty acids. Activation
energies for cis to trans geometrical isomerization of ethylenic double bonds of
linoleic and α-linolenic acids are 178 kJ/mole and 144–148 kJ/mole respec-
tively (Greyt and Kellens, 2005). Exposure to elevated temperatures during the
deodorization triggers the formation of trans isomers of polyunsaturated
linoleic and α-linolenic acids. It has been shown that the consumption of trans
isomers of α-linolenic acid increases the ratios of plasma LDL-cholesterol to
HDL-cholesterol and total cholesterol to HDL-cholesterol in healthy men
(Vermunt et al., 2001).
B. Geometrical isomerization of linoleic acid and α -linolenic acid

The geometrical isomerization of oleic acid is negligible during deodorization
(Schwartz, 2000) and trienoic acid such as α-linolenic acid isomerizes signifi-
cantly faster than dienoic fatty acids such as linoleic acid (see Figure 2). It has
been shown that geometrical isomerization of both linoleic and α-linolenic
FORMATION OF TFA DURING DEODORIZATION 67
5 Total TFA
C18:3 trans
TFA content (%)
4 C18:2 trans
C18:1 trans
3
0
200 220 240 260
Deodorisation temperature (°C)
Figure 2. Formation of trans isomers of oleic, linoleic and α-linolenic acids during low-erucic
rapeseed oil deodorization (adapted from Bertoli and co-workers, 1998)
HO
O O
HO HO

HO
Figure 3. Representation of the geometrical isomers formed from linoleic acid during deodorization.
The star symbol and plain arrows indicate the main isomers formed under conventional deodorization
conditions.
acids follow first order kinetics (Wolff, 1992a). Geometrical isomerization of

linoleic acid leads mainly to the formation of an equimolar amount of the
mono-trans isomers cis-9,trans-12 acid and trans-9,cis-12 18:2 acid as shown
in Figure 3 (Wolff, 1992a; Pudel and Denecke, 1997). Under drastic
deodorization conditions (> 260°C, not used at plant scale), traces amounts of
all-trans linoleic (trans-9,trans-12 18:2) acid are formed (Pudel and Denecke,
1997).
The isomerization pathway of α-linolenic acid, present in rapeseed and
soybean oils at 6–11%, is more complex and have been described by Wolff
O
HO
68
O O O
HO HO HO

O O
HO HO
HO

HO
Figure 4. Representation of the geometrical isomers formed from α-linolenic acid during deodorization. The star symbol and plain arrows indicate the main
isomers formed under conventional deodorization conditions.
(1992a, see Figure 4). Out of the 8 geometrical isomers that can be produced by
geometrical isomerization of α-linolenic acid, only 4 isomers are found in
conventionally refined vegetable oils (Wolff, 1992). The kinetics of isomeriza-
tion of α-linolenic acid during deodorization are shown in Figure 5. The main
isomers formed are the three mono-trans (trans-9,cis-12,cis-15; cis-9,cis-
12,trans-15; and cis-9,trans-12,cis-15) and the di-trans isomer trans-9,cis-
12,trans-15 (Figures 4 and 5). Less cis-9,trans-12,cis-15 18:3 acid isomer is
produced (about 6–7 times less, Wolff 1992a) compared to the other mono-
trans isomers having a trans double bond (Δ 9 or Δ15) at the extremity of the
methylene interrupted system (Wolff 1992a). This fact clearly shows that some
conformations are thermodynamically preferred. The level of trans-9,trans-
12,cis-15, cis-9,trans-9,trans-15 and trans-9,trans-12,trans-15 18:3 acid
isomers is usually very low or not detectable (Wolff 1992a; Wolff 1993).
It has been observed that the amount of geometrical isomers derived from
linoleic acid increases linearly with deodorization time (Pudel and Denecke,
1997) but below 220°C very small amounts of trans linoleic are formed. It has
been observed at high temperatures, for example 275°C (conditions that are
never applied in industrial refining), that about 10% of (all-cis) linoleic acid is
isomerized into trans isomers (Pudel and Denecke, 1997). Also, below 220°C,
the isomerization rate of α-linolenic acid is low and the level of geometrical
isomers formed increases linearly with time (Pudel and Denecke, 1997).
Bertoli and co-workers (1998) showed that the final TFA content in rapeseed
oil stripped below 220°C for 6 hours did not exceed 1% (see Figure 5). At
higher temperature, it appears that geometrical isomerization of polyunsatu-
rated fatty acids becomes exponential; at 255°C, about half of the initial
quantity of the α-linolenic acid is isomerized (see Figure 5).
10 2
c,c,c linolenic acid (%)
8 1.6
t, linolenic acid (%)
6 1.2
4 0.8
2 0.4
0 0
180 200 220 240 260 280
Temperature (°C)
Figure 5. Influence of deodorization temperature on the formation of α-linolenic trans isomers

(deodorization time 180 min) õ c,c,c linolenic; t,c,c linolenic; K c,t,c
linolenic; c,c,t linolenic; ¸ t,c,t linolenic (adapted from Pudel & Denecke, 1997).
C. Geometrical isomerization of long-chain polyunsaturated fatty

acids during marine oil deodorization
Fish oil is widely used as a dietary source of long-chain polyunsaturated fatty

acids in food products of dietary supplements. Raw materials used to prepare
refined fish oils are often of poor quality and adequate refining is needed to
obtain high quality fish oils. The two main long-chain polyunsaturated fatty
acids found in fish oils are eicosapentaenoic (cis-5,cis-8,cis-11,cis-14,cis-17
20:5, EPA) and docosahexadecenoic (cis-4,cis-7,cis-10,cis-13,cis-16,cis-19
22:6, DHA) acids. These two n–3 long-chain polyunsaturated fatty acids are
found at different levels and proportions depending on origin of the fish.
The deodorization step is critical to remove undesired volatile compounds in
order to obtain high quality oils. Few research groups had studied the formation
of geometrical isomers from EPA or DHA in heat-treated fish oil or pure long-
chain polyunsaturated fatty acids (Sébédio 1989; Sébédio et al., 1993; Sébédio
and De Rasilly 1993 and Wijesundera et al., 1989). In these studies, it was
clearly found that geometrical isomers of EPA and DHA are readily formed at
high temperature. However, the first study showing the profile of degradation
products formed during fish oil deodorization was published by Fournier and
co-workers (2006a). The loss of long-chain polyunsaturated fatty acids during
fish oil deodorization is extremely important depending on the temperature
used (Figure 6). The number of ethylenic double bonds influences the sensitiv-
ity of long-chain polyunsaturated fatty acids toward thermal degradation
(Figure 6). It was found that three different types of degradation products are
formed during the deodorization of fish oil: polar compounds (a generic
category comprising triacylglycerol polymers, oxidized glycerides,
diacylglycerols, monoacylglycerols and free fatty acids); cyclic fatty acid
monomers (CFAM) and geometrical isomers. The relative distribution of these
different degradation products depends on the temperature used during the
deodorization since the activation energy for inter (formation of polymers) or
intra (formation of CFAM) molecular oligomerization differs from the energy
required for geometrical isomerization. Distribution of polar compounds,
CFAM and geometrical isomers formed during deodorization of tuna oil
performed at 180°C, 200 and 220°C for 2 h is shown in Figure 7. The main
degradation products found are polar compounds which mainly consist in
triacylglycerol polymers (Fournier and co-workers, 2006a). It is important to
note that the nutritional effects of fish oil triacylglycerol polymers have never
been studied. The geometrical isomers of EPA and DHA are produced in
smaller quantities than polar compounds. CFAM are formed significantly at
temperatures above 220°C (Figure 7).
The nature of the geometrical isomers formed from EPA and DHA has been
studied in detail recently by two research groups (Fournier and co-workers,
2006a,b; Fournier and co-workers, 2007; Mjøs and Solvang, 2006; Mjøs, 2005;
100
ARA
90
EPA
w3 DPA
80
w6 DPA
70 DHA
60
Relative %
50
40
30
20
10
0
Control 180 220 250
Deodorization Temperature (°C)
Figure 6. Relative decrease in concentration of the main long-chain polyunsaturated fatty acids found
in tuna oil submitted to deodorization at various temperatures for 2 hours (adapted from Fournier et al.,
2006a).
Mjøs, 2008). The theoretical number of geometrical isomers formed from EPA
(25 = 32) or DHA (26 = 64) are important but not all isomers are formed during
heat treatment of these long-chain polyunsaturated fatty acids. The chromato-
graphic separation of EPA and DHA geometrical isomers can be achieved
using the high-polarity open-tubular capillary columns conventionally used for
separation of fatty acid methyl ester derivatives (Fournier and co-workers,
2006a,b; Fournier and co-workers, 2007; Mjøs and Solvang, 2006; Mjøs 2005)
Interesting separations have also been obtained on long columns with
polyethylene glycol stationary phase (Mjøs 2008).
Identification of geometrical isomers formed during deodorization has been
achieved using elaidinized pure methyl EPA and DHA as previously done by
Wolff for the identification of geometrical isomers of α-linolenic acid (1992b).
It has been shown that mono- and di-trans isomers are mainly formed during
fish oil deodorization (Fournier and co-workers, 2006a,b; Mjøs and Solvang,
2006). Further isomerization can occur when EPA or DHA are exposed to
temperatures above 200°C as reported by Fournier and co-workers (2006a,b)
and illustrate in Figure 8. The main results obtained from these different studies
is that deodorization temperatures above 200°C lead to an important (> 10%)
400
Cyclic fatty acid monomers (CFAM)
350 Geometrical isomers

LC-PUFA degradation products (mg/g of oil)
Polar compounds
300
250
200
150
100
50
0
Control 180 220 250
Deodorization Temperature (°C)
Figure 7. Distribution of degradation products formed from long-chain polyunsaturated fatty acids
(LC-PUFA) during deodorization of fish (tuna) oil at various temperatures for 2 hours. Results are
expressed as mg per g of oil (adapted from Fournier et al., 2006a).
loss of all cis EPA and DHA and the formation of geometrical isomers (mainly
mono-trans) (Fournier and co-workers, 2006a; Mjøs and Solvang, 2006).
Analysis of commercially available refined marine oil samples revealed that
levels of geometrical isomers rarely exceed 1% (Fournier et al., 2007). The
effects of these trans fatty acids on health are not known. However, the
consumer exposure levels are low and marine oils are incorporated at low
levels in food formulations (Kolanowski and Laufenberg, 2006). The forma-
tion of polar compounds seems to be more important than that of geometrical
isomers in the case of fish oil, which differs from the case of vegetable oils
(Figure 7).
D. Technical solutions to prevent excessive isomerization of

polyunsaturated fatty acids during deodorization
It is technically feasible to prevent the loss of polyunsaturated fatty acids by
controlling the deodorization conditions. The current trend in refining is to
deodorize at relatively low temperatures (200–240°C) under low vacuum
conditions (< 3 mbar) and shorter residence time in order to limit both the
geometrical isomerization of polyunsaturated fatty acids and the loss of valuable
Di-trans EPA
Tri-trans DHA
Ag-TLC Fraction 4
Mono-trans EPA trans-11 + trans-14
trans-17 Di-trans DHA

Mono-trans DPA
Ag-TLC Fraction 3
Mono-trans DHA
All-cis EPA trans-10
trans-13 + trans-16
trans-19
All-cis DPA
Ag-TLC Fraction 2
All-cis DHA
Ag-TLC Fraction 1
40.0 51.0
Retention Time (min)
Figure 8. Partial gas chromatograms of Ag-TLC fractions of fatty acid methyl ester derivatives of
eicosapentaenoic (cis-5,cis-8,cis-11,cis-14,cis 17 20:5, EPA) and docosahexaenoic (cis-4,cis-7,cis-
10,cis-13,cis-16,cis 19 22:6, DHA) geometrical acid isomers obtained from fish oil deodorized at 220°C.
Fractions 1, 2, 3 and 4 correspond respectively to the 6 cis, 5 cis, 4 cis and 3 cis polyenoic fractions. DPA
stands for docosapentaenoic acid and, as an example, trans 19 indicates the cis-4,cis-7,cis-10,cis-13,cis
16,trans-19 22:6 acid isomer. Figure adapted from Fournier et al., 2006b.
minor components such as tocopherols (Evrard et al., 2007). It is worth noting

that the level of TFA in, for example, French commercial refined rapeseed and
soybean oils does not exceed 1% (Morin, 2005). Appropriate technology is
nowadays available to limit excessive time/temperature exposure and efficient
steam/oil contact. As an example, DeSmet Ballestra (Zaventem, Belgium)
developed the ‘Qualistock’ deodorizer that integrates in a compact system, a
new steam injection concept in which deaeration, heat exchanging, heating,
deodorizing/stripping, cooling and scrubbing all take place in one modular
single vessel. This system allows different options such as deodorization at two
different temperatures, and deep or shallow bed deodorization. Likewise, the
development of more efficient vacuum systems (dry condensing) allows the
reduction of the deodorization temperature without affecting the stripping
efficiency (Greyt & Kellens, 2005).
E. Conclusion
The formation of geometrical isomers from all-cis polyunsaturated fatty acids
happens when oils are deodorized at high temperature. The main consequence
of this phenomenon is the loss of nutritive value. The development of analytical
methods in the early 1990s provided a better understanding of the isomers
formed from linoleic and α-linolenic acids and measures were taken to better
control the deodorization operation at the plant. The formation of trans isomers
of EPA and DHA during the deodorization of fish oil was investigated recently.
Overall, it can be concluded that the residual content of geometrical isomers
formed from polyunsaturated fatty acids during deodorization is nowadays
very low and does not seem to represent a public health issue.
References
Ackman, RG, Hooper, SN and Hooper, DL (1974) Linolenic acid artifacts from the
deodorisation of oils. J. Am. Oil Chem. Soc., 51, 42–49.
Ackman, RG and Mag, TK (1998) Trans fatty acid and the potential for less in technical
products. In: Trans Fatty Acids in Human Nutrition (J-L Sébédio and WW Christie, eds),
Oily Press, Bridgwater, UK, pp.35–36.
Berdeaux, O, Fournier, V, Lambelet, P, Dionisi, F, Sébédio, J-L and Destaillats, F (2007).
Isolation and structural analysis of the cyclic fatty acid monomers formed from
eicosapentaenoic and docosahexaenoic acids during fish oil deodorisation. J. Chrom. A,
1138, 216–224.
Bertoli, C, Delvecchio, A, Durand, P, Gumy, D, Bellini, A and Stancanelli, M (1998)
Formation of trans fatty acids during deodorisation of low erucic acid rapeseed oil. In:
World Conference on Oilseed and Edible OIls Processing (SS Köseolu, KC Rhee and RF
Wilson, eds), AOCS Press, Champaign, Illinois, USA, pp.67–71.
Fournier, V, Destaillats, F, Juaneda, P, Dionisi, F, Lambelet, P, Sébédio, J-L and Berdeaux,
O (2006a) Thermal degradation of long-chain polyunsaturated fatty acids (LC-PUFAs)
during deodorisation of fish oil. Eur. J. Lipid Sci. Technol., 108, 33–42.
Fournier, V, Juanéda, P, Destaillats, F, Dionisi, F, Lambelet, P, Sébédio, J-L and Berdeaux,
O (2006b) Analysis of eicosapentaenoic and docosahexaenoic acid geometrical isomers
formed during fish oil deodorisation. J. Chrom. A, 1129, 21–28.
Fournier, V, Destaillats, F, Hug, B, Golay, P-A, Joffre, F, Juanéda, P, Sémon, E, Dionisi, F,
Lambelet, P, Sébédio, J-L and Berdeaux, O. (2007) quantification of eicosapentaenoic
(EPA) and docosahexaenoic (DHA) acid geometrical isomers formed during fish oil
deodorisation by gas-liquid chromatography, J. Chrom. A, 1154, 353–359.
Kolanowski, W and Laufenberg, G (2006) Enrichment of food products with polyunsaturated
fatty acids by fish oil addition. Eur. Food Res. Technol., 222, 472–477.
Mjøs, SA (2005) Properties of trans isomers of eicosapentaenoic acid and docosahexaenoic
acid methyl esters on cyanopropyl stationary phases. J Chrom A, 1100, 185–192.
Mjøs, SA, and Solvang, M (2006) Geometrical isomerisation of eicosapentaenoic and

docosahexaenoic acid at high temperatures. Eur. J. Lipid Sci. Technol., 108, 589–597.
Mjøs, SA (2008) Retention behavior of trans isomers of eicosapentaenoic and docosahexaenoic
acid methyl esters on a polyethylene glycol stationary phase. Eur. J. Lipid Sci. Technol.,
110, 547–553.
Morin, O (2005) Acides gras trans: récents développements. OCL, 12, 414–421.
Pudel, F and Denecke, P (1997) Influences on the formation of trans fatty acids during
deodorisation of rapeseed oil. Oléagineux Corps Gras Lipides/OCL, 4, 58–61.
Schwarz, W (2000) Formation of trans polyalkenoic fatty acids during vegetable oil refining.
Eur. J. Lipid Sci. Technol., 102, 648–649.
Sébédio J-L, (1989) Concentration d’acides gras polyinsaturés en oméga-3 à partir d’huiles
de poissons. Transformation de l’EPA (20:5n–3) et du DHA(22:6n–3) au cours des
traitements thermiques. Ichtyophysiologica acta, 12, 49–59.
Sébédio, J-L, Ratnayake WMN, Ackman RG and Prevost J (1993) Stability of polyunsatu-
rated omega-3 fatty acids during deep fat frying of Atlantic mackerel (Scomber scombrus
L.). Food Res. Int., 26, 163–172.
Sébédio, J-L and De Rasilly, A (1993) Analysis of cyclic fatty acids in fish oil concentrates.
17th Nordic Lipid Symposium, 212–216.
Vermunt SH, Beaufrère B, Riemersma RA, Sébédio JL, Chardigny JM and Mensink RP
(2001) Dietary trans alpha-linolenic acid from deodorised rapeseed oil and plasma lipids
and lipoproteins in healthy men: the TransLinE Study. Br. J. Nutr., 85, 387–392.
Wijesundera RC, Ratnayake WMN and Ackman RG (1989) Eicosapentaenoic acid geometri-
cal isomer artifacts in heated fish oil esters. J Am Oil Chem Soc., 66, 1822–1830.
Wolff, RL (1992a) Trans-polyunsaturated fatty acid in French edible rapeseed and soyabean
oils. J. Am. Oil Chem. Soc., 69, 106–110.
Wolff, RL (1992b) Resolution of linolenic acid geometrical isomers by gas-liquid chroma-
tography on a capillary column coated with a 100% cyanopropyl polysiloxane film
(CPTMSil 88). J. Chrom. Sci., 30, 17–21.
Wolff, RL (1993) Heat-induced geometrical isomerisation of α-linolenic acid: effect of
temperature and heating time on the appearance of individual isomers. J. Am. Oil Chem.
Soc., 70, 425–430.
CHAPTER 4
Chemical synthesis of monounsaturated trans
fatty acids
ZEPHIRIN MOULOUNGUI1,2 AND LAURE CANDY1,2
1
Université de Toulouse; INP; LCA (Laboratoire de Chimie AgroIndustrielle);
ENSIACET, Toulouse, France.
2
INRA; LCA; Toulouse, France.
The US Food and Drug Administration defines trans fatty acids (TFA) as fatty
acids containing one or more non-conjugated double bonds in trans configura-
tion. This definition excludes conjugated linoleic acids (CLA) that have at least
one trans ethylenic double bond.
Pure trans fatty acids are currently not available in large quantities and their
study is therefore limited. In order to improve our understanding of TFA, it is
necessary to achieve large-scale production of individual isomers in high
purity and, for each TFA, a different chemical operating process has to be
developed. All the processes are built according to an integrated approach,
each step providing a value-added molecule. The common process for TFA
synthesis can be broken down into three steps as follows.
• Double bond creation in specific geometrical configuration and position:
the Wittig olefination between a carbonyl compound and a phosphonium
ylide is the preferred method that leads to cis-trans fatty acid mixtures.
• Cis-trans isomerization: as Wittig olefination provides geometrical isomers
mixtures, it is necessary to convert the cis fatty acids into trans fatty acids
by double bond isomerization.
• Fractionated crystallization: a purification process in dry or solvent condi-
tions that allows high purity to be achieved. Such technology is well-known
and already developed at industrial scale for stearin/olein purification.
Each step will be developed in this chapter with the presentation of the
chemical reaction, the key parameters of multi-step processes, the technologies
and technical barriers, the limits, and the products obtained with the objective
of helping readers in their choice of reaction pathways.
A. Unsaturated fatty acids synthesis

The chemical synthesis of unsaturated fatty acids requires the introduction of
77
–
X
PPh 3 + X CH R Ph 3P+ CH R Ph 3P+ C
–
R Ph 3P C R
base
R' R' R' R'
Phosphonium salt Ylide
Figure 1. Ylides synthesis from triphenylphosphine and an alkyl halide.
double bonds in specific positions and with a specific geometrical configura-

tion. The various strategies develop in two different ways (Gravier-Pelletier
et al., 1990). In one case, the Wittig reaction an aldehyde or ketone is treated
with a phosphorus ylide (also called phosphorane) to give an olefin by the
creation of a C–C double bond. In the other, alkynes are introduced by
palladium coupling and semi-reduction of triple bonds. In palladium coupling
vinylic halide or conjugated vinylic halide are directly coupled with 1-alkyne
or conjugated alkyne in the presence of an amine; a catalytic amount of
palladium (Pd0)-cuprous iodide affords generally good yield of the desired
conjugated enyne or dienyne under mild conditions with complete preservation
of the geometry. Triple bonds can be reduced into Z-olefins by catalytic semi-
hydrogenation or into E-olefins by hydride reduction.
This chapter focuses on the Wittig reaction to introduce double bonds by
homologation reaction.
1. The Wittig olefination reactions

The Wittig reaction is undoubtedly one of the most useful methods for
selectively constructing carbon–carbon double bonds. The stereochemistry of
the newly formed double bond depends on several factors such as the nature of
the carbonyl compound, the type of phospho reagent and the conditions of the
reaction (Maryanoff & Reitz, 1989; Vedejs & Peterson, 1994).
Phosphorus ylides (nucleophilic species) are usually prepared by treatment
of phosphonium salts with a base and phosphonium salts are prepared from a
phosphine and an alkyl halide (Figure 1). The bases used are generally strong
(butyllithium, sodium amide, sodium hydride or sodium alkoxide) but could be
weak if the salt is acidic enough.
The Wittig type phospho-reagents fall into three categories as follows.
• Non-stabilized ylides: when R and R' are hydrogen or alkyl, the ylides are
classified as non-stabilized. These types of species are highly reactive,
prepared in situ and unstable in the presence of water. They generally show
Z stereochemistry.
• Semi-stabilized ylides: bearing substituents with a small conjugation ca-
pacity such as C–C double or triple bonds, phenyl, propargyl or fluorine. As
they are highly reactive, they can be used at low temperatures. They give a
mixture of Z-and E-configurations.
CHEMICAL SYNTHESIS OF MONOUNSATURATED TRANS FATTY ACIDS 79
• Stabilized ylides: ylides bearing substituents with a large conjugation

capacity. When an electron withdrawing group (COR, CN, CO2R) is present
in the α position, the charge of the carbon is spread by resonance, stabilizing
the ylide. Due to the lower reactivity of these ylides, the reactions are
carried out at room temperature at least and lead to predominant
E configuration.
The ylide reacts with the carbonyl to form an oxaphosphetane that decom-
poses into an olefin of Z or E stereochemistry (Figure 2). It was long assumed
that a betaine intermediate was the initially formed species but little evidence
for its formation has come forth. On the other hand, much evidence for the
intermediary of an oxaphosphetane has been obtained (Vedejs & Snoble,
1973). An important advantage of the Wittig reaction is that the position of the
double bond is always certain. The aldehyde or ketone may be aliphatic,
alicyclic, aromatic or heteroaromatic ; it may contain double or triple bonds and
Figure 2. The Wittig reaction between a phosphonium ylide and an aldehyde.

Figure 3. Olefination by the Wittig-Horner reaction.
various functional groups, such as OH, OR, NR2, aromatic nitro or halo, acetal
or even ester group. The stereochemistry of the product is thought to be
influenced by the reversibility of the formation of the isomeric erythro and
threo oxaphosphetanes which undergo specific loss of triphenylphosphine
oxide to give the trans (E) and cis (Z) respectively (Abell & Edmonds, 2004).
Factors that enhance the reversibility of this step favour the threo intermediate.
2. The Horner-Emmons, Wadsworth-Emmons and Wittig-Horner

olefination reactions
The Wittig reaction has also been carried out with other types of ylides, the
most important being prepared from phosphonates. This method is called the
Horner-Emmons, Wadsworth-Emmons or Wittig-Horner reaction (Figure 3).
The centre that is deprotonated possesses a mesomer effect group (for example
alkenyl, CN, COOR) that will allow the stabilization of the negative charge and
thus the formation of the E olefin. Strong bases should be used to activate the
phosphonate because the phosphorus atom is not initially charged.
The carbanionic phosphonates are formed by the action of strong bases on
alkylphosphonates, which are usually prepared from a trialkyl phosphite and an
alkyl halide (Figure 4).
O
P(OR1)3 + R2CH2X R2CH2P(OR1)2 + R1X
O O
-
R2CH2P(OR1)2 R2CH2P(OR1)2
base
Figure 4. Synthesis of ylides from phosphonates.
These ylides have the following advantages:

• they are more reactive than the corresponding phosphoranes;
• they are more likely to react with ketones; and
• the by-product of the synthesis is a phosphate salt that is water soluble and
so easily removed from the medium, unlike PR3O.
The stereochemistry of the reaction depends on several parameters: structure

of the phosphonium salt, presence of metal cations and experimental condi-
tions. The literature data show that the majority of olefins obtained from
homogeneous Wittig-Horner reactions are trans selective.
3. The Wittig olefination under phase transfer catalysis conditions

In the 1970s, it was discovered that Wittig reactions can be carried out under
phase transfer catalysis (PTC) conditions and that the syntheses are then easily
performed. NaOH and KOH, either in concentrated aqueous solution or in solid
form, are the bases generally used in PTC conditions, contrary to classical
conditions that mostly use n-BuLi, sodium or potassium alkoxides, or NaH.
This base modification uses milder temperature conditions (20 to 70°C)
compared to the low temperatures (–100°C to 0°C) used in classical conditions.
The PTC reactions can be achieved in liquid-liquid or solid-liquid systems
(Pascariu et al., 2003).
In terms of the stereochemistry of the reaction, the conclusions for PTC
olefinations are the same as those obtained under classical conditions (Anderson
& Henrick, 1975; Vinczer et al., 1988; Thompson & Heathcock, 1990). The
stereochemistry is essentially influenced by the base, the reaction temperature
and the solvent and the effects of each of these are outlined below.
With regards to the base, reaction is slow for a stabilized ylide and the
thermodynamic product is preferentially formed leading to the threo
oxaphosphetane that decomposes into the E isomer. For non-stabilized ylides,
the stereoselectivity depends on the base counter-ion (M+) used to obtain the
ylide. If the counter-ion is large, the oxaphosphetane formation is rapid. The
reaction proceeds under kinetic control and the erythro oxaphosphetane is
favoured. If the counter-ion is small, the reaction proceeds under thermody-
namic control leading to a slower oxaphosphetane formation. The formation of
the threo oxaphosphetane is favoured leading to E olefin. In summary, working
in conditions free of lithium salt will help the formation of Z olefin.
Low reaction temperatures enhance Z isomer formation. Studies have dem-
onstrated that working at low temperatures decreases the reversibility of the
reaction. The kinetic ratios are favoured compared to thermodynamic ratios.
Complementary to Figure 2, there is a third simultaneous reaction sequence
taking place at low temperature. It consists of a [π²s + π²a] cycloaddition to form
an oxaphosphetane which rearranges to the sterically more crowded cis inter-
mediate. It decomposes through an open chain zwitterion to give the Z olefin.
At higher temperatures, the reaction starts with the formation of the erythro
betaine and oxaphosphetane which are partially converted to the thermody-
namically more stable threo forms (Figure 5).
Polar protic solvents stabilize the betaines and zwitterions species, thus
ensuring the possibility of isomerization. The E isomer is then favoured with
+ - + R1CHO
PR3 CH R2
Low temperature High temperature

O- O
-
H R2 H R2
+
O PR3 Betaines
R1 H H R1
+
PR3 +
PR3
R1 R2
O PR3 O PR3
Oxaphosphetanes
R1 R2 R1 R2
+ +
O PR3 O PR3
- -
R2
R1 R2 R1 R2
R1 R2 Zwitterions R1
Figure 5. Temperature influence on the stereochemistry of the Wittig olefination (Vinczer et al.,
1988).
Table 1. Usual solvents in the Wittig olefination and some of their physical and
empirical parameters: bp, boiling point; εr, dielectric constant; µ, dipole moment; ENT,
normalized acid Lewis solvent parameter (Zalewski & Kokocinske, 1989).
Solvent bp (°C) εr µ(10–30 C.m) ENT
Water 100.00 78.30 5.90 1.00

Methanol 64.70 32.70 5.70 0.76
Acetonitrile 81.60 36.00 13.70 0.46
Dimethylsulphoxide 189.00 46.70 13.70 0.44
Dimethylformamide 152.30 37.00 13.00 0.40
Dichloromethane 39.80 8.93 5.20 0.31
Tetrahydrofuran 66.00 7.58 5.80 0.21
1,4-Dioxane 101.30 2.21 1.50 0.16
Benzene 80.10 2.28 0.00 0.11
Cyclohexane 80.70 2.02 0.00 0.01
n-Hexane 68.70 1.88 0.00 0.01
R1CHO
Ph 3P CHR R1HC CHR
+ - NaOH
Ph 3P CH2RX
+ -
Ph 3P CH2ROH Ph 3PO + RCH3
Figure 6. Wittig reaction in phase transfer catalysis (PTC) conditions using aqueous sodium hydroxide.
the use of polar solvents. Table 1 provides physical and empirical parameters
of usual solvents for the Wittig synthesis. The knowledge of these parameters
could ease the choice for controlling the stereoselectivity.
a. Liquid-liquid phase transfer catalysis (PTC) systems

For the Wittig reaction, ylides generation from alkyltriphenylphosphonium
salts in the presence of an aqueous sodium hydroxide solution has been widely
studied (Märkl & Merz, 1973; Tagaki et al., 1974). Increasing the NaOH
concentration increases the yield of olefins but phosphonium salts are also
consumed in a concurrent anion exchange reaction followed by decomposition
reaction of phosphonium salt into phosphine and alkylmethane (Figure 6).
A variety of solvents can be used in this PTC system, including THF,
benzene, CH2Cl2 and CHCl3. The Z/E ratio in PTC reactions is similar to that
observed for classical reactions.
For Wittig-Horner reactions, aqueous sodium hydroxide and CH2Cl2 or
benzene can also be used with the help of tetrabutylammonium iodide as
catalyst (Piechucki, 1974; Mikolajczyk et al., 1976). Under mild conditions
(aqueous solution of potassium hydrogen carbonate or potassium carbonate),
the influence of the reaction time and the temperature was studied for reaction
with long carbon-chain aldehydes (Villieras & Rambaud, 1983) (Table 2).
b. Solid-liquid phase transfer catalysis (PTC) systems

In addition to solid NaOH or KOH, other bases such as potassium carbonate or
potassium tertbutoxide in CH2Cl2, benzene or methanol have been used to
synthesize for, example, trans-stilbene (Boden, 1975), hydroxycinnamic esters
(Dupin & Chenault, 1985) and steroids (Cui et al., 2002). Mild bases have
proven interesting compared to NaOH because they increase the yields. Using
Table 2. Influence of the reaction temperature and reaction time on the Wittig Horner
reaction using a phase transfer catalysis (PTC) liquid-liquid system (Villieras &
Rambaud, 1983). T, temperature.
Reagent Catalyst T Time Yield Products

(°C) (%)
O O K2CO3 100 10 min 64 O H COOC2H5
EtO P CH2 COOC2H5 + C7H15 KHCO3 100 1 h 76 EtO P OH +

EtO H K2CO3 20 20 h 81 EtO C7H15 H
Table 3. Influence of the aliphatic chains of the aldehyde and of the phosphonium salt
on the Wittig reaction in 1,4-dioxane and methanol (Moussaoui et al., 2006).
Br– Ph3P+ CH2 R' + R CHO R CH CH R'
R R' Yield in olefin products (%)

Base = K2CO3 Base = NaOH
1,4-Dioxane Methanol 1,4-Dioxane Methanol
CH3 C2H5 72 68 30 20
n-C4H9 74 68 28 20
C2 H 5 C2H5 70 68 28 –
n-C4H9 70 66 30 –
C3 H 7 C2H5 68 60 26 18
n-C4H9 72 62 26 –
n-C7H15 C2H5 64 54 22 –
n-C4H9 70 54 28 –
n-C8H17 C2H5 58 50 26 18
n-C4H9 60 52 26 18
Figure 7. Self-aldol condensation of linear aliphatic aldehydes.
anhydrous 1,4-dioxane and methanol, Moussaoui and co-workers (2006) com-

pared K2CO3 and NaOH for the synthesis of olefins from aliphatic aldehydes
(Table 3). The lower yields measured with NaOH could be due to the formation
of secondary products (Canizzaro reaction, aldol condensation, ylide hydroly-
sis) that are inhibited in the presence of carbonate ions (Le Bigot et al., 1982).
The inhibition of the aldol reaction (Figure 7) is ascribed to the absence of acid-
base interactions between the enolizable hydrogen and the carbonate anion. In
such conditions, the anion is not active enough to react.
In addition, the reactivity decreases with increasing chain length of the
+ -
M2CO3 solid + 2 (C4H9)4N Br ((C4H9)4N+)2 CO32- + 2 MBr
O O
O O O O
M2CO3 solid + M+ CO3
2-
O O O O
O O 2
Figure 8. Use of phase transfer catalysts with carbonates in the Wittig olefination.
aldehyde. This could be due to the donating effect of the alkyl group placed on
the carbonyl bond, increasing thereby the electron density on the carbon of the
carbonyl group which is detrimental to the attack of the ylide. On the other
hand, there is very little effect of phosphonium chain length, confirming
previous studies (Vinczer et al., 1988).
The reaction temperature has also an influence on the yield. Increasing
temperature results in the enhancement of the basicity by the weakening of the
anion-cation link of the base.
In solid-liquid PTC systems, reactions using bases are usually carried out in
the presence of phase transfer catalysts such as crown ethers or tetra-
alkylammonium salts which are beneficial for reactivity in organic solvent. The
catalyst leads to an increase of the basicity of the anion via ionic exchange
equilibria in the case of tetra-alkylammonium salts or via complexation of the
metal cations with crown ethers (Figure 8).
In solid-liquid systems, another synthesis option involves transforming
triphenylalkylphosphonium bromides, chlorides or iodides into more reactive
phosphonium fluorides by the use of NaF or KF coupled with dibenzo-18-crown-
6 ether as catalyst in organic solvent (Kossmehl & Nuck, 1979) (Figure 9).
For Wittig-Horner reactions, solid KOH or NaOH, K2CO3, Ba(OH)2 and
Cs2CO3 can be used (Dehmlow & Barahona-Naranjo, 1981; Sinisterra et al.,
1991). For mild bases, the determination of the carbanionic structures and the
effect of water on carbanion formation have been investigated for the conver-
sion of furfural and benzaldehyde into α,β-unsaturated compounds by the
reaction of triethyl-phosphonoacetate in the presence of Ba(OH)2.H2O,
K2CO3.1.5H2O and Cs2CO3.3H2O (Mouloungui et al., 1989). The quantity of
water needed to accelerate the reaction depends upon the nature of the cation.
Apparently, water decreases the reticulation energy of the crystalline structure
at the interface level. The interaction between water and the solid base
corresponds to the solid-liquid equilibrium in a binary system. The reactions
under the specified conditions proceed through three distinct steps: phosphonate
adsorption on the base active centre and carbanion formation; reaction between
carbanion and carbonyl substrate with the formation of the final product on the
surface of the base catalyst; and desorption of the reaction product.
R3R4C=O
R1R2C=CR3R4 (R)3P-C-R1R2 (R)3P=CR1R2
-R3PO
- HF + KF
-
(R)3P+CHR1R2X- (R)3P+CHR1R2F
- + + -
F K K X Liquid phase
Solid phase
KF KX
R1 = H R3 = p-CH3-C6H4 X = Cl, Br, I

R2 = CH2C6H5 R4 = CF3
= dibenzo-18-crown-6
Figure 9. Wittig reaction in phase transfer catalysis conditions; in situ synthesis of phosphonium
fluorides (Kossmehl & Nuck, 1979).
4. Wittig type olefination applied to unsaturated fatty acid synthesis

Wittig type olefination has already proven its interest as the key step for the
synthesis of several natural products (Nicolaou et al., 1997; Odinokov, 1999)
and especially unsaturated fatty acids (Table 4). Because of the use of strong
and mild bases, the terminal carboxylic acid function, either on the aldehyde or
the phosphonium salt, is protected beforehand by esterification to prevent it
from being saponified (concurrent reaction). Each Wittig reaction will then be
followed by the hydrolysis of methyl or ethyl esters into carboxylic acid.
The operating conditions are usually harsh: toxic solvents (THF, DMSO,
DMF) and strong bases (NaH, NaOCH3, LiHMDS) are required. Using such
base/solvent pairs implies experimental constraints concerning temperature
and moisture content. Working at low temperatures and ensuring completely
anhydrous conditions for reagents, solvents and materials are not ideal condi-
tions for large-scale production. Only a few studies have led to unsaturated
fatty acids production larger than 10 g and only two studies provided more than
500 g. The fatty acid isomers obtained were all 80 to 95% cis configured.
Taking into account the example of vaccenic acid synthesis (DeJarlais &
Emken, 1978; Duffy et al., 2006), a new operating process has been developed
to provide large quantities of vaccenic acid (Mouloungui et al. unpublished
data). This process is based on solid-liquid phase transfer catalysis conditions
in highly apolar cyclohexane solvent with potassium carbonate and 18-crown-
Figure 10. Wittig reaction in phase transfer catalysis conditions for large-scale production of ethyl
vaccenate (Mouloungui et al., 2007)
6-ether as catalyst (Figure 10). Triphenylphosphonium salt and potassium

carbonate are insoluble in cyclohexane contrary to the PTC catalyst and the
aldehyde. The non-stabilized ylide is obtained at reflux with the help of the
catalyst. The 18-crown-6-ether modifies the reticulation energy of potassium
carbonate by solvating K+ and, in consequence, activating the carbonate CO2– 3
.
Once the ylide is activated, an interfacial reaction between the aldehyde and the
ylide proceeds through three distinct steps: adsorption of the aldehyde on the
ylide surface; reaction between the carbanion and the carbonyl substrate with
the formation of the final product; and desorption of the reaction product
(cyclohexane soluble). The use of a non-polarizing solvent favours ionic
reactions at the solid-liquid interface. Its importance lies in the neutral condi-
tions of the reaction.
Both phosphonium salt and triphenylphosphine oxide are not soluble in
cyclohexane and this facilitates their removal. Using these operating condi-
tions at pilot scale, 315 g of ethyl vaccenate (yield: 82%; Z:E 84:16) were
obtained. The stereochemistry is in agreement with the use of an apolar solvent.
As all of the experimental conditions lead to principally cis configured fatty
acid esters, the production of pure trans unsaturated fatty acids need two
additional steps: the hydrolysis of the fatty ester function followed by a cis-
trans isomerization.
B. Cis-trans isomerization
Fatty acid double bonds can be found in two different configurations, namely
cis or Z, and trans or E (Figure 11). The olefination by Wittig reaction mainly
R2
R1 R2 R1
Z, cis E, trans
Figure 11. Unsaturated fatty acid geometrical isomers.
88
Table 4. Examples of unsaturated fatty acid synthesis via the Wittig reaction. T, temperature; α, ylide activation temperature; β, reaction
temperature.
Reference Phosphonium salt Aldehyde Solvent T Time Product Yield % Product

base (°C) (h) (cis/trans) (g)
O O
(DeJarlais & 0 α,β 16 Methyl vaccenate 70 /
Ph3P+, I– C8H17
Emken, 1978) (91/9)
H3CO C10H20
H
O O
(Tucker et al., + – C8H17
DMF 0 α,β 16 Methyl oleate 76 <1
Ph3P , I
1965) H3CO C8H16 NaOCH3 (92/9)
H
O H3C O
(Foglia & Vail, 40 α 3 Methyl 13- 87 70
1993) Ph3P+, I– 20 β methyloctadec-11- (95/5)
H3CO C10H20
H3C H enoate
(Prakash et al., (CH2)3COOCH3 0α 2 (Z,Z,Z,Z) methyl 5,8, 60 <1
1989) C5H11 Ph3P+, Br– H –78 β 11,14 eicosatetra-
enoate
O
O
(Tranchepain α,β unsaturated THF/ –78 α,β 4 Methyl-13-benzoyl- 55 <1
Ph3P+, Br–
et al., 1989) H3CO C8H16
Hexane/ 25 β oxy-9(Z),11(E)- (85/15)
HMPA octadecadienoate
O O
(Labelle et al., +
Ph3P , I – C8H17 –78 α,β 2 (Z,Z,Z,E)methyl 62 <1
1990) nBuLi/ 20 β 5,8,11,13
H3CO C10H20 H HMDS eicosatetraenoate
O
(Genard & Patin, + – –78 α,β 6 Methyl n-methyl- 80 <1
Ph3P , Br H3COOC
1991) C7H14 C7H14 H –10 β octadec-9-enoate (97/3)
O
O
(Duffy et al., –78 α,β 48 Vaccenic acid 70 10
C6H13
2006) Ph3P+, Br– 20 α,β (82/18)
HO C10H20 H
O O
(Duffy et al., THF/ –78 α,β 16 Vaccenic acid 89 5
+ – C6H13
2006) Ph3P , Br Toluene 20 α,β (88/12)
HO C10H20 H KHMDS
O O
(Klein et al., C9H19
DMSO 20 α,β 24 Ethyl 6-hexa- 35 40
Ph3P+, Br–
2002) C2H5O C6H12
NaHMDS decenoate (95/5)
H
(Tyrwhitt-Walker, O H3C O CH2Cl2/ Reflux 48 Ethyl 8-methyl 67 870

1985) + – 1,4-diox- -6-nonenoate
Ph3P , Br
C2H5O C6H12 ane K2CO3
H3C H
O
CH3 O
(Le Pivert et al., Ph3P , Br
+ – THF 70 α 14 Methyl 14-methyl- 62 570
1995) H3CO C10H20 NaH 20 β pentadec-11-enoate
H3C H
O O
(Mouloungui et al., + –
Cyclo- Reflux 12 Ethyl octadec- 50 570
Ph3P , Br C7H15
2007) C2H5O C9H18
hexane 10-enoate (86/14)
H
O
K2CO3
O
Ethyl vaccenate 82 315
CHEMICAL SYNTHESIS OF MONOUNSATURATED TRANS FATTY ACIDS
C6H13
Ph3P+, Br– (84/16)
C2H5O C10H20 H
89
70
60
50
Melting point (°C)
40
cis
30 trans
20
10
0
0 2 4 6 8 10 12 14 16 18
Double Bond Position
Figure 12. Melting points of 18:1 fatty acid positional and geometrical isomers (Barve & Gunstone,
1971).
affords cis configured molecules. As our objective is trans fatty acid, the
olefination should be followed by an isomerization reaction.
The energy required to activate the transition from one configuration to
another is relatively high (30 kcal/mole) and both configurations are thermally
stable. A cis-trans equilibrium is reached and isomers can be separated
according to their different physical and chemical properties and most often
according to their melting points. As illustration, we can represent octadece-
noic acid melting points versus the position and the configuration of the double
bond (Barve & Gunstone, 1971) (Figure 12).
Cis-trans isomerization involves a rotation at the double bond level. Isomeri-
zation consists in using experimental conditions that cause loss of sp2 character
at the ethylenic carbons. Several reactants have been found to catalyse geo-
metrical isomerization of double bonds, for example selenium, nitrous acids,
sulphinic acids, and thiyl and phosphonyl radicals.
Selenium produces cis-trans isomerization through π complexes (Figure 13) .
Yield is limited by a 0.70 trans / 0.30 cis equilibrium reached under very
strong conditions (190–210°C) (Fitzpatrick & Orchin, 1957). Positional isomeri-
zation and the formation of highly polar substances are drawbacks.
The main isomerization technique rests upon nitrous oxides (NO and NO2),
commonly named ‘nitrous vapours’. Molar fractions of 0.75 trans / 0.25 cis
are obtained at equilibrium. Isomerization proceeds through a radical mecha-
nism but also through an electrophilic addition-elimination mechanism
(Litchfield et al., 1965). Nitric acid reacts with sodium nitrite to produce
nitrous acid in situ which decomposes in the presence of stronger nitric acid.
Se6 3Se2
cis + Se2 cis Se2 cis Se2

(π ) (σ)
trans + Se2 trans Se2 trans Se2

(π ) (σ)
Figure 13. Geometrical isomerization mechanism using selenium (Fitzpatrick & Orchin, 1957).
2 NaNO 2 + 2 HNO 3 2 HNO2 + 2 NaNO 3
2 HNO 2 H2O + N2O3

HNO3
N2O3 • NO
2 + NO
• NO ((aqueous phase) • NO
2 2 (organic phase)
NO2 R2
• NO • •
2 + NO2 +
R1 R2 R1 R2 R1
• NO
2
R1 R2 R1 R2
or
O 2N NO2 O2N ONO
Figure 14. Geometrical isomerization mechanism using nitrous oxides (Litchfield et al., 1965).
Nitrous acid anhydride (N2O3) then decomposes to give the free radical •NO2.
After transition from aqueous to fatty acid phase, the radical adds to the double
bond of the fatty acid, creating a freely rotating C–C bond. This addition is
reversible and either a cis or a trans double bond can be formed when nitrogen
dioxide leaves. The addition of a second •NO2 molecule produces side reaction
products and a fall in yield (Figure 14).
Another isomerization method, leading to 77–80% trans ethylenic com-
pounds, uses p-toluenesulphinic acid (Gibson & Strassburger, 1976; Snyder &
Scholfield, 1982). This acid acts under reflux in 1,4-dioxane. The mechanism
has not been fully established yet. The homolytic fission of the acid depends on
O O O O
2 ArSO2H Ar S S Ar + H2O Ar S • + Ar S•
O O
Figure 15. Geometrical isomerization mechanism using p-toluenesulphinic acid (Gibson & Strassburger,
1976).
Radical source X•
X • + RSH XH + RS •
R2
RS
RS • + • + RS
R1 R2 R2
R1 R1
A•
A• + RSH AH + RS•
2 RS • RSSR
Figure 16. Geometrical isomerization mechanism using thiols (Chatgilialoglu et al., 2005).
a duplication reaction with water elimination. One of the two species formed
can add to the double bond and remove it (Figure 15).
Finally, thiols, mainly 2-mercaptoethanol but also thioacetic acid, alkylthiols
and cysteine chlorhydrate undergo homolytic fission in the presence of an
initiator (e.g. peroxide) (Figure 16). Formation of additional substances is
unavoidable (Sgoutas & Kummerow, 1969). Molar fractions of 0.84 trans /
0.16 cis are obtained at equilibrium, using 2-mercaptoethanol under photo-
lytic conditions (Chatgilialoglu et al., 2005).
The usual experimental conditions are summarized in Table 5. Experiments
were done first on fatty acids and fatty acid methyl esters and then applied to the
triacylglycerols of linseed oil (Wolff, 1992) and of borage, rice and olive oils
(Samadi et al., 2004).
All the described methods allow the transition cis → trans. In fact, for all of
these methods, the equilibrium is more rapidly reached starting from cis isomer
than from trans isomer. So, during the same reaction, the conversion of cis
isomer has finished whereas that of trans has just begun. That is why the trans
→ cis conversion is more difficult and must pass through a stabilized reaction
intermediate at the level of the ethylenic bond: epoxide, glycol, thiocarbonate
or dihalide. The stereochemistry of addition and elimination will determine the
configuration of the reaction product.
Table 5. Comparison of the experimental conditions used for cis-trans geometrical isomerization. T, temperature.
Reference Raw material Catalyst Solvent Time (h) T (°C) Final trans/cis ratio
(Fitzpatrick & Orchin, 1957) oleic acid selenium solvent-free 1.0–1.5 190–210 70/30
(Gibson & Strassburger, 1976) methyl oleate p-toluenesulfinic acid 1,4-dioxane 0.2–1.0 Reflux 79/21
(Litchfield et al.,1965) oleic acid HNO3/NaNO2 solvent-free 0.2–0.5 40–65 75/25
(Chatgilialoglu et al.,2005) methyl oleate 2-mercaptoethanol tert-butanol 0.5–1.0 20 84/16
+ di-tert-butyl-ketone Photolytic
CHEMICAL SYNTHESIS OF MONOUNSATURATED TRANS FATTY ACIDS
93
In addition to geometrical isomerization, the ethylenic bond can undergo a

shift or positional isomerization, called ‘migration’ for monounsaturated acids
and ‘conjugation’ for polyunsaturated acids. Conjugation leads to other trans-
formations: intramolecular cyclization, cycloaddition and dimerization.
Common metal catalysts (Rh, Pt, Ni, Pd, Ru) typically give positional isomeri-
zation, especially conjugation, in addition to geometrical isomerization (Hsu
et al., 1989; Larock et al., 2001; Bernas et al., 2004). Strong alkaline condi-
tions will also lead to conjugation in addition to geometrical isomerization
(Nichols et al., 1951; Berdeaux et al., 1998; Yang & Liu, 2004). These
catalysts are mostly used for the synthesis of conjugated linoleic acid (CLA).
C. Fractionation techniques
Investigation of the clinical effects of individual trans fatty acids requires that
they be available at a purity of at least 90%, and this is why the different fatty
acids in a mixture should be separated. That is the aim of the two fractionation
techniques, thermal fractionation and fractional crystallization. Thermal sepa-
ration aims to separate different chain length fatty acids and is based on the
difference in their boiling points. Fractional crystallization leads to the separa-
tion of fatty acids having the same chain length but with different unsaturation
(saturated vs unsaturated, cis vs trans) and is based on the difference in their
melting points. The purification of trans fatty acids is therefore achieved using
fractional crystallization. The difference in melting point between cis and trans
18:1 fatty acids illustrates this choice (Figure 12).
Fractional crystallization has already been studied widely and optimized to
fractionate oils into high melting stearin fractions and lower melting olein
fractions. This was the original aim of the purification technique and its
principal application. Contrary to fatty acids which are characterized by a
single melting point, triacylglycerols exist as three polymorphic crystalline
forms: α (metastable), β' (intermediate) and β (stable), differing from each
other in the hydrocarbon chain packing. Table 6 summarizes the melting points
for the three different crystalline forms for several homogeneous triacyl-
glycerols.
Table 6. Melting points of the different crystal modifications of homogeneous

triacylglycerols (Kodali et al., 1987).
Triacylglycerol Melting point of modification (°C)

α β' β
Tripalmitin 45 56.5 63.5

Tristearin 55 65 73.1
Triolein –32 –12 4.9
Trielaidin 15 42.0
60
50
Melting point (°C)
40
30 cis
trans
20
10
0
0 2 4 6 8 10 12 14 16
Double Bond Position
Figure 17. Homogeneous triacylglycerol β melting point versus double bond position in the 18:1 chain
(Hagemann et al., 1975).
Fractional crystallization of fatty acids uses techniques developed for stearin/

olein separation and allows separation of saturated fatty acids from unsaturated
fatty acids. This is also true for separation of cis and trans homogeneous
triacylglycerols. In fact, Figure 17 shows the evolution of β form melting
points for 18:1 homogeneous triacylglycerols versus the double bond location
in the chain. On average, trans melting points are 10°C to 30°C higher than cis
melting points for the same positional isomers of the triacylglycerol. The gap
between cis and trans melting points is similar in the case of fatty acids
(Figure 12).
1. Fractional crystallization in solvent

At first, fractional crystallization was achieved in a solvent, with regulation of
the temperature so that saturated fatty acids crystallized while unsaturated fatty
acids were left in solution. The object of these studies was, in 90% of the cases,
the separation of stearic acid from oleic acid. Despite numerous attempts, only
one single process (‘Emersol’ process, Demmerle, 1947) was capable of large-
scale industrial application. In this process, the fatty acids are dissolved in
water-miscible solvents (acetone, ethanol or methanol) containing no more
than 15% water. The most desirable solvent was proven to be 90% methanol.
The solution, containing no more than 30% fatty acid in solvent, is then
gradually cooled in a set of crystallizers. The miscella of saturated crystallized
and dissolved unsaturated fatty acids reaches a rotating filter which continu-
ously separates these two phases. This separation is followed by concentration
FatMixture
Fat Mixturetoto be
be Solvent:
crystallized
crystallized acetone and hexane
Solvent + Fat
Crystallization:
Nucleation, crystal growth
Separation
Separation
Lowmelting
Low melting fraction
fraction
in solvent
solvent Highmelting
High melting fraction
fraction
in
Concentration:
Solvent evaporation
Solvent evaporation
Solvent distillation
Low
Low melting
melting High melting
High melting
Solvent
Solvent fraction
fraction fraction
fraction
Figure 18. Principle of fractional crystallization in solvent.
of each phase to remove solvent. Repetition of this cycle and the processing of
materials finally allows a very high quality product to be obtained (Figure 18).
Recent studies on solvent fractionation demonstrate that, occasionally, a
phase separation occurs before crystallization. The liquid phases are solutions
of the material in solvent but they differ in their concentrations. This is known
as ‘oiling out’ (Smith et al., 2007). This liquid-liquid phase separation prior to
crystallization raises the nucleation temperature and has an influence on the
separation of the different molecules of the mixture.
Apart from obtaining a high-quality product, the other advantages of using a
solvent are faster nucleation and growth, easier filtration and easier heat
transfer. However, the Emersol process also presents several drawbacks. The
efficiency of the filter depends on the characteristics of the fatty acid to be
crystallized. Moreover, the use of solvents such as methanol requires that
safety precautions be taken. Alternative technological solutions have been
developed to overcome these drawbacks.
2. Dry fractionation
In contrast to the classical fractionation in solvent, the second technology is
TEMPERATURE
TEMPERATURE
Liquid
T1 a b c
Solid +
Liquid
Solid
A COMPOSITION B
Figure 19. Schematic diagram of a two-compound binary mixture.
called dry fractionation because it is solvent-free. It involves cooling a fatty

acid mixture (or the fat) in a crystallizer followed by separation of the
unsaturated fatty acid from the solidified saturated fatty acid cake. The
fractionation process comprises: complete melting of the fatty acid mixture;
slow cooling; gentle or no agitation to help the development of large crystals;
and separation into liquid and solid fractions with different physical and
chemical compositions. When a liquid fat is cooled, a solid phase separates,
whose composition and amount depend principally on the temperature. Fig-
ure 19 shows a schematic phase diagram of a binary mixture of two compounds
completely miscible in the solid state. Holding the solution at temperature T1
results in the formation of a solid phase (crystals) of composition c in a liquid
of composition a. The fraction of solid phase is ab/ac.
To obtain crystallization, it is necessary to increase the concentration of the
molecules to be crystallized to a level above the saturated-solution concentra-
tion at given temperature. In practice, this is not sufficient to cause crystallization
because some solutions, called ‘supersaturated’, can exist indefinitely with
concentrations above the saturation level without forming any crystals (Fig-
ure 20). Then, three zones exist around the saturation curve: stable zone (no
crystallization), metastable zone (possible crystallization with assistance such
as stirring or seeding) and unstable zone (spontaneous crystallization). The
boundary between metastable and unstable zones depends on process variables
such as cooling rate and agitation. To understand the existence of the metastable
zone, crystallization should be considered as a two-step process: nucleation
followed by crystal growth (Timms, 2005; Smith et al., 2007).
Once the crystallization has ended, the desired solid molecules (crystals)
10
0.5
2.5
10
1.0
8
12.5
Solid content index (%)
6 Saturation
curve
4 Metastable Stable
0
20 40 60 80
Temperature (F)
Figure 20. Saturation-supersaturation diagram for crystallization of partially hardened soybean oil.
Effect of cooling rate on metastable/unstable boundary (dashed line). Numbers are rate of cooling (°F/
min) at 120 rev/min (Singh, 1976).
must be separated from the liquid molecules. The separation step can be
achieved by centrifugation, vacuum filtration or pressing (Gibon & Tirtiaux,
2002). This last technique is nowadays the preferred choice in new fractionation
plants.
3. Fractionation in presence of a surfactant (‘Lanza’ process)

The second alternative technology is the separation using surfactant solutions
(chlorosulphonated alcohols C8–C12, sodium dodecylsulphate) which only wet
the solid particles (i.e. the crystallized particles). Known as the Lanza process,
this technology has been developed for palm oil fractionation. The wetting
agent wets the surface of the fat crystals that have been precipitated from the
melted phase. The wetted crystals become hydrophilic and sediment into the
aqueous phase. Large oil droplets that are totally free of crystals are formed;
these droplets coalesce to form finally a continuous oil phase. The process,
applied to olein/stearin separation in fats, is given in Figure 21.
A centrifugation step allows separation: on one hand, saturated fatty acids
and aqueous solution, and on the other, unsaturated fatty acids. Breaking up the
Fat
Crystallization Surfactantsolution
Surfactant solution
Mixture
Mixture
Separation
Separationby
bycentrifugation
centrifugation
Olein Pseudo
Pseudo-solution
-solution ofofstearin
stearin
Washing Separation
Separation by heating
heating
Olein
Surfactant
Surfactant
Stearin solution
solution
Figure 21. Principle of the Lanza process applied to stearin/olein fractionation.
emulsion provides saturated fatty acids and surfactant solution. This separation
uses an Alfa-Laval ‘Hyfran’ system. This process is simple, compact safe and,
above all, the surfactant solution wets all the solid particles, irrespective of the
size and form of the crystals.
4. Comparison of the processes

Table 7 compares the three processes in the case of palm oil fractionation. The
Lanza process concerns a 100 tonne/d plant whereas the dry and solvent
fractionation processes were studied for a 200 tonne/d installations. Finally,
taking into account that dry fractionation and solvent fractionation are the most
usual processes, a list of the advantages and drawbacks of the solvent process
will help the choice (Table 8).
D. Conclusions and outlook

An important research focus for lipid nutrition is to understand the biological
properties of fatty acids. The health implications of the different TFA isomers
in modulating cardiovascular disease risk factors are not completely eluci-
dated. There is therefore a need for pure TFA isomers for biomedical studies.
Table 7. Comparison of fractionation processes for the case of palm oil (Bockisch,
1998).
Consumption per tonne for Lanza fractionation

Steam (4 bar) (kg) 150
Electrical energy (kWh) 30
Cooling water (m3) 0.2
Na-Laurylsulfate (kg) 0.2–0.6
Electrolyte, MgSO4, Na2SO4 (kg) 0.2–0.7
Consumption per tonne for double-stage fractionation
Dry fractionation Solvent fractionation
Steam (6 bar) (kg) 85 400
Electrical energy (kWh) 36 35
Water (m3) 1.2 1
Solvent (m3) solvent-free 8
Yield of olein (%) 60 70
Yield of stearin I (%) 40 18
Yield of stearin II (%) 40 12
Table 8. Advantages and disadvantages of solvent fractionation process.
Advantages Disadvantages
• Lower temperature of crystallization • Larger industrial set-up

• Shorter crystallization times • Higher investments
• Easier separation • Higher production cost
• Products with clearer specifications and • Risks of explosion and combustion
higher quality
Human clinical studies usually require significant quantities of pure com-

pounds and therefore methodologies must be developed to allow production on
a large scale (> 1 kg). Wittig olefination is the key step for such production and
recent studies demonstrated the opportunity for working under mild condi-
tions: mild bases and non-toxic solvents at temperatures between 20 and
100°C. Controlling the reaction parameters leads to manufacture of the target
molecule with the desired structure. As intermediates, synthetic fatty acids may
be esterified into mono-, di- or triacylglycerols whose final homogeneous
structures differ from those of natural molecules. Finally, apart from Wittig
reaction, isomerization and fractionation techniques can be applied to vegeta-
ble oils to change the natural cis structures into new trans structures, thus
providing new polymorphic materials directly usable as value-added mol-
ecules or as new reagents in organic chemistry.
References
Abell, AD and Edmonds, MK (2004) The Wittig and related reactions. In: Organophospho-
rus reagents: a practical approach in chemistry, Oxford University Press, Oxford, UK,
99–127.
Anderson, RJ and Henrick, CA (1975) Stereochemical control in Wittig olefination synthesis.
Preparation of the pink bollworm sex pheromone mixture, Gossyplure. J. Am. Chem.
Soc., 97, 4327–4334.
Barve, JA and Gunstone, FD (1971) Fatty acids. 33. Synthesis of all the octadecynoic acids
and all the trans-octadecenoic acids. Chem. Phys. Lipids, 7, 311–323.
Berdeaux, O, Voinot, L, Angioni, E, Juaneda, P and Sebedio, JL (1998) A simple method of
preparation of methyl trans-10,cis-12- and cis-9,trans-11-octadecadienoates from me-
thyl linoleate. J. Am. Oil Chem. Soc., 75, 1749–1755.
Bernas, A, Kumar, N, Maki-Arvela, P, Holmbom, B, Salmi, T and Murzin, DY (2004)
Heterogeneous catalytic production of conjugated linoleic acid. Organic Process Re-
search & Development, 8, 341–352.
Bockisch, M (1998) Fats and oils handbook, AOCS Press, Champaign, Illinois, USA.
Boden, RM (1975) A mild method for preparing trans-alkenes; crown ether catalysis of the
Wittig reaction. Synthesis, 784.
Chatgilialoglu, C, Samadi, A, Guerra, M and Fischer, H (2005) The kinetics of Z/E
isomerization of methyl oleate catalyzed by photogenerated thiyl radicals. Chemphyschem,
6, 286–291.
Cui, JG, Lin, CW, Zeng, LM and Su, JY (2002) Synthesis of polyhydroxysterols (III):
synthesis and structural elucidation of 24-methylenecholest-4-en-3β,6α-diol. Steroids,
67, 1015–1019.
Dehmlow, EV and Barahona-Naranjo, S (1981) Applications of phase transfer catalysis. Part
17. Wittig reactions in various two-phase systems. J. Chem. Res., Synopses, 5, 142.
DeJarlais, WJ and Emken, EA (1978) Synthesis of di-, tetra-, and hexadeuterated 11-
octadecenoates. Journal of Labelled Compounds and Radiopharmaceuticals, 15, 451–460.
Demmerle, RL (1947) Emersol process. Journal of Industrial and Engineering Chemistry,
39, 126–131.
Duffy, PE, Quinn, SM, Roche, HM and Evans, P (2006) Synthesis of trans-vaccenic acid and
cis-9-trans-11-conjugated linoleic acid. Tetrahedron, 62, 4838–4843.
Dupin, JFE and Chenault, J (1985) Phase transfer catalyzed Wittig-Horner synthesis.
Preparation of hydroxycinnamic esters from ortho hydroxy aromatic aldehydes. Obten-
tion of hydroxycinnamic acids. Synthetic Communications, 15, 581–586.
Fitzpatrick, JD and Orchin, M (1957) The selenium catalyzed cis-trans isomerization of 9-
octadecenoic (oleic-elaidic) acids. J. Am. Chem. Soc., 79, 4765–4771.
Foglia, TA and Vail, PD (1993) An efficient, large-scale synthesis of triisopentadecanoin.
Organic Preparations and Procedures International, 25, 209–213.
Genard, S and Patin, H (1991) Studies related to fatty acid desaturation. Part I. Preparation
of methyl-substituted nonanols, nonyl bromides and total synthesis of new branched oleic
acids. Bulletin de la Societe Chimique de France, 397–406.
Gibon, V and Tirtiaux, A (2002) Latest trends in dry fractionation. Lipid Technology, 14, 33–
36.
Gibson, TW and Strassburger, P (1976) Sulfinic acid catalyzed isomerization of olefins. J.
Org. Chem., 41, 791–793.
Gravier-Pelletier, C, Dumas, J, Le Merrer, Y and Depezay, JC (1990) Methods for the total
synthesis of acyclic hydroxylated fatty acids. Prog. Lipid Res., 29, 229–276.
Hagemann, JW, Tallent, WH, Barve, JA, Ismail, IA and Gusntone, FD (1975) Polymorphism
in single-acid triglycerides of positional and geometric isomers of octadecenoic acid.
J. Am. Oil Chem. Soc., 52, 204–207.
Hsu, N, Diosady, LL and Rubin, LJ (1989) Catalytic behavior of Palladium in the hydrogena-
tion of edible oils. II. Geometrical and positional isomerization characteristics. J. Am. Oil
Chem. Soc., 66, 232–236.
Klein, D, Ernst, H and Koppenhofer, J (2002) Process for the preparation of cis-6-
hexadecenoic acid. US patent 2002/0107412 A1.
Kodali, DR, Atkinson, D, Redgrave, TG and Small, DM (1987) Structure and polymorphism
of 18-carbon fatty acyl triacylglycerols: effect of unsaturation and substitution in the 2-
position. J. Lipid Res., 28, 403–413.
Kossmehl, G and Nuck, R (1979) Phase-transfer-catalyzed Wittig reactions without base
addition. Chemische Berichte, 112, 2342–2346.
Labelle, M, Falgueyret, JP, Riendeau, D and Rokach, J (1990) Synthesis of two analogues of
arachidonic acid and their reactions with 12-lipoxygenase. Tetrahedron, 46, 6301–6310.
Larock, RC, Dong, X, Chung, S, Reddy, CK and Ehlers, LE (2001) Preparation of conjugated
soybean oil and other natural oils and fatty acids by homogeneous transition metal
catalysis. J. Am. Oil Chem. Soc., 78, 447–453.
Le Bigot, Y, Delmas, M and Gaset, A (1982) A simplified Wittig synthesis using solid/liquid
transfer processes. II. The use of potassium carbonate for the synthesis of alkenes from
aromatic and aliphatic aldehydes. Synthetic Communications, 12, 107–112.
Le Pivert, M, Poisson, S, Laur, J, Coustille, JL and Pages, X (1995) Synthesis of a branched
fatty acid triglyceride at the laboratory and pilot plant level. Oleagineux, Corps Gras,
Lipides, 2, 369–374.
Litchfield, C, Harlow, RD, Isbell, AF and Reiser, R (1965) Cis-trans isomerization of oleic
acid by nitrous acid. J. Am. Oil Chem. Soc., 42, 73–78.
Märkl, G and Merz, A (1973) Styrene derivatives by aromatic aldehyde reaction with
nonstabilized phosphonium ylides in an aqueous system. Synthesis, 5, 295–297.
Maryanoff, BE and Reitz, AB (1989) The Wittig olefination reaction and modifications
involving phosphoryl-stabilized carbanions. Stereochemistry, mechanism, and selected
synthetic aspects. Chemical Reviews, 89, 863–927.
Mikolajczyk, M, Grzejszcak, S, Midura, W and Zatorski, A (1976) Organosulfur compounds.
X. Horner-Wittig reactions in a two-phase system in the absence of a typical phase-
transfer catalyst. Synthesis, 6, 396–398.
Mouloungui, Z, Candy, L and Destaillats, F (2007) A simplified Wittig method for large scale
molecule production. Unpublished results.
Mouloungui, Z, Delmas, M and Gaset, A (1989) The Wittig-Horner reaction in weakly
hydrated solid/liquid media: structure and reactivity of carbanionic species formed from
ethyl (diethyl phosphono)acetate by adsorption on solid inorganic bases. J. Org. Chem.,
54, 3936–3941.
Moussaoui, Y, Said, K and Ben Salem, R (2006) Anionic activation of the Wittig reaction
using a solid-liquid phase transfer: examination of the medium-, temperature-, base- and
phase-transfer catalyst effects. ARKIVOC, 1–22.
Nichols, PL, Jr., Herb, SF and Riemenschneider, RW (1951) Isomers of conjugated fatty
acids. I. Alkali-isomerized linoleic acid. J. Am. Chem. Soc., 73, 247–252.
Nicolaou, KC, Harter, MW, Gunzner, JL and Nadin, A (1997) The Wittig and related
reactions in natural product synthesis. Liebigs Annalen/Recueil, 1283–1301.
Odinokov, VN (1999) Olefination of aliphatic aldehydes in the synthesis of mono-, di-, and
trienic pheromones of linear structure. Chemistry of Natural Compounds, 35, 260–285.
Pascariu, A, Ilia, G, Bora, A, Iliescu, S, Popa, A, Dehelean, G and Pacureanu, L (2003) Wittig
and Wittig-Horner reactions under phase transfer catalysis conditions. Central European
Journal of Chemistry, 1, 491–542.
Piechucki, C (1974) Horner reaction in an aqueous two-phase system using phase-transfer
catalysis. Synthesis, 869–870.
Prakash, C, Saleh, S, Sweetman, BJ, Taber, DF and Blair, IA (1989) A synthon for C-20
trideuterated eicosanoids: preparation of [2H3]-arachidonic acid. J. Labelled Compd.
Radiopharm., 27, 539–551.
Samadi, A, Andreu, I, Ferreri, C, Dellonte, S and Chatgilialoglu, C (2004) Thyil radical-
catalyzed isomerization of oils: an entry to the trans lipid library. J. Am. Oil Chem. Soc.,
81, 753–758.
Sgoutas, DS and Kummerow, FA (1969) Cis-trans isomerization of unsaturated fatty acid
methyl esters without double bond migration. Lipids, 4, 283–287.
Singh, G (1976) Analysis of crystallization systems with applications to continuous fractional
crystallization of fatty acid triglycerides AIChE Symposium Series, 72, 100–109.
Sinisterra, JV, Barrios, J, Mouloungui, Z, Delmas, M and Gaset, A (1991) New cesium
carbonate (Cs2CO3) catalyst for the Wittig-Horner synthesis of acrylates in interfacial
solid liquid conditions. Bulletin des Societes Chimiques Belges, 100, 267–275.
Smith, KW, Cain, FW, Favre, L and Talbot, G (2007) Liquid-liquid phase separation in
acetone solutions of palm olein: implications for solvent fractionation. Eur. J. Lipid Sci.
Technol., 109, 350–358.
Snyder, JM and Scholfield, CR (1982) Cis-trans isomerization of unsaturated fatty acids with
p-toluenesulfinic acid. J. Am. Oil Chem. Soc., 59, 469–470.
Tagaki, W, Inoue, I, Yano, Y and Okonogi, T (1974) Wittig reaction in benzene-aqueous
alkaline solution. Tetrahedron Letters, 30, 2587–2590.
Thompson, SK and Heathcock, CH (1990) Effect of cation, temperature, and solvent on the
stereoselectivity of the Horner-Emmons reaction of trimethyl phosphonoacetate with
aldehydes. J. Org. Chem., 55, 3386–3388.
Timms, RE (2005) Fractional crystallisation — the fat modification process for the 21st
century. Eur. J. Lipid Sci. Technol., 107, 48–57.
Tranchepain, I, Le Berre, F, Dureault, A Le Merrer, Y and Depezay, JC (1989) Total
enantiospecific syntheses of 13(S)-hydroxy 9Z, 11E-octadecadienoic (coriolic) acid and
13(S)-N-tosylamino analogue. Tetrahedron, 45, 2057–2065.
Tucker, WP, Tove, SB and Kepler, CR (1965) The synthesis of 11,11-Diteurerooleic acid.
Journal of Labelled Compounds, 7, 137–143.
Tyrwhitt-Walker, GTG (1985) Unsaturated carboxylic acid. GB patent 2153826.
Vedejs, E and Peterson, MJ (1994) Stereochemistry and mechanism in the Wittig reaction.
Topics in Stereochemistry, 21, 1–157.
Vedejs, E and Snoble, KAJ (1973) Direct observation of oxaphosphetanes from typical Wittig
reactions. J. Am. Chem. Soc., 95, 5778–5780.
Villieras, J and Rambaud, M (1983) Wittig-Horner reactions in heterogeneous media. 2. A
convenient synthesis of α,β˜-unsaturated esters and ketones using weak bases in water.
Synthesis, 300–303.
Vinczer, P, Juvancz, Z, Novak, L and Szantay, C (1988) Synthesis of pheromones. IV.
Chemistry of the Wittig reaction. I. Effects of reaction conditions on the stereoselectivity
and yield of the Wittig reaction. Acta Chimica Hungarica, 125, 797–820.
Wolff, RL (1992) Resolution of linolenic acid geometrical isomers by gas-liquid chromatog-
raphy on a capillary column coated with a 100% cyanopropyl polysiloxane film
(CPTMSil 88). J. Chromatogr. Sci., 30, 17–22.
Yang, T-S and Liu, T-T (2004) Optimization of production of conjugated linoleic acid from
soybean oil. J. Agric. Food Chem., 52, 5079–5084.
Zalewski, R and Kokocinske, H (1989) Non-hierarchical classification of organic solvents
using characteristic vector analysis of physical and empirical solvent parameters. Journal
of Physical Organic Chemistry, 2, 232–242.
CHAPTER 5
Analysis of trans fatty acids of partially
hydrogenated vegetable oils and dairy products
W.M. NIMAL RATNAYAKE1 AND CRISTINA CRUZ-HERNANDEZ2
1
Nutrition Research Division, Food Directorate, Health Products and Food
Branch, Health Canada, Ottawa, Ontario, Canada
2
A. Introduction
It is now well recognized that the consumption of trans fats, especially
industrially produced trans fats via partial hydrogenation of vegetable oils
(PHVO), increases the risk of coronary heart disease and that trans fats may do
even more harm than saturated fats (for a review see Mozaffarin et al., 2006).
During the last five years, professional health organizations have taken notice
of the negative effects of industrially produced trans fats and some govern-
ments agencies have taken actions to reduce the trans content in processed
foods. In 2002, the Panel on Macronutrients of the US National Academies of
Science, Institute of Medicine, recommended that trans fat consumption be as
low as possible (Institute of Medicine, 2002). In recent years, the Food and
Agricultural Organization/World Health Organization (FAO/WHO) (2002)
and the American Heart Association (Lichtenstein et al., 2006) have recom-
mended that the trans fat content in the human diet be less than 1% of total
energy. In 2003, Denmark became the first country to set an upper limit on the
percentage of trans fats in foods, prohibiting the sale of foods containing more
than 2% industrially produced trans (as percent total fat) (Stender & Dyerberg,
2003). In December 2005, Canada became the first country to regulate the
mandatory labelling of trans fats on pre-packaged foods (Health Canada,
2003). Subsequently, the USA introduced mandatory declaration of trans fats
in foods (Food and Drug Administration, 2006).
These actions aimed at reducing the trans fat content in foods have prompted
a renewed interest in reliable analytical methods for accurate determination of
the total trans fat content, the proportions of individual trans isomers and the
total fatty acid profile of dietary fats. Several good methods are now available
for the determination of trans fats. The advances in the methods up to the mid
1990s were comprehensively reviewed by Firestone & Sheppard (1992), the
105
British Nutrition Foundation (1995) and Ratnayake (1998). Recent develop-

ments in chromatographic techniques for the analysis of trans fats have been
reported by Ratnayake (2004), Cruz-Hernandez et al. (2004) and Sebedio and
Ratnayake (2007), whereas Mossoba et al. (2004) have reviewed recent ad-
vances in the infrared methodologies.
Analysis of fatty acid profile of dietary fats encompasses many steps, such as
fat extraction, conversion of the extracted fat into fatty acid methyl esters
(FAME) and the analysis of FAME by capillary gas chromatography (GC). To
obtain detailed information on individual cis-and trans-positional isomers, the
GC analysis is complemented by other chromatographic techniques, such as
gas chromatography-mass spectrometry (GC-MS), silver nitrate-thin layer
chromatography (AgNO3-TLC), or silver ion-high performance liquid chroma-
tography (Ag+- HPLC). The following chapter reviews these procedures for
analysis of trans fats of PHVO and ruminant fats.
B. Analysis of partially hydrogenated oils
1. Methylation of oils, fats and lipids

Since capillary GC is the primary tool used in the determination of fatty acid
profile of fats and since fatty acids are always determined by GC as their FAME
derivatives, the first step in the analysis involves conversion of fat into FAME.
Several procedures are available for methylation of fats. For a good description
of the methods, the readers may refer to the various articles written by WW
Christie and available on the internet (Christie, 2006a).
In general, for all routine analysis of fat, oil or lipid samples, FAME are
prepared by transesterification using hydrogen chloride (HCl), sulphuric acid
(H2SO4), or boron trifluoride (BF3) in methanol (MeOH). In a typical proce-
dure, about 10–20 mg of the fat sample is dissolved in toluene (1 ml) in a 10-ml
screw capped (Teflon-lined) centrifuge tube. Fourteen percent BF3-MeOH
(1 ml) (alternatively 5% HCl-MeOH or 2% H2SO4-MeOH) is added and the
mixture is heated at 100°C for 60 minutes. The reaction mixture is diluted with
water (3 ml) and the FAME extracted with toluene, hexane or diethyl ether.
After concentrating under nitrogen, the FAME sample is ready to be analysed
by GC. Although transesterification with BF3-MeOH is a convenient and a very
popular procedure, caution needs to be exercised using this reagent. Christie
has indicated that BF3-MeOH has a limited shelf-life, even when refrigerated,
and that the use of old and overly concentrated (>7%) solutions might result in
the production of artefacts and the loss of appreciable amounts of highly
unsaturated fatty acids (Christie, 2006a; Christie, 1989). These acidic methyl-
ation procedures should not be used for measurement of conjugated linoleic
acid (CLA) isomers which are normally present in fat from ruminant animals
(see section B for methylation procedures for ruminant fats).
ANALYSIS OF TRANS FATTY ACIDS OF PHVO 107
For pure oil samples, such as refined vegetable oils, which contain primarily
triacylglycerols with no other lipid classes (free fatty acids, phospholipids,
sterol esters, wax esters etc.), base-catalysed transesterification is recom-
mended for simplicity and speed (Christopherson & Glass, 1969; Christie,
1989): sodium or potassium hydroxide or methoxide in methanol are the most
commonly used transesterification bases. Treatment of 19 parts of 10% solu-
tion of the pure fat sample in pertroleum ether or hexane with 1 part of either 2N
methanolic potassium hydroxide or sodium hydroxide at room temperature
results in the almost instantaneous formation of FAME. The solution may be
analysed by GC at once.
2. Extraction and transmethylation of fat from food and tissue samples

For the analysis of the fatty acid profile of food or animal tissue samples, the
additional step of fat extraction is required. The various extraction procedures
used were also recently reviewed by Christie (2006b). In many lipid laborato-
ries in North America, the AOAC Official method 996.06 (AOAC, 2005a),
initially developed by House et al. (1994), is routinely used for extraction and
methylation of fat from foods samples. This official method is applicable to a
variety of food samples and it complies with the definition of fat of the current
labelling regulations in Canada (Health Canada, 2003) and the USA (Food and
Drug Administration, 2006). The procedure involves extraction of the fat from
the food samples by acid- or base-hydrolysis, which provides a complete
breakdown of the food matrix, followed by ether extraction, transesterification
of fatty acids to FAME using BF3-MeOH, and capillary GC quantitation. Foods
excluding dairy products and cheese are hydrolysed using concentrated HCl,
whereas dairy products are hydrolysed using ammonium hydroxide. For cheese,
a combination of alkali and acidic hydrolysis is used. Pyrogallic acid is added
to minimize oxidative degradation of fatty acids during analysis. A
triacylglycerol, triundecanoin (11:0), is added as the internal standard, mixed
with the weighed food sample prior to hydrolysis. It is thus hydrolysed and
transesterified along with the fat from the food sample and participates in all
phases of the determination. A commercial reference mixture of FAME is used
to provide retention time checks and quantitative response factors. Individual
fatty acids are measured with respect to the 11:0 internal standard, and each
fatty acid is converted to its triacylglycerol equivalent prior to the summation
of triacylglycerols to give total fat. Factors for conversion of FAME to their
corresponding triacylglycerol equivalents are given in a tabulated form in the
AOAC Official Method 996.06. Saturated, TFA, cis-monounsaturated and
polyunsaturated fatty acids are also measured as subclasses of fat from the GC
traces. The method is applicable to a wide variety of foods including various
types of cereals, meat and meat products, dairy products and cheeses (House,
1997). However, it is important to note that the use of acid hydrolysis as
stipulated in the AOAC 996.06 (AOAC 2005a) could isomerize CLA isomers
present in dairy products. Therefore, this official method is not suitable for
analysis of ruminant fat containing CLA isomers. Special procedures, which
are discussed in section B of this chapter, need to be used.
The AOAC official method 996.01 also meets the current definition of fat for
food labelling purposes (AOAC, 2005b). This official method is similar to
AOAC 996.06, but is intended for cereal products containing 0.5–13% total fat.
The primary difference between the two official methods is in the recom-
mended internal standard; AOAC 996.01 recommends tritridecanoin
(triacylglycerol 13:0) as the internal standard, whereas AOAC 996.06 recom-
mends triacylglycerol 11:0. Considering the lower volatility of 11:0 FAME
compared to 13:0 FAME, it might be an advantage to use triacylglycerol 13:0
as the internal standard. Ali et al. (1997) demonstrated that the AOAC acid
hydrolysis/GC procedure (996.01) is not only applicable to cereal products, but
also to many other foods, including salad oils, partially hydrogenated fats,
margarines, fish fillets and chocolate bars.
3. Cis and trans isomers of partially hydrogenated vegetable oils

The carbon-carbon double bond (i.e., the ethylenic bond) of unsaturated fatty
acids can exist either in the cis or the trans configuration. In the cis form, the
hydrogen atoms of the ethylenic bond are on the same side, while in the trans
form, they are on opposite sides. The carbon-carbon double bonds of all the
major naturally occurring unsaturated fatty acids in the human diet are in the cis
configuration. Because of this cis configuration, the naturally occurring un-
saturated fatty acids have melting points lower than those of saturated fatty
acids. Consequently, vegetable oils such as canola (rapeseed), soybean, corn,
sunflower and marine oils that are rich in unsaturated fatty acids are liquid at
room temperature. These unsaturated liquid oils do not have the oxidative
stability, physical and functional properties required for preparation of hard-
margarines, shortenings, frying fats and processed foods such as cakes, french
fries, cookies, crackers, and doughnuts. However, these oils could be used after
converting to solid fats via partial hydrogenation. During partial hydrogena-
tion, the cis double bond of the unsaturated fatty acid in the original oil migrates
to a new position in the fatty acid chain (i.e., positional isomerization) and at the
same time, some of the cis double bonds (both old and new cis double bonds)
are converted to the trans configuration (i.e., geometrical isomerization). Thus,
partial hydrogenation results in the formation of a complex mixture of new cis
and trans isomeric fatty acids. The TFA have melting points higher than those
of cis-unsaturated fatty acids and are less susceptible to auto-oxidation, there-
fore, partially hydrogenated oils could be used for formulating foods that
require a stable solid fat.
In PHVO, octadecenoic acid (18:1) represents the major fraction with trans
Table 1. Distribution of positional 18:1 trans isomers of some partially hydrogenated

canola oil (PHCO) and soybean oil (PHSO) basestocks hydrogenated to different iodine
values (IV)†.
18:1 trans isomer PHCO PHSO

IV 92 IV 80 IV 64 IV 109 IV 86 IV 64
4t – – 0.1 – 0.1 0.1
5t – – 0.1 – 0.1 0.1
6t – 0.2 0.5 0.1 1 0.6
7t – 1 3.8 1.2 1.2 1.4
8t 6 9.9 13.4 8.5 7.6 8.8
9t 38.6 28.7 23 21.1 18.6 22.5
10t 20.8 25.5 23.3 23.7 26.9 21.8
11t 14.9 16.1 15.7 23.4 23.8 18.2
12t 8.6 9.4 9.7 9.7 9.9 12.5
13t 5.3 5.1 5.7 7.5 6.3 8.1
14t 2.6 2.2 3.1 3.4 2.8 3.5
15t 2.2 0.9 1.3 1.4 1.1 1.4
16t 1 1 0.3 – 0.6 1
Total 18:1 t 18.9 38.3 39.8 13.8 27 36.7
Total trans fatty acids 25.3 40.1 4.4 17.7 30.6 37.4
† Iodine value is a measure of degree of unsaturation; the value falls with the decreases in the proportion
of unsaturated fatty acids.
and unnatural cis isomeric fatty acids. The position of the double bond of these
dietary trans 18:1 isomers, counted from the carboxylic carbon, usually varies
from 4 to 16 carbon atoms of the fatty acid molecule. Very often in PHVO, the
trans-18:1 isomers form a Gaussian distribution that centres around the 9 or 10
double bond. The distribution depends on the fatty acid composition of the
starting vegetable oils and on the extent of hydrogenation. For example, in
mildly hydrogenated canola oil, elaidic acid (i.e., 9t-18:1) is the major isomer
whereas 10t-18:1 is the major isomer in moderately and heavily hydrogenated
canola oils (Table 1). On the other hand, 10t-18:1 and vaccenic acid (11t-18:1)
are the major isomers in mildly and moderately hydrogenated soybean oils and
9t-18:1 is the major isomer in heavily hydrogenated soybean oil (Table 1).
In addition to the trans-18:1 isomers, PHVO contain several cis-octadece-
noic isomers (cis-18:1), whose double bond position generally ranges from 6 to
16 (Table 2). The naturally occurring cis isomer, namely oleic acid (i.e., 9c-
18:1) is always the predominant isomer followed by 10c-18:1 and cis-vaccenic
acid (11c-18:1).
Very often, the human diet contains a number of positional and geometrical
isomers of linoleic (18:2n–6) and α-linolenic acids (18:3n–3). Both partially
hydrogenated and non-hydrogenated unsaturated vegetable oils are sources of
these isomers (Ratnayake & Pelletier, 1992; Ackman et al.,1974). They are
often detected in large quantities in mildly hydrogenated vegetable
Table 2. Distribution of positional 18:1 cis isomers of some partially hydrogenated

canola oil (PHCO) and soybean oil (PHSO) base stocks hydrogenated to different iodine
values (IV).
18:1 cis isomer PHCO PHSO

IV 92 IV 80 IV 64 IV 109 IV 86 IV 63
6c 1 0.2 1.4 0.1 0.4 1.3
7c – 0.4 3.6 0.1 0.4 0.5
8c 2.6 2.5 11.1 1.2 2.3 2.1
9c 78 74 37.7 72.2 59.2 42.8
10c 5.6 5.2 18.2 3.4 7.2 11.6
11c 7.2 7.3 13.1 7.6 8.5 11.6
12c 4.2 7.4 8.2 12.6 18.3 18.9
13c 0.5 1 3.6 1.3 0.5 4.9
14c 0.3 0.5 2.1 0.7 1.6 2.7
15c 2.2 0.6 0.7 0.8 1.2 2.4
16c 0.6 0.9 0.3 – 0.4 1.3
Total 18:1 c 59.4 40.9 29.2 30 44.6 30.5
Table 3. Composition (wt% of total fatty acids) of linoleic acid isomers in some
hydrogenated canola oil basestocks.
18:2 isomer Mildly Moderately Heavily

hydrogenated hydrogenated hydrogenated
canola oil (IV92) canola (IV 80) canola oil (IV 64)
9c,13t + 8t,13c 1.94 0.71 0.08

9c,12t 1.09 0 0
8c,13t 0.54 0.23 0
9t,12c 1 0 0
10t,15c + 9t,15c 1.11 0.17 0
tt 0.31 0.47 0.1
9t,12t 0.36 0.28 0
8c,13c 0.31 0.27 0.1
9c,13c 0.06 0 0
9c,14c 0.03 0 0
9c,15c 1.27 0.48 0.1
oils (up to 6% of total fatty acids) and refined, but unhydrogenated oils (up to
3%), whereas they are hardly detectable in heavily hydrogenated oils. PHVO
contain at least fifteen different isomers of linoleic acid (Ratnayake & Pelletier,
1992), the major ones being 9c,13t-18:2, 9c,12t-18:2 and 9t,12c-18:2 (Table 3).
The isomers present in non-hydrogenated oils or in many common dietary oils
are the results of exposure of the fat to some form of heat treatment such as
steam deodorization, stripping during refining of oils (Ackman et al., 1974), or
simple heating in deep fat frying (Grandgirard et al., 1984). In these processes,
the double bonds do not shift in position, but isomerize from cis to trans
resulting in the formation of small amounts of geometric trans isomers.

α-Linolenic acid is more prone to isomerization than linoleic acid, whereas
oleic acid is hardly isomerized. In many non-hydrogenated dietary oils, the two
mono-trans isomers of linoleic (i.e., 9c,12t-18:2 and 9t,12c-18:2) are present at
similar levels and very often higher than the all-trans isomer, 9t,12t-18:2. Eight
geometric isomers are possible for α-linolenic acid, but usually only four are
present in industrially refined oils or oils subject to mild heat treatments. They
have been identified as 9t,12c,15t-18:3; 9c,12c,15t-18:3; 9c,12t,15c-18:3 and
9t,12c,15c-18:3 (Ackman et al.,1974, Grandgirard et al., 1984).
4. Analysis of cis and trans isomers of partially hydrogenated vegetable

oils
a. Capillary GC separation of oleic acid isomers

Analyses of cis and trans isomers of oleic acid are best carried out by GC using
100 m long flexible fused silica capillary columns coated with highly polar
cyanopolysiloxane stationary phases, containing various polar substituents.
Varieties of cyanopolysiloxane capillary columns are available from chroma-
tographic suppliers and marketed under trade designations such as SP-2560,
SP-2340, CP Sil 88, BPX 70 and HP-88.
In cyanopolysiloxane capillary columns, the trans isomer always comes out
before the corresponding cis isomer. Within the trans or cis isomeric group, the
positional isomers with the double bond closest to the carboxylic group (i.e.,
low Δ value isomers) elute before the positional isomer with the double bond
furthest away from the carboxylic group (i.e., the high Δ value isomers). For
example, as shown in Figure 1, the four trans-18:1 isomers elute before their
cis counterparts and both cis and trans isomers elute in the order 7-18:1, 9-18:1,
11-18:1 and 12-18:1. However, overlaps could occur between some cis and
trans isomers with different double bond positions. For example, in Figure 1,
the 12t-18:1 isomer partially overlaps with 7c-18:1. These overlaps are com-
mon when analysing PHVO that contain an assortment of positional cis and
trans 18:1 fatty acid isomers.
As a general rule, with any type of cyanopolysiloxane capillary column, the
trans isomers from 4t-18:1 to 11t-18:1 are readily separated from all the cis-
18:1 isomers. But after the peak for 11t-18:1 (i.e., trans isomers from 12t-18:1
to 16t-18:1), cis-trans overlap problems start to occur. The extent of overlaps
is quite severe on capillary columns shorter than 100 m. For example, it has
been observed that in the analysis of a PHVO using a 60 m SP-2340 capillary
column with isothermal column temperature operation at either at 160°C or
185°C, the 13t,14t and 15t- 18:1 isomers were completely overlapped with the
major cis 18:1 peak which included five isomers (6c to 10c-18:1) (Ratnayake
& Beare-Rogers, 1990). In addition, there were two other cis-trans overlaps;
the 12t-18:1 isomer only partially separated from this major cis 18:1 peak and
9c
9t
7c 12c
12t 11t
11c
7t
22.5 25 min
Figure 1. Gas chromatogram of a standard FAME mixture of 7t, 9t-, 11t-, 12t-, 7c-, 9c-,11c- and 12c-
18:1. Analysis on a SP 2560 fused silica capillary column (100 m × 0.25 mm i.d × 20 m film thickness),
operated isothermally at 180°C. Hydrogen carrier gas flow rate 1 ml/min.
16t-18:1 overlapped with 14c-18:1. These overlapping problems have also

been observed for 50 m CP-Sil 88 (Wolff et al., 1998). Similar overlapping
problems occur even with 100 m capillary columns when the analysis is
performed using a column temperature program (Ratnayake, 1998). These cis-
trans isomer overlaps place a constraint on accurate determination of the total
TFA content. It is evident that considering solely the trans isomer peak eluting
before the oleic acid peak will lead to considerable under reporting of TFA. In
some fatty acid analysis, the results for total trans would be lower by about 30%
(corresponding to the masked 12t to 16t-18:1 isomers) (Wolff et al., 1998).
The best separation of fatty acids of PHVO with minimum overlaps of the
cis-and trans-18:1 isomers is achieved using 100 m SP-2560 or CP-Sil 88
capillary columns (or other cyanosilicone columns of equivalent polarity) with
column oven temperature operated isothermally at 180°C using hydrogen as
the carrier gas with a flow rate of 1.0 ml/min (Ratnayake et al., 2006). These
operating conditions have now been included in the American Oil Chemists’
Society’s recommended official method (Ce 1h-05) for the analysis of the fatty
acid composition of fats containing TFA (AOCS, 2005). Figures 2 and 3 show
the 18:1 and 18:2 FAME regions of the chromatogram for a hard margarine fat
sample analysed on 100 m SP-2560 and CP-Sil 88 capillary columns, respec-
tively. Note that the 12t-18:1 peak is readily resolved from the oleic acid peak
(9c-18:1). In addition, the 13t-18:1 and 14t-18:1 isomer pair, which always
18:0
9c-18:1
10t-18:1
11t-18:1
18:2n-6
(13+14)t +(6-8)c-18:1
10c+15t-18:1
6t-8t-18:1
9t-18:1
11c-18:1
12c-18:1
9c,13t+8t,12c-18:2
12t-18:1
14c-18:1
15c-18:1
16t-18:1
9c,12t-18:2
9c,15c-18:2
9t,12c-18:2
9t,12t-18:2
13c-18:1
4t-18:1
5t-18:1
tt-18:2
24 26 28 30 min
Figure 2. The C18 region of gas chromatogram of FAME from a margarine sample analysed on a SP
2560 fused silica capillary column (100 m × 0.25 mm id × 20 m film thickness), operated isothermally
at 180°C. Hydrogen carrier gas, flow rate 1 ml/min.
elute together in any GC analysis, is sufficiently separated from 9c-18:1 to be

quantified. As shown in Figure 2, there should always be a distinct valley
between the 13t +14t-18:1 peak and 9c-18:1 peak, though it does not need to be
a large valley to get correct results. Furthermore, the 16t-18:1 isomer is also
separated from the cis-18:1 isomers. Usually, the 15t-18:1 is the only trans-
18:1 isomer that remains unresolved. Quantitatively, this is not an important
drawback, since 15t-18:1 is a minor dietary component. Its level very often is
less than 2% of total trans-18:1 isomers (Table 1). It should be noted here that
the peak for 13t+14t-18:1 contains three cis-18:1 isomers, namely 6c-18:1, 7c-
18:1 and 8c-18:1. These overlaps are of minor importance in many partially
hydrogenated oils, where the total amount of these three isomers usually
account for less than 2% of total fatty acids. However, these isomers are
important in some partially hydrogenated oils, for example the heavily hydro-
genated canola oil base stocks (Table 1). For such samples, the complete fatty
acid composition needs to be determined by combining GC analysis with
AgNO3-TLC or Ag+-HPLC (described in Section A.5).
The column operating temperature has a considerable influence on the cis-
trans isomer separation (Ratnayake et al., 2002). As discussed above, with
both 100 m SP-2560 and CP-Sil 88 capillary columns, isothermal operation at
180°C produced the fewest overlapping peaks of the cis and trans isomers.
Isothermal operations of the column above or below 180°C, or column
counts
CP-S11 88 (100 m × 0.25 mm)
14000 180°C isothermal
18:0 9
18:2N-6
12000
10000 10
11
12
8000 6 13
5 21
14
8 26 18:3N-3
15 20
16 19 25
6000 4 7 27
17 24
3
23 29
4000 18 34+20:1
22 20:0 32
1 2 28 30 31 33
2000
24 26 28 30 32 34 36 38 40 min
Figure 3. The C18 region of a GC chromatogram of FAME from a margarine sample analysed using a
CP Sil-88 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
180°C. Hydrogen carrier gas, flow rate 1 ml/min. Peak identification: 1, 4t-18:1; 2, 5t-18:1; 3, (6t-8t)-
18:1; 4, 9t-18:1; 5, 10t-18:1; 6, 11t-18:1; 7,12t-18:1; 8, (13t+14t)-18:1 + (6c-8c)-18:1; 9, 9c-18:1; 10,
(15t+10c)-18:1; 11, 11c-18:1; 12, 12c-18:1; 13, 13c-18:1; 14, 16t-18:1; 15,14c-18:1; 16, 15c-18:1; 17-
20, tt-18:2; 21, 9t,12t-18:2; 22, (9c,13t+8t,12c)-18:2; 23, 9c,12t-18:2; 24, 8c,13c-18:2; 25, 16c-18:1; 26,
9t,12c-18:2; 27, (9t,15c+10t,15c)-18:2; 28, 9c,13c-18:2; 29, 9c,14c-18:2; 30, 9c,15c-18:2; 31, 9t,12c,15t-
18:3; 32, 9c,12c,15t-18:3; 33, 9c,12t,15c-18:3, 34; 9t,12c,15c-18:3 (Ratnayake, 2004) (Reproduced with
the kind permission of J. Assoc. Off. Analyt. Chem. Int.).
temperature programming, greatly affects the cis-trans isomer separation and

the relative elution order of some of the 18:1 isomers, the 11c-20:1, as well as
the geometric isomers of α-linolenic acid. For instance, isothermal operation at
190°C results in overlap of 16t-18:1 isomer with the 13c-18:1 isomer peak
(peaks 13 and 14 in Figures 4 and 5). A minor resulting advantage is the
improved separation of the isomer pair 13t+14t-18:1 from the oleic acid peak,
but the peak for 13t+14t-18:1 still contains the 6-8c-18:1 peaks.
Operating the column temperature below 180°C also creates a few extra
overlap problems. For example, with isothermal operation at 170°C, as shown
in Figures 6 and 7, the resolution of 13t+14t-18:1 (peak 8) from oleic acid (peak
9) is lost; there is no distinct valley between these two peaks and therefore,
accurate measurement of the areas of these two peaks is difficult. In addition,
12000 SP-2560 (100 m û 0.25 mm)

190°C isothermal
18:0
18:2n-6
10000 9 + 10
8000 25
5
6 18:3n-3
12
6000 4 8 13+14 19
7 11 18 26
3
20 25 27 34+11c-20:1
17 30
4000 16 24
21 28 29 20:0
1 2 15 23 31 32 33
22
2000
20 22 24 26 28 min
Retention Time (min)
SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
190°C. Hydrogen carrier gas flow rate 1 ml/min. For peak identifications see Figure 3. (Ratnayake, 2004)
(Reproduced with the kind permission of J. Assoc. Off. Analyt. Chem. Int.).
the 16t-18:1 (peak 14) now overlaps with the 14c-18:1 isomer (peak 15). The
only advantage at lower operating temperature is the improved separation of
the individual trans-18:1 isomers compared to that at 180°C or other higher
temperatures. These overlap problems are also seen with column temperature
programme analysis (Figure 8).
b. Capillary GC separation of isomers of linoleic and linolenic isomers

The best separation of the various isomers of 18:2n–6 and 18:3n–3 present in
both refined, unhydrogenated liquid vegetable oils (see Figure 9) and PHVO
(see Figure 10) is obtained when the column temperature is operated isother-
mally at 180°C. In general, the 18:2 isomers elute in the order trans,trans <
trans,cis < cis,trans followed by cis,cis (Figures 9 and 10). The 9c,13t-18:2,
which is the major 18:2 isomer formed during partial hydrogenation of vegeta-
ble oils (Ratnayake & Pelletier, 1992), has a retention time close to that of
9t,12t-18:2 and in some GC analyses, especially at column temperatures above
CP-Sil 88 (100 m × 0.25 mm)

190°C isothermal
14000
18:2n-6
18:0 9+10
12000
11
10000
5 6 12 18:3n-3
13+14
8000
4 15
7 16 28
3 8 17 30
6000 27 29
18 20 26
25 34
19
11c-20:1
4000 21 24
2 23 20:0 31
1 22 32 33
2000
20 22 24 26 28 30 min
Retention time (min)
CP Sil 88 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
counts SP-2560 (100 m × 0.25 mm)

18:0 170°C isothermal
7000 18:2n-6
6000
5 18:3n-3 + 11c-20:1
6
4 10
5000 8 14+15 21
11 20
7
19
12 18
3 17 26
4000 13 25
16 24 27 20:0 32 33
22 28 29 30 31 34
1 2 23
3000
30 32.5 35 37.5 40 42.5 45 47.5 50 52.5 min

SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
CP-Sil 88 (100 × 0.25 mm)

170°C isothermal
18:0
9000
18:2n-6
9
8000
7000
6000 18:3n-3
56
5000 8 10
4 21
7 11 20 11c-20:1
14+15
3 12 19 28
4000 18 25
13 17 27
24
30
16 23 26
3000 34
12 29 20:0
31
32
22 33
2000
35 40 45 50 55 min
CP Sil 88 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
(13+14)t-18:1+ (6-8)c-18:1
12c-18:1 10c-18:1+15t-18:1
9c-18:1
10t-18:1
FID1
11t-18:1
PA
11c-18:1
18:0
9t-18:1
20
6t-8t-18:1
18
12t-18:1
16
14
14c-18:1
16t-18:1
15c-18:1
12
13c-18:1
10
5t-18:1
4t-18:1
8
6
4
38 40 42 44
SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness). Analysis performed by the
following column temperature program. Initial temperature 45°C (hold 4 min), programmed to 175°C at
13°C/min, hold at 175°C for 27 min, programmed to 215°C at 4°C/min, hold at 215°C for 3 min.
Hydrogen carrier gas flow rate 1 ml/min.
18:2n-6
9c-18:1
11c-18:1
9c,12t-18:2
9t,12t-18:2
9t,12c-18:2
13t-18:1
25 30 min
min
Figure 9. The 18:2 region of a GC chromatogram of FAME from a refined canola oil sample analysed
using a SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally
at 180°C. Hydrogen carrier gas flow rate 1 ml/min.
9c,13t-18:2 (major) + 8t,12c-18:2 (minor)
13c-18:1
18:2n-6
9t,15c+10t,15c-18:2
20:0
14c-18:1
9c,12t-18:2
16t-18:1
16c-18:2
9c,13c-18:2
15c-18:1
9t,12c-18:2
9c,14c-18:2
8c,13c-18:2
9c,15c-18.2
8c,13t-18:2
tt-18:2
ct/tc-18:2
9t,12t-18:2
28 30 32 min
Figure 10. The 18:2 region of a GC chromatogram of FAME from a margarine sample analysed using
a SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally at
180°C. Hydrogen carrier gas flow rate 1 ml/min.
18:3n-3
20:0
9c,12c,15t-18:3
9t,12c,15c-18:3
9c,12t,15c-18:3
11c-20:1
9t,12c,15t-18:3
32.5 35 37.5 min
Figure 11. The 18:3 region of a GC chromatogram of FAME from a refined canola oil sample analysed
using a SP 2560 capillary column (100 m × 0.25 mm i.d. × 20 µm film thickness), operated isothermally
at 180°C. Hydrogen carrier gas flow rate 1 ml/min.
190°C or with column temperature programs, these two isomers are not well
resolved. In addition, the unavailability of a commercial 9c,13t-18:2 standard
means misidentification of this isomer is common in such instances. It is
recommended that a FAME mixture from a PHVO with a well-established fatty
acid composition be used as the secondary standard, especially for the identi-
fication of the peaks in the 18:2 region of the chromatogram. Such a secondary
standard mixture is now available from the American Oil Chemists’ Society
(Champaign, Illinois, USA).
The four common geometric isomers of α-linolenic acid, i.e., 9t,12c,15t-
18:3; 9c,12c,15t-18:3; 9c,12t,15c-18:3 and 9t,12c,15c-18:3, which are also
commonly found in refined liquid vegetable oils give peaks that can be readily
recognized in GC analyses on cyanosilicone capillary columns and are eluted
from the column in the order stated above (Figure 11).
c. Capillary GC resolution of eicosenoic and α-linolenic acid isomers

The 11c-eicosenoic acid (11c-20:1) is a natural monounsaturated fatty acid
present in some animal fats (for example, lard) and vegetable oils such as high-
erucic rapeseed oil, peanut oil and canola oil. This C20-monounsaturated fatty
acid elutes in the 18:3 region of the chromatogram and its relative retention
time with respect to the geometrical isomers of 18:3 varies with the column
temperature. Generally, on the SP-2560 capillary column, as for the 18:1 cis
and trans isomer separation, isothermal operation at 180°C gives the best
separation of 11c-20:1 from α-linolenic acid and its geometric isomers
(Ratnayake et al., 2002) (Figure 11). At column temperature above 180°C, the
11c-20:1 isomer co-elutes with 9t,12c,15c-18:3 (Figure 4) and below 180°C it
overlaps with α-linolenic acid (Figure 6). In contrast, the CP-Sil 88 column
operated at 180°C does not cleanly separate 11c-20:1 from the α-linolenic
isomers, but co-elutes it with 9t,12c,15c-18:3 (Figure 3). However, isothermal
operations of CP-Sil 88 column at 10°C above or below 180°C give baseline
separation of 11c-20:1 from α-linolenic acid and its geometrical isomers. At
190°C, 11c-20:1 elutes before 9t,12c,15c-18:3 and 9c,12t,15c-18:3 isomers
(Figure 5), whereas at 170°C, it elutes after 9t,12c,15c-18:3 but before
α-linolenic acid (Figure 7). It is recommended that pure standards of 11c-20:1
and 18:3n–3 be used for correct location of these fatty acids in the analysis of
dietary fat samples.
d. Effect of the type of GC carrier gas and flow rate on cis and trans
isomer resolution and fatty acid quantification
Recently, Ratnayake et al. (2006) tested the performances of hydrogen and
helium as carriers at different flow rates on the separation and quantification of
FAME of dietary fats containing TFA. Their data suggested that hydrogen
carrier gas at 1.0 ml/min and isothermal column temperature operation at
180°C are optimal for the analysis of fatty acid profiles of vegetable and animal
fats and oils containing TFA.
Table 4 shows the normalized area percentages concentration for total trans,
trans-18:1, trans-18:2, trans-18:3, conjugated linoleic acid (CLA), 16:0 and
18:0 of a margarine sample when analysed on different columns (100 m SP-
2560 and CP-Sil 88) with different carrier gases (hydrogen and helium) and
flow rates (0.6, 0.8 and 1.0 ml/min). The results were essentially identical at all
different conditions for both columns. Even though the quantification differ-
ences were minimal, the data demonstrate a few advantages of using hydrogen
as the carrier gas at a flow rate of 1 ml/min. The use of hydrogen at the same
flow rate as helium results in shortening the GC run time by 20 minutes. For
example, on the 100 m SP-2560 column with hydrogen carrier and a flow rate
of 1 ml/min, the 24:0 peak, usually the last to elute in many dietary fats, comes
off in about 60 minutes. With helium as the carrier gas at the same flow rate, the
same peak elutes at over 80 minutes. This time-saving could be of benefit to
most lipid analytical laboratories where run time is very important. In addition,
hydrogen carrier gas gives a better resolution of the critical pairs of 13t+14t-
18:1 and 9c-18:1, 16t-18:1 and 14c-18:1 and of 11c-20:1 and 9c,12c,15c-18:3.
Furthermore, hydrogen produced sharper peaks than helium. It was also noted
that there was no loss of resolution between flow rates of 0.6, 0.8 and 1.0 ml/
min in the early-eluting peaks for either hydrogen or helium (Ratnayake et al.,
Table 4. Effect of helium and hydrogen carrier gas flow rates on the fatty acid
composition (% total fatty acids) of a partially hydrogenated margarine sample on 100 m
CP-Sil 88 and SP-2560 columns (Ratnayake et al. 2006). CLA, conjugated linoleic acids.
Fatty acid CP-Sil 88 SP-2560

H2 H2 H2 He H2 He
0.6 ml/ 0.8 ml/ 1.0 ml/ 1.0 ml/ 1.0 ml/ 1.0 ml/
min min min min min min
Total trans 34.41 34.33 34.37 34.30 33.98 34.09

18:1 trans 28.06 28.01 28.04 27.91 27.92 27.85
18:2 trans 6.09 6.07 6.08 6.09 5.87 6.02
18:3 trans 0.26 0.25 0.25 0.30 0.19 0.22
CLA 0.55 0.51 0.56 0.54 0.32 0.41
16:0 11.79 11.76 11.78 11.65 11.91 11.96
18:0 11.52 11.51 11.51 11.08 10.74 10.98
2006). However, for the separation of the peaks in the 18:3 region, a flow rate
1.0 ml/min provided the most satisfactory separation.
5. Analysis of trans isomers by capillary GC in combination with silver-ion

chromatography
As previously discussed, it is very clear that direct GC analysis using 100 m SP-
2560 or CP-Sil 88 capillary columns gives adequate separation of almost all the
cis/trans-18:1 isomers encountered in partially hydrogenated vegetable oils.
The only drawback is the minor overlaps of 15t-18:1 with 9c-18:1 or 10c-18:1
isomers, and 13t+14t-18:1 with 6c+7c+8c-18:1. Quantitative data on these
isomers can be obtained by using GC analysis in combination with fractionation
of the cis-and trans-18:1 isomers by preparative AgNO3-TLC or Ag+-HPLC.
Both are very powerful techniques which fractionate the fatty acids (as FAME)
on the basis of the number, the configuration and to some extent, the positions
of the double bonds. The fractionation of fatty acids in silver ion chromatogra-
phy is based primarily on the ability of the carbon-carbon double bonds to form
a complex with the Ag+. Since saturated fatty acids do not have carbon-carbon
double bonds, they are less retained than unsaturated fatty acids. Within the
unsaturated fatty acids, the highly unsaturated fatty acids are more retained
than less unsaturated fatty acids. Cis isomers form a stronger Ag+/double bond
complex than trans isomers and therefore the cis isomers are retained more
strongly than the trans isomers. Furthermore, polyunsaturated fatty acids with
the double bonds separated by more than two methylene units (CH2, i.e, non-
methylene-interrupted fatty acids) are more strongly retained than the
corresponding polyunsaturated fatty acids with double bonds separated by one
CH2 unit (i.e., methylene interrupted fatty acids). In this way saturated,
monounsaturated, methylene-interrupted polyunsaturated fatty acids and non-
methylene-interrupted polyunsaturated fatty acids as well as their cis and trans

isomers can be separated. Isolation and subsequent analyses of these fractions
by capillary GC allow complete and accurate determination of the entire fatty
acid profile including the cis and trans isomers. Comprehensive reviews on
silver ion chromatography and theory have been recently written by Nikolova-
Damyanova (2003) and Adlof & List (2007). In this section of the chapter we
will describe the application of AgNO3-TLC and Ag+-HPLC for analysis of
trans-18:1 fatty acids.
a. Silver nitrate thin layer chromatography

An important step in the AgNO3-TLC procedure is the preparation of a clean
TLC plate with a uniform layer of silver nitrate. This is easily achieved using
commercial pre-coated silica gel preparative TLC plates (20 cm × 20 cm,
thickness 0.5 mm) (Ratnayake, 2004, Sebedio & Ratnayake, 2007). The proce-
dure is as follows: commercial plates are first cleaned of dust particles by
developing in a TLC tank containing either ethyl acetate or chloroform/
methanol (1:1, v/v). The cleaned plate is then activated in an oven at 105–
110°C for one hour. After allowing the plate to come to room temperature, the
plate is placed horizontally (the side containing the silica gel facing down-
wards) in a glass tray containing 5% solution of silver nitrate in acetonitrile for
30 minutes. The plate is activated again by heating in an oven at 110°C for
about one hour. It is advised to use the plate immediately after its preparation;
if not it should be stored in a dessicator over drying agents in a dark place.
For separation of the various fatty acid classes without any band overlaps,
overloading of the plate with FAME sample should be avoided. Ideally about
4–5 mg of the FAME sample produces good separation with minimal overlaps.
The procedure used in the authors’ laboratory is as follows: the FAME sample
(dissolved in hexane) is applied to the plate using a TLC streaker or a
disposable glass pipette and developed using either 100% toluene or a mixture
of hexane and diethyl ether (90:10, vol/vol). The plates are normally developed
at room temperature in a dark place to minimize oxidation of highly unsaturated
fatty acids. The separated bands are made visible by spraying the plate with a
0.1% solution of 2',7'-dichlorofluorescein in ethanol and examining under UV
light (234 nm). Figure 12 shows a typical separation of a FAME mixture from
a margarine made from partially hydrogenated canola oil. The various bands
are scraped off and extracted with diethyl ether or a 1:1 mixture of hexane-
chloroform and then analysed by GC using 100 m SP-2560 or CP-Sil 88
column.
AgNO3-TLC gives a clean separation of the trans-18:1 isomers from the cis-
18:1 isomers and other fatty acids (Figure 12). Isolation and analyses of these
two bands by GC permit quantification of the levels of trans- and cis-18:1
isomers. Figure 13 shows the GC traces of the trans-and cis-18:1 fractions. In
practice, a convenient means of quantification of the trans-18:1 fraction is to
Figure 12. AgNO3-TLC fractionation of FAME from a margarine. Development in toluene. Bands
were visualized after spraying with 0.1% solution of 2',7'-dichlorofluorescein in ethanol and examining
under UV light (234 nm).
treat the trans-18:1 isomers with the double bond closer to the carboxyl group
(from Δ6 to Δ11) as the internal standard. This is possible because, as discussed
earlier, the 6t-18:1 to 11t-18:1 isomers are always well resolved from the cis-
18:1 isomers. In this method, the proportion of the trans-18:1 isomers from Δ12
to Δ16 that overlap with the cis-18:1 isomer peak on capillary GC is calculated
by comparing the 18:1 region of the GC chromatogram of the isolated trans-
18:1 with that of the parent FAME prior to AgNO3-TLC fractionation. This
calculation is done with respect to the well-separated trans-18:1 isomers (i.e.,
sum of 6t-18:1 to 11t-18:1). The total trans-18:1 content is then calculated by
summing the proportion of the trans-18:1 isomers (12t-18:1 to 16t-18:1) that
overlaps with the cis isomers with the well-separated trans-18:1 isomers (from
6t-18:1 to 11t-18:1). This also allows calculation of the total cis-18:1 content.
This approach eliminates the errors resulting from sample application, scraping
losses, incomplete extraction, weighing of small quantities of internal standard
and isolated bands. The AgNO3-TLC fractionation offers several other advan-
tages, especially the separation of the linoleic acid isomers into trans,trans and
cis,trans fractions (see Figure 12) that can be used to provide confirmatory
evidence for the identification of the 18:2 isomers. Ratnayake & Pelletier
(1992) used AgNO3-TLC fractionation in combination with GC mass
spectrometry (GC-MS) to determine the structures of linoleic acid isomers of
PHVO.
As an alternative to glass plates, pre-coated plastic-backed sheets with a
0.2 mm layer of silica gel impregnated with 10% AgNO3 in acetonitrile can be
10t 11t 18:1 trans fraction
12t 13t+14t
9t
6t+7t+8t
18:0 15t
16t
4t 5t
24 26 min
9c 18:1 cis fraction
11c
10c
13c
6c+7c+8c 15c
12c 14c
24 26 28 min
Figure 13. GC chromatograms of the trans- and cis-18:1 AgNO3-TLC bands of the margarine sample
shown in Figure 13. Analysis on a SP 2560 fused silica capillary column (100 m × 0.25 mm id × 20 µm
film thickness), operated isothermally at 180°C. Hydrogen carrier gas flow rate 1 ml/min.
used (Uberth & Henninger, 1992). The development of the sheet is similar to
that of glass plates with n-hexane-diethyl ether (9:1, v/v), which is followed by
spraying with 0.05% rhodamine B in ethanol. The separated bands are cut off
with a pair of scissors, and the cuttings are placed in a small bottle containing
5 ml diethyl ether. Extraction is complete in 30 minutes. These AgNO3-sheets
give separation of trans and cis isomers similar to that of conventional AgNO3-
TLC fractionation.
b. Silver ion high performance liquid chromatography

Ag+-HPLC is another powerful technique for the isolation of trans and cis fatty
acid isomers, especially 18:1 isomers before GC analysis. The chromatography
is usually performed using commercial silver ion HPLC column (for example,
ChromSpher Lipids Column, 250 mm × 4.6 mm i.d., stainless steel, 5 µm
particle size), an isocratic solvent system composed of acetonitrile in hexane
and UV detection (Adlof & List, 2007). A typical chromatogram for the FAME
from a partially hydrogenated vegetable oil sample is shown in Figure 14
Figure 14. Analysis of partially hydrogenated vegetable oil FAME by Ag+-HPLC. Sample size, 20 µg;
flow rate, 1.0 ml/min; solvent system, 0.15% acetonitrile in hexane; ultraviolet detection at 206 nm.
Fractions: A, saturates; B, trans-18; C, cis-18:1; D, 18:2 (Adlof et al. 1995). (Reproduced with kind
permission of J. Am. Oil Chem. Soc.).
(Adlof et al., 1995). The separation was performed with a solvent system of
0.15% acetonitrile in hexane at a flow rate of 1 ml/min and UV detection at
204 nm. The separation is completed in 45 minutes. The saturated, trans-18:1
isomers, cis-18:1 isomers and the 18:2 were completely separated without any
overlaps. Collection of the trans-18:1 isomers with the saturated fatty acid
fraction and subsequent analysis by GC would allow quantitation of all the
trans-18:1 isomers with respect to the total saturated fatty acids or just with
respect to methyl stearate. By using 0.08% acetonitrile in hexane and injecting
a very small sample size (< 0.5 µg), individual cis and trans-positional FAME
isomers can be resolved (Figure 15, Adlof et al., 1995). The elution order of the
positional isomers is the reverse of those obtained by capillary GC. The high
delta value isomers elute before the low delta value isomers in Ag-HPLC.
The Ag+-HPLC method offers some advantages over AgNO3-TLC, includ-
ing rapid analysis (45 min/run) and complete separation of cis- and trans-18:1
isomers with no cross contamination. However, a severe draw back is the
difficulty of obtaining consistent separation pattern from one day to another.
Δ8 + Δ9
Δ10
Δ11
Δ8 + Δ9
Δ12
Δ13 Δ10
Δ11
Δ12
Δ14
Δ13
Figure 15. Analysis of partially hydrogenated vegetable oil FAME by Ag+-HPLC. Sample size,
0.4 µg; flow rate, 1.0 ml/min 0.08% acetonitrile in hexane; ultraviolet detection at 206 nm. Fractions: A,
saturates; B, trans-18:1; C, cis-18:1 (Adlof et al., 1995). (Reproduced by permission of Journal of the
American Oil Chemists’ Society).
This might be due to the sensitivity of resolution of FAME with slight

variations in the solvent system. Adlof & List (2007) have noted that the
presence of water can significantly affect resolution and they recommended
that solvents should be anhydrous or, if laboratory conditions are humid, stored
over drying agents.
6. GC-MS analysis of fatty acids

A major challenge in the GC analysis of fatty acid profile has been the correct
identification of all the fatty acid peaks. Since almost all the fatty acids found
in the human diet have already been characterized and since it is rare to find any
novel dietary fatty acids nowadays, the GC peak identification for routine
analysis is primarily based on comparison of the GC retention times with
commercially-available standards. Alternatively, peaks could be identified by
comparing the fatty acid profiles with those published in the literature. Some
official methods, for example the AOCS Official Method Ce-1h 05 (AOCS,
Figure 16. Mass spectra of Δ8, Δ9 and Δ10 trans 18:1 DMOX positional isomers; a gap of 12 amu was
observed between the pairs of adjacent ions at m/z 182 and 194; 196 and 208; and 210 and 222,
respectively (Mossoba et al. 1997). (Reproduced by permission of J. Am. Oil Chem. Soc.).
Figure 17. Mass spectra of Δ13 and Δ14 trans 18:1 DMOX positional isomers (Mossoba et al. 1997).
(Reproduced by permission of J. Am. Oil Chem. Soc.).
2005), have figures of GC FAME profile of representative dietary fats and

these are also useful for peak identifications. Nevertheless, it is important to
occasionally confirm the peak identifications. This could be easily achieved by
GC-MS analysis.
Fatty acids are generally analysed by GC as their methyl esters, but unfortu-
nately unsaturated FAME suffer rearrangement of the double bonds under
electron impact ionization conditions and give mass spectra with extensive
fragmentation which cannot be interpreted to locate the original double bond
positions. This problem can be overcome if the fatty acids are first transformed
into a nitrogen-based derivative, which stabilizes some fragments related to the
bond position (Christie et al., 2007). Among the different derivatives, dimethyl
oxazoline (DMOX) has been used extensively as these derivatives give spectra
that are easier to interpret. Furthermore, the chromatographic resolution of the
DMOX derivatives of isomeric fatty acids is similar to that for FAME on the
same cyanopropyl polysiloxane polar capillary GC columns.
The application of GC/MS analysis of DMOX derivatives for the identifica-
tion of the position of double bond of trans-18:1 positional isomers of a
partially hydrogenated vegetable oil was demonstrated by Mossoba et al.
(1997). In their analytical procedure, the trans-18:1 fraction (as FAME) was
isolated by Ag+-HPLC and then converted to DMOX derivative using 2-amino-
2-methyl-1-propanol. The DMOX derivatives were analysed by GC-MS in the
electron ionization mode. A mass interval of 12 atomic mass units (amu)
instead of 14 amu between two neighbouring even mass homologous frag-
ments containing n–1 and n carbon atoms in the acid moiety indicates a double
bond between n and n+1 carbons in the chain. The mass spectral data of the
DMOX derivatives (Figures 16 and 17) confirmed the double position for nine
individual trans 18:1 positional isomers, each having a double bond at a carbon
located between C8 and C16. The double bond positions of trans-polyunsatu-
rated fatty acids have also been determined by GC-MS analysis of DMOX
derivatives (Ratnayake et al., 1994).
C. Analysis of ruminant fats
1. Ruminant Fats
Cow’s milk usually contains 3.4 to 5.1% fat (Jensen, 2002). The lipid fraction
is represented by 96 to 98% triacylglycerol. Minor lipid components in milk
include diacylglycerols, monoacylglycerols, phospholipids, free fatty acids,
sterols and sterol esters (Gurr, 2002). Triacylglycerols are present in the form
of small globules surrounded by a thin membrane that is composed of protein
and phospholipids. More than 400 different fatty acids have been identified
(Jensen et al., 1991), with the potential to create over 6 million triacylglycerol
molecular species. This makes the lipid fraction of milk one of the most
complex matrices known to food science (Jensen, 2002). Milk fatty acids
include saturated fatty acids of different chain-length groups, namely short-
chain, medium-chain, long-chain, and branch-chain fatty acids, and also a
variety of cis-and trans-monounsaturated FA, polyunsaturated FA, geometric
and positional trans-polyunsaturated fatty acids isomers, CLA, hydroxy fatty
acids and cyclic fatty acids (Christie, 1995; Walstra, 1999). Although only 14
fatty acids – including 14:0, 16:0 and 18:1 – represent about 96% of the total
fatty acid content, milk fat also has an exceptionally high content of short-chain
fatty acids, especially butyric acid (4:0) which makes milk fat a distinctive food
product (Christie, 1995; Gurr, 2002). Butyric acid is considered to play a role
in cancer prevention (German, 1999). It is also of considerable commercial
importance as a raw material in the manufacture of esters of lower alcohols
which are used as flavouring agents. Another minor, but very important group
of fatty acids present in ruminant fats is the CLA isomers. Studies conducted
over the last 10 years, especially with small laboratory animals, have shown
that CLA isomers, in particular rumenic acid (9c,11t-18:2), can provide many
health benefits including anti-carcinogenic, anti-obesogenic, and anti-athero-
genic effects (Beppu et al., 2006; Bhattacharya et al., 2006; De La Torre et al.
2006; Belury et al., 2002).
The fatty acid profile of dairy fats depends to a large extent on the diet of the
cow, the vast microflora activity present in the rumen, and the cellular fat
synthesis (Chilliard et al., 2001; 2000). If diets contain a high content of
polyunsaturated fatty acids protected by protein, the final product (milk or
meat) will also contain them (Beauleiu & Palmquist, 1995). Fatty acids such as
trans-18:1 are produced as a result of incomplete bio-hydrogenation of dietary
polyunsaturated fatty acids in the rumen. It has recently been demonstrated that
feeding of cows with grain diets results in milk and meat with unusually high
levels on trans-18:1 fatty acids (Bauman & Griinari, 2003).
Over the last 25 years, a number of good analytical methods have appeared
in the literature for the analysis of fatty acid profile in dietary fats (Albertyn
et al., 1982; Precht & Molkentin 1996; Ratnayake, 1998; Kramer et al., 1999;
Precht & Molkentin, 1999; 2000; Precht et al., 2001; Ratnayake, 2001; Wolff
et al., 2001; Destaillats & Angers, 2002; LeDoux et al., 2002; Wolff and
Precht, 2002; Christie, 2003; Ratnayake, 2004; and Aldai et al., 2005). Some of
these methods, however, are not suitable for analysis of ruminant fats. Because
of the presence of a very complex mixture of a large number of fatty acids, the
analysis of ruminant fats requires special analytical procedures. The objective
of this section is to propose a comprehensive methodology to completely
characterize the fatty acid profile of ruminant fats with enough accuracy and
precision to detect all the TFA and CLA isomers. The methodology is a
combination of complementary techniques, including lipid extraction, meth-
ylation of the extracted lipids to FAME and subsequent analysis of FAME
using various forms of chromatography such as Ag+-TLC, Ag+-HPLC and GC.
2. Lipid extraction
There are certain basic principles and precautions that should be followed
during extraction of lipids from dairy products and ruminant meat. The
extraction should ideally be carried out using a solvent system that yields both
neutral and polar lipids. Since CLA is naturally present in both meat and dairy
products, it is also important to avoid acid digestion, which can lead to
isomerization of CLA isomers. When cheese is analysed, samples should be
entirely ground prior to lipid extraction in order to ensure a complete solvent
interaction with the sample. Butter should be melted, homogenized and sub-
sampled before extraction.
Total lipids can be extracted in two stages from dairy samples using a
chloroform/methanol/water mixture (Bligh & Dyer, 1959). Samples are first
well homogenized for 2 minutes in the mono-phase mixture chloroform/
methanol/water (1:2:0.8; by vol.), taking into account the water that is already
present in milk (Kramer et al., 2001). For this reason, adding water last when
preparing the mono- and biphasic systems is suggested. A solvent to sample
ratio greater than 15:1 is recommended to ensure better recovery (Kramer
et al., 2001; Cruz-Hernandez et al., 2006). After the addition of chloroform/
water (1:1; v/v) samples are again homogenized in the bi-phase system
(chloroform/methanol/water; 2:2:1.8, by vol.) for an additional 2 minutes. The
lipid fraction (including neutral and polar lipids) will be partitioned into the
chloroform phase, which is at the bottom of the flask. The lipid/chloroform
phase is collected and the lipids are then recovered by removing the chloroform
using a rotary evaporator as described by Cruz-Hernandez et al. (2006). In
order to obtain a dry lipid sample, the addition of a couple of drops of acetone
is recommended at the end of the extraction to completely remove all traces of
water. Small amounts of solvent can be removed by flushing with nitrogen and
working in a fume hood. Finally, the total lipid content is determined by weight
differences, and lipids should be stored in chloroform at very low temperatures
(preferably below –20°C) until further analysis.
For analysis of meat, the fresh sample should be rinsed immediately with
water and frozen (either on dry ice or dropped into liquid N2), and stored at
–70°C until analysed. Sample preparation is slightly different from dairy
samples. Meat samples should be pulverized at dry ice temperature in order to
restrain the action of lipases and phospholipases during homogenization of
tissues (Kramer and Hulan, 1978, Kramer et al., 1999). Briefly, the stored
frozen meat at –70°C (2 to 5 g) is directly pulverized using a stainless steel
mortar and pestle kept at dry ice temperature. The content is then transferred to
a beaker containing 30 ml chloroform/methanol (2:1) at room temperature. The
mixture is gently homogenized, allowed to stand for one hour, and then filtered
through a sintered glass funnel. Subsequently, the residue is washed two more
times with chloroform/methanol (2:1). The combined solvents are removed
using a rotary evaporator as described above to obtain total lipids. This same
procedure is also recommended for the extraction of lipids from tissues of
laboratory animals.
Other solvent systems recommended for matrices such as milk or tissues are
those described by Folch et al. (1957). These systems utilize different ratios of
chloroform/methanol/ water or alternatively use mixtures containing either
hexane/isopropanol (Hara & Radin, 1978) or dichloromethane/methanol
(Marmer & Maxwell, 1981; Maxwell et al., 1986).
In meat fat, some lipid classes such as phospholipids and other individual
lipid classes are also important to consider. Analyses of individual lipid classes
could be helpful in understanding different functional properties of meat fat
(Kramer et al.,1998, Yurawecz et al., 1999, Cruz-Hernandez et al., 2006).
3. Methylation
Although base methylations are believed to be the best derivatization method
for the analysis of CLA isomers, other lipids present in the matrix must be taken
into consideration. Base catalysts specifically transesterify acyl lipids, but they
are not capable of transesterifying N-acyl and alk-1-enyl acyl lipids (Kramer &
Zhou 2001; Cruz-Hernandez et al., 2004; Kramer et al., 2004). For example, if
free fatty acids are also present and need to be analysed, then the methylation
has to be carried out using an acid catalyst such as HCl/methanol (Kramer
et al., 2001; Cruz-Hernandez et al., 2006).
Methylation using sodium methoxide, NaOCH3 (e.g., 0.5N methanolic base
#33080, Supelco Inc, Bellefonte, PA, USA) is recommended for the analysis of
CLA-containing lipids because the methylation is complete within 15 to
20 minutes and this reagent does not isomerize CLA isomers (Kramer et al.,
1997). A water-free system is preferred for the methylation of milk fat lipids
(Christie, 1999; Chouinard et al., 1999). Briefly, 2 mg of total fat is added to a
2 ml autosampler vial, the solvent is removed with a stream of N2, 1.7 ml of
hexane and 40 µl of methyl acetate are added and thoroughly mixed before the
addition of 100 µl of NaOCH3. The vial is securely capped, the contents shaken
and allowed to react for 20 minutes at room temperature with occasional
mixing. The vial is cooled at –20°C for 10 minutes, followed by addition of
60 µl of oxalic acid (0.5 g in 15 ml diethyl ether) and thoroughly shaken. The
vial is centrifuged to settle the Na oxalate precipitate, and the upper phase is
passed through a Pasteur pipette (5.75 inch) column containing a glass wool
plug (glass wool previously washed with 1:1 chloroform/methanol and dried),
and a 2 cm bed of anhydrous Na2SO4, directly into a 2 ml autosampler vial. The
FAME solution can be used directly for GC analyses.
Reagents other than NaOCH3, such as TMS-DAM (trimethylsilyl diazo-
methane), can also be used for methylation of lipids containing CLA isomers.
This reagent does not isomerize CLA. However, the primary drawback is that
very often this reagent is not pure and, consequently, the FAME produced
should be purified by TLC, which is a time-consuming procedure. When
working with dairy products, purification of the FAME leads to loss of some
fatty acids (Kramer et al., 2001).
4. Capillary GC of FAME from dairy and meat fat

As for the PHVO, 100 m capillary columns coated with cyanosilicone liquid
separate and resolve most of the total FAME from milk fat. Complementary
techniques are essential to resolve some FAME.
Temperature programs from 45 to 215°C for GC analysis have resulted in
good separation and resolution of ruminant fat FAME (Kramer et al., 2001;
Cruz-Hernandez et al., 2006). Identification of many fatty acids can be based
on the comparison of retention times with commercial standards. Since com-
mercial standards are not available for all the fatty acids, it is necessary to
prepare some standards in the laboratory or make tentative identifications. The
short-chain fatty acids present in dairy fats can be corrected for mass discrep-
ancy using the correction factors published by Wolff et al. (1995). The method
that we routinely use for milk and meat lipid analysis has been described
elsewhere (Cruz-Hernandez et al., 2006). Briefly, we use a gas chromatograph
equipped with a splitless injection port, a flame ionization detector (FID), and
fitted with either a 100 m × 0.25 mm i.d. SP-2560 fused silica capillary column
(Supelco, Bellefonte, PA, USA), or a 100 m CP-Sil 88 fused capillary column
(Varian Inc, Mississauga, ON, USA), and a software system for data handling.
Operating conditions included: injector at 250°C and detector at 275°C; H2 as
carrier gas (1 ml/min) and H2 for the FID (40 ml/min); N2 make-up gas (100 ml/
min); and purified air (400 ml/min). The following temperature program was
employed to resolve the FAME isomers: an initial temperature of 45°C that is
held for 4 min; programmed at 13°C/min to 175°C and held for 27 minutes;
then programmed at 4°C/min to 215°C and held for 35 minutes (Kramer et al.,
2001; 2002). FAME are injected directly into the GC and identified by
comparison with a GC reference FAME standard (#463) spiked with the four
positional CLA isomers (mixture #UC-59M), and long-chain saturated FAME
21:0, 23:0, and 26:0. These standards could be obtained from Nu-Chek-Prep
Inc (Elysian MN, USA) or other chemical suppliers.
a. Influence of temperature program and sample load on the elution

pattern of FAME
FAME from dairy products or meat fat exhibit different patterns of separation
when compared with those of PHVO, depending on the chromatographic
conditions used. One such difference is the column temperature (Kramer et al.,
2001 and Cruz-Hernandez et al., 2006). Special attention is required when 18:1
and CLA isomers are analysed. Temperatures such as 170 or 175°C make a
Figure 18. A partial GC chromatogram of a dairy fat 18:1 FAME region using a 100 m CP Sil 88
capillary column, hydrogen as a carrier gas, and a typical temperature program from 45 to 215°C.
capillary column, hydrogen as a carrier gas, and an isothermal temperature at 180°C.
remarkable difference in the way isomers such as 13t to 15t are separated from
9c 18:1 (Precht & Molkentin, 1996 and 2003). Different elution patterns for cis
and trans C18 isomers were recognized when the column temperature was
operated at 150°C versus 130°C (Precht & Molkentin 2003). Effects of column
temperature on the relative order of elution of 18:3 and 20:1 isomers have also
been evaluated (Wolff & Precht, 1998). In Figures 18 and 19 different column
temperatures are compared. Figure 18 shows the typical separation obtained
with the ramped temperature program already described. Figure 19 shows the
typical separation obtained with 180°C isothermal as recommended in this
chapter for PHVO. As mentioned at the beginning of this chapter, 180°C
capillary column, hydrogen as a carrier gas, and a typical temperature program from 45 to 215°C. High
sample load.
isothermal temperature is appropriate for the analysis of the fatty acid profile,
including all the trans isomers of PHVO (Ratnayake, 2001). Isothermal
analysis at 180°C is also suitable for the analysis of trans isomers of milk fat.
However, it may not be appropriate for the complete analysis of the fatty acid
profile of dairy fats. At 180°C, the short-chain fatty acids, especially those from
4:0 to 8:0 would elute with the solvent front. In addition, the CLA isomers are
not adequately separated (Kramer et al., 2001; Cruz-Hernandez et al., 2004).
The GC sample load is an important parameter in GC analysis that is not
usually taken into consideration. The 18:1 cis and trans isomers will resolve
better if a low sample load is used, but a higher sample load is better for the
analysis of 16:1, CLA and polyunsaturated fatty acids. Therefore, any sample
solution that is too concentrated should be appropriately diluted. Generally,
two sample loads are analysed: a 1 µl solution containing about 1 to 2 µg is
injected to determine FAME present at low concentration, and a 3:1 or 5:1
dilution is necessary to gain a better resolution of FAME of the 18:1 region. In
the case of the CLA region, it will not be possible to separate the two major
isomers of CLA (7t,9c- and 9c,11t-CLA). CLA isomers are best analysed by
combining GC analysis with Ag+-HPLC fractionation. The procedure is dis-
cussed later in this section.
Figure 20 shows a chromatogram for the 18:0, 18:1 and 18:2 FAME regions
of a dairy sample. It illustrates why it is not possible to resolve some of the
trans-18:1 isomers from the cis-18:1 isomers at high sample loads. Especially,
the 12t-, 13-14t/6-8c-18:1 isomers overlap with the peak for 9c-, 10c-,15t-18:1.
At low sample load (Figures 18 and 19), there is an improved separation of 13-
14t/6-8c, appearing as a leading shoulder to the 9c peak. In Figures 18 and 19,
many of the different isomers present in the 18:1 region can be observed. The
predominant trans isomer is 11t.
b. Analysis of the 18:1 FAME region

Typical separations of the 18:1 isomers in PHVO have been shown at the
beginning of this chapter. Below, we will show chromatographic separations
for dairy products as an example of the typical separations obtained from dairy
and meat samples under the described conditions.
A number of differences were evident in the 18:1 region of the GC chroma-
togram. The best resolution of this region was observed when the total FAME
were analysed using a temperature program described in Figure 18. The 18:0 to
18:2n–6 region shown in Figure 18 contains peaks for cis-18:1, trans-18:1, and
18:2 cis/trans. Ten positional 18:1 isomers are clearly visible with 9c as the
predominant isomer. Among the 18:1 peaks resolved are: 4t, 5t, 6-8t, 9-11t, 12t,
13-14t/6-8c, 9c/15t/10c, 11-13c, 14c/16t, 15c and 16c. The region from 4t to
11t contains only trans isomers and there are no cis-18:1 isomers. Among the
cis isomers, the region from 11c to 13c isomers contains only cis isomers with
no overlaps with trans isomers, whereas with other isomers there are some cis
and trans isomer overlaps (Figure 18). The actual proportions of these overlap-
ping isomers can only be measured by complementing GC analysis with
Ag+-TLC or Ag+-HPLC fractionation (Cruz-Hernandez et al., 2004).
5. Analysis of monounsaturated FAME by combining AgNO3-TLC and GC

Several papers have described the advantages and disadvantages of different
kinds of columns and chromatographic conditions for GC analysis of
monounsaturated FAME (Molkentin & Precht, 1994, 2000; Kramer et al.,
1998, 2002; Precht & Molkentin 2003). Even when a good resolution of many
fatty acids is accomplished by GC, others are not separated or resolved, as is the
case with 18:1 isomers where overlaps are observed as previously mentioned in
this chapter. Under-estimation of these peaks is sometimes reported due to
inaccurate or incomplete analysis and consequently, the real isomers present
are incorrectly reported. The AgNO3-TLC/GC method has been highly recom-
mended and in many cases proved to be the best way to obtain a complete
analysis of the 18:1 isomers. Eleven t-18:1 isomers have been separated by
such techniques (Wolff et al., 1998; Precht & Molkentin, 1999; Kramer et al.,
2001). To complete the separation and resolution of trans-18:1 isomers, a
longer GC temperature program is recommended as described later in this
section (Precht & Molkentin, 1999; Kramer et al., 2001). Total dairy or meat
FAME can be fractionated into saturates, trans- and cis-monounsaturated fatty
acids plus CLA using AgNO3-TLC (Wolff et al., 1995; Precht & Molkentin,
1996; Kramer et al., 2001). Although less frequently used, separations can also
be performed using Ag+-HPLC columns (Bauman et al., 2000).
A B
5
6
Origin
Figure 21. AgNO3-TLC separation of the total milk fat FAME and FAME standards using hexane/
diethyl ether (90:10) as a developing solvent. Section A - Dairy fat FAME: 1, SFA; 2, trans-MUFA; 3,
cis-MUFA; 4, 18:2; 5, PUFA 1; 6, PUFA 2. Section B - FAME Standards: 1, 21:0; 2, 11t-18:1; 3, 9c-
18:1; 4, 18:2n–6; 5, 20:4n–6; 6, 20:5n–3.
Figure 21 shows a typical AgNO3-TLC separation of FAME from a cheese

sample and FAME standards. The bands corresponding to saturated fatty acids,
trans-monounsaturated fatty acids, and cis- monounsaturated fatty acids plus
CLA, are identified, scraped off and collected in Pasteur pipette (5.75 inch)
columns containing a glass wool plug. The silica gel is eluted with chloroform,
chloroform/methanol (1:1) and finally with methanol. The solvent is removed
under a stream of N2. To remove the dye and dissociate any remaining Ag+/
polyunsaturated fatty acid complex, lipids are partitioned in a Bligh and Dyer
solvent ratio in which water is substituted with 5N HCl solution. The chloro-
form phase is removed and taken to dryness using a stream of N2. Samples are
dissolved in hexane and analysed by GC using a stepwise isothermal tempera-
ture program starting at 120°C (Kramer et al., 2001). The temperature program
is as follows: Set initial temperature at 120°C and hold for 200 minutes,
increase to 150°C at 15°C/min and hold for 70 minutes, increase to 175°C at
15°C/min and hold for 60 minutes, increase to 220°C at 15°C/min and hold for
50 minutes. The cis and trans fractions were analysed at high and low sample
loads to resolve the minor and major FAME constituents.
In order to separate the overlapping cis and trans 18:1 isomers shown in
Figure 18, the combination of AgNO3-TLC and GC analysis is used. Figure 22
shows the cis (lower) and trans (upper) fractions from AgNO3-TLC after
resolution using a GC short temperature program (45 to 215°C). This tempera-
ture program is used to compare the positional isomers after AgNO3-TLC
Figure 22. A partial GC chromatogram of the trans-and cis-8:1 FAME fractions isolated by AgNO3-
TLC and analysed by a 100 m CP Sil 88 capillary column, hydrogen as carrier gas, and a typical
temperature program from 45 to 215°C.
separation with the total fat (Figure 18) and to identify some of the peaks that
were not there before the separation. The drawback of the short temperature
program is that 6t to 11t and 12t to 14t peaks elute close together (Figure 22,
upper) and it is difficult to separate them. Overlaps are also observed for the cis
isomers (Figure 22, lower). The relative concentration of 9c is too high, and
this was the reason for its co-elution with 10c.
6. Calculation of monounsaturated FAME

Total fatty acid composition was obtained analysing total milk FAME by GC
using the 100 m CP Sil 88 column as described in the previous section. A
sample load of total FAME was prepared to give baseline resolution between
11t- and 12t-18:1, as well as the 11c- complex and 12c-18:1 (Figure 18).
The sum proportion of 4t- to 11t-C18:1 (X%) determined by GC of total
FAME was used to calculate the total trans-18:1 isomers (Y%) and the content
of each of the trans-18:1 isomers. The percent Y was calculated by multiplying
X% by the ratio of the area response (4t- to 16t-18:1)/(4t- to 11t-18:1) obtained

from the GC separation (stepwise isothermal condition starting at 120°C) of the
total trans FAME fraction isolated by AgNO3-TLC (Kramer et al., 2001; Cruz-
Hernandez et al., 2006). The individual trans-18:1 isomers were calculated
using the relative concentrations obtained from the Ag+-TLC/GC results and
total trans-18:1 (Y%). To calculate the trans monounsaturated fatty acids other
than 18:1, the 16t-18:1 was used as a reference isomer.
7. Separation and identification of CLA isomers using GC and Ag+-HPLC

Both of these analyses are necessary to be able to resolve and identify all the
CLA isomers. The mixture of four positional CLA isomers serves as our CLA
standard (Figure 23), and is always added to the GC FAME mixture (for
example Mixture #463 available from Nu-Chek Prep). For identification, 21:0
FAME is added to the #463 standard FAME mixture (Kramer et al., 2002). To
investigate the minor CLA isomers by GC, a higher sample load of total FAME
is used, considering that the total CLA isomers generally comprise only 0.5 to
4% of total FAME (Figure 24). GC and Ag+-HPLC are used as complementary
techniques to perform a complete analysis of the CLA. The CLA isomers have
been separated using an HPLC system (Model 1100, Agilent Technologies,
Palo Alto, CA, USA) equipped with a quaternary pump (G1311A), autosampler
(G1313A), diode array detector (G1315A), and a Hewlett-Packard ChemStation
software system (Version A.07). Three ChromSpher 5 Lipids (4.6 mm i.d. ×
250 mm stainless steel; 5 µm particle size) analytical silver-ion impregnated
columns (Varian Inc., Mississauga, ON, Canada) are used in series. The mobile
phase, consisting of hexane containing 0.1% acetonitrile and 0.5% diethyl
Figure 23. A partial GC chromatogram of the CLA standard FAME region using a 100 m CP Sil 88
capillary column, hydrogen as carrier gas, and a typical temperature program from 45 to 215°C.
Figure 24. A partial GC chromatogram of the CLA FAME region of a milk fat using a 100 m CP Sil
88 capillary column, hydrogen as carrier gas, and a typical temperature program from 45 to 215°C.
ether, is supplied isocratically at a flow rate of 1.0 ml/min. The mobile phase
should be continuously mixed using a magnetic stirring bar. The diode array
detector is operated at 233 nm for the identification of conjugated FAME, at
205 nm for unsaturated (non-conjugated) FAME, and at 268 nm for conjugated
trienoic acids if present (Cruz-Hernandez et al., 2004). The elution order of
the CLA isomers was described previously (Sehat et al., 1998; Eulitz et
al., 1999; Kramer et al., 1999). The elution order of the CLA isomers by GC
(c/t < c,c < t,t) and Ag+-HPLC (t,t < c/t < c,c) are different. The identification
of the individual CLA isomers has been established previously (Eulitz et al.,
1999; Kramer et al., 1999; Sehat et al., 1998). The columns are cleaned daily
by flushing them with 1% acetonitrile in hexane (v/v) for one hour, followed by
flushing with the mobile phase for one hour.
8. Identification of all CLA isomers.

Analysis of the CLA region requires the combination of GC and Ag+-HPLC.
Identification and quantification of CLA isomers have been described else-
where (Cruz-Hernandez et al., 2004). The CLA peaks are quantified by GC
analyses of total FAME. There are three GC peaks in the CLA region that are
composed of unresolved CLA isomers: the ‘9c,11t-CLA’ peak may contain
7t,9c-, 9c,11t- and 8t,10c-CLA; the ‘9c,11c-CLA’ peak may also contain
11t,13c-CLA; and the combined t,t-CLA peak may contain the t,t-CLA isomers
from 7,9- to 10,12-CLA (Figure 25). On the other hand, there are several CLA
isomers that are well resolved by GC, such as 10c,12t-; 9t,11c-; 10t,12c-;
10c,12c-; 11c,13c-; 11t,13t-CLA. The identification of 21:0 poses a challenge,
9c11t
12,14 10,12 9t11c

10t12c
11,13 11t13c 7t9c
9,11 8t10c
25 30 35 40 min
Figure 25. A partial Ag+-HPLC separation of the cis/trans and trans/trans CLA region of a milk fat
sample using three Ag+-HPLC columns in series and UV detection at 233 nm.
since its concentration in natural dairy fats is similar to concentrations of the

minor CLA isomers, and it elutes anywhere between 11c,13t- to 10c,12c-CLA
depending on the GC column and the temperature program used (Kramer et al.,
1997, 1998; Roach et al., 2000; Precht & Molkentin 2000a) as previously
discussed. Therefore, the addition of 21:0 FAME and the four positional CLA
isomer mixture such as #463 from Nu-Chek Prep to the GC standard is
recommended.
D. Summary
Satisfactory analyses of the fatty acid profile of fats containing TFA are
achieved by GC using capillary columns coated with highly polar
cyanopolysiloxane stationary phases. In capillary GC methods, the key limita-
tion has been the incomplete separation of trans-18:1 isomers from their cis
isomers. However, almost complete separation of cis and trans-18:1 isomers of
PHVO origin is attainable with the use of 100 m cyanopolysiloxane capillary
columns operated isothermally at 180°C using hydrogen as the carrier gas at a
flow rate of 1 ml/min. Under these conditions, there is very little overlap of cis
and trans isomers. For obtaining detailed data on the individual cis and TFA
isomers, chromatographic analysis in conjunction with either AgNO3-TLC or
Ag+-HPLC has to be used.
Dairy and meat fats are more challenging lipid matrices to analyse. Milk fat
contains more than 400 different fatty acids, hence it is important to have
comprehensive analytical techniques to better evaluate their complex matrices.
In the past, the separation obtained by any kind of GC analysis was considered
satisfactory, and most researchers were not interested in separating and identi-
fying TFA isomers. Detailed fatty acid analysis is not accomplished with a
single chromatographic technique, and many factors (i.e., extraction, methyl-
ation, GC parameters, and unavailability of commercial standards) influence
the outcome of the analysis. GC remains, by far, the most convenient method
to characterize fatty acids, as explored in this chapter. Recent studies have
described the advantages and disadvantages of different kinds of columns and
chromatographic conditions (Molkentin and Precht, 1994 and 2000; Molkentin
and Precht 1995; Kramer et al., 1998; Kramer et al. 2002; Precht and Molkentin
2003). Even when a good resolution of many fatty acids is accomplished by
GC, some fatty acids are not separated or resolved. That is the case with trans-
18:1 isomers where overlaps of isomers are observed. Underestimation of these
peaks is sometimes reported due to inaccurate or incomplete analysis and as a
consequence, the real isomers present are incorrectly reported. The AgNO3-
TLC/GC method has been highly recommended and in many cases proved to be
the best way to obtain a complete analysis of the 18:1 isomers. Eleven t-18:1
isomers have been separated by such techniques (Wolff et al., 1998; Precht and
Molkentin, 1999; Kramer et al., 2001). To complete the separation and resolu-
tion of trans-18:1 isomers, a longer GC temperature program is also
recommended (Precht and Molkentin, 1999; Kramer et al., 2001).
References
Ackman, RG, Hooper, SN and Hooper, DL (1974) Linolenic acid artifacts from the
deodorization of the oils. J. Am. Oil Chem. Soc., 51, 42–49.
Adlof, R and List, G (2007) Analysis and characterization of trans isomers by silver-ion
HPLC. In: Trans Fats in Foods (List, G, Ratnayake, WMN and Kricthevsky, D, eds),
AOCS Press, Champaign, Ilinois, USA, pp.193–222.
Adlof, RO, Copes, LC and Emken EA (1995) Analysis of the monoenoic fatty acid
distribution in hydrogenated vegetable oils by silver-ion high-performance liquid chro-
matography. J. Am. Oil Chem. Soc., 72, 571–574.
Albertyn, DE, Bannon, CD, Craske, JD, Hai, NT, O’Rourke, KL and Szonyi, C (1982)
Analysis of fatty acid methyl esters with high accuracy and reliability. I. Optimization of
flame-ionization detectors with respect to linearity. J. Chromatog., 247, 47–61.
Aldai, N, Murray, BE, Nájera, AI, Troy, DJ and Osoro, K. (2005) Review. Derivatization of
fatty acids and its application for conjugated linoleic acid studies in ruminant meat lipids.
J. Sci. Fd Agric., 85, 1073–1083.
Ali, LH, Angyal, G, Weaver, CM and Rader, JL (1997) Comparison of capillary column gas
chromatographic and AOAC gravimetric procedures for total fat and distribution of fatty
acids in foods. Food Chem., 58, 149–190.
AOAC 2005 (a). AOAC Official Method 996.06. Fat (total, saturated, and unsaturated) in
foods, hydrolytic extraction gas chromatographic method, Revised 2001. In: Official
Methods of Analysis of AOAC International, 18th Edition (Horwitz, W, ed.).
AOAC 2005 (b). AOAC Official Method 996.01. Fat (total, saturated, and unsaturated) in
products, acid hydrolysis capillary gas chromatographic method, first action. In: Official
Methods of Analysis of AOAC International 18th Edition (Horwitz, W, ed).
AOCS 2005. Determination of cis-, trans-, saturated, monounsaturated and polyunsaturated
fatty acids in vegetable or non-ruminant animal oils and fats by capillary GLC method.
In: Official and Recommended Practices of the AOCS, 5th ed., Revisions and Corrections,
AOCS Press, Champaign, Official Method Ce 1h-05, approved 2005, revised 2005.
Bauman, DE and Griinari, JM (2003) Nutritional regulation of milk fat synthesis. Ann. Rev.
Nutr., 23, 203–227.
Bauman, DE, Barbano, DM, Dwyer, DA and Griinari JM (2000) Technical note: production
Beaulieu, AD and Palmquist, DL (1995) Differential effects of high fat diets on fatty acid
composition in milk of Jersey and Holstein cows. J. Dairy Sci., 78, 1336–1344.
Belury, MA (2002) Dietary conjugated linoleic acid in health: physiological effects and
mechanisms of action. Annual Review of Nutrition, 22, 505–531.
Beppu, F, Hosokawa, M, Tanaka, L, Kohno, H, Tanaka, T, Miyashita, K (2006) Potent
inhibitory effect of trans 9, trans 11 isomer of conjugated linoleic acid on the growth of
human colon cancer cells. The Journal of Nutritional Biochemistry, 17, 830–836.
Bhattacharya, A, Banu, J, Rahman, M, Causey, J and Fernandes, G (2006) Biological effects
of conjugated linoleic acids in health and disease (Review.) J. Nutr. Biochem., 17, 789–
810.
Bligh, EG and Dyer, WJ (1959) A rapid method of total lipid extraction and purification, Can.
J. Biochem. Physiol., 37, 911–917.
British Nutrition Foundation. Trans Fatty Acids, The Report of the British Nutrition
Foundation Task Force, July 1995. Holborn House, London, UK.
Chilliard, Y, Ferlay, A, Mansbridge, RM and Doreau, M (2000) Ruminant milk fat plasticity:
nutritional control of saturated, polyunsaturated, trans and conjugated fatty acids.
Annales de Zootechnie, 49, 181–205.
Chilliard, Y, Ferlay, A and Doreau, M (2001) Effect of different types of forages, animal fat
or marine oils in cow’s diet on milk fat secretion and composition, especially conjugated
linoleic acid (CLA) and polyunsaturated fatty acids. Livestock Production Science, 70,
31–48.
Chouinard, PY, Corneau, L, Barbano, DM, Metzger, LE. and Bauman, DE (1999) Conju-
gated linoleic acids alter milk fatty acid composition and inhibit milk fat secretion in dairy
cows. J. Nutr., 129, 1579–1584.
Christie, WW (1989) Gas Chromatography and Lipids. The Oily Press, Bridgwater, UK.
Christie, WW (1995) Composition and structure of milk lipids. In: Advanced Dairy Chemistry,
Volume 2 (Fox, P.F., ed.) 2nd ed., Chapman and Hall, London, UK, pp. 1–36.
Christie, WW (1999) The analysis of evening primrose oil. Industrial Crops and Products,
10, 73–83.
Christie, WW (2003) Lipid Analysis. Isolation, separation, identification and structural
analysis of Lipids. 3rd Edition, The Oily Press, Bridgwater, UK.
Christie, WW (2006a) Methylation of fatty aids. Online: www.lipidlibrary.co.uk/topics/
methests/index.htm
Christie, WW (2006b) Preparation of lipid extracts from tissues. Online:
www.lipidlibrary.co.uk/topics/extract2/index.htm
Christie, WW (2007) Mass spectrometry of fatty acid derivatives. Online:
www.lipidlibrary.co.uk/topics/masspec.html
Christopherson, SW and Glass, RL. (1969) Preparation of milk fat methyl esters by
alcoholysis in an essentially non-alcoholic solution. J. Dairy Sci., 52, 1289–1290.
Cruz-Hernandez, C, Deng, Z, Zhou, J, Hill, AR, Yurawecz, MP, Delmonte, P, Mossoba, MM,
Dugan, MER and Kramer, JKG (2004) Methods for analysis of conjugated linoleic acids
and trans-18:1 isomers in dairy fats by using a combination of gas chromatography,
silver-ion thin-layer chromatography/gas chromatography, and silver-ion liquid chroma-
tography. J. Assoc. Off. Analyt. Chem. Int., 87, 545–562.
Cruz-Hernandez, C, Kramer, JKG, Kraft, J, Santercole, V, Or-Rashid, M, Deng, Z, Dugan,
MER, Delmonte, P and Yurawecz, MP (2006) Systematic analysis of trans and conju-
gated linoleic acids in the milk and meat of ruminants, chapter 4. Pages 45–93 in Advances
in Conjugated Linoleic Acid Research. Volume III. (M.P Yurawecz, J.K.G. Kramer, O.
Gudmundsen, M.P. Pariza, and S. Banni, eds) AOCS Press, Champaign Ilinois, USA.
De La Torre, A, Debiton, E, Juaneda, P, Durand, D, Chardigny, JM, Barthomeuf, C, Bauchart,

D and Gruffat, D (2006) Beef conjugated linoleic acid isomers reduce human cancer cell
growth even when associated with other beef fatty acids. Brit. J. Nutr., 95, 346–352.
Destaillats, F and Angers, P (2002) Base-catalyzed derivatization methodology for FA
analysis. application to milk fat and celery seed lipid TAG, Lipids, 37, 527–532.
Eulitz, K, Yurawecz, MP, Sehat, N, Fritsche, J, Roach, JAG, Mossoba, MM, Kramer, JKG,
Adlof, RO and Ku, Y (1999) Preparation, separation, and confirmation of the eight
geometrical cis/trans conjugated linoleic acid isomers 8,10- through 11,13-18:2, Lipids
34, 873–877.
FAO/WHO (2002) Food and Agriculture Organization of the United Nations. WHO/FAO
Expert Consultation on Diet, Nutrition, and the Prevention of Chronic Diseases. WHO
Technical Report Series 916, Geneva.
Firestone, D and Sheppard, D (1992) Lipid analysis. In Advances in Lipid Methodology —
One, pp. 273–322 (Christie, WW, ed.) The Oily Press, Bridgwater, UK, pp.273–322.
Folch, J, Lees, M. and Sloane Stanley, GH (1957) A simple method for the isolation and
purification of total lipids from animal tissues. J. Biol. Chem., 226, 497–509.
Food and Drug Administration (2006) FDA acts to provide better information to consumers
on trans fats 2.005 (Accessed March 17, 2006 at www.health.gov/dietaryguidelines/
dga2005/document/
German, JB (1999) Butyric acid: a role in cancer prevention. Nutrition Bulletin, 24, 2993–
2999.
Grandgirard, A, Sebedio, JL, and Fleury, J (1984) Geometrical isomerization of linolenic acid
during heat treatment of vegetable oils. J. Am. Oil Chem. Soc., 61, 1563–1568.
Gurr, MI, Harwood, JL and Frayn, KN (2002) Lipid Biochemistry: An Introduction, 5th
edition, Blackwell Science Inc., Malden, Massachusetts, USA.
Hara, A and Radin, NS (1978) Lipid extraction of tissues with a low-toxicity solvent.
Analytical Biochemistry, 90, 420–426.
Health Canada (2003) Regulations Amending the Food and Drug Regulations (Nutrition
Labelling, Nutrient Content Claims and Health Claims). Department of Health, Canada
Gazette, Part 11. January 1, 2003.
House, SD, Larson, PA, Johnson, RR, DeVries, JW and Martin, DL (1994) Gas chromato-
graphic determination of total fat extracted from food samples using hydrolysis in the
presence of antioxidant. J. Assoc. Off. Analyt. Chem. Int., 77, 960–965.
House, SD (1997) Determination of total, saturated, and monounsaturated fat in foodstuffs
by hydrolytic extraction and gas chromatogrpahic quantitation. Collaborative study. J.
Assoc. Off. Analyt. Chem. Int., 80, 555–563.
Institute of Medicine (2002) Report of the panel on macronutrients of the standing committee
on the scientific evaluation of dietary reference intakes. Dietary reference intakes of
energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. Food
and Nutrition Board, Institute of Medicine. The National Academies Press, Washington,
DC, USA. pp. 8-1 to 8-97.
Jensen, RG, Ferris AM and Lammi-Keefe CJ (1991) The composition of milk fat. J. Dairy
Sci., 74, 3228–3243.
Jensen, RG (2002) The composition of bovine milk lipids: January 1996 to December 2000,
J. Dairy Sci., 85, 295–350.
Kramer, JKG and Hulan, HW (1978) A comparison of procedures to determine free fatty
acids in rat heart. J. Lipid Res., 19, 103–106.
Kramer, JKG, Fellner, V, Dugan, MER, Sauer, FD, Mossoba, MM and Yurawecz, MP (1997)
Evaluating acid and base catalysts in the methylation of milk and rumen fatty acids with
special emphasis on conjugated dienes and total trans fatty acids. Lipids, 32, 1219–1228.
Kramer, JKG, Sehat, N, Dugan, MER, Mossoba, MM, Yurawecz, MP, Roach, JAG, Eulitz,
K, Aalhus, JL, Schaefer, AL and Ku, Y (1998) Distribution of conjugated linoleic acid
(CLA) isomers in tissue lipid classes of pigs fed a commercial CLA mixture determined
by gas chromatography and silver ion-high performance liquid chromatography. Lipids,
33, 549–558.
Kramer, JKG, Sehat, N, Fritsche, J, Mossoba, MM, Eulitz, K, Yurawecz, MP and Ku, Y
(1999) Separation of conjugated linoleic acid isomers. In: Advances in Conjugated
Linoleic Acid Research, Volume 1. (Yurawecz, M.P., Mossoba, M.M., Kramer, J.K.G.,
Pariza, M.W. and Nelson, G.J., eds), pp.83–109, AOCS Press, Champaign, Illinois, USA.
Kramer, JKG and Zhou, J (2001) Conjugated linoleic acids and octadecenoic acids: extraction
and isolation of lipids. Eur. J. Lipid Sci. Technol., 103, 594–600.
Kramer, JKG, Cruz-Hernandez C and Zhou J (2001) Conjugated linoleic acids and octade-
cenoic acids: analysis by GC. Eur. J. Lipid Sci. Technol., 103, 600–609.
Kramer, JKG, Blackadar, CB and Zhou, J (2002) Evaluation of two GC columns (60-m
Supelcowax 10 and 100-m CP Sil 88) for analysis of milk fat with emphasis on CLA,
18:1, 18:2 and 18:3 isomers, and short- and long-chain FA. Lipids. 37, 823–835.
Kramer, JKG, Cruz-Hernandez, C, Deng, Z, Zhou, J, Jahreis, G and Dugan, MER (2004) The
analysis of conjugated linoleic acid and trans 18:1 isomers in synthetic and animal
products. Am. J. Clin. Nutr., 79 (Suppl.), 1137S–1145S.
LeDoux, M, Rouzeau, A, Bas, P and Sauvant, D (2002) Occurrence of trans-C18:1 fatty acid
isomers in goat milk: effect of two dietary regimens. J. Dairy Sci., 85, 190–197.
Lichtenstein, AH, Appel, LH, Brands, M, Carnethon, M, Daniels, S, Franch, HA, Franklin,
B, Kris-Etherton, P, Harris, WS, Howard, B, Karanja, N, Lefevre, M, Rudel, L, Sacks,
F, Van Horn, L, Winston, M, and Wylie-Rosett, J (2006) Diet and lifestyle recommen-
dations revision 2006: a scientific statement from the American Heart Association
Nutrition Committee (AHA Scientific Statement). Circulation 114, 82–96.
Marmer, WN and Maxwell, RJ (1981) Dry column method for the quantitative extraction and
simultaneous class separation of lipids from muscle tissue. Lipids 16, 365–371.
Maxwell, RJ, Mondimore, D and Tobias, J (1986) Rapid method for the quantitative
extraction and simultaneous class separation of milk lipids. J. Dairy Sci., 69, 321–325.
Molketin, J and Precht, D (1994) Optimized analysis of trans-octadecenoic acids in edible
fats. Chromatographia, 41, 267–272.
Molkentin, J and Precht, D (2000) Occurrence and biochemical characteristics of natural
bioactive substances in bovine milk lipids. Brit. J. Nutr., 84, Suppl. 1, S47–S53.
Mossoba, MM, McDonald, RE, Roach, JAG, Fingerhurt, DD, Yurawecz, MP and Sehat, N
(1997) Spectral confirmation of trans monounsaturated C18 fatty acid positional isomers.
J. Am. Oil Chem. Soc., 74, 125–130.
Mossoba, MM, Yurawecz, MP, Delmonte, P and Kramer, JKG (2004) Overview of
infrared methodologies for trans fat determination. J. Assoc. Off. Analyt. Chem. Int.,
87, 540–544.
Mozaffarin, D, Katan, MB, Ascherio, A, Stampfer, MJ and Willett, WC (2006) Trans fatty
acids and cardiovascular disease. New Engl. J. Med., 354, 1601–1613.
Nikolova-Damyanova, B (2003) Lipid analysis by silver ion chromatography. In: Advances
in Lipid Methodology-Five (Adlof, R.O., ed.). The Oily Press, Bridgwater, UK, pp.43–
123.
Precht, D and Molkentin, J (1996) Rapid analysis of the isomers of trans-octadecenoic acid
in milk fat, International Dairy Journal, 6, 791–809.
Precht, D and Molkentin, J (1999) C18:1, C18:2 and C18:3 trans and cis fatty acid isomers
including conjugated cis Δ9, trans Δ11 linoleic acid (CLA) as well as total fat composition
of German human milk lipids. Nahrung 43, 233–244.
Precht, D and Molkentin, J (2000) Identification and quantitation of cis/trans C16:1 and
C17:1 fatty acid positional isomers in German human milk lipids by thin-layer chroma-
tography and gas chromatography/mass spectrometry. Eur. J. Lipid Sci. Technol., 103,
102–113.
Precht, D and Molkentin, J (2003) Overestimation of linoleic acid and trans-C18:2 isomers
in milk fats with emphasis on Δtrans 9, Δtrans 12-octadecadienoic acid. Milchwissenschaft,
58, 30–34.
Precht, D and Molkentin, J (2000a) Frequency and distributions of conjugated linoleic acid
and trans fatty acid contents in European bovine milk fats. Milchwissenschaft, 55, 687–
691.
Precht, D, Molkentin, J, Destaillats, F and Wolff, RL (2001) Comparative studies on
individual isomeric 18:1 acids in cow, goat, and ewe milk fats by low-temperature high-
resolution capillary gas-liquid chromatography. Lipids, 36, 827–832.
Ratnayake, WMN (1998) Analysis of trans fatty acids. In: Trans fatty acids in human
nutrition (Sebedio, JL and Christie, WW eds), The Oily Press, Bridgwater, UK, pp.115–
161.
Ratnayake, WMN (2001) Analysis of dietary trans fatty acids. Journal of Oleo Science, 50,
339–352.
Ratnayake, WMN (2004) Overview of methods for the determination of trans fatty acids by
gas chromatography, silver-ion thin-layer chromatography, silver-ion liquid chromatog-
raphy, and gas chromatography/mass spectrometry. J. Assoc. Off. Analyt. Chem. Int., 87,
523–539.
Ratnayake, WMN and Beare-Rogers, JL (1990) Problems of analyzing C18 cis- and trans-
fatty acids of margarine on the SP-2340 capillary column. J. Chromatogr. Sci., 28,
633–639.
Ratnayake, WMN and Pelletier G (1992) Positional and geometrical isomers of linoleic acid
in partially hydrogenated oils. J. Am. Oil Chem. Soc., 69, 95–105.
Ratnayake, WMN, Chen, ZY, Pelletier, G and Weber, D (1994) Occurrence of 5c,8c,11c,15t-
eicosatetraenoic acid and other unusual polyunsaturated fatty acids in rats fed partially
hydrogenated canola oil. Lipids 29, 707–714.
Ratnayake, WMN, Plouffe, LJ, Pasquier, E and Gagnon, C (2002) Temperature-sensitive
resolution of cis- and trans-fatty acid isomers of partially hydrogenated vegetable oils on
SP-2560 and CP-Sil 88 capillary columns. J. Assoc. Off. Analyt. Chem. Int., 85, 1112–
1118.
Ratnayake, WMN, Hansen, SL, Kennedy, MP (2006) Evaluation of the CP-Sil 88 and SP-
2560 GC columns used in the recently approved AOCS Official Method Ce 1h-05:
determination of cis-, trans-, saturated, monounsaturated, and polyunsaturated fatty acids
in vegetable or non-ruminant animal oils and fats by capillary GLC method. J. Am. Oil
Chem. Soc., 83, 475–488.
Roach, JAG, Yurawecz, MP, Kramer, JKG, Mossoba, MM, Eulitz, K and Ku, Y (2000) Gas
chromatography-high resolution selected-ion mass spectrometric identification of trace
21:0 and 20:2 fatty acids eluting with conjugated linoleic acid isomers. Lipids 35, 797–
802.
Sebedio, JL and Ratnayake, WMN (2007) Analysis of trans mono- and polyunsaturated fatty
acids. In: Trans Fatty Acids (Hamilton R., Hamm, W. and Dijkstra, A.J. eds), Blackwell
Publishers, New York, USA, pp.102–131.
Sehat, N, Yurawecz, MP, Roach, JAG, Mossoba, MM, Kramer, JKG and Ku, Y (1998)
Silver-ion high-performance liquid chromatographic separation and identification of
conjugated linoleic acid isomers. Lipids, 33, 217–221.
Stender, S and Dyerber, J (2003) The influence of trans fatty acids on health. A report from
the Danish Nutrition Council. 4th edition, Publication no. 34.
Uberth, F and Henninger, M (1992) Simplified method for the determination of trans
monoenes in edible fats by TLC-GLC. J. Am. Oil Chem. Soc., 69, 829–831.
Walstra, P, Geurts, TJ, Noomen, A, Jellema, A and van Boekel, MAJS (1999) Dairy
Technology, Principles of Milk Properties and Processes, Marcel Dekker Inc, New York,
USA.
Wolff, RL, Bayard, CC and Fabien, RJ (1995) Evaluation of sequential methods for the
determination of butterfat fatty acid composition with emphasis on trans-18:1 acids.
Application to the study of seasonal variations in French butters. J. Am. Oil Chem. Soc.,
72, 1471–1483.
Wolff, RL and Precht, D (1998) Comments on the resolution of individual trans-18:1 isomers
by gas-liquid chromatography, J. Am. Oil Chem. Soc., 75, 421–422.
Wolff, RL, Combe, NA, Precht, D, Molkentin, J, and Ratnayake, WMN (1998) Accurate
determination of trans-18:1 isomers by capillary gas-liquid chromatography on cyanoalkyl
polysiloxane stationary phases. a Oleagineux, Corps Gras Lipids 5, 295–300.
Wolff, RL, Precht, D, Nasser, B and El Kebbaj, MS (2001) Trans- and cis-octadecenoic acid
isomers in the hump and milk lipids from Camelus dromedarius. Lipids, 36, 1175–1178.
Wolff, RL and Precht, D (2002) A critique of 50-m CP Sil 88 capillary columns used alone
to assess trans-unsaturated fatty acids in foods: the case of the TRANSFAIR study,
Lipids, 37, 627–629.
Yurawecz, MP, Mossoba, MM, Kramer, JKG, Pariza, MW and Nelson, GJ (1999) Advances
in Conjugated Linoleic Acid Research, Volume 1 (Yurawecz, M.P., Mossoba, M.M.,
Kramer, J.K.G., Pariza, M.W. and Nelson, G.J., eds), AOCS Press, Champaign, Illinois,
USA.
CHAPTER 6
Replacement of partially hydrogenated oils in
food products: a technological challenge
GUILLERMO NAPOLITANO1 AND FRANCESCA GIUFFRIDA2
1
2
A. Introduction
Vegetable fats with tailor-made melting properties are used in several food
applications, such as margarine, ice cream, bakery, confectionery, frying and
culinary. There are four basic ways to modify edible fats for obtaining the
necessary food functionality: fractionation, interesterification, hydrogenation
and blending.
Due to economic reasons, the hydrogenation process is the most commonly
used method to produce fats with the desired functionality and long shelf-life.
A decade ago, it was estimated that about one-third of the edible oils produced
worldwide were hydrogenated (Fitch-Haumann, 1994). Nevertheless, the hy-
drogenation process is under attack because it produces high levels of harmful
trans fatty acids (TFA) (up to 50% of total fat depending on the starting oil and
the hydrogenation conditions). As a consequence, efforts have been made to
reduce or eliminate TFA in food products.
The food industry is using or developing four strategies to reduce or
eliminate TFA in its products. These are modification of the hydrogenation
process; interesterification; modification of the food formulation; and the use
of oil with modified composition.
These four strategies are not used alone, but often in combination. Some of
the most important challenges faced by the food industry in replacing TFA in
foods are to: develop formulation options that provide the same functionality
without increasing costs; maintain relatively low levels of saturated fatty acids;
and comply with food regulations (Hunter, 2005).
B. Modification of the hydrogenation process

Wilhelm Norman patented the process for hydrogenation of oleic acid to
produce stearic acid in 1903. This was followed by the development of the
147
industrial hydrogenation of edible oils in the 1920s in Europe and the USA,
which has remained basically unchanged until recently. By the mid-1990s
approximately 25 million tonnes of fats, oils and fatty acids were being
hydrogenated every year for the food, cosmetics and lubricant industries (Fitch
Haumann, 1994). Despite the recent decline in the demand for partially
hydrogenated vegetable oils containing high concentrations of TFA, hydro-
genation of edible vegetable oils is still an important process among other oil
modification technologies.
Partial hydrogenation in the food industry is traditionally used to produce
semi-solid, oxidatively-stable fats from vegetable oils for multiple food appli-
cations. The partial hydrogenation of vegetable oils, mainly from soybean,
palm, rapeseed (canola) and cottonseed oils, is the major source of TFA in our
food supply (for example Van Poppel et al., 1998; Wijesundera et al., 2007).
The concentration of TFA in partially hydrogenated vegetable fats can vary
from about 3% in lightly hydrogenated (brushed) oils, to more than 50%,
depending on the type of base oil, process conditions and degree of hydrogena-
tion. In many cases, the desired functionality of the hydrogenated fat, such as
a particular solid fat content and melting profile, is to a large extent a function
of the concentration and type of triacylglycerols containing TFA. For example,
high levels of elaidic acid (9-trans-18:1), with a melting point of 43°C are
frequently required in many applications in order to achieve the desired
hardness or melting profile. In those cases a reduction of the concentration of
TFA will be accompanied by a reduction of the desired functionality of the fat.
This is typically the case in many high TFA content confectionery products
(Timms, 2003) and in culinary fats derived from the partial hydrogenation of
palm, soybean and cottonseed oils. On the other hand, in cases where the
desired functionality is mainly the function of the content and distribution of
saturated and cis unsaturated fatty acids in the triacylglycerols, changes in the
parameters and conditions of hydrogenation that favour the reduction of
unsaturation, and at the same time minimizing trans isomerization, are of
interest.
Formation of trans fatty acids during the hydrogenation process is described
in Chapter 2.
C. Interesterified low-TFA fats

Interesterification of edible oils is a relatively old modification technique,
rarely used in the past but revitalized now due to the high demand for low-TFA
fat alternatives. The edible oil industry has developed many different types and
variations of interesterification process directed to modify the physical, chemi-
cal and functional properties of fats and oils. Most of them are designed to
obtain specific nutritional and textural characteristics. The ability of
interesterification to modify the melting and crystallization characteristics of a
REPLACEMENT OF PARTIALLY HYDROGENATED OILS 149
fat without changing its fatty acid composition was rapidly recognized as an
important tool for the development of low-TFA alternatives (Rozendaal &
Macrae, 1997).
In the context of applications to the edible fat and oil industry,
interesterification involves one form or another of rearrangement of the fatty
acids in the triacylglycerol molecules. Natural fats and oils have not only
specific fatty acid compositions, but also specific fatty acid distribution in the
triacylglycerol structure. This specific distribution is the response of organisms
to the need to maintain appropriate cellular function and to adjust to environ-
mental variables (Ucciani & Debal 1996). Vegetable fats and oils tend to have
most of the unsaturated fatty acids esterifying the position sn–2 of the
triacylglycerol, whereas saturated fatty acid occupies positions sn–1 and sn–3.
The two basic types of fatty acid rearrangement with respect to applications
for the edible fat and oil industry are chemical interesterification and enzymatic
interesterification.
1. Chemical interesterification
Chemical interesterification of lipids proceeds spontaneously at temperatures
> 250°C. However, chemical interesterification of edible fats is achieved at
much lower temperatures, usually 80–100°C and in the presence of sodium
methoxide powder (0.05–0.1% w/w) as a catalyst. Industrial chemical
interesterification of fats and oils produces a statistical or random distribution
of the fatty acids over the sn–1, 2 and 3 positions of the glycerol molecules as
shown schematically in Figure 1.
The effect of chemical interesterification of a fat or oil varies depending on
the degree of order and arrangement of the fatty acids present in the initial
material. The effect on melting point, for example on refined, bleached and
deodorized (RBD) palm oil, is relatively small, but for other oils with a more
Figure 1. Schematic reactions in the chemical and enzymatic interesterification of mono-acid

triacylglycerols.
marked non-random distribution of fatty acids, such as cocoa butter, the effect
of chemical interesterification is more pronounced.
The reduction in the melting point obtained by chemical interesterification
of a blend of vegetable oil with a fully hydrogenated vegetable oil depends on
the ratio of the components in the original blend and on the decrease in the
concentration of high melting point components, mainly di- and tri-saturated
triacylglycerols. Reduction in the melting point of 20–25ºC after chemical
interesterification of vegetable oil blends (sunflower and rapeseed oils) with
fully hydrogenated vegetable oils is possible (Schmidt et al., 1996).
In a special case of interesterification (directed chemical interesterification),
the temperature of the process can be controlled by inducing the crystallization
of the newly formed tri-saturated triacylglycerols of palmitic and stearic acids.
These solid tri-saturated triacylglycerols no longer take part in the
interesterification reaction, remaining as such and significantly increasing the
melting point and solid fat content of the final product (Sreenivasan, 1978).
2. Enzymatic interesterification
Enzymatic interesterification typically produces triacylglycerols with a non-
random distribution, selectively cleaving and re-attaching the fatty acids in the
sn–1 and sn–3 positions, leaving the fatty acids in the sn–2 position virtually
untouched (Figure 1). The enzymes used for this type of interesterification are
lipases isolated from yeast and other fungi such as Candida antarctica,
Thermomyces lanuginose and Rhizomucor miehei, which are immobilized in
an inert substrate. Enzymatic interesterification is now widely used for the
production of low-TFA plastic fats for numerous applications, but it is not a
new technique. It was patented by in 1977 and its use was very limited
until immobilized enzymes became affordable for large-scale industrial use.
Initially, lipase-catalysed interesterification was used for high-value speciality
structured lipids. Still today, enzymatic interesterification is required for
special applications where positional specificity of the fatty acids in the
triacylglycerol molecule is important or necessary to achieve the intended
functionality. Examples are the production of symmetric triacylglycerols of the
type SUS or USU, such as cocoa butter equivalents (Timms, 2003) and breast
milk substitutes (Lien et al., 1997), respectively.
The operational costs of enzymatic interesterification are higher than those
of chemical interesterification. In recent years, however, efficiency of the
immobilized enzymes has increased significantly making enzymatic
interesterification more competitive. More recently, the availability of cost-
effective immobilized enzymes has allowed its use for the production of plastic
fats of massive consumption for margarines, spreads and shortenings that
traditionally used partially hydrogenated vegetable oils.
Additional advantages of enzymatic interesterification are those associated
with much lower reaction temperatures, such as reducing chemical degradation

of the interesterified fat (especially oxidative degradation) and side reactions
(such as the production of alkylketones), avoidance of flammable catalysts
(e.g., sodium methoxide), and a reduction of environmental effluents (Lampert,
2000; Cook, 2000).
Compared to enzymatic interesterification, chemical interesterification pro-
ceeds very fast to completion, so intermediate interesterified stages are not
possible. On the other hand, enzymatic interesterification can be stopped
before completion and therefore it offers the possibility to produce fats with
different degrees of conversion. Although most changes occur within the first
3–4 hours, different degrees of conversion are feasible and can be monitored by
measuring free fatty acids, diacylglycerols, solid fat, crystal polymorph (Zhang
et al., 2004), or by using spectroscopic methods (Chang et al., 2005).
Despite the many fundamental differences between chemical and enzymatic
interesterification, there is a significant overlap in the range of products that
can be obtained by these competing techniques.
3. Applications
Chemical and enzymatic interesterification are applied to edible fats and oils
for many different purposes. Most applications of interesterification to edible
fats are aimed at modifying the melting point of blends or single fats by
reducing the concentration of tri-saturated or tri-unsaturated triacylglycerol
species (Sonntag, 1982). Interesterification is also performed to induce the
formation of a preferred crystal polymorphism, or to increase the miscibility
between ‘incompatible’ fats (e.g., lauric and non-lauric fats), thus reducing an
unwanted eutectic softening effect (Noor Lida et al., 2006). Due to the current
health and nutritional trends, those changes in physicochemical characteristics
of fats and oils induced by interesterification are extensively used for produc-
ing the ‘zero’- and low-TFA fats and oils now required in many food products.
These mainly include margarines, shortening, spreads, confections and many
other products that require a fat phase with the plasticity provided by the
previously ubiquitous high-TFA partially hydrogenated vegetable oils.
The versatility of interesterified palm and palm kernel oil blends for the
formulation of low-TFA fat products have been summarized and reviewed by
Berger and Idris (2005). These authors covered a number of common applica-
tions for the production of reduced-TFA or low-TFA foods using both chemical
and enzymatic interesterification (e.g. shortenings, margarines and spreads),
but also referred to less common fat products obtained by interesterification,
such as frying fats, low-fat spreads and whipped cream substitutes.
Most low-TFA margarines and spreads are currently produced by
interesterification of vegetable oils with fully hydrogenated vegetable oils as
base stocks. The purpose of interesterification for manufacturing low-TFA
margarines is not only to obtain plastic fat with the suitable spreading and
melting characteristics, but also to promote the formation of small and stable β'
crystals that confer the desired smoothness and integrity of the product, in
blends that otherwise would tend to form coarse β crystals.
Soybean and canola oils are used as common base stocks for manufacturing
shortening and margarines in the USA. Interesterification blends of palm or
canola oils with lauric fats are preferentially used in Europe. Fully hydrogen-
ated palm or canola oils interesterified with a lauric fat can produce a range of
spread products of different hardness and low in TFA (< 2%). The
interesterification also significantly sharpens the melting profile of the original
physical blend, producing a fat with more than 90% solids at temperatures
below 20°C and much less solids at ambient and near-body temperatures
(Allen, 1996).
In an effort to improve the oxidation stability of interesterified hard stocks
used for the production of shortenings, soybean oil was replaced by high- or
mid-oleic/low-linoleic varieties of sunflower and canola oils (Lampert, 2000).
Blends of high-oleic canola oil and fully hydrogenated soybean oil (55% and
45% respectively) can produce low-TFA fats that would reproduce the func-
tionality required for baking applications of a conventional all-purpose
high-TFA (>20%) shortening. Furthermore, it is important to note that the
TFA + SFA (SFA = saturated fatty acids) of these interesterified blends re-
mains well below those of the conventional all purpose shortenings (~35% vs
55% respectively).
Enzymatic interesterification of a blend of palm stearin and coconut oil
(75:25 w/w) was used to study the production of TFA-free margarine in a
bench-scale reactor using the Rhizomucor miehei lipase immobilized in an ion
exchange resin (Zhang et al., 2000). Manipulating the reaction conditions,
such as temperature, reaction time and especially water content, it was possible
to obtain an enzymatically interesterified fat that was similar in characteristics
to that obtained by chemical interesterification.
The enzyme stability and retention of activity was a major concern during the
early stages of development of enzymatic interesterification. Rapid lost of
enzymatic activity was a major factor limiting the use and cost competitiveness
of enzymatic interesterification of commodity oils. However, significant
progress has been made in the immobilization and retention of enzyme activity.
Zhang et al. (2000) measured effective and stable activity of immobilized
Rhizomucor miehei lipase after re-using it for a total of ten cycles. Effective
activity for the enzymatic interesterification of palm stearin and soybean oil of
immobilized lipase suitable for margarine production was maintained for
21 days in a continuous reactor with an immobilized lipase from Candida
antartica (Osorio et al., 2005).
Vanaspati, a popular alternative to ghee in India and South Asia, is manu-
factured with partially or fully hydrogenated vegetable oils and can contain
high levels of TFA. Chemically interesterified blends of palm stearin with

‘exotic fats’ (i.e. mango and mahua oils) were studied for the production of
low-TFA vanaspati (Khatoon & Reddy, 2005). The authors claim that differ-
ent interesterified blends of palm stearin with the exotic fats have melting
characteristics which are similar to those of the partially hydrogenated fats
used for the same applications. Although melting characteristics are impor-
tant, texture and crystal stability are also fundamental parameters, which
need to be measured in order to predict the functionality of fats in new
applications.
Ronne and co-workers (2005) investigated the production of butter-based
products obtained by lipase interesterification of butter fat with rapeseed oil.
The primary purpose of this work was to reduce the level of nutritionally
questionable saturated fatty acids present in milk fat, i.e. lauric, myristic and
palmitic acids. The physical properties of the resulting interesterified blends
were not reported. Furthermore, an important drawback of this approach was
the release of short-chain fatty acids from butter fat and the development of
unacceptable off-flavours.
Partially hydrogenated soybean, cottonseed and palm oils have been used as
the raw materials of choice for manufacturing cost effective cocoa butter
replacers (CBR). These fats have a much higher compatibility with cocoa
butter compared to cocoa butter substitutes based on lauric fat, allowing the use
of more cocoa liquor in the formulations with a positive impact on the flavour
of the ‘chocolate’ confection. CBR, however, have one of the highest concen-
trations of trans fatty acids in foods, and levels of 45% in some CBR are not
uncommon (Timms, 2003). Chemical interesterification of blends of canola
and palm stearin (50:50) and cottonseed oil and palm stearin (60:40) have been
proposed as suitable low-TFA confectionery fats (Karabulut et al., 2004).
Based on the melting characteristics reported for these interesterified fats (35–
40% solid fat content at 20°C) it can be predicted that they are rather soft to be
used as compound coatings, and one can assume that the application of these
fats may be limited to soft confectionery fillings and creams.
Solvent fractionated tallow stearins and oleins were interesterified with soft
oils and hard exotic fats (e.g. sal fat, Shorea robusta) respectively, in order to
obtain various types of edible products (Bhattacharyya et al., 2000). The tallow
stearin was chemically or enzymatically interesterified (Rhizomucor miehi)
with soybean, sunflower or rice bran oils. These various interesterified fats
were proposed for use as hard base stock for manufacturing non-hydrogenated
shortenings, margarines, vanaspati substitutes and cocoa butter alternatives.
The characterization of the resulting interesterified blends was based on their
fatty acid composition, melting point and solid fat index (SFI), but no reference
has been made to the crystallization and the textural attributes of the resulting
fats. Therefore, it is difficult to assess and predict the potential applications and
functionality of these interesterified blends. Furthermore, the use of tallow as
a major component in the fat raw material may not represent, or may not be
perceived as, a significant nutritional advantage for the replacement of high-
TFA hydrogenated fats.
Reduced fat spreads made with partially hydrogenated oils can contain up to
20% TFA. Chemically interesterified ternary blends of sunflower, palm and palm
kernel oils and oleins were studied for the production of tub and block spreads
(Noor Lida & Md. Ali, 1998). It is generally accepted that a solid fat content (SFC)
value between 4° and 10°C determines the spreadability of the product at
refrigeration temperature, and that an SFC < 30% at 10°C is necessary to achieve
spreadability at that temperature. The SFC at refrigerator temperatures of the
tertiary blends reported here ranged from about 10% to 38%, depending on their
composition, making most of them suitable for the intended applications.
A study was conducted to follow the changes in physical properties of
enzymatically interesterified hard stocks during a storage time of 12 weeks
(Zhang et al., 2005). These hard stocks were intended to be used in the
formulation of table margarines and the parameters measured included hard-
ness, dropping point, crystal form and colour. Margarine stability increased
with the degree of conversion and, not surprisingly, margarine produced with
the fat fully converted had overall physical properties similar to that produced
with chemically interesterified fat.
Fattahi-far and co-workers (2006) studied the use of chemical inter-
esterification of hydrogenated and non-hydrogenated tea seed oil as a hard
stock for the production of table margarines. Physical and sensory properties of
these experimental margarines (i.e., melting point, SFC, texture, colour, fla-
vour, spreadability) were favourable compared with commercial types. It was
claimed that interesterification is viewed as a natural process for the production
of low-TFA foods and tea seed oil is considered a healthy ingredient. One
wonders whether the positive health attributes of the tea seed oil can be
maintained after subjecting the oil to such a drastic process of hydrogenation
and chemical interesterification.
According to food legislation in different countries, interesterification,
whether chemical or enzymatic, does not need to be declared as an oil
modification technique. However, there are some concerns over long-term
acceptance of chemical interesterification. These concerns are related to envi-
ronmental (water effluents) and safety issues for plant workers (flammable
catalyst) and consumers (side-reaction products).
D. Modification of the food formulation

Although the use of partially hydrogenated oils is decreasing dramatically
worldwide, they still play an important role in the food industry. The major
effects of oil hydrogenation in the physicochemical properties of foods are:
• increased melting point and manipulation of melting profile;
• textural improvements by modification of crystal polymorphs; and

• increased oxidative stability by reducing the levels of PUFA present in the
native oils.
The main uses of partially hydrogenated fats are in frying, production of

table margarines, and culinary applications.
Partially hydrogenated fats used for frying are often semi-liquid oils pre-
ferred for their convenient handling and oxidative stability. An alternative to
hydrogenated oils in frying applications is the use of high-oleic or mid-oleic
variants of seed oils developed by using conventional breeding and genetic
modification techniques, described later in this chapter.
Many important characteristics of table margarines, such as, spreadability,
resistance to oil exudation, mouth feel, and flavour release, depend on the solid
fat content provided by partial hydrogenation of oils. Therefore, the replace-
ment needs to reproduce particular melting profiles and crystallization patterns
in order to achieve the optimal structure and sensory characteristics of the
original product. Trans-free fats for margarines can be designed by
interesterifying fully hydrogenated and soft oils (see previous section in this
chapter). An example of a trans-free oil/fat blend for table margarines is a blend
of palm stearin, palm kernel oil and sunflower seed oil in the ratio of 60:20:20
respectively (Yusoff & Dian, 1995).
In culinary products, partially hydrogenated fats provide the right crystalli-
zation behaviour for manufacture, an appropriate melting point, and a long
shelf-life. In order to decrease or eliminate TFA, manufacturers replace par-
tially hydrogenated fats with non-hydrogenated (trans free) fats, often
represented by unsaturated oil. Unfortunately, this simple substitution does not
work well and formulations frequently require major modifications. For exam-
ple, bouillon cubes require oils with a specific level of solids at different
temperatures to achieve the desired functionality. Solid fat comes either from
TFA or saturated fatty acids. Replacing partially hydrogenated fats with soft
oils rich in unsaturated fatty acids reduces the solid fat content of the fat phase
at all temperatures, making the product softer, prone to oil staining of the
packaging and increases the susceptibility to oxidation.
One way to compensate for the use of softer oils in culinary products is to
increase the proportion of dry ingredients and powders, thus providing more
absorption surface for the larger proportion of oil or liquid fraction in the new
fat. Some negative aspects of low-TFA fats, such as an indirect reduction of the
flavour intensity that is possibly due to coating of sensory buds in the mouth,
could be corrected by enhancing the flavour components in the formulation
(spices, flavourings). Moreover, some low-TFA fats tend to have a higher
crystallization rate at critical temperatures, compared to their partially hydro-
genated counterparts, producing premature solidification of the fat during
production or consumption.
E. Oils with modified composition
In order to obtain oils more resistant to oxidation and with an improved health
profile, many efforts have been put into ‘designing’ food oils, for instance
agronomically suitable crops, with a high content of oleic acid and low content
of polyunsaturated and saturated fatty acids.
Several oilseeds with modified fatty acid compositions have been developed
(List, 2004). Breeding programmes have used available germplasm to develop
crops, for example sunflower and canola, that have a more desirable oil content
or fatty acid composition (McKeon, 2005). Mutagenic processes and genetic
engineering have been intensively used to produce genetically-modified seed
oils (e.g. sunflower and canola seed oils).
Currently the four genetically-engineered (transgenic) crops that have been
adopted are all oilseed crops: sunflower, soybean, corn, cotton and canola. The
mutagenic approach is geared to eliminating genes; as a consequence this
approach decreases levels of undesired fatty acids or increases levels of a
desired fatty acid (Spiller et al., 1998). An example of induced mutagenesis
applied to oil crops is the conversion of the non-edible high α-linolenate
oilseed, linseed, to a high-linoleate variety. The high α-linolenate seeds were
treated with the mutagen ethyl methane sulphonate to disrupt the fatty acid
desaturase responsible for the conversion of linoleate to α-linolenate. By
crossing single mutants, a population of double mutants was obtained that
yielded linseed oil with similar acyl composition to that of sunflower oil
(Murphy, 2006).
Crop genetic engineering can lead to undesirable consequences, such as the
need for chemical treatments, by introducing the genes encoding herbicide
tolerance and/or insect resistance (Murphy, 2006 ), and it can introduce
desirable traits such as high levels of monounsaturated and low levels of
saturated fatty acids to improve nutritive value (Murphy, 2006). The most
common approach involves understanding plant lipid metabolism followed by
identification and cloning of the enzyme producing desired fatty acids and
further expression in the breeding varieties.
There follow some examples of oils with modified composition used in the
food industry.
1. High-oleic, non-genetically-modified sunflower oil

Sunflower oil contains about 70% linoleic acid. By reducing the linoleic acid
content and increasing the concentration of oleic acid both shelf life and
oxidation stability are significantly extended. The origin of the high-oleic acid
gene in most breeding programmes for sunflower oil date back to the early
1970s with an open pollinated release, known as ‘Pervenets’, from what was
then the Soviet Union. Conventional plant breeding techniques allowed trained
breeders to introduce this high-oleic gene into developmental lines, thereby

fixing the oleic content trait, with minimal variation, regardless of climatic
conditions. Crossing these lines with other types which displayed strong
agronomic traits allowed production of high-oleic hybrids now accepted by
growers in most countries (Oilseeds International Ltd, San Francisco, USA;
web page: www.oilseedssf.com/products/prod_olsun.html).
To maintain the proper structure and other textural attributes in foods, the
reduction of TFA commonly results in a proportional increase in saturated fatty
acids. In low-TFA formulations, partially hydrogenated vegetable oils are
widely replaced by palm oil fractions and interesterified oils, often containing a
fully hydrogenated component. Lauric oils are also used with the purpose of
providing solid structural fat, but they are restricted mostly to confectionery
products and margarines. Unfortunately, all these low-TFA fat alternatives
contain high levels of potentially unhealthy saturated fatty acids, i.e. lauric,
myristic and palmitic. There is increasing evidence indicating that stearic acid
does not have the same cholesterolaemic and atherosclerogenic effect of the
shorter chain fatty acids (C12–C16) (Crupkin & Zambelli, 2008), making oils rich
in stearic acid a healthy low-TFA structural fat alternative. An interesting new
development is ‘Nutrisun’, a high-stearic and high-oleic, non-GM sunflower oil
produced by Advanta Semillas (Mar del Plata, Argentina). This oil contains
approximately 18% stearic acid. Stearin fractions of Nutrisun can be obtained,
with stearic acid contents of up to 45% (Advanta, personal communication).
2. High-oleic, non-genetically-modified canola oil

In the mid-1990s, Dow AgroSciences LLC developed a new line of canola seed
that was naturally-bred to yield an oil having high stability without the need for
hydrogenation. Natreon canola oil, derived from canola bred through classical
plant breeding, has > 70% oleic acid and < 3% linolenic acid content
(www.dowagro.com/omega9oils). This fatty acid profile gives the oil its
characteristic flavour stability in a variety of applications, including frying,
baking, spray oils, bottled oil, and salad dressings.
3. Low-linolenic, high-stearic, non-genetically-modified soybean oil

Initially, modification of soybean oil composition through plant breeding was
targeted at reduction of linolenic acid to improve the oxidative stability of non-
hydrogenated soybean oil. Later, efforts were made to both reduce linolenic
acid and increase stearic acid in order also to improve oil plasticity. In other
words, these modified soybean oils are actually high-stearate and low-lino-
lenic. Therefore, without hydrogenation, such modified oils would be suitable
not only for frying applications but also for making low-trans shortening and
tub margarine (Liu et al., 1999).
4. High-oleic, genetically-modified soybean oil

α-Linolenic acid content in soybean oil ranges from 5 to 12%, making the oil
susceptible to oxidation and flavour reversion. As a consequence, soybean oil
is often partially hydrogenated to reduce drastically the content of linolenic and
linoleic acids and extend the shelf-life of deep-fried products. In order to
reduce the level of linolenic acid without the need for hydrogenation, a low-
linolenate trait was introduced by traditional breeding into soybean a line
already genetically-engineered to carry the gene for glyphosate resistance with
herbicide tolerance (Mounts et al., 1994). This low-linolenic soybean, named
‘Vistive’ (Monsanto, St Louis, Missouri, USA), is now available with < 3%
linolenic acid compared to 8% in traditional soybean oil (www.monsanto.com).
A high-oleic acid soybean variety has been under development for years by the
Du Pont Corporation (Delaware, USA) (Kinney et al., 1997; Liu et al. 1999) and
will be commercialized in 2009. High-oleate soybean oil was obtained by
suppressing the oleyl desaturase, the enzyme catalysing the conversion of
linoleic to oleic acid. In addition, Du Pont Corporation has in the pipeline a high-
oleate/stearate soybean oil which will be commercialized in 2012.
5. Issues related to transgenic oilseed

Transgenic crops with modified fatty acid composition show a higher oxidative
and thermal stability (Smith et al., 2007) and sometimes an improved healthy
fatty acid profile. Nevertheless, the commercial development of genetically-
modified crops is limited by a negative public perception, resulting from food
safety and environmental concerns. Transgenic crop approval in the USA
requires the action and approval of three Federal agencies: the Food and Drug
administration, the Environmental Protection Agency, and the Department of
Agriculture’s Animal and Plant Health Inspection Service (Thomson et al.,
2003). Finally, in the European Union the use of non-approved genetically-
modified organisms is not permitted. The use of approved genetically-modified
organisms is permitted with mandatory labelling for amounts above 0.9% by
weight of genetically-modified material in the ingredient.
F. Conclusions
Current efforts to replace or reduce TFA in food products focus on the
following strategies: modification of the hydrogenation process; use of
interesterified fat; modification of the food formulation; and use of oil with
modified composition
Currently, there is no hydrogenation technology available that could elimi-
nate TFA formation. Modification of the food formulation is not always
possible and would affect taste and texture with consequent consumer com-
plaints. Breeding programmes take a long time; traditional breeding techniques
can, for example, require up to 10 years from breeder crosses to commercial

seed release.
Production of oils from genetically-modified organisms is still expensive
and their use is not permitted worldwide. In addition, public concern over
genetically-modified food sources represents a barrier to their commercializa-
tion.
Alternative oils and fats are available or are in development to replace TFA
in food. Nevertheless, the decision on which alternatives to use is complex,
involving consideration of health effects, quality, availability, process modifi-
cations and cost. Therefore, the balance between fat composition and functional
properties, and re-design of the manufacturing process and supply chain, will
drive the identification of the appropriate partially hydrogenated oil replace-
ment in food production.
G. Sources of further information and advice

Several valuable reviews are already cited in the text and listed in the Refer-
ences section; however, we wish to highlight the excellent article “Food
application of trans fatty acid substitutes” by Wassel and Young (2007).
References
Albright, LF (1987) Hydrogenation of triglycerides. In: Hydrogenation: Proceedings of the
AOCS Colloquium (Hastert, R, ed.) American Oil Chemists’ Society, Champaign
Illinois, USA, p.12.
Allen, DA (1996) Interesterification. A vital tool for the future? Lipid Technology, 11, 15.
An, W, Hong, JK, Pintauro, PN, Warnr, K and Neff, WE (1998) The electrochemical
hydrogenation of edible oils in a solid polymer electrolyte reactor: I. Reactor design and
operation. J. Am. Oil. Chem. Soc., 75, 917–925.
Berben, P (2002) Hydrogenation of fats and oils: trends in catalyst development. Lipid
Technology Newsletter, 12, 133–137.
Berger, KG and Idris, NA (2005) Formulation of zero-trans acids shortening and margarines
and other food fats with products of the oil palm. J. Am. Oil. Chem. Soc., 82, 775–782.
Bhattacharyya, S, Bhattacharyya, DK and De, BK (2000) Modification of tallow fractions in
the preparation of edible fat products. Eur. J. Lipid Sci. Technol., 102, 323–328.
Chang, T, Lai, X, Zhang, H, Sondergaard, I and Xu, X (2005) Monitoring lipase-catalyzed
interesterification for bulky fat modification with FT-IR spectroscopy. J. Agric. Food
Chem., 53, 9841–9847.
Cook, R (2000) Chemical interesterification – Process and safety management. SCI Lecture
Papers Series 116. Ebortec, Whale Bridge Park, UK.
Crupkin, M, and Zambelli, A (2008) Detrimental impact of trans fats on human health: stearic
acid-rich fats as possible substitutes. Comprehensive Reviews in Food Science and Food
Safety, 7 (3), 271–279.
Dijkstra, AJ (2002) Hydrogenation and fractionation. In: Fats in Food Technology (Rajah,
KK, ed.) CRC Press, Boca Raton, Florida, USA, pp.123–157.
Dijkstra, AJ (2006) Revisiting formation of trans isomers during partial hydrogenation of
triacylglycerol oils. Eur. J. Lipid Sci. Technol., 108, 249–264.
Dijkstra, AJ, Toke, ER, Kolonitis, P, Recseg, K, Kovari, K, and Poppe, L (2006) The base
catalyzed, low temperature interesterification mechanism revisited. Eur. J. Lipid Sci.
Technol., 107, 912–921
Dow Agrosciences, information available at: www.dowagro.com
Drozdowski, B and Szukalska, E (2000) Effect of rapeseed oil hydrogenation conditions on
trans isomer formation. Eur. J. Lipid Sci. Technol., 102, 642–645.
Farr, WE (2005) Hydrogenation: processing technologies. In: Bailey’s Industrial Oil and Fat
Products. Sixth edition, volume 5 (Shahidi, F, ed.) Wiley-Interscience, Hoboken, New
Jersey, USA, pp.385–96.
Fattahi-far, E, Sahari, MA and Barzegar, M (2006) Interesterification of tea seed oil and its
application in margarine production. J. Am. Oil. Chem. Soc., 83, 841–845.
Fitch-Haumann, B (1994) Tools: hydrogenation, interesterification. Inform, 5, 668–678.
Hastert, RC (2000) Hydrogenation. In: Introduction to Fats and Oils Technology (O’Brien,
RD, Farr, WE and Wan, PJ, eds) American Oil Chemists’ Society, Champaign Illinois,
USA, pp.179–193.
Hunter, JE (2005) Dietary levels of trans-fatty acids: basis for health concerns and industry
efforts to limit use. Nutrition Research, 25, 499–513.
Karabulut, I, Turan, S and Ergin, G (2004) Effects of chemical interesterification on solid fat
content and slip melting point of fat/oil blends. Eur. Food Res. Technol., 218, 224–229.
Kellens, M (2000) Oil modification processes. In: Edible Oil Processing (Hamm, W and
Hamilton, RJ, eds) CRC Press. Boca Raton, Florida, USA, pp.129–173.
Khatoon, S and Reddy, SRY (2005) Plastic fats with zero trans fatty acids by interesterification
of mango, mahua and palm oils. Eur. J. Lipid Si. Technol., 107, 786–791.
King, JW, Russell, LH List, GR and Snyder, JM (2001) Hydrogenation of vegetable oils using
mixtures of supercritical carbon dioxide and hydrogen. J. Am. Oil. Chem. Soc., 78, 107–
113.
Kinney, AJ (1997) In: Physiology, Biochemistry and Molecular Biology of Plant Lipids
(Williams JP, Khan MU and Lem NW, eds) Kluwer Academic, Dordrecht, The Nether-
lands, 298–300.
Koritala, S, Friedrich, JP and Dutton, HJ (1980) Selective hydrogenation of soybean oil. X.
Ultra high pressure and low pressure. J. Am. Oil. Chem. Soc., 57, 1–5.
Lampert, D (2000) Processes and products of interesterification. In: Introduction to Fats and
Oils Technology, Second edition (O’Brien, RD, Farr, W E and Wan, PJ, eds) American
Oil Chemists’ Society, Champaign Illinois, USA, pp.208–234.
Lien, E, Boyle, FG, Yuhas, R, Tomarelli, RM, Quinlan, P (1997) J. Pe d. Gastroenterol. Nutr.,
25, 167–174.
List, GR (2004) Decreasing trans and saturated fatty acids content in food oils. Food
Technology, 58 (1), 23–31.
List, GR, Warner, K, Pintauro, P and Gil, M (2007) Low-trans shortening and spread fats
produced by electrochemical hydrogenation. J. Am. Oil. Chem. Soc., 84, 497–501.
Liu, KS (1999) Soy oil modification: products, applications. Inform, 10 (9), 868–878.
Macher, M-B, Högberg, J, Moller, P, and Härröd, M (1999) Partial hydrogenation of fatty
acids methyl esters at supercritical conditions. Fett/Lipid, 101, 301–305.
McKeon, TA (2003) Genetically modified crops for industrial products and processes and
their effect on human health. Trend in Food Science & Technology, 14, 229–241.
McKeon, TA (2005) Transgenic oils. In: Bailey’s Industrial Oil & Fat Products, Sixth
Edition, Volume 3 (F Shahidi, ed.) Wiley-Interscience, J. Wiley & Sons, Inc., Hoboken,
New Jersey, USA.
Moller, P, Harrod, M (1997) Hydrogenation of oils at supercritical conditions. Abstract of
88th American Oil Chemists’ Society Annual Meeting & Expo, Seattle, Washington,
USA, p.8.
Monsanto, information available at: www.monsanto.com
Mounts, TL, Warner, K, List, GR, Neff, WE and Wilson, RF (1994) J. Am. Oil. Chem. Soc.,
71, 495–499.
Murphy DJ (2006) Plant breeding to change lipid composition for use in food. In: Modifying
Lipids for Use in Food. (Gunstone, FD, ed.) CRC Press, Cambridge, UK.
National Sunflower Association, information available at: www.sunflowernsa.com
Noor Lida, HMD and Ali, AR (1998) Physico-chemical characteristics of palm-based oil
blends for the production of reduced fat spreads. J. Am. Oil. Chem. Soc., 75, 1625–
1631.
Noor Lida, HMD, Sundram, K and Idris, NA (2006) DSC study on the melting properties of
palm oil, sunflower, and palm kernel olein blends before and after chemical
interesterification. J. Oil Am. Chem. Soc., 83, 739–745.
Osorio, NM, Gusmao, JH da, Fonseca, MM and Ferreira-Diaz, S (2005) Lípase-catalyzed
interesterification of palm stearin with soybean oil in a continuous fluidised-bed reactor.
Eur. J. Lipid Sci. Technol., 107, 455–463.
Ronne, TH, Lars, SP and Xu, X (2005) Triglyceride selectivity of immobilized Thermomyces
lanuginose lipase in interesterification. J. Am. Oil. Chem. Soc., 82, 737–743.
Ronne, TH, Yang, T, Mu, H, Jacobsen, C and Xu, X (2005) Enzymatic interesterification of
butterfat with rapeseed oil in a continuous packed bed reactor. J. Agric. Food Chem., 53,
5617–5624.
Rozendaal, A and Macrae, AR (1997) Interesterification of oils and fats. In: Lipid Technolo-
gies and Applications (Gunstone, FD and Padley, FB, eds) Marcel Deckker, New York,
USA, pp.223–263.
Rubin, LJ, Koseoglu, SS, and Graydon, WF (1986) Hydrogenation of canola oil in the
presence of nickel and methyl benzoate-chrome-carbonyl complex. J. Am. Oil Chem.
Soc., 63, 1551–1557.
Schmidt, S, Hurtova, S, Zemanovic, J, Sekretar, S, Simon, P and Ainsworth, P (1996)
Preparation of modified fats from vegetable oil and fully hydrogenated vegetable oil by
randomization with alkali catalysis. Food Chemistry, 55, 343–348.
Schööm, N-H (1995) Is a low trans content attainable by conventional hydrogenation of
vegetable oils. In: Oil-Fats-Lipids (Castenmiller, WAM, ed.) PJ Barnes & Associates,
Bridgwater, UK, pp.155–158.
Smith, SA, King RE, Min DB (2007) Oxidative and thermal stabilities of genetically modified
high oleic sunflower oil. Food Chemistry, 102, 1208–1213
Sonntag, NOV (1982) Fat splitting, esterification and interesterification. In: Bailey’s Indus-
trial Oil and Fat Products, Volume 2, Fourth Edition (Swern, D, ed.) John Wiley & Sons,
New York, USA, pp.97–173.
Sreenivasan, B (1978) Interesterification of fats. J. Am. Oil Chem. Soc., 55, 796–805.
Sundram, K, Karapaiah, T and Hayes, KC (2007) Stearic acid-rich interesterified fat and
trans-rich fat raise the LDL/HDL ration and plasma glucose relative to palm olein in
humans. Nutrition & Metabolism, 4 (3), online: doi:10.1186/1743-7075-4-3
Timms, R (2003) Confectionery Fats Handbook. Properties, Production and Application.
See Chapter 4. The Oily Press, Bridgwater, UK.
Ucciani, E and Debal, A (1996) The effect of triglyceride positional distribution on fatty acid
absorption in rats. chemical properties of fats. In: Oils and Fats Manual (Karlesking, A
and Wolff J-P, eds) Lavoisier Publishing, Paris, France, pp.325–332.
Van Poppel, G, Van Erp-Baart, MA, Kafatos, DA and Aro, A (1998) Trans fatty acids in foods
in Europe: The Transfair study. J. Food Com. Anal., 11, 112–136.
Warner, K, Neff, WE, List GR and Pintauro, P (2000) Electrochemical hydrogenation of
edible oils in solid polymer electrolyte reactor. Sensory and compositional characteristics
of low trans soybean oils. J. Am. Oil. Chem. Soc., 77, 1113–1118.
Wassel, P and Young, NWG (2007) Food applications of trans fatty acid substitutes.
International Journal of Food Science and Technology, 42, 503–517.
Wijesundera, C, Richard, A and Ceccato, C (2007) Industrially produced trans fat in foods
in Australia. J. Am. Oil Chem. Soc., 84, 433–442.
Wright, AJ, Mihele, AL and Diosady, LL (2003) Cis selectivity of mixed catalyst systems in
canola oil hydrogenation. Food Res. Intl., 36, 77–804.
Yusoff MSA and Dian NHM (1995) Trans free formulation: a short review. Palm Oil
Developments, 22, 33–37.
Zhan, H, Jacobsen, C and Adler-Nilsen, J (2005) Storage stability study of margarines
produced from enzymatically interesterified fats compared to margarines produced by
conventional methods. Eur. J. Lipid Sci. Technol., 107, 530–539.
Zhang, H, Smith, P and Adler-Nilssen, J (2004) Effect of degree of enzymatic interesterification
on the physical properties of margarine fats: solid fat content, crystallization behaviour,
crystal morphology, and crystal network. J. Agric. Food Chem., 52, 44323–4431.
Zhang, H, Xu, X, Mu, H Nilsson, J, Adler-Nissen, J and Höy, C-E (2000) Lipozyme IM-
catalyzed interesterification for the production of margarine fats in a 1 kg scale stirred
tank reactor. Eur. J. Lipid Sci. Technol., 102, 411–418.
CHAPTER 7
Metabolism of trans fatty acid isomers
JEAN-LOUIS SÉBÉDIO1 AND WILLIAM W. CHRISTIE2
1
UMR 1019, Unité de Nutrition Humaine, Plateforme d’exploration du
métabolisme, INRA centre de Theix, St Genes Champanelle, France
2
Scottish Crop Research Institute (and Mylnefield Lipid Analysis), Invergowrie,
Dundee, Scotland
A. Introduction
Trans fatty acids (TFA) include monounsaturated and polyunsaturated fatty
acids having methylene interrupted double bonds but also isomers having
conjugated double bonds such as conjugated isomers of linoleic acid or CLA.
These may be natural fatty acids but they may also be formed during techno-
logical treatments such as hydrogenation, refining, or frying of oils (Sebedio &
Juaneda, 2007). This review will only deal with isomers not having a conju-
gated double bond system. A comprehensive review on the metabolism of
non-conjugated isomers was recently published by Emken (2007). Earlier
reviews deal with trans monoenes (Hølmer, 1998) and trans polyenes (Sébédio
& Chardigny, 1998) separately.
Trans monounsaturated fatty acids are formed during hydrogenation of oils
to produce margarine and shortening (Dutton, 1979), and they are also present
in ruminant meat and milk as a result of biohydrogenation in the rumen (Precht
et al., 2001; Wolff, 1995). These quantitatively and nutritionally are the most
important sources of TFA in the human diet. The position of the double bond
of these dietary trans 18:1 isomers, counted from the carboxylic carbon,
usually varies from Δ4 to Δ16 carbon atoms of the fatty acid molecule. Very
often in partially hydrogenated vegetable oils, the trans 18:1 isomers form a
Gaussian distribution that centres on the Δ9 or Δ10 double bond. The distribu-
tion depends on the fatty acid composition of the starting vegetable oils and on
the extent of hydrogenation. The trans 18:1 isomer distribution of ruminant fats
is distinctly different from that of partially hydrogenated vegetable oils. In the
former, 11t-18:1 is always the major isomer, whereas 9t-18:1 and 10t-18:1
isomers occur in relatively small amounts (Wolff et al., 1998).
There is also evidence that TFA can be produced within animal tissues by
thiyl-radical catalysed cis/trans isomerization in vitro and in vivo (Zambonin
163
et al., 2006), but the contribution from this source in humans is not known.
Similarly, isomerization of one of the double bonds in polyunsaturated fatty
acids and catalysed by nitrous oxide is possible (Balazy, 2000; Zghibeh et al.,
2004).
Dietary fats may also contain a number of positional and geometrical
isomers of linoleic and α-linolenic acids, which are frequently present in low
concentrations in both partially hydrogenated and non-hydrogenated dietary
fats (Ratnayake, 2004). The linoleic and α-linolenic acid isomers present in
non-hydrogenated fats or in many common food fats are the result of exposure
of these polyunsaturated fatty acids to heat treatment; such as steam
deodorization (Ackman et al., 1974), or deep fat frying (Grandgirard et al.,
1984). α-Linolenic acid is more prone to isomerization than linoleic acid,
whereas oleic acid is hardly isomerized at all. In many non-hydrogenated
dietary fats, usually the two mono-trans isomers of linoleic (i.e., 9c,12t-18:2
and 9t,12c-18:2) are present at similar levels and generally higher than the all-
trans isomer, 9t,12t-18:2. Eight geometric isomers are possible for α-linolenic
acid, but usually only four are present in industrially refined oils or oils subject
to mild heat treatments. They have been identified as 9t,12c,15t-18:3, 9c,12c,15t-
18:3, 9c,12t,15c-18:3 and 9t,12c,15c-18:3 (Ackman et al., 1974; Grandgirard
et al., 1984) Geometrical isomers of eicosapentaenoic acid (EPA) and of
docosahexaenoic acid (DHA) are also formed during the deodorization process
of fish oils. In this case, the mono-trans isomers are the major TFA formed
(Fournier et al., 2006; Fournier et al., 2007)
Besides the trans polyunsaturated fatty acids resulting from technological
processing of fats and oils, some unusual trans polyunsaturated fatty acids are
present in certain seeds from which it is possible to extract oil. For example, one
fatty acid of this type is columbinic acid (5-trans,9-cis,12-cis-18:3), which is
linoleic acid with addition of a trans Δ5 bond (Houtsmüller, 1981). However,
these are rarely if ever encountered in the diet.
B. Metabolism of trans isomers in vivo
1. Absorption and incorporation into tissues

Fat comprises about 40% of the energy intake in the human diet in Western
countries, and a high proportion of this is in the form of triacylglycerols. The
process of fat digestion is begun in the stomach by gastric or lingual lipases, but
this is relatively unimportant in comparison to the reaction with pancreatic
lipase, which occurs in the duodenum. Bile acids released from the gall bladder
act to emulsify the hydrophobic triacylglycerols enabling attack by the hydro-
lytic enzyme pancreatic lipase. The process of hydrolysis is regiospecific and
results in the release of the fatty acids in free form from the 1(3) positions of the
triacylglycerols and formation of 2-monoacyl-sn-glycerols. The free fatty acids
METABOLISM OF TRANS FATTY ACID ISOMERS 165
and 2-monoacyl-sn-glycerols are rapidly taken up by the intestinal cells, via

specific carrier molecules, and are esterified into triacylglycerols before they
are packaged into chylomicrons and transported to the liver in the form of
chylomicrons. This process has been reviewed by Mu & Hoy (2004).
In general, there is ample evidence to suggest that nearly all isomeric cis and
trans mono- and dienoic C18 fatty acids in ruminant fats and partially hydrogen-
ated vegetable oils, when fed as part of a mixed diet, are efficiently absorbed
and incorporated into chylomicrons. Emken (2007) discounted experimental
results that came to a different conclusion on the grounds that the experimental
protocols compromised absorption. Possible exceptions are dietary
triacylglycerols containing fatty acids with double bonds in the Δ2 to Δ7
positions, which are hydrolysed more slowly by pancreatic lipase. Dietary
mono- and dienoic fatty acids rarely contain fatty acids with double bonds in
these positions, but polyunsaturated fatty acids, such as those from fish oils do.
Only limited data dealing with the absorption of trans polyunsaturated fatty
acids are available (Emken, 1983; Ono & Frederickson, 1964; Trus et al.,
1991) However, studies using the all trans-18:2 and the 18:3 isomers did not
shown any differences in their absorption compared to the cis isomers.
Once they reach the liver, chylomicron remnant triacylglycerols are taken
up, repackaged and exported into the circulation in the form of very-low-
density and low-density lipoproteins. In the process, dietary TFA have an effect
on plasma LDL cholesterol concentrations and therefore on the risk of coronary
heart disease, but this aspect is discussed elsewhere in this book. During
triacylglycerol synthesis in the intestines and liver, the fatty acids are esterified
in a stereospecific manner in which saturated fatty acids are located predomi-
nantly in the primary positions and unsaturated in position 2. Similarly,
phospholipids in the plasma contain mainly unsaturated fatty acids in posi-
tion 2. However, the experimental evidence suggests that there is little or no
discrimination between isomers of cis and trans monoenes in the process of
triacylglycerol synthesis (Emken, 1984, 1986). Following incorporation into
lipoprotein fractions, triacylglycerols are transported to the peripheral tissues,
where they are broken down by the enzyme lipoprotein lipase and taken up into
cells. During transport in plasma, fatty acids in position 2 of phosphatidylcholine
are transferred to cholesterol to form cholesterol esters by the enzyme
lecithin:cholesterol acyl transferase, which discriminates strongly against trans-
monoenes in comparison to the corresponding cis isomers (Sgoutas, 1970).
Similarly, a human feeding study has shown a strong discrimination against
incorporation of trans isomers of linolenic acids in human plasma cholesteryl
esters (Sebedio et al., 2000).
There is no evidence of discrimination between cis and trans monoenes in
the uptake of TFA by tissues, which is reportedly proportional to the dietary
intake (Baylin et al., 2002). Dietary TFA are incorporated into most lipid
classes of the liver and peripheral tissues with little or no effect on lipid class
distributions when these are fed as part of a normal diet. The reviews by
(Hølmer, 1998) and Emken (2007) cover the literature exhaustively, and some
general conclusions can be drawn. For example, much of the TFA are found in
triacylglycerols in tissues, with somewhat less in phospholipids, especially in
liver, kidney, testes, heart and adipose tissue. Very little TFA is found in testes
or the brain. With the exception of the adrenals, cholesterol esters tend to have
the lowest concentrations of TFA. Amongst the main glycerophospholipid
classes, there is greater trans monoene incorporation into phosphatidyl-
ethanolamine, phosphatidylserine and phosphatidylinositol than into
phosphatidylcholine.
In a study with rats on a diet containing 15% partially hydrogenated
safflower oil (50%) for 5 weeks, Wood (1979) observed that all positional
isomers of TFA were incorporated into triacylglycerols of all tissues examined
in proportions similar to those fed, with a slight discrimination in favour of the
8- and to some extent the 9-18:1 isomers. In addition, some trans-hexadec-
enoates, especially trans-8-16:1, were found in triacylglycerols, indicating
chain shortening of trans-10-18:1. In general, trans-12, 13- and 14-18:1
isomers were preferentially incorporated. Comparable results have been ob-
tained by others (see Hølmer, 1998, and Emken, 2007).
Acyltransferases are responsible for triacylglycerol, phospholipid and cho-
lesterol esterification, the specificities of these controlling tissue fatty acid
composition, membrane fluidity and cell function. In all phospholipid classes,
trans monoenoic fatty acids like the saturated fatty acids tend to be concen-
trated in the primary positions relative to position 2, in contrast to cis monoenes.
For example in one study, 11-, 12-, 13- and 14-18:1 were the main trans
isomers found in position sn–1 of the phosphatidylcholine of liver, heart and
plasma, with evident discrimination against trans-10-18:1. In contrast, small
amounts only of trans-9-, 11- and 12-18:1 were present in position sn-2, where
cis-9- and 11-18:1 were the main monoenoic components (Reichwald-Hacker
et al., 1979). In a series of papers published between 1979 and 2002, Emken
and colleagues studied the specificity of the incorporation of specific cis and
trans positional isomers of monoenoic fatty acids into the phosphatidylcholine
and phosphatidylethanolamine of human plasma (reviewed by Emken, 2007),
and some comparable data were obtained by Wood (1979) for phospholipids in
other tissues. Relative to oleate, they observed discrimination against 8t- and
10t-18:1 and the preferential incorporation of 9t- and 12t-18:1.
In relation to trans-polyenes, it was demonstrated that di-trans 18:2 was
incorporated preferentially into position 1 of phosphatidylcholine (Selinger &
Holman, 1965). Comparable results were obtained by Privett et al. (1966).
Furthermore, trans-monoenes, di-trans 18:2, palmitic and stearic acids were
incorporated mainly at the sn–1 and sn–3 positions of the triacylglycerols while
the 9-cis,12-trans-18:2, like linoleic acid was incorporated in the sn-2 position.
A preferential incorporation of 9-cis,12-cis,15-trans-18:3 into rat mitochon-
dria cardiolipin was demonstrated by Wolff et al. (1993). 9-Cis,12-cis,15-

trans-18:3 accumulates in cardiolipin at a higher level than in other
phospholipids (11 times in liver and 5–7 times in heart), representing 22–24%
of the total fatty acids in cardiolipin. Furthermore, the 15-trans-18:3 is esteri-
fied to both the 1 (1") and 2 (2") positions of liver mitochondria cardiolipin,
with a marked selectivity for positions 1 (1"). The selectivity ratio (1(1")/2(2"))
is similar to what was found for linoleic acid, its structural analogue. It
therefore appears that the trans-15 ethylenic bond would be perceived as a
single bond by the enzymatic systems that ensure acylation of cardiolipin.
Interestingly, Waku and Lands (1968) showed that the acylation of 9-cis,12-
trans-18:2 in phosphatidylcholine (position 2) was close to that of 18:1. On the
other hand, the acylation rate of 9-trans,12-cis-18:2 was similar to that of
linoleic acid. These data enhance the hypothesis that a trans double bond
located closer to the methyl end of the carbon chain may be recognized as a
saturated bond by acylation enzymes. That is not the case when the trans
double bond is closer to the carboxyl end.

Desaturation of fatty acids, i.e. insertion of a cis-double bond into the structure,
increases their fluidity and can change completely their biological properties
and function. In relation to trans-monoenes, by far the most widely studied
reaction is the desaturation of vaccenic acid (11t-18:1) to the conjugated
isomer 9-cis,11-trans-18:2. Similarly 7-trans-18:1 can be converted to 7-trans,
9-cis-18:2. However, this is the topic of a separate chapter in this book, so need
not be discussed at length here. Although the process does not appear to have
been studied systematically, elongation of trans-18:1 fatty acids to form the
homologous C20 trans monoenes has been observed in many studies in vivo (cf.
Wood, 1979).
The metabolism of the mono-trans isomers of linoleic acid was studied as
early as 1963 by Blank and Privett who showed that methyl 9-cis,12-trans-
linoleate had been converted to a trans eicosatetraenoic acid, which was not
fully identified. Further work in the field allowed characterization of the
different trans polyunsaturated fatty acids formed by elongation and desaturation
of the mono-trans isomers of linoleic acid (Privett et al., 1967; Berdeaux,
1996; Berdeaux et al., 1996; Ratnayake et al., 1994). All the studies demon-
strated that at least one trans isomer of linoleic acid, 9-cis,12-trans-18:2, could
be desaturated and elongated into a trans isomer of arachidonic acid, 5-cis,8-
cis,11-cis,14-trans-20:4. However, these studies did not agree as to whether
the other mono-trans isomer, 9-trans,12-cis-18:2, could be converted into a
trans isomer of arachidonic acid. Thus, Beyers and Emken (1991) and Berdeaux
et al. (1996) showed that only the 9-trans,12-cis-18:2 was elongated to 11-
trans,14-cis-20:2, while Ratnayake et al. (1994) suggested that both the
mono-trans isomers of 9,12-18:2 could be elongated into trans isomers of

11,14-20:2.
Studies on mice using isotopically labelled 18:2 isomers (9-cis,12-cis; 9-
cis,12-trans and 9-trans,12-cis) showed that the metabolites were incorporated
into liver, plasma, heart and brain lipids (Beyers & Emken, 1991). Analysis of
the metabolites showed a similar pattern in plasma and in liver lipids. For the
brain, the levels of the all cis metabolites were about four times higher than the
9-cis,12-trans-18:2 metabolites especially for the 20:4. However, the amount
of 5-cis,8-cis,11-cis,14-trans-20:4, 5-cis,8-cis,11-trans,14-cis-20:4 and 5-cis,8-
cis,11-cis,14-cis-20:4 in the brain were 31, 15 and 10 times lower than their
concentrations in the liver lipids. In contrast, the data for the heart were
somewhat different, with the concentration of the 20:4 metabolite of 9-cis,12-
trans-18:2 being 2.3 times that of the 9-cis,12-cis-18:2 metabolites.
Furthermore, as shown by Berdeaux et al. (1996) for the rat, the long-chain
trans 20:4 can be incorporated into different liver lipid classes. However no
modifications of the fatty acid composition of the phospholipid classes were
induced by small quantities of isomers of linoleic acid in the diet (0.6% of the
energy). The 14-trans-20:4 isomer was only detected in three phospholipid
classes, phosphatidylethanolamine, phosphatidylinositol and phosphatidyl-
choline, the highest quantity (0.8%) being found in phosphatidylinositol. It
should be noted that this phospholipid class also contained the highest amount
of 20:3(n–9) (2.8%). The lowest level of 14-trans-20:4 (0.2%) was observed in
phosphatidylethanolamine, the phospholipid which contained the lowest quan-
tity of 20:3(n–9). The phosphatidylcholine showed an intermediate situation. It
therefore appears that 14-trans 20:4 behaves like its structural analogue, the
trans Δ14 bond being recognized as a single bond by enzymatic systems as
already noted for the 18:3 isomers (Wolff et al., 1993).
Ratnayake et al. also showed (1994) that the 9-cis,12-trans-18:2 and 9-
trans,12-cis-18:2 were not the only trans 18:2 isomers which were desaturated
and elongated into trans isomers of arachidonic acid. Thus, 9-cis,13-trans-
18:2, which is the major trans polyunsaturated fatty acid in partially
hydrogenated vegetable oils, was shown to be desaturated and elongated to 5-
cis,8-cis,11-cis,15-trans-20:4. This fatty acid was detected in several tissues of
rats fed a partially hydrogenated canola oil.
The trans,trans-18:2 isomer was reported initially by Knipprath and Mead
(1964) to be converted into a di-trans isomer of eicosatetraenoic acid, but the
work of Privett et al. (1967) demonstrated that no such conversion occurred.
Studies carried out by Karney and Dhopeshwarkar (1978) on the developing
rat brain also demonstrated a lack of direct conversion of the trans, trans-
18:2 isomer. However, analysis of the brain polyunsaturated fatty acids after
intracranial injection showed that 9-trans,12-trans-18:2 was desaturated into
6-cis,9-trans,12-trans-18:3 followed by an elongation to form 8-cis,11-
trans,14-trans-20:3. These authors concluded that the conversion of
9-trans,12-trans-18:2 was blocked at the Δ5 desaturation step rather than at

the Δ6 desaturation one.
Some trans 18:3 isomers can also be desaturated and elongated into trans
isomers of eicosapentaenoic (EPA) and docosahexaenoic acids (DHA). In vivo,
among the different isomers formed during heat treatment of oils (9-trans,12-
cis,15-cis-18:3, 9-cis,12-cis,15-trans-18:3 and to a lesser extent
9-trans,12-cis,15-trans-18:3), the 9-cis,12-cis,15-trans-18:3 seems to be the
preferential substrate for the conversion into 5-cis,8-cis,11-cis,14-cis,17-trans-
20:5, 7-cis,10-cis,13-cis,16-cis,19-trans-22:5 and into 4-cis,7-cis,10-cis,13-cis,
16-cis,19-trans-22:6 (Grandgirard et al., 1989; Grandgirard, 1992). First evi-
dence of this conversion was reported by Grandgirard et al.(1989) by feeding
rats with a heated linseed oil which contained high levels of geometrical
isomers of linolenic acid. The mono-trans geometrical isomers of 20:5(n–3)
and 22:6(n–3) were identified in liver lipids and their structures were elucidated
using partial hydrazine reduction, oxidative ozonolysis and Fourier-transform-
infrared spectroscopy.
These trans (n–3) long-chain polyunsaturated fatty acids were found to be
incorporated into almost all lipid classes, including phospholipids of different
tissues, such as liver, heart, kidneys, adrenals and testes. The incorporation of
the different trans long-chain PUFA is quite different among the different
tissues that have been considered. However, while the 17-trans isomer of EPA
is incorporated in all the neutral lipid classes and in most of the phospholipid
classes except sphingomyelin, the trans isomers of 22:5(n–3) and 22:6(n–3)
were never detected in triacylglycerols or phosphatidylcholine.
Other 18:3 geometrical isomers can also be converted into long-chain trans
polyunsaturated fatty acids (Chardigny et al., 1996). For example, this is the
case for the 9-trans,12-cis,15-cis-18:3 and of 9-trans,12-cis,15-trans-18:3,
which are converted into 5-cis,8-cis,11-trans,14-cis,17-cis-20:5 and 5-cis,8-
cis,11-trans,14-cis,17-trans-20:5, respectively. Other trans 20:5 isomers were
detected. This indicates that the other 18:3 isomers (9-cis,12-trans,15-trans
and 9-cis,12-trans,15-cis) may be converted to a lesser extent into long-chain
metabolites (Piconneaux, 1987). In any case, the conversion is better if the
geometry of the Δ9 ethylenic bond is cis (Brenner, 1971). This confirms the
previous data found for the (n–6) family (Privett et al., 1967).
The trans long-chain (n–3) polyunsaturated fatty acids may also be incorpo-
rated into brain and retinal lipids (Grandgirard et al., 1994; Chardigny et al.,
1994). Feeding rats for 8 weeks with an oil containing a mixture of 18:3
geometrical isomers resulted in the incorporation of trans 20:5 and 22:6
isomers. The incorporation of trans 22:6 was found to be time-dependent,
whereas trans 20:5 only increased strongly after 8 weeks of diet, while being
almost constant between 2 and 6 weeks. The incorporation of trans 22:6
reached 0.8% of the total fatty acids after 8 weeks of feeding which represents
almost 4% of the total 22:6 fatty acids. This incorporation is comparable to
Retinal phospholipids
cis DHA (dark squares) trans DHA (white squares)
35 1.4
30 1.2
% of total fatty acids

25 1.0
20 0.8
15 0.6
10 0.4
5 0.2
0 0
0 6 12 18 21 months of diet
Cerebral phospholipids
cis DHA (dark squares) trans DHA (white squares)
35 1.4
30 1.2

25 1.0
20 0.8
15 0.6
10 0.4
5 0.2
0 0
0 6 12 18 21 months of diet
Figure 1. Incorporation of cis and trans isomers of DHA into retinal and cerebral phospholipids of rats
fed for 21 months with cis and trans isomers of 18 :3n–3 (n=6). Adapted from Acar et al. (2006).
what was observed for the liver (4.5% of the total 22:6). Along with the
retina, synaptosomes and brain microvessels were shown to contain the
highest levels of trans 22:6 (Grandgirard et al., 1994). This fatty acid was
also observed in myelin and sciatic nerve but to a lesser extent. However, the
ratios of trans 22:6 to cis 22:6 were similar in all the tissues studied. When
the diet was deficient in linolenic acid, the incorporation of trans 22:6 was
doubled. However, the decrease in the amount of 22:6 in the tissue was not
balanced by the increase in 19-trans-22:6. Instead, a large increase in 22:5
(n–6) was observed, as has previously been described in (n–3) fatty acid
deficiency (Bourre et al., 1984).
When comparing incorporation of trans PUFA isomers in rat retina and brain
during a 21 months feeding study with trans 18:3, Acar et al. (2006) showed
that, after 6 months, the incorporation of trans DHA was higher in retinal
phospholipids since trans DHA represented 0.5% of total fatty acids whereas
it was only 0.2% in cerebral cortex. After 21 months of feeding, the level of
trans DHA in retina was 1.2% of the total fatty acids while that in the cerebral
cortex increased to 0.3% (Figure 1). Concomitantly, the amount of DHA
decreased in retina from 32% to 26% after 21 months, while it was fairly
constant in the cerebral cortex (14% of the total fatty acids). At the same time,
the amount of 22:5n–6 doubled in the retina while only increasing slightly in
the cerebral cortex. Consequently, the n–6/n–3 ratio increased in the retina of
trans 18:3 fed animals compared to controls, which were fed cis 18:3. On the
contrary, no significant modification of this ratio occurred in the cerebral
cortex. These results suggest that the mechanisms leading to the incorporation
of cis and trans fatty acids are quite different in the retina compared to the brain,
the retina being more susceptible to modification of dietary lipids.
3. β-oxidation
In experiments in which 1-14C-labelled oleic, elaidic and palmitic acids were
fed to rats, the rates of catabolism as measured by the respiratory CO2 were the
same (Coots, 1964), and similar results were found by others after intravenous
injections as albumin complexes (Ono & Fredrickson, 1964). Subsequently, in
experiments with fasted rats, uniformly 14C-labelled oleate in a soybean oil
vehicle was reportedly catabolized to CO2 to a slightly greater extent than was
elaidate but the differences were trivial (Anderson & Coots, 1967). Also in the
latter work, the 14CO2 production from mono-trans 18:2 (70%) and 9-trans,12-
trans-18:2 (68%) was found to be slightly higher than that from linoleic acid
(64%, P < 0.10). This difference was correlated with a preferential esterifica-
tion of cis,cis-18:2 (linoleic acid) into liver phospholipids. It is well established
from a number of feeding experiments with animal models that partial β-
oxidation of trans-18:1 to trans-16:1 fatty acids occurs, but this process
appears not to have been studied in a formal manner (see Wood, 1979).
More recently, the oxidative metabolism of linoleic and α-linolenic acids
and of their synthetic mono-trans isomers, 9-cis,12-trans-18:2, 9-trans,12-cis-
18:2 and 9-cis,12-cis,15-trans-18:3, 9-trans,12-cis,15-cis-18:3, was studied in
fasting rats by Bretillon et al. (1998b). A single dose of 18.5 MBq of each 1-14C
labelled fatty acid (260 mg) was orally given to the animals. The 14CO2 expired
was monitored at regular intervals over 24 h. The 14CO2 production 24 h after
oral administration of the fatty acid ranged from 55.5% to 68.7% of the
radioactivity administered for the 18:2 isomers and from 69.7% to 73.5% for
40
A
35
production per 100g of rat
9c, 12t
30 9t, 12c
9c, 12c
25
20
15
2
14CO
10
0
0 2 4 6 8 10 12 14 16 18 20 22 24
Time
Figure 2. 14CO2 recovery after oral administration of [1 14C] linoleic acid and its mono trans isomers.
Adapted from Bretillon et al. (1998).
the 18:3 fatty acids. From 6 to 24 h, 14CO2 recovery was significantly higher
after oral administration of 9-cis,12-trans-18:2 than after giving both of the
other octadecadienoic isomers (Figure 2). The 14C retention per gram of tissue
in the liver and in the heart was significantly lower after feeding 9-cis,12-trans-
18:2 than after administration of both other 18:2 isomers. The 14C retention per
gram of tissue in the muscle was significantly lower after administration of both
trans 18:2 isomers in comparison to linoleic acid.
Neither 14CO2 recoveries nor 14C retentions were significantly different after
administration of the three octadecatrienoic acids. The difference observed in
14
CO2 recovery within the dienes was probably not due to a higher specificity
of the enzymes involved in the β-oxidation sequence for the Δ12 trans double
bond. Indeed, due to the labelling of the fatty acids on the carboxyl end, 14C
values recorded in the CO2-trapping agent were only due to the first cycle of β-
oxidation. No major differences were found for the 18:3 geometrical isomers,
neither for 14C recovery in rat tissues nor for 14CO2 expired.

All unsaturated fatty acids have the potential to inhibit desaturation of di- and
polyenoic fatty acids by mechanisms that include substrate inhibition, product
inhibition, and competitive inhibition. The nature and extent of such inhibition
is dependent mainly on the position and configuration of double bonds in the

inhibitor, with monoenes tending to have lower inhibitory effects than polyenes.
In addition, cis and trans monoenoic fatty acids in particular appear to have
differing effects on desaturation of fatty acids of the n–6 and n–3 families.
However, the observed effects were small when monoenoic fatty acids were
fed as part of a diet containing higher levels of linoleate.
For example, in one study in which partially hydrogenated soybean oil, i.e.
enriched in both cis and trans monoenes, was fed to rats for up to 4 weeks, the
arachidonate levels in the phospholipids from many different tissues were
substantially reduced (Lawson et al., 1983). In fact, synthesis of 20:5(n–3) and
20:3(n–9) was accentuated by this diet, suggesting that the inhibition was
specific for fatty acids of the n–6 family. In comparable experiments in which
concentrates of cis and trans isomers were fed separately, the trans-18:1
concentrate was found to reduce the levels of 20:4(n–6) and 20:3(n–9), and to
increase those of 20:5(n–3) in liver phospholipids. In contrast, The cis-18:1
concentrate reduced the levels of 20:4(n–6), 20:5(n–3) and even of 22:6(n–3)
(Lawson et al., 1985). Much smaller effects were observed in comparable
experiments when elaidic acid per se was fed, suggesting that other trans
isomers may have a greater influence on desaturation in vivo (Astorg &
Chevalier, 1988). In experiments in which particular fatty acids were injected
into rat brain, oleate strongly inhibited desaturation and elongation of linoleate,
while elaidate, and a trans-monoene from margarine were stimulatory (Cook,
1981). However, the effects may be dependent on the levels of linoleate in the
diet, and Zevenbergen et al. (1988), for example, found small effects only of
dietary trans monoenoic fatty acids on mitochondrial fatty acid composition
when higher levels of linoleate were fed. Comparable results have been found
in a number of other studies, including a more recent one with pigs (Kummerow
et al., 2004) (reviewed by Emken, 2007). While effects of longer-chain dietary
monoenoic fatty acids on levels of polyunsaturated fatty acids have been
observed in experiments with partially hydrogenated fish oils, the separate
effects of trans as opposed to cis monoenes have yet to be determined.
Selinger and Holman (1965) showed that the di-trans 18:2 isomer, when
given to rats that have been fed a fat-free diet together with linoleic and
linolenic acids, decreased the levels of the polyunsaturated fatty acids formed
from these precursors. Subsequently, Privett et al. (1967) reported that dietary
mono-trans 18:2 did not impair the conversion of linoleic into arachidonic acid
using animals previously fed a fat-free diet. However, they confirmed that di-
trans linoleate impaired the conversion of oleic acid into 20:3(n–9) and of
linoleic into arachidonic acid (Privett et al., 1967). Anderson et al. (1975) also
showed that the presence of the mono-trans 18:2 isomers in the diet did not
affect the level of arachidonic acid in the liver and that the di-trans 18:2 isomer
exerted an inhibition of arachidonic acid synthesis only at ratios of di-trans-
18:2/linoleic acid that would be much higher than those observed in the normal
human diet. More recent studies of Berdeaux et al. (1996) also confirmed that
when fed at low levels in the diet, the mono-trans isomers of linoleic acid did
not induce any modification in the arachidonic acid content in different organs.
However, in the presence of linoleic acid at 1.1% of the calories, dietary di-
trans 18:2 may affect essential fatty acid metabolism by reducing the Δ6
desaturase activity (Schimp et al., 1982).
When fed to animals in the form of heat-isomerized linseed oil (HLO), 18:3
isomers resulted in a decrease in 20:4(n–6) and in the 20:4(n–6)/18:2(n–6) ratio
in liver phospholipids compared to what was observed for the animals fed the
fresh linseed oil (LO) (Blond et al., 1990). In vitro assays using the liver
microsomes of the rats fed the HLO and the LO diets showed that the Δ6
desaturase activity [18:2 → 18:3(n–6)] was not significantly altered. In con-
trast, the Δ5 desaturase activity [20:3 → 20:4(n–6)] was higher in the HLO
group compared to the LO group. However, the 18:3 geometrical isomers fed
to animals as a purified fraction were shown to increase the Δ6 desaturase
activity of linolenic acid when rats were fed a (n–3) deficient diet (Blond et al.,
1995). This may indicate that the presence of trans-18:3 in the diet induces
increased rates of 18:3(n–3) desaturation, related to a requirement of the liver
for cis (n–3) fatty acids.
C. Metabolism of trans isomers in vitro

A number of different models have been used to study the metabolism of TFA
in vitro, as well as their interference with the metabolism of essential fatty
acids. For example, these include liver microsomes (Berdeaux et al., 1998),
cultured glioma cells (Cook & Emken, 1990), brain homogenates (Cook,
1981), bovine endothelial cells (Loi et al., 2000) and perfused rat liver (Bretillon
et al., 1998a).
1. Incorporation into lipid classes

The position and configuration of the double bond in monoenoic fatty acids has
little effect on conversion to coenzyme A esters in vitro, the first step in the
esterification of long-chain fatty acids (Normann et al., 1981). In general,
trans-octadecenoyl coenzyme esters were shown to be good substrates for
acyl-CoA:phospholipid acyltransferases to form phosphatidylcholine or
phosphatidylethanolamine with microsomal preparations in vitro. All isomers
except the 10t- and 12t-18:1 were incorporated into position 1 preferentially,
competing for this position with saturated fatty acids. The natural 9c- and 11c-
18:1 isomers were found to be poor substrates for acylation in position 1 but
were rapidly esterified to position 2 (Reitz, et al., 1969; Marchand & Beare-
Rogers, 1978; Okuyama et al., 1972).
Using a perfused rat liver model and synthetic 14C-labelled trans-18:2 and
18:3-isomers, Bretillon et al. (1998a) demonstrated a major difference in the

acylation of these acids. About 30% of the radioactivity in the liver was
recovered in the triacylglycerols after infusion of cis,cis-linoleic acid, while
this value was higher than 50% after perfusion of the trans isomers (9c,12t and
9t,12c). For the 18:3, the 14C–label was partitioned equally between
phospholipids and triacylglycerols after perfusion with the three isomers
(9c,12c,15c; 9c,12c,15t-; and 9t,12c,15c-18:3).
For the 18:2 isomers, the lower incorporation of cis,cis-linoleic acid into the
triacylglycerols was balanced by a higher incorporation of this fatty acid into
phosphatidylcholine (PC) compared to both trans-18:2 isomers. Furthermore,
the 14C retention in phosphatidylinositol (PI) was significantly lower after the
infusion with the 9t,12c isomer. For the 18:3 fatty acids, a higher 14C retention
was found in PC but only after infusion of the 9c,12c,15t isomer. No differ-
ences were found for PI. Results of phospholipase A2 digestion of PC showed
a marked selectivity for incorporation of 9c,12c,15t-18:3 into the sn-2 position
as was observed for linoleic acid. A similar finding, which could be explained
by their comparable structural features, was obtained by Wolff et al. (1993) in
vivo. This indicates that these fatty acids would be recognized in the same
manner by the acyl-CoA:phospholipid acyltransferase.

As discussed briefly above, desaturation of trans monoenoic fatty acids is the
subject of a separate chapter in this book, so need not be discussed at length
here. In a study of chain elongations of trans-octadecenoic acid isomers using
rat liver microsomes in vitro, 7- and 9-trans-18:1 were converted to trans-9 and
trans-11-eicosenoic acids, respectively, at about 40% of the rate of conversion
of cis-9-octadecenoic to cis-11-eicosenoic acid (Kameda et al., 1980). Even
lower rates of conversion of the Δ8, Δ10, Δ11, and Δ12 trans isomers were
observed, but no elongation of the Δ4, Δ5, Δ6, Δ13, Δ14, and Δ15 trans isomers
was detected.
The Δ6 desaturation of 18:2 isomers in vitro has been studied by Cook (1981)
and Berdeaux et al. (1998), using liver microsomes, and by Bretillon et al.
(1998a) using a perfused liver model. Studies on microsomes (Berdeaux et al.,
1998) showed that the 9c,12t-18:2 isomer was a much better substrate for
desaturase than the 9t,12c isomer. In contrast, the 9t,12c-18:2 was better
elongated than 9c,12c while the amount of product formed from 9c,12t was
lower than was produced from the 9c,12c-18:2. Products of desaturation and
elongation were identified by a radio-GC detector as illustrated in Figure 3.
Furthermore, both mono-trans isomers at low concentrations only inhibit
slightly the Δ6 desaturation of the all-cis 18:2 isomer. Cook and Emken (1990)
also showed using cultured glioma cells that for dienoic isomers the presence
of a Δ12 trans bond inhibited the formation of both 20:4n–6 and 20:5n-3.
cnts
120 18:3 Δ9c,12t
100
80
60
18:3 Δ6c,9c,12t
40
20
0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
min
cnts
500
18:2 Δ9t,12c
400
300
200
100 20:2 Δ11t,14c
0
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
min
Figure 3. Top: GC analysis (radioactive signal) of the FAME from liver microsomes (5 mg protein)
incubated for 15 min with [1 14C] 9c, 12t 18:2. Bottom: GC analysis (radioactive signal) of the FAME
from liver microsomes (5 mg protein) incubated for 30 min in anaerobic conditions with [1 14C] 9t, 12c
18:2. Adapted from Berdeaux et al. (1998).
Similarly, studies of the Δ6 desaturation of trans isomers of α-linolenic acid

in vitro showed that the 9-trans isomer was less converted by microsomes
under desaturation conditions than the 15-trans isomer (Chardigny et al.,
1995), which has its Δ9 double bond with a cis configuration. The relative
conversions were 1.00, 0.54 and 0.15 for 9-cis,12-cis,15-cis-18:3, 9-cis,12-
cis,15-trans-18:3 and 9-trans,12-cis,15-cis-18:3, respectively.
In the perfused rat liver model (Bretillon et al., 1998a), data indicate that for
both 18:2 and 18:3, the geometry of the double bond greatly affected the
conversion of the fatty acids as already reported for the study using liver
microsomes in vitro. By comparison with linoleic acid, the trans geometry in
the Δ12 position greatly increased the desaturation of the precursor while the
trans geometry in the Δ9 position increased the elongation. For the 18:3
isomers, by comparison with linolenic acid the trans geometry in the Δ15
position only decreased the desaturation of the fatty acid while the trans
geometry in the Δ9 position both decreased the desaturation and increased the
elongation of the precursor into some 20:3 end products. The data on elonga-
tion together with those obtained on the Δ6 desaturation are well correlated
with what is observed in vivo (see above) (Privett et al., 1967; Berdeaux et al.,
1996; Beyers & Emken, 1991).
Bovine endothelial cells were chosen by Loi et al. (2000) to study the
metabolism of the 20:5 isomers formed by desaturation and elongation of the
different 18:3 geometrical isomers as previously described (20:5,11t; 20:5,17t,
and 20:5,11t,17t). Data showed that trans 20:5 can be incorporated in endothe-
lial cells and further elongated into trans-22:5 and trans-24:5 isomers. For
example, the 20:5 Δ17t which is the major trans 20:5 isomer found in human
tissues was as much converted into two unidentified fatty acids in the same
manner as for the elongation of 20:5n–3 into 22:5n–3 and 24:5n–3. The effect
on prostacyclin synthesis is described below.
3. β-Oxidation
Animals degrade fatty acids by β-oxidation systems that operate separately in
mitochondria and peroxisomes. These organelles utilize different enzyme
systems, which have some specificity for particular fatty acid structures. The
mitochondrial system deals with most of the more conventional fatty acids and
is the more important quantitatively. It has been estimated that about 70 to 85%
of the oxidation of dietary fatty acid occurs by this route (Masters and Crane,
1984; Osmundsen et al., 1991). Oxidation usually goes nearly, if not entirely,
to completion with carbon dioxide as the end product. In contrast, the peroxi-
somal β-oxidation system oxidizes those fatty acids longer than 24 carbons
together with some less abundant substrates including branched-chain and
dicarboxylic fatty acids. In this instance, chain-shortening of fatty acids by only
two or three β-oxidation cycles tends to occur so that trans-18:1 fatty acids
yield trans-14:1 and 12:1 products. The latter can then be subjected to
elongation with formation of 16:1 isomers. Emken (2007) has comprehen-
sively reviewed the enzymology of the process in respect to TFA.
Somewhat conflicting results of the relative rates of oxidation of oleate and
elaidate have been obtained in liver perfusion experiments, for which no
convincing explanation is available. On the other hand, experiments with
mitochondrial preparations from rat heart, strongly suggested that elaidic acid
was oxidized more slowly than oleic acid, a finding that was attributed to the
fact that part of the normal mechanism of β-oxidation involves a cis-3-12:1
intermediate, which is isomerized prior to the next step. In contrast, a trans
intermediate requires other enzymatic reactions before oxidation can continue.
In a classic study, the cis and trans isomers of 4-18:1 to 16-18:1, all of which
can be present in partially hydrogenated soybean oil, were compared as
substrates for β-oxidation in the form of the CoA esters by rat heart and liver
mitochondria in vitro (Lawson & Holman, 1981). With a few exceptions only,
both heart and liver mitochondria oxidized the cis isomers, especially 9- and
11-18:1, significantly more rapidly than their respective trans isomers. The cis
isomers were found to be catabolized in a similar manner by heart and liver,
with the even-positional cis isomers being oxidized much more slowly than
adjacent odd-positional isomers, which were in fact oxidized at approximately
the same rate. Very different results were obtained with the trans isomers, on
the other hand. In this instance, liver mitochondria tended to oxidize the even-
positional trans isomers much more rapidly than the adjacent odd-positional
isomers, and at almost the same rate as stearoyl-CoA. This was only true for the
trans isomers in which the double bond was located near the middle of the acyl
chain with heart mitochondria. Again, these results have been interpreted in
terms of the activities of two key enzymes in the β-oxidation pathway, the 3-
hydroxyacyl-CoA epimerase and Δ3-cis-Δ2-trans-enoyl-CoA isomerase.
However, it was recently demonstrated that when elaidoyl-CoA was incu-
bated with rat mitochondria from liver or heart a major metabolite,
5-trans-tetradecenoyl-CoA accumulated in the mitochondrial matrix, although
no analogous metabolites were detected with oleoyl-CoA or stearoyl-CoA as
substrates. Following its conversion to 5-trans-tetradecenoylcarnitine, this
partially degraded fatty acid was able to escape from mitochondria (Yu et al.,
2004). Whether the magnitude of this reaction is sufficient to explain the
reduced rate of conversion of elaidate to CO2 relative to oleate by mitochon-
dria, which is observed in vitro, is not yet known.
Increased insulin production in islet cells exposed to trans as opposed to cis-
monoenoic fatty acids may have been a consequence of the reduced rate of
oxidation of the former (Alstrup et al., 2004).
It is well established that peroxisomes have a preference for the oxidation of
the less common fatty acids, and it appears that trans monoenoic fatty acids
may fall into this category. Certainly rat liver peroxisomes in vitro have a
distinct preference for the oxidation of trans in comparison to cis fatty acids.
For example, a recent study demonstrated that the rate of peroxisomal oxida-
tion of elaidic acid was about 2.5-fold greater than that of oleic acid. This
appears to reflect the substrate specificity of the peroxisomal β-oxidation
apparatus as opposed to the specificity of entry into the peroxisomal compart-
ment (Guzmán et al., 1999). Peroxisomal oxidation is especially important for
the oxidation of the longer-chain TFA of the kind found in partially hydrogen-
ated fish oils (Flatmark et al., 1988).
In order to explain the preferential oxidation of 9c,12t-18:2 compared to
linoleic acid and to the 9t, 12c-18:2 isomer, Bretillon and colleagues (1998a,
1998b) studied all the different steps involved in the mitochondrial oxidation.
In a first step, hepatic mitochondria from rats were incubated in the presence of
the three radio-labelled fatty acids studied (9c,12c-, 9c,12t- and 9t,12c-18:2).
Data revealed that during the 40 min incubation, fatty acid oxidation ranged
from 60 to 80%, the 9c,12t-18:2 being more oxidized than the other two
isomers as was also demonstrated in vivo. The measure of the acyl-CoA
synthase activity indicated that this enzyme had more affinity for the mono-
trans isomers compared to linoleic acid. However the carnitine palmitoyl
transferase I had the same affinity for 9c,12c-acyl CoA, the 9c,12t-acyl CoA
and the 9t,12c-acyl CoA. These results did not permit an explanation of the
difference between the two TFA geometrical isomers.

It has been recognized for more than 40 years that the metabolism of essential
fatty acids is inhibited by other unsaturated fatty acids both in vivo and in vitro.
In particular, linoleic acid strongly inhibits the desaturation and elongation of
α-linolenic acid and vice versa, but monoenoic fatty acids appear to have minor
effects only. With rat liver microsomes, when single trans-18:1 isomers were
used in studies of the desaturation of palmitic acid to palmitoleic (Δ9-desaturase),
linoleic to γ-linolenic (Δ6-desaturase) and eicosa-8,11,14-trienoic to arachi-
donic acid (Δ5-desaturase), some marked inhibitory effects were observed in
each instance, which were specific to particular isomers (Mahfouz et al.,
1980). However, when stearic acid was added to the incubation medium, a
condition that was arguably more physiological, weak inhibition of the Δ9-
desaturase only was observed.
As described above, numerous studies have been carried out to determine the
effects of feeding trans polyunsaturated isomers on the metabolism of other
unsaturated fatty acids, but much less literature is available on the in vitro
effects of these fatty acid isomers. In vitro experiments using cultured glioma
cells demonstrated that the presence of a trans ethylenic bond at the Δ12
position in a dienoic acid competitor inhibited the formation of 5c,8c,11c,14c-
20:4 from 9c,12c-18:2 (Cook & Emken, 1990). It was also shown earlier that
dietary 9c,12t-18:2 and 9t,12c-18:2 did not modify the levels of 5c,8c,11c,14c-
20:4 in rat liver lipids but induced changes in the Δ6 desaturation of the
9c,12c-18:2 by liver microsomes in vitro (Berdeaux et al., 1996). Furthermore,
assays using liver microsomes in vitro from rats fed a fat-free diet showed that
9c,12t-18:2 was a more effective inhibitor than 9t,12c-18:2 (Berdeaux et al.,
1998).
Blond et al. (1990) used liver microsomes of rats fed a heated oil which
contained geometrical isomers of linolenic acid and showed that hepatic Δ6
desaturase activity was not significantly modified while the Δ5 desaturase
activity was increased. However, great care must be taken in interpreting these
results considering that rats were fed a heated fat which usually contains a
variety of components. For the 18:3 geometrical isomers, it was shown that the
9t,12c,15c-18:3 was a weak inhibitor of the conversion of linolenic into
stearidonic acid in vitro as were other geometrical isomers having a trans bond
in the Δ9 position as 9t-18:1 and 9t,12c-18:2 (Chardigny et al.,, 1995).
D. Effects on eicosanoid production

The eicosanoids comprise a large number of distinct fatty acids derived
primarily from arachidonic acid by the action of cyclooxygenases and
lipoxygenases, which have a wide range of functions in tissues at low levels,
especially in relation to inflammation. Dietary fatty acids, including those with
trans double bonds, could potentially impact upon eicosanoid metabolism by
inhibiting the relevant enzymes or by reducing the availability of the substrate
fatty acids.
Two studies with aortal pieces and platelets from rats in vitro suggested that
there was no effect of trans monoenoic fatty acids on eicosanoid production,
provided that there was a sufficiency of linoleic acid in the diet (> 2% of the
energy requirement) (Zevenbergen et al., 1989; Mahfouz & Kummerow, 1999).
In a human study, trans monoenes and stearic acid were found to have
comparable effects upon prostacylin production (Turpeinen et al., 1998).
Similarly, the enzymes involved in thromboxane production in human volun-
teers were not affected significantly by dietary elaidic acid, although
trans-linoleate isomers did have an appreciable inhibitory effect (Stachowska,
2004).
In the late 1970s, studies looking at the effect of trans 18:2 isomers were
carried out by the group of Kinsella (for details see the review by Sebedio &
Chardigny, 1998). For example it was shown that the 9t,12t-18:2 can alter the
production of arachidonic acid metabolites in rat platelets [TxB 2, PGF2α,, 12-
hydroxyeicosatetraenoic acid (12-HETE )] only when the dietary 9-trans,12-
trans-18:2 level is similar or greater than the level of linoleic acid.(Hwang
et al., 1978) Most of the work carried out during that period was dealing with
the di-trans isomer of linoleic acid. The major drawback of these studies was
200
(A) a
pmol 6-keto PGF1α/mg cell
150
a
O-diene
100
H-diene
50
a
a
0
Basal Stimulated
Figure 4. Prostacyclin release by endothelial cells from endogenous arachidonic acid (mean ± SD).
Cells were incubated with 100 µM of O diene or H diene for 3 days. Values having the same superscript
are significantly different. O diene is a diene fraction isolated from a soybean oil and H diene is a diene
fraction isolated from a partially hydrogenated soybean oil. Adapted from Kummerow et al. (2007).
that only very minor quantities of this di-trans isomer have been detected in
commercial fats and oils while the major isomers in the human diet were the
9c,12t- and the 9t,12c-18:2.(Entressangles, 1986). Consequently, the large
quantities of the fatty acids used in these studies did not reflect the human diet.
Effect of trans dienoic isomers present in hydrogenated fat on prostacyclin
release by endothelial cells in the presence of high level of linoleic acid was
recently reported by Kummerow et al. (2007). A mixture of fatty acids isolated
from partially hydrogenated soybean oils (diene fraction) were shown to inhibit
the release of prostacyclin from endothelial cells compared to cells incubated
with a diene fraction isolated from a soybean oil (Figure 4). The diene fraction
isolated from soybean oil contained 96% of linoleic acid, 4% 18:1 and no trans
18:2 isomers while the mixture isolated from partially hydrogenated soybean
contained 69.4% of linoleic acid, 3.7% 18:1 and 26.7% of a mixture of trans
18:2 isomers. The data were explained by an inhibition of the conversion of
linoleic to arachidonic acid, as the quantity of arachidonic acid decreased and
that of linoleic acid increased in the cells incubated with the diene fraction
isolated from the partially hydrogenated oil. Some effects of a trans-monoene
fraction on prostacyclin release were also observed, but the effect was much
smaller than that for the trans-diene fraction.
Mono-trans isomers of 18:2(n–6) and 18:3(n–3) are converted into higher
metabolites (see above), including C20 and C22 PUFA containing one (or two)
trans double bonds, which may then be incorporated in phospholipids. Conse-
quently, the effects of geometrical isomers of C20 PUFA on eicosanoid produc-

tion have been assessed, generally using fully synthetic molecules (Chardigny
et al., 1995; Berdeaux et al., 1996; Loi et al., 1998; Loi et al., 2000) but also
fatty acids purified from rat liver lipids fed the trans precursor (i.e. 18:3 trans
isomers) (O’Keefe et al., 1990).
The effects of 14-trans-20:4 on the metabolism of arachidonic acid in
eicosanoids was reported by Berdeaux et al. (1996) using rat platelets. It was
shown that this structural analogue of 20:3(n–9) induced an inhibition of the
conversion of 20:4(n–6) into thromboxane B2 or hydroxyheptadecatrienoic
acid (HHT) (cyclooxygenase pathway) and increased the production of 12-
hydroxyeicosatetraenoic acid (12-HETE) through the 12-lipoxygenase pathway.
These data were well correlated to platelet aggregation and were different from
those obtained with 20:3(n–9) (Figure 5). Mead acid [all-cis-20:3(n–9)] only
induced a slight decrease in the 12-HETE production. These data confirm a
specific effect of the trans double bond located at the Δ14 position. Moreover,
using radiolabelled 14-trans-20:4, it was shown that this was metabolized by
platelets into two metabolites, One of them is probably a product of the 12-
lipoxygenase pathway, as its production is lowered when platelets are
pre-incubated with baicalein, a selective 12-lipoxygenase inhibitor (Berdeaux,
1996). The production of this unknown metabolite is greatly enhanced when
arachidonic acid or 12-hydroperoxyeicosatetraenoic acid are present in the
incubation medium. The data suggest that a sufficient ‘peroxide tone’ is needed
to produce this unknown metabolite. They enhance the hypothesis about the
origin of the metabolite, which might be a trans isomer of 12-HETE. Berdeaux
et al. (1996) demonstrated that 14-trans-20:4 is also metabolized into another
product, but to a lesser extent. This second unknown metabolite might be
produced by the cyclooxygenase pathway, but its exact origin has not been
reported.
For the trans isomers of long-chain (n–3) PUFA, O’Keefe et al. (1990),
using washed human platelets, demonstrated that 19-trans-22:6 appeared to be
an inhibitor of the cyclooxygenase pathway, as assessed by the thromboxane
TxB2 and hydroxyheptadecatrienoic acid production. On the other hand, 17-
trans-20:5 inhibited the 12-lipoxygenase pathway. At the same time,
17-trans-20:5 was less of an anti-aggregant than eicosapentaenoic acid, whereas
19-trans-22:6 and the natural all-cis acid had similar anti-aggregant effects.
These specific effects of 19-trans-22:6 on the cyclooxygenase pathway and of
17-trans-20:5 on the lipoxygenase pathway were not understood, as they were
not well correlated to the platelet aggregation response. Later, the effects of
these isomers were studied after incorporation of the specific fatty acid into
human platelet lipids (Chardigny et al., 1995). Thrombin and collagen stimu-
lation of platelets enriched in 20:5 or 22:6 PUFA showed that the occurrence of
a trans double bond at the (n–3) position decreased the anti-aggregant effect of
both 20:5 and 22:6 fatty acids. However, the metabolism of arachidonic acid in
Figure 5. Left: inhibition of rat platelet aggregation by increasing amount of arachidonic acid,
20:3n–9, or 14t 20:4. Right: effect of 14t 20:4 on [1 14C] arachidonic acid metabolism by rat platelets.
Adapted from Berdeaux et al. (1996).
hydroxyheptadecatrienoic acid , 12-HETE, and TxA2 was apparently not modi-

fied.
Further studies on the different trans 20:5 isomers were carried out using rat
platelets, which were shown to be deprived of the trans C20 and C22 n–3 isomers
(Loi et al., 1998). With regards to the metabolism of arachidonic acid, the three
trans 20:5 isomers studied (Δ11, Δ11,17 and Δ17) reduced the formation of the
cyclooxygenase metabolites and enhanced that of 12-HETE. Furthermore, the
Δ17 isomer was metabolized by the cyclooxygenase and 12-lipoxygenase
pathways into compounds that were not further identified.
These trans-20:5 isomers can also be incorporated in endothelial cell
phospholipids (Loi et al., 2000) and further elongated into trans-22:5 and
trans-24:5. Their presence in the cells affects the availability of arachidonic
acid and then inhibits the prostacylin synthesis as a competition between
arachidonic acid and the trans-20:5 isomers at the level of the cyclooxygenase
pathway which seems to take place considering that data (Figure 6) indicated
that the Δ17 isomer was metabolized by this enzyme.
A few years ago, the presence of trans isomers of arachidonic acid was
reported in human plasma following inflammation, for example (Balazy, 2000;
Zghibeh et al., 2004). These trans arachidonic acid isomers may be formed
within the phospholipids of biological membranes by a free radical process
initiated by NO2. Four trans arachidonic acid isomers have been identified so
far, the trans Δ5, Δ8, Δ11 and Δ14 (Zghibeh et al., 2004). Using rat liver
microsomes, Roy et al. (2005), demonstrated that the trans Δ5 arachidonic acid
isomer could be converted into a complex mixture of three classes of oxidized
products amongst which were four epoxides (EET), their respective hydrolysis
products (dihydroxyeicosatrienoic acids) and several HETE. The identifica-
tion of the metabolites formed from trans arachidonic acid should warrant
(dpm) C20:4 n-6

10000
A
8000
6000
4000
6-keto PGF1α
2000
PGD2
0
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 ml
(dpm)
C20:5 Δ17t
5000
B
4000
3000
2000
P1
1000
CP2
0
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 ml
Figure 6. HPLC profile of radiolabelled eicosanoids from endothelial cells incubated with A: [1 14C]
arachidonic acid, and B: [18 14C] 20:5 Δ17t. Adapted from Loi et al. (2000).
further studies to uncover the potential role of such molecules as mediators of

the activity of TFA.
E. Human studies
In this part of the chapter, we will not describe the effects of trans monoethylenic
fatty acids on lipoprotein metabolism in humans as this is dealt with elsewhere
in this volume, but we will discuss the presence and the transformation of these
monoethylenic isomers in the tissues as well as their effects on the metabolism
of other fatty acids. The reader will find most of the earlier literature in reviews
by Emken (2007) and Holmer (1998). To summarize, TFA have been detected
in different tissues such as adipose, liver, heart, aorta and milk, as a conse-
quence of their presence in the human diet (Aitchison et al., 1977; Jensen,
1999). Feeding deuterated fatty acid isomers has shown that plasma cholesteryl
esters content exhibited a 4 to 1 ratio of oleic to elaidic acid while all plasma
phospholipids selectively incorporated elaidic compared to oleic acid.
Studies of human milk composition of two cohorts comprising 103 mothers
and 769 mothers respectively showed that milk TFA were inversely related to
milk linoleic and linolenic acid contents (Innis & King, 1999; Szabo et al.,
2007). Furthermore, in the larger cohort, milk TFA were also inversely related
to the availability of long-chain PUFA (arachidonic and docosahexaenoic
acids) in mature human milk. It was also reported that there could be a PUFA
deficiency in infants due to the presence of TFA (reviewed by Holmer, 1998).
For example, the effects of TFA, including trans mono- and di-enoic isomers
on the biosynthesis of long-chain polyunsaturated fatty acids in premature
infants were studied earlier by Koletzko (1992). Trans octadecenoic acid and
total trans isomers, including 9-cis,12-trans-18:2, 9-trans,12-cis-18:2 and 9-
trans,12-trans-18:2 in plasma lipid fractions were not related to either linoleic
or linolenic acids but these were correlated inversely to the long-chain (n–3)
and (n–6) polyunsaturated acids. These data may indicate a potential impair-
ment of essential fatty acid metabolism.
More recent data confirmed the earlier observations as high exposure to TFA
is related to lower levels of docosahexaenoic acid, and consequently, TFA may
have adverse effects on growth and development through interfering with
essential fatty acid metabolism (Decsi, 2003; Innis, 2006). Results of two birth
cohorts in the Netherlands demonstrated that birth length and head circumfer-
ence were significantly and negatively related to the trans-18:1 concentration
in phospholipids isolated from umbilical plasma, umbilical arterial walls, or
umbilical venous walls (Hornstra et al., 2006).
Different studies have also shown that TFA are not only incorporated in
human tissue but also that a conjugated linoleic acid isomer, the 9c,11t-18:2
(CLA) can be synthesized from vaccenic acid, the 11t-18:1 isomer present in
large quantities in ruminant milk and fat (Turpeinen et al., 2002; Mosley et al.,
2006, Kuhnt et al., 2006a). Vaccenic acid was shown to be converted into
CLA, the conversion rate being 20% on average. This indicates that this
endogenous synthesis is likely to contribute significantly to the total amount of
CLA in the body (Turpeinen et al., 2002).
Using 11t vaccenic-1-13C acid, Mosley et al., (2006) confirmed this in vivo
synthesis of rumenic acid in a population of lactating women. However, it
would be necessary to carry out a larger study to define more accurately the
contribution of the Δ9 desaturase activity in mammary and non-mammary
tissues to the endogenously synthesized 9c,11t-18:2. The present results seem
to indicate that about 10% of the CLA present in milk would come from a
synthesis in the mammary gland. In a recent study, Kuhnt et al. (2006a)
evaluated this in humans, the endogenous conversion of vaccenic acid into
9c,11t-CLA contributes approximately 25% to the human CLA pool. Further-
more, a conversion of another trans isomer the 12t-18:1 into 9c, 12t-18:2 was
not detected.
Interestingly, the urinary excretion of 8-isoprostaglandin F2α, corresponding
to arachidonic acid oxidation, increased significantly for the volunteers in-
cluded in the intervention study of Turpeinen et al. (2002). A similar effect on
the production of 8-isoprostaglandin F2α was observed by Kuhnt et al. (2006b)
after feeding a large quantity of a mixture of 11t- and 12t-18:1 isomers.
Association between higher intake of TFA and higher concentration of urine
F isoprostanes was also found in the Study of Women’s health Across the
Nation (SWAN), which included 1610 participants (Tomey et al., 2007).
Geometrical linoleic acid isomers were reported in human tissues, including
milk (Boatella et al., 1993; Chardigny et al., 1995, Chen et al., 1995, Koletzko
et al., 1988) adipose tissue (Adlof & Emken, 1986; Boué et al., 2000; Baylin
et al., 2002), kidney, heart and liver (Adlof & Emken, 1986). Adlof and Emken
(1986) reported that adipose tissue had the highest concentration of trans 18:2
(9-cis,12-trans and 9-trans,12-cis) while the amount present in the brain was
too small to be quantified. Only traces of the di-trans isomers were detected.
While 18:3 geometrical isomers were reported in human milk (Chardigny
et al., 1995), none of the 18:3 isomers were detected, neither by Boué (2000)
nor by Adlof and Emken, in the other human tissues. Longer-chain (n–3)
polyunsaturated fatty acids, especially the metabolites of 9-cis,12-cis,15-trans-
18:3, were also detected in human platelets (Chardigny et al., 1993).
The metabolism in humans of the 12-cis,15-trans-18:2, which can be formed
from linolenic acid by catalytic hydrogenation, was studied by Emken
et al.(1987) using triacylglycerols containing deuterium-labelled fatty acids.
Analyses performed on two human subjects indicated that the turnover for
triacylglycerols, cholesterol esters, phosphatidylethanolamine and
phosphatidylcholine of all the deuterated fatty acids tested (16:0, 18:0, 18:1,
9-cis,12-cis-18:2 and 9-cis,15-trans-18:2) were similar. Furthermore, the select-
ivity ratios for the deuterium-labelled isomers incorporated into the
triacylglycerol and free fatty acid classes were small compared to what was
observed for the phospholipids. It appeared that the 9-cis,15-trans-18:2 must
have been utilized mainly for energy, considering that most of the selectivity
ratios for the phospholipids are negative and that no large positive ratios for the
neutral lipids were observed.
To our knowledge, only one intervention study using linolenic acid geo-
metrical isomers has been carried out on humans (Sebedio et al., 2000;
Vermunt et al., 2001). After a wash out period without TFA, human volunteers
either received or not some isomerized rapeseed oil which contained trans
isomers of linolenic acid at the level of 0.6 % of the total energy for 6 weeks.
At the end of the nutritional intervention, parameters connected to cardiovas-
cular risk factors were measured. It was shown that the trans isomers of
linolenic acid may increase the ratios of plasma LDL-cholesterol to HDL-
Figure 7. Fatty acid composition of platelets from human volunteers who received (high trans) or not
(low trans) 0.6% energy as trans isomers of linolenic acid for six weeks. T0–T6: 6 weeks wash out period
and T6–T12: experimental period. Adapted from Sebedio et al. (2000).
–5
–10
Expired 13CO2
12cis-18:2
–15 12trans-18:2
–20
–25
Time (min)
–30
0 600 1200 1800 2400 3000
Figure 8. 13CO2 enrichment (δ ‰ vs PDB) after ingestion of linoleic acid (12 cis 18:2) or the mono-
trans isomer (12 trans 18:2). Adapted from Bretillon et al. (2001).
cholesterol and total cholesterol to HDL-cholesterol, leading to a more athero-

genic profile. This study also demonstrated that humans may also metabolize
one of the trans 18:3 isomers present in the diet, the 9-cis,12-cis,15-trans-18:3
which seems to be the preferential substrate for the conversion into 5-cis,8-
cis,11-cis,14-cis,17-trans-20:5. However, the conversion did not seem to
operate beyond that point as no 7-cis,10-cis,13-cis,16-cis,19-trans-22:5 and 4-

cis,7-cis,10-cis,13-cis,16-cis,19-trans-22:6 were observed (Figure 7).
Furthermore feeding to humans 18:3 isomers at 0.6% energy did not seem to
affect the conversion of n–6 and n–3 fatty acids as no variation of 20:4n–6 and
22:6n–3 were observed under these conditions.
The effects of these 18:3 isomers on the Δ6 desaturation of linoleic acid was
further studied by Scrimgeour et al., (2001) using 3-oleyl, 1,2-(U-13C)-labelled
linoleic acid (15 mg twice daily for 10 days) in 7 healthy men from the
TransLine study (Sebedio et al., 2000). The data indicated that a diet rich in
trans α-linolenic acid at 0.6% energy does not inhibit the conversion of linoleic
acid to dihomo-γ-linolenic and arachidonic acid in healthy middle-aged men
consuming a diet rich in linoleic acid.
Utilization of synthetic triacylglycerols containing 1-13C linoleic, linolenic
or geometrical isomers of these essential fatty acids showed that all these
isomers could be used for energy production (Bretillon, 2001). Cumulative
oxidation over 8 h of linoleic acid, 9-cis,12-trans-18:2, α-linolenic acid, and 9-
cis,12-cis,15-trans-18:3 were, respectively, 14%, 25% 24%, and 23% of the
oral load showing that isomerization increases postprandial oxidation of
linoleic acid but not of linolenic acid in men (Figure 8).
References
Acar, N, Bonhomme, B, Joffre, C, Bron, AM, Creuzot-Garcher, C, Doly, M and Chardigny,
JM (2006) The retina is more susceptible than the brain and the liver to the incorporation
of trans isomers of DHA in rats consuming trans isomers of alpha-linolenic acid. Reprod.
Nutr. Dev., 5, 515–525.
Ackman RG, Hooper SN and Hooper DL (1974) Linolenic acid artefacts from the deodorisation
of oils. J. Am. Oil Chem. Soc., 51, 42–49.
Adlof, RO and Emken, EA (1986) Distribution of hexadecenoic, octadecenoic and octadeca-
dienoic acid isomers in human tissue lipids. Lipids, 21, 543–547.
Aitchison, JM, Dunkley, WL, Canolty, NL and Smith, LM (1977) Influence of diet on trans
fatty acids in human milk. Am. J. Clin. Nutr., 30, 2006–2015.
Alstrup, KK, Brock, B and Hermansen, K (2004) Long-term exposure of INS-1 cells to cis
and trans fatty acids influences insulin release and fatty acid oxidation differentially.
Metabolism, 53, 1158–1165.
Anderson, RL and Coots, RH (1967) The catabolism of the geometric isomers of uniformly
14C-labeled δ9-octadecenoic acid and uniformly 14C-labeled δ9,12-octadecadienoic acid
by the fasting rat. Biochim. Biophys. Acta, 144, 525–531.

Anderson, RL, Fullmer, CS and Hollenbach, EJ (1975) Effects of the trans isomers of linoleic
acid on the metabolism of linoleic acid in rats. J. Nutr., 105, 393–400.
Astorg, PO and Chevalier, J (1988) Effects of elaidic acid on polyunsaturated fatty acids of
heart and liver phosphatidylcholine and phosphatidylethanolamine in rats fed high or low
levels of linolenic acid. Nutrition Reports International, 38, 885–895.
Balazy, M (2000) Trans-arachidonic acids: new mediators of inflammation. J. Physiol.
Pharmacol., 51, 597–607.
Baylin, A, Kabagambe, EK, Siles, X and Campos, H (2002) Adipose tissue biomarkers of
fatty acid intake. Am. J. Clin. Nutr., 76, 750–757.
Berdeaux, O (1996) Les isomers monotrans de l’acide linoleique. Effets nutritionnels,

synthèse Métabolisme. PhD thesis, Université de Bordeaux I, France.
Berdeaux, O, Sébédio, JL, Chardigny, JM, Blond, JP, Mairot, T, Vatèle, JM, Poullain, D and
Noël, JP (1996) Effects of trans n–6 fatty acids on the fatty acid profile of tissues and liver
microsomal desaturation in the rat. Grasas y Aceitas, 47, 86–99.
Berdeaux, O, Blond, JP, Bretillon, L, Chardigny, JM, Mairot, T, Vatele, JM, Poullain, D and
Sebedio, JL (1998). In vitro desaturation or elongation of mono-trans isomers of linoleic
acid by rat liver microsomes. Mol. Cell. Biochem., 185, 17–25.
Beyers, EC and Emken, EA (1991) Metabolites of cis,trans, and trans,cis isomers of linoleic
acid in mice and incorporation into tissue lipids. Biochim. Biophys. Acta, 1082, 275–284.
Blank, ML and Privett, OS (1963) Studies on the metabolism of cis,trans isomers of methyl
linoleate and linolenate. J. Lipid Res. 4, 470–476.
Blond, JP, Chardigny, JM, Sébédio, JL and Grandgirard, A (1995) Effects of dietary 18:3 n–
3 trans isomers on the δ6 desaturation of α-linolenic acid. J. Food Lipids, 2, 99–106.
Blond, JP, Henchiri, C, Precigou, P, Grandgirard, A. and Sebedio, JL (1990) Effect of 18:3
n–3 geometrical isomers of heated linseed oil on the biosynthesis of arachidonic acid in
rat. Nutr. Res., 10, 69–79.
Boatella, J, Rafecas, M, Codony, R, Gibert, A, Rivero, M, Tormo, R, Infante, D and Sánchez-
Valverde, F (1993) Trans fatty acid content of human milk in Spain. J. Pediatr.
Gastroenterol. Nutr., 16, 432–434.
Boué, C, Combe, N, Billeaud, C, Mignerot, C, Entressangles, B, Thery, G, Geoffrion, H,
Brun, JL, Dallay, D and Leng JJ (2000) Trans fatty acids in adipose tissue of French
women in relation to their dietary sources. Lipids, 35, 561–566.
Bourre, JM, Pascal, G., Durand, G, Masson, M, Dumont, O and Piciotti, M (1984) Alterations
in the fatty acid composition of rat brain cells (neurons, astrocytes, and oligodendrocytes)
and of subcellular fractions (myelin and synaptosomes) induced by a diet devoid of n–
3 fatty acids. J. Neurochem., 43, 342–348.
Brenner, RR (1971) The desaturation step in the animal biosynthesis of polyunsaturated fatty
acids. Lipids, 6, 567–571.
Bretillon, L (1998) Approches du metabolisme des acides gras trans polyinsaturés chez
l’homme et chez le rat. PhD thesis, University of Bordeaux, No.1984, 231 pages.
Bretillon, L, Chardigny, JM, Noel, JP and Sebedio, JL (1998a) Desaturation and chain
elongation of [1-C14]mono-trans isomers of linoleic and alpha-linolenic acids in perfused
rat liver. J. Lipid Res., 39, 2228–2236.
Bretillon, L, Chardigny, JM, Sebedio, JL, Poullain, D, Noel, JP and Vatele, JM (1998b)
Oxidative metabolism of [1-C14] mono-trans isomers of linoleic and alpha-linolenic acids
in the rat. Biochim. Biophys. Acta, 1390, 207–214.
Bretillon, L, Chardigny, JM, Sébédio, JL, Noel, JP, Scrimgeour, CM, Fernie, CE, Loreau, O,
Gachon, P and Beaufrère, B (2001) Isomerization increases the postprandial oxidation of
linoleic acid but not α-linolenic acid in men. J. Lipid Res., 42, 995–997.
Chardigny, JM, Blond, JP, Saget, L, Eynard T, Sebedio, JL, Berdeaux, O, Poullain, D, Vatele,
JM and Noel JP (1995) Influence of a Δ9 trans ethylenic bond on the Δ6 desaturation of
linolenic acid. Oils-Fats-Lipids 1995 (WAM Castenmiller, ed.) Proceedings of the 21st
World Congress of the International Society for Fat Research, The Hague, PJ Barnes &
Associates, Bridgwater, UK, pp.219–221.
Chardigny, JM, Bron, A, Sébédio, JL, Juaneda, P and Grandgirard, A (1994) Kinetics of
incorporation of n–3 trans fatty acid in rat retina. Nutr. Res., 14, 909–917.
Chardigny, JM, Sébédio, JL, Grandgirard, A, Martine, L, Berdeaux, O and Vatèle, JM (1996)
Identification of novel trans isomers of 20:5n–3 in liver lipids of rats fed a heated oil.
Lipids, 31, 165–168.
Chardigny JM, Sebedio, JL, Juaneda, P, Vatele, JM, and Grandgirard, A (1993) Occurrence
of n–3 trans polyunsaturated fatty acids in human platelets. Nutr. Res. 13, 1105–1111.
Chardigny, JM, Wolff, RL, Mager, E, Sebedio, JL, Martine, L, Juaneda P and Grandgirard,
A (1995) Trans mono and polyunsaturated fatty acids in human milk. Eur. J. Clin. Nutr.,
49, 523–531.
Chen, ZY, Pelletier, G, Hollywood, R and Ratnayake WMN (1995) Trans fatty acid isomers
in Canadian human milk. Lipids, 30, 15–21.
Cook, HW (1981) The influence of trans acids on desaturation and elongation of fatty acids
in developing brain. Lipids, 16, 920–926.
Cook, HW and Emken EA (1990) Geometric and positional fatty acid isomers interact
differently with desaturation and elongation of linoleic and linolenic acids in cultured
glioma cells. Biochem. Cell Biol., 68, 653–660.
Coots, RH (1964) A comparison of the metabolism of elaidic, oleic, palmitic, and stearic acids
in the rat. J. Lipid Res., 5, 468–472.
Decsi, T (2003) Nutritional relevance of trans isomeric fatty acids in human milk. Acta
Paediatr., 92, 1369–1371.
Dutton, HJ (1979) Hydrogenation of fats and its significance In: Geometrical and Positional
Fatty Acid Isomers (EA Emken and HJ Dutton, eds), AOCS Press, Champaign, Illinois,
USA, pp.1–16.
Emken, EA (1983) Biochemistry of unsaturated fatty acid isomers. J. Am. Oil Chem. Soc.,
60, 995–1004.
Emken, EA (1984). Nutrition and biochemistry of trans and positional fatty acid isomers in
hydrogenated oils. Ann. Rev. Nutr., 4, 339–376.
Emken, EA (1986) Absorption of deuterium-labeled cis- and trans-octadecenoic acid
positional isomers in man. In: Fat Absorption, Volume 1 (A K uksis, ed), CRC Press, Boca
Raton, Florida, USA, pp.159–175.
Emken, E (2007) Metabolism of trans and cis fatty acid positional isomers compared to non-
isomeric fatty acids. In: Trans Fats in Foods (GR List, WMN Ratnayake and D
Kritchevsky, eds), AOCS Press, Champaign, Illinois, USA, pp. 59–95.
Emken, EA, Rohwedder, WK, Adlof, RO, Rakoff, H and Gulley, RM. (1987) Metabolism in
humans of cis-12,trans-15-octadecadienoic acid relative to palmitic, stearic, oleic and
linoleic acids. Lipids, 22, 495–504.
Entressangles, B (1986) Survey of edible trans fatty acids. Rev. Fr. Corps Gras, 33, 47–58.
Flatmark, T, Nielsson, A, Kvannes, J, Eikhom, TS, Fukami, MH, Kryvi, H and Christiansen,
EN (1988) On the mechanism of induction of the enzyme systems for peroxisomal β-
oxidation of fatty acids in rat liver by diets rich in partially hydrogenated fish oil. Biochim.
Biophys. Acta, 962, 122–130.
Fournier, V, Destaillats, F, Hug, B, Golay, PA, Joffre, F, Juaneda, P, Semon, E, Dionisi, F,
Lambelet, P, Sebedio, JL, and Berdeaux, O (2007) Quantification of eicosapentaenoic
and docosahexaenoic acid geometrical isomers formed during fish oil deodorization by
gas-liquid chromatography. J. Chromatogr. A, 1154, 353–359.
Fournier, V, Juaneda, P, Destaillats, F, Dionisi, F, Lambelet, P, Sebedio, JL, and Berdeaux,
O (2006) Analysis of eicosapentaenoic and docosahexaenoic acid geometrical isomers
formed during fish oil deodorization. J. Chromatogr. A, 1129, 21–28.
Grandgirard, A (1992) Transformations des lipides au cours des traitements thermiques.
Effets nutritionnels et toxicologiques. les Cahiers de l’ENSBANA, 8, 49–67.
Grandgirard, A, Bourre, JM, Julliard, F, Homayoun, P, Dumont, O, Piciotti, M and Sébédio,
JL (1994) Incorporation of trans long-chain n–3 polyunsaturated fatty acids in rat brain
structures and retina. Lipids, 29, 251–258.
Grandgirard, A, Piconneaux, A, Sébédio, JL, O’Keefe, SF, Sémon, E and Le Quéré, JL (1989)
Occurrence of geometrical isomers of eicosapentaenoic and docosahexaenoic acids in
liver lipids of rats fed heated linseed oil. Lipids, 24, 799–804.
Grandgirard, A, Sebedio, JL and Fleury, J (1984) Geometrical isomerization of linolenic acid
during heat treatment of vegetable oils J. Am. Oil Chem. Soc., 61, 1563–1568.
Guzmán, M, Klein, W, Gómez del Pulgar, T and Geelen, MJH (1999) Metabolism of trans
fatty acids by hepatocytes. Lipids, 34, 381–386.
Hølmer, G (1998) Biochemistry of trans monoenoic fatty acids. In: Trans Fatty Acids in
Human Nutrition (JL Sébédio and WW Christie, eds), Oily Press, Bridgwater, UK,
pp.163–189.
Hornstra, G, van Eijsden, M, Dirix, C and Bonsel, G (2006) Trans fatty acids and birth
outcome: Some first results of the MEFAB and ABCD cohorts. Atheroscler. Suppl., 7,
21–23.
Houtsmüller, UMT (1981) Columbinic acid, a new type of essential fatty acid. Prog. Lipid
Res., 20, 889–896.
Hwang, DH and Kinsella, JE (1978) The effects of trans linoleic acid on the concentration
of serum prostaglandin F2 alpha and platelet aggregation. Prostaglandins Med., 1, 121–
30.
Innis, SM (2006) Trans fatty acid intakes during pregnancy, infancy and early childhood.
Atheroscler. Suppl., 7, 17–20.
Innis, SM and King DJ (1999) Trans fatty acids in human milk are inversely associated with
concentrations of essential all cis n–6 and n–3 fatty acids and determine trans but not n–
6 and n–3 fatty acids in plasma lipids of breast-fed infants. Am. J. Clin. Nutr., 70, 383–390.
Jensen, RG (1999) Lipids in human milk. Lipids, 34, 1243–1271.
Kameda, K, Valicenti, AJ and Holman, RT (1980) Chain elongation of trans-octadecenoic
acid isomers in rat liver microsomes. Biochim. Biophys. Acta, 618, 13–17
Karney, RI and Dhopeshwarkar, GA (1978) Metabolism of all trans 9,12-[1-
14C]octadecadienoic acid in the developing brain. Biochim. Biophys. Acta, 531, 9–15.
Knipprath, WG and Mead, JF (1964) The metabolism of trans,trans-octadecadienoic acid:

incorporation of trans,trans-octadecadienoic acid in the C20 polyunsaturated acids of the
rat. J. Am. Oil Chem. Soc., 41, 437–440.
Koletzko, B (1992) Trans fatty acids may impair biosynthesis of long-chain polyunsaturates
and growth in man. Acta Paediatr., 81, 302–306.
Koletzko, B, Mrotzek M and Bremer, HJ (1988) Fatty acid composition of mature human milk
in Germany. Am. J. Clin. Nutr., 47, 954–959.
Kuhnt, K, Kraft, J, Moeckel, P and Jahreis, G (2006a) Trans-11-18:1 is effectively Δ9
desaturated compared with trans-12-18:1 in humans. Br. J. Nutr., 95, 752–761.
Kuhnt, K, Wagner, A, Kraft J, Basu, S and Jahreis, G (2006b) Dietary supplementation with
11trans- and 12trans-18:1 and oxidative stress in humans. Am. J. Clin. Nutr., 84, 981–
988.
Kummerow, FA, Mahfouz, MM and Zhou, Q (2007) Trans fatty acids in partially hydrogen-
ated soybean oil inhibit prostacyclin release by endothelial cells in presence of high level
of linoleic acid. Prostaglandins Other Lipid Mediators, 84,138–153.
Kummerow, FA, Zhou, Q, Mahfouz, MM, Smiricky, MR, Grieshop, CM and Schaeffer, DJ
(2004) Trans fatty acids in hydrogenated fat inhibited the synthesis of the polyunsaturated
fatty acids in the phospholipid of arterial cells. Life Sci., 74, 2707–2723.
Lawson, LD and Holman RT (1981) β-Oxidation of the geometric and positional isomers of
octadecenoic acid by rat heart and liver mitochondria. Biochim. Biophys. Acta, 665, 60–65.
Lawson, LD, Hill, EG and Holman, RT (1983) Suppression of arachidonic acid in lipids of
rat tissues by dietary mixed isomeric cis and trans octadecenoates. J. Nutr., 113, 1827–
1835
Lawson, LD, Hill, EG and Holman, RT (1985) Dietary fats containing concentrates of cis or
trans octadecenoates and the patterns of polyunsaturated fatty acids of liver
phosphatidylcholine and phosphatidylethanolamine. Lipids, 20, 262–267.
Loï, C, Chardigny, JM, Berdeaux, O, Vatele, JM, Poullain, D, Noel, JP and Sebedio, JL
(1998) Effects of three trans isomers of eicosapentaenoic acid on rat platelet aggregation
and arachidonic acid metabolism. Thromb. Haemost., 80, 656–661.
Loï, C, Chardigny, JM, Almanza, S, Leclere, L, Ginies C, and Sébédio, JL (2000) Incorpo-
ration and metabolism of dietary trans isomers of linolenic acid alter the fatty acid profile
of rat tissues. J. Nutr., 130, 2550–2555.
Mahfouz, MM and Kummerow, FA (1999) Hydrogenated fat high in trans monoenes with
an adequate level of linoleic acid has no effect on prostaglandin synthesis in rats. J. Nutr.,
129, 15–24.
Mahfouz, MM, Johnson, S and Holman, RT (1980) The effect of isomeric trans-18:1 acids
on the desaturation of palmitic, linoleic and eicosa-8,11,14-trienoic acids by rat liver
microsomes. Lipids, 15, 100–107.
Marchand, CM and Beare-Rogers, JL (1978) The acylation of 1-palmityglycerol 3-phosphate
with cis and trans C-16 to C-22 monoenoic fatty acids in rat liver microsomes. Lipids, 13,
329–333.
Masters, C and Crane, D (1984) The role of peroxisomes in lipid metabolism. Trends
Biochem. Sci., 9, 314–317.
Mosley, EE, McGuire, MK, Williams, JE and McGuire, MA (2006) Cis-9,trans-11 conju-
gated linoleic acid is synthesized from vaccenic acid in lactating women. J. Nutr., 136,
2297–2301.
Mu, H and Høy, CE (2004) The digestion of dietary triacylglycerols. Prog. Lipid Res., 43,
105–133.
Normann, PT, Thomassen, MS, Christiansen, EN and Flatmark, T (1981) Acyl-CoA
synthetase activity of rat liver microsomes substrate specificity with special reference to
very-long-chain and isomeric fatty acids. Biochim. Biophys. Acta, 664, 416–427.
O’Keefe, SF, Lagarde, M, Grandgirard, A and Sebedio, JL (1990) Trans n–3 isomers exhibit
different inhibitory effects on arachidonic acid metabolism in human platelets compared
to the respective cis fatty acids. J. Lipid Res., 31, 1241–1246.
Okuyama, H, Lands, WEM, Gunstone, FD and Barve, JA (1972) Selective transfers of trans-
ethylenic acids by acyl coenzyme A:phospholipid acyltransferases. Biochemistry, 77,
4392–4398.
Ono, K and Fredrickson, DS (1964) The metabolism of 14C-labeled cis and trans isomers of
octadecenoic and octadecadienoic acids. J. Biol. Chem., 239, 2482–2488.
Osmundsen, H, Bremer, J and Pedersen JI (1991) Metabolic aspects of peroxisomal β-
oxidation. Biochim. Biophys. Acta, 1085, 141–158.
Piconneaux, A (1987) Etude de la désaturation et de l’élongation in vivo d’isomères
géométriques de l’acide linolenique. PhD thesis, University of Dijon, France.
Precht D, Molkentin J, Destaillats F, and Wolff, RL (2001) Comparative studies on individual
isomeric 18:1 acids in cow, goat, and ewe milk fats by low-temperature high-resolution
capillary gas-liquid chromatography. Lipids, 36, 827–32.
Privett, OS, Nutter, LJ and Lightly, FS (1966) Metabolism of trans acids in the rat: influence
of the geometric isomers of linoleic acid on the structure of liver triglycerides and
lecithins. J. Nutr., 89, 257–264.
Privett, OS, Stearns, EM and Nickell, EC (1967) Metabolism of the geometric isomers of
linoleic acid in the rat. J. Nutr., 92, 303–310.
Ratnayake WMN (2004) Overview of methods for the determination of trans fatty acids by
gas chromatography, silver-ion thin-layer chromatography, silver-ion liquid chromatog-
raphy, and gas chromatography/mass spectrometry. J. AOAC Int., 87, 523–539.
Ratnayake, WMN, Chen, ZY, Pelletier, G and Weber, D (1994) Occurrence of 5c,8c,11c,15t-
eicosatetraenoic acid and other unusual polyunsaturated fatty acids in rats fed partially
hydrogenated canola oils. Lipids, 29, 707–714.
Reichwald-Hacker, I, Grosse-Oetringhaus, S, Kiewitt, I and Muhkerjee, KD (1979) Incorpo-
ration of positional isomers of cis- and trans-octadecenoic acids into acyl moieties of rat
tissue lipids. Biochim. Biophys. Acta, 575, 327–334.
Reitz, RC, El-Sheikh, M, Lands, WEM, Ismail, IA and Gunstone, FD (1969) Effects of
ethylenic bond position upon acyltransferase activity with isomeric cis-octadecenoyl

coenzyme A thiol esters. Biochim Biophys Acta, 176, 480–490.
Roy, U, Joshua, R, Stark, L and Balazy, M (2005) Cytochrome P450/NADPH-dependent
biosynthesis of 5,6-trans-epoxyeicosatrienoic acid from 5,6-trans-arachidonic acid.
Biochem. J., 390, 719–727.
Schimp, JL, Bruckner, G and Kinsella, JE (1982) The effects of dietary trilinoelaidin on fatty
acid and acyl desaturases in rat liver. J. Nutr., 112, 722–735.
Scrimgeour, CM, Macvean, A, Fernie, C, Sébédio, JL and Riemersma, RA (2001) Dietary
trans α-linolenic acid does not inhibit Δ5 and Δ6-desaturation of linoleic acid in man. Eur.
J. Lipid Sci. Technol., 103, 341–349.
Sébédio, JL and Chardigny, JM (1998) Biochemistry of trans polyunsaturated fatty acids. In:
Trans Fatty Acids in Human Nutrition (JL Sébédio and WW Christie, eds), Oily Press,
Sebedio, JL and Juaneda, P (2007) Isomeric and cyclic fatty acids as a result of frying In: Deep
Frying: Chemistry, Nutrition, and Practical Applications, 2nd Edition (M Erickson, ed.),
AOCS Press, Champaign, Illinois, USA, pp.57–86.
Sébédio, JL, Vermunt, SHF, Chardigny, JM, Beaufrère, B, Mensink, RP, Armstrong, RA,
Christie, WW, Niemela, J, Hénon, G and Riemersma, RA (2000) The effect of dietary
trans α-linolenic acid on plasma lipids and platelet fatty acid composition: the TransLinE
study. Eur. J. Clin. Nutr., 54, 104–113.
Selinger, Z and Holman, RT (1965) The effects of trans,trans-linoleate upon the metabolism
of linoleate and linolenate and the positional distribution of linoleate isomers in liver
lecithin. Biochim. Biophys. Acta, 106, 56–62.
Sgoutas, DS (1970) Effect of geometry and position of ethylenic bond upon acyl coenzyme
A-cholesterol-O-acyltransferase. Biochemistry, 9, 1826–1833.
Stachowska, E, Dolegowska, B, Olszewska, M, Gutowska, I and Chlubek, D (2004) Isomers
of trans fatty acids modify the activity of platelet 12-P lipoxygenase and cyclooxygenase/
thromboxane synthase. Nutrition, 20, 570–571.
Szabo, E, Boehm, G, Beermann, C, Weyermann, M, Brenner, H, Rothenbacher, D and Decsi,
T (2007) Trans octadecenoic acid and trans octadecadienoic acid are inversely related to
long chain polyunsaturates in human milk: results of a large birth cohort study. Am. J.
Clin. Nutr., 85, 1320–1326.
Tomey, KM, Sowers, FR, Li, X, McConnell, DS, Crawford, S, Gold, EB, Lasley, B and
Randolph, JF (2007) Dietary fat subgroups, zinc, and vegetable components are related
to urine F2α-isoprostane concentration a measure of oxidative stress in midlife women.
J. Nutr., 137, 2412–2419.
Trus, M, Grandgirard, A and Sébédio, JL (1991) Utilisation digestive des isomères trans de
l’acide linolénique chez le rat. Reprod. Nutr. Develop., 31, 294.
Turpeinen, AM, Mutanen, M, Aro, A, Salminen, I, Basu, S, Palmquist, DL and Griinari, JM
(2002) Bioconversion of vaccenic acid to conjugated linoleic acid in humans. Am. J. Clin.
Nutr., 76, 504–510.
Turpeinen, AM, Wübert, J, Aro, A, Lorenz, R and Mutanen, M (1998) Similar effects of diets
rich in stearic acid or trans-fatty acids on platelet function and endothelial prostacyclin
production in humans. Arterioscler. Thromb. Vasc. Biol., 18, 316–322.
Vermunt, SHF, Beaufrère, B, Riemersma, RA, Sébédio, JL, Chardigny, JM and Mensink, RP
(2001) Dietary trans α-linolenic acid from deodorised rapeseed oil and plasma lipids and
lipoproteins in healthy men: the TransLinE Study. Brit. J. Nutr., 85, 387–392.
Waku, K and Lands, WEM (1968) Control of lecithin biosynthesis in erythrocyte membranes.
J. Lipid Res., 9, 12–18.
Wolff, RL (1995) Content and distribution of trans-18:1 acids in ruminant milk and meat fats:
their importance in European diets and their effect on human milk. J. Am. Oil Chem. Soc.,
72, 259–272.
Wolff, RL, Combe, NA, Entressangles, B, Sébédio, JL and Grandgirard, A (1993) Preferen-
tial incorporation of dietary cis-9,cis-12,trans-15 18:3 acid into rat cardiolipins. Biochim.
Biophys. Acta, 1168, 285–291.
Wolff, RL, Precht, D and Molkentin, J (1998) Occurrence and distribution profiles of trans-
18:1 fatty acids in edible fats of natural origin. In: Trans Fatty Acids in Human Nutrition
(JL Sébédio and WW Christie, eds), Oily Press, Bridgwater, UK, pp.1–33.
Wood, R (1979) Distribution of dietary geometrical and positional isomers in brain, heart,
kidney, liver, lung, muscle, spleen, adipose and hepatoma. In: Geometrical and Positional
Fatty Acids Isomer, (EA Emken and HJ Dutton, eds) AOCS Press,, Champaign, Illinois,
USA, pp.213–281.
Yu, W, Liang, YW, Ensenauer, RE, Vockley, J, Sweetman, L and Schulz, H (2004). Leaky
β-oxidation of a trans-fatty acid: incomplete β-oxidation of elaidic acid is due to the
accumulation of 5-trans-tetradecenoyl-coa and its hydrolysis and conversion to 5-trans-
tetradecoylcarnitine in the matrix of rat mitochondria. J. Biol. Chem., 279, 52160–52167.
Zambonin, L, Ferreri, C, Cabrini, L, Prata, C, Chatgilialoglu, C and Landi, L (2006)
Occurrence of trans fatty acids in rats fed a trans-free diet: A free radical-mediated
formation? Free Rad. Biol. Med., 40, 1549–1556.
Zevenbergen, JL and Haddeman, E (1989) Lack of effects of trans fatty acids on eicosanoid
biosynthesis with adequate intakes of linoleic acid. Lipids, 24, 555–563.
Zevenbergen, JL, Houtsmuller, UM and Gottenbos, JJ (1988) Linoleic acid requirements of
rats fed trans fatty acids. Lipids, 23, 178–186.
Zghibeh, CM, Gopal, VR, Poff, CD, Falck, JR and Balazy, M (2004) Determination of trans-
arachidonic acid isomers in human blood plasma. Anal. Biochem., 332, 137–144.
CHAPTER 8
Biosynthesis and biological activity of rumenic
acid: a natural CLA isomer
ADAM L. LOCK1, JANA KRAFT1, BETH H. RICE1 AND DALE E.

BAUMAN2
1
Department of Animal Science, University of Vermont, Burlington, Vermont,
USA
2
Department of Animal Science, Cornell University, Ithaca, New York, USA
A. Introduction
Nutritional quality is increasingly an important consideration in food choices
because of the growing consumer awareness of the link between diet and
health. Many foods contain micro components that have beneficial effects
beyond those associated with their traditional nutrient content, and these are
often referred to as ‘bioactive’ or ‘functional food’ components. These func-
tional food components can play an important role in health maintenance and
the prevention of chronic diseases. Cancer and coronary heart disease (CHD)
are two examples, and epidemiological studies suggest that the risk for these
diseases can be substantially reduced by diet modification. Functional food
components from plants are often highlighted in popular press articles as the
basis for lowered risk of these diseases resulting from consumption of fruits
and vegetables. Foods of animal origin also contain functional food compo-
nents, but these have been less extensively investigated. In the case of dairy
products, there has been a general perception that a food containing saturated
fat is unlikely to be beneficial to health. Yet, over the last decade, evidence has
accumulated that the form of dietary fat is very important in determining the
relative risk to diseases like cancer and CHD, and that milk-derived fat may
offer significant health benefits compared to some common sources of dietary
fats (National Research Council, 1996; Parodi, 2004; Bauman et al., 2006).
One such component in foods derived from ruminants is cis-9,trans-11 con-
jugated linoleic acid.
The term conjugated linoleic acids (CLA) refers to a mixture of positional
and geometric isomers of linoleic acid (cis-9,cis-12 octadecadienoic acid) with
a conjugated double bond system. The presence of CLA isomers in ruminant fat
195
is related to the biohydrogenation of polyunsaturated fatty acids (PUFA) in the

rumen, with over 20 different isomers of CLA identified in milk fat (Kraft
et al., 2003; Lock & Bauman, 2004). The interest in health benefits of CLA has
its genesis in the research by Pariza and associates who first demonstrated that
CLA isomers are bioactive food components when their search for mutagens in
cooked meat instead identified CLA as an anti-mutagen (see review by Pariza,
1999). As a result of this discovery, research on CLA isomers has increased
exponentially over the last 20 years and a number of potential health benefits
of CLA isomers have been reported. Particularly noteworthy is the fact that
CLA is bioactive when supplied as a natural food component in the form of
CLA-enriched milk fat as discussed later in this review.
The presence of CLA isomers in ruminant milk has been known for more
than 70 years. Scientists at the University of Reading, UK, first demonstrated
that fatty acids obtained from summer butter differed from those obtained from
winter butter by exhibiting a much stronger spectrophotometric absorption at
230 nm (Booth et al., 1933). It was subsequently concluded that the absorption
at this wavelength was due to a conjugated double bond pair (Moore, 1939).
Parodi (1977) was the first to identify cis-9,trans-11 octadecadienoic acid as a
fatty acid in milk fat that contained the conjugated double bond pair. As
analytical techniques improved it was discovered that milk fat and body fat
from ruminants contained many isomers of CLA that differ by position (e.g., 7-
9, 8-10, 9-11, 10-12, 11-13) or geometric orientation (cis-trans, trans-cis,
cis-cis, and trans-trans) of the double bond pair. The range of CLA isomers and
their levels in milk and dairy products is summarized in Table 1. Cis-9,trans-
11 is the major CLA isomer in ruminant fat, representing about 75 to 90% of the
total CLA. The common name ‘rumenic acid’ has been proposed and is
commonly used for this isomer because of its unique relationship to ruminants
(Kramer et al., 1998). In most situations the second most common isomer is
trans-7,cis-9 CLA, representing about 10% of total CLA. The exception is
trans-11,cis-13 CLA which is significantly increased in milk from cows
grazing mountain pastures, rich in linolenic acid (Kraft et al., 2003). Each of
the other CLA isomers is at a low concentration when present, generally
representing less than 0.5% of the total CLA in ruminant fat (Palmquist et al.,
2005; Bauman & Lock, 2006). Since the unique biological effects of specific
fatty acids are related to their specific structure and the presence of a certain
configuration of double bonds at specific positions in the fatty acid, Figure 1
contrasts the structure of linoleic acid with rumenic acid and trans-11 18:1
(vaccenic acid). In the following sections we will elaborate the state of
knowledge for rumenic acid as a bioactive food component; the discussion will
also include vaccenic acid because of its special relationship to rumenic acid in
ruminants and ruminant-derived foods.
The dietary source of CLA, specifically rumenic acid, in humans has been
examined in several countries and ruminant-derived food products are the
BIOSYNTHESIS AND BIOLOGICAL ACTIVITY OF RUMENIC ACID 197
Table 1. Range of positional and geometric isomers of conjugated linoleic acids (CLA)
in milk and dairy products. Adapted from Lock & Bauman (2004).
Isomer % of total CLA isomers
trans-7,cis-9 1.2–8.9
trans-7,trans-9 < 0.1–2.4
trans-8,cis-10 < 0.1–1.5
trans-8,trans-10 0.2–0.4
cis-9,trans-11 72.6–91.2
trans-10,cis-12 < 0.1–1.5
cis-11,trans-13 0.2–4.7
trans-11,cis-13 0.1–8.0
cis-12,trans-14 < 0.01–0.8
cis, cis isomers 0.1–4.8
Figure 1. Chemical structures of (A) linoleic acid (cis-9,cis-12 18:2), (B) rumenic acid (cis-9,trans-11
conjugated linoleic acid), and (C) vaccenic acid (trans-11 18:1). Arrows indicate location of double
bonds. Adapted from Bauman et al. (2006).
major source (see review by Parodi, 2003). As illustrated by data from the USA,
dairy products account for about 70% of total food intake of CLA and ruminant
meats account for another 25% (Ritzenthaler et al., 2001). Differences in food
preferences result in a wide range in individual daily intakes. In general, the
average dietary intake of rumenic acid ranges from 100 to 300 mg/d (Parodi,
2003). This intake estimate, however, does not include the contribution of
endogenous synthesis of rumenic acid from dietary vaccenic acid. Based on
kinetic data obtained from studies with humans (Turpeinen et al., 2002; Kuhnt
et al., 2006), Parodi (2003) has emphasized that the endogenous synthesis of
rumenic acid needs to be considered to obtain an estimate of the ‘effective
rumenic acid’ provided by ruminant-derived food products.
In this chapter we will first review the origin of rumenic acid present in milk
fat, developing the interrelationships between biohydrogenation intermediates
produced in the rumen, synthesis of rumenic acid in tissues and its presence in
milk fat. Secondly, we will highlight the nutritional and physiological factors
that affect the level of rumenic acid in milk fat and discuss milk quality
considerations for dairy products that have a naturally enhanced content of
rumenic acid. Finally, we will review the biological effects of rumenic acid in
relation to human health. This will include the significance of rumenic acid in
dairy products and its potential as a functional food component that may benefit
health maintenance and disease prevention.
B. The ruminant dimension
1. Origin of rumenic acid

As discussed in Chapter 1, biohydrogenation of PUFA is the major transforma-
tion that dietary lipids undergo in the rumen. A comprehensive discussion of lipid
metabolism in the rumen and its effects on the production of CLA and trans 18:1
isomers is also provided in a recent review by Palmquist et al. (2005). In relation
to rumenic acid formation in the rumen, the initial step in rumen biohydrogenation
of linoleic and linolenic acid involves an isomerization of the cis-12 double bond
to a trans-11 configuration, resulting in a conjugated dienoic or trienoic fatty acid
(Figure 2). Next, is a reduction of the cis-9 double bond resulting in a trans-11
octadecenoic acid. Therefore, rumenic acid is an intermediate formed only
during the biohydrogenation of linoleic acid. The conversion of rumenic acid to
vaccenic acid (trans-11 18:1) is catalysed by a reductase produced by some
rumen bacteria. The structure of vaccenic acid is given in Figure 1, and it is
noteworthy that this fatty acid is an intermediate in the biohydrogenation of both
linoleic and linolenic acid (Figure 2). The final step is a further reduction of
vaccenic acid, producing stearic acid. Reduction of vaccenic acid to stearic acid
is generally the rate-limiting step and, as a consequence, there is often an
accumulation of vaccenic acid in the rumen (Harfoot & Hazlewood, 1997).
Rumen Mammary gland
1. Increase C18 PUFA Precursors

Linolenic acid Linoleic acid
cis-9,cis-12,cis-15 18:3 cis-9,cis-12 18:2
cis-9,trans-11,cis-15 18:3 cis-9,trans-11 CLA cis-9,trans-11 CLA
2. Maintain trans-11 pathways

Δ9-desaturase
trans-11,cis-15 18:2
trans-11 18:1 trans-11 18:1
Vaccenic acid Vaccenic acid
3. Inhibit final step 4. Increase desaturation
18:0
Figure 2. Pathways for ruminal and endogenous synthesis of rumenic acid (cis-9,trans 11-CLA) in the
lactating dairy cow. Pathways for biohydrogenation of linoleic and linolenic acids yielding vaccenic acid
(trans-11 18:1) are shown in the rumen box and endogenous synthesis by Δ9-desaturase is shown in the
mammary gland box. Potential strategies to increase milk fat content of CLA are indicated in the white
boxes. Adapted from Bauman & Lock (2008).
As analytical techniques have improved, we have gained an appreciation of

the complexity of the biohydrogenation processes occurring in the rumen. In
addition to the major pathways involving rumenic acid and vaccenic acid as
intermediates, there must be many additional pathways. A remarkable range of
trans-18:1 and CLA isomers are produced during biohydrogenation and their
outflows from the rumen based on limited data from growing cattle and
lactating cows are shown in Table 2. This range of CLA and trans-18:1 isomers
is not accounted for by known pathways of rumen biohydrogenation. It appears
that the type of diet, rather than level of intake, is a major factor affecting
biohydrogenation and diet-induced changes in the rumen environment can
shift the biohydrogenation pathways resulting in dramatic changes in the fatty
acid intermediates leaving the rumen and subsequently available for incorpo-
ration into milk and body fat.
Initially, it was assumed that the rumenic acid in milk fat and body fat of
ruminants originated from incomplete biohydrogenation in the rumen. This
hypothesis was based on the fact that rumenic acid was the major CLA isomer
in milk fat (Table 1) and the first intermediate in the major biohydrogenation
pathway for linoleic acid (Figure 2). A close linear relationship was also
observed between the levels of vaccenic acid and rumenic acid in milk fat (e.g.
Jiang et al., 1996; Jahreis et al., 1997; Griinari & Bauman, 1999), consistent
Table 2. Range of double bond positions in trans-18:1 and conjugated linoleic acid
(CLA) isomers and their ruminal outflow in growing and lactating cattle. Adapted from
Bauman & Lock (2006).
Trans-18:1 Conjugated 18:2

Isomer Ruminal outflow (g/day) Isomer Ruminal outflow (g/day)
trans-4 0.5–0.7 trans-7,cis-9 < 0.01

trans-5 0.4–0.6 trans-7,trans-9 < 0.01–0.05
trans-6-8 0.4–6.7 trans-8,cis-10 0.01–0.02
trans-9 0.8–6.2 trans-8,trans-10 < 0.01–0.10
trans-10 1.7–29.1 cis-9,cis-11 < 0.01–0.01
trans-11 5.0–121.0 cis-9,trans-11 0.19–2.86
trans-12 0.5–9.5 trans-9,trans-11 0.22–0.55
trans-13 + 14 6.5–22.9 trans-10,cis-12 0.02–0.32
trans-15 3.2–8.5 trans-10,trans-12 0.05–0.06
trans-16 3.1–8.0 cis-11,trans-13 0.01–0.10
trans-11,cis-13 0.01–0.46
trans-11,trans-13 0.09–0.40
cis-12,trans-14 < 0.01–0.05
trans-12,trans-14 0.08–0.19
with the concept that these two fatty acid intermediates had escaped complete
biohydrogenation in the rumen and were subsequently absorbed from the
digestive tract and used for milk fat synthesis. Lines of evidence, however,
highlighted a number of inconsistencies with this premise. First, the kinetics of
rumen biohydrogenation are such that rumenic acid represents only a transitory
product, and vaccenic acid is the major biohydrogenation intermediate that
accumulates in the rumen (Harfoot & Hazlewood, 1997). Second, nutrition
studies demonstrated that increases in the milk fat content of rumenic acid
occurred when linseed oil and other dietary sources of linolenic acid were fed
(e.g., Kelly et al., 1998; Lock & Garnsworthy, 2002). As previously discussed,
rumenic acid is not an intermediate in the biohydrogenation of linolenic acid,
but the biohydrogenation of both linoleic and linolenic acids produces vaccenic
acid as an intermediate. Third, the ratio of vaccenic acid to rumenic acid is
> 50:1 in rumen fluid but only about 3:1 in milk fat. Based on these considera-
tions, Griinari and Bauman (1999) proposed that endogenous synthesis could
be an important source of the rumenic acid found in milk fat, with synthesis
involving the enzyme Δ9-desaturase and vaccenic acid as the substrate (see
Figure 2). Previous investigations with Δ9-desaturase from rat liver established
that while the preferred reaction was the conversion of stearic acid to oleic acid,
this enzyme could also desaturate positional isomers of trans-octadecenoic
acids (Mahfouz et al., 1980; Pollard et al., 1980).
The first study to show directly that milk fat CLA could originate via
endogenous synthesis was Griinari et al. (2000); they infused 12.5 g/d of
Table 3. Quantitative studies of endogenous conjugated linoleic acids (CLA) synthesis

in lactating dairy cows. Adapted from Bauman et al. (2006).
Diet and approach1 Milk fat CLA Endogenous Reference

(mg/g fatty acid) synthesis
estimate
Inhibition of Δ9-desaturase2
Hay/concentrate TMR 4 64% Griinari et al. (2000)
Hay/concentrate TMR 7 ⎫
plus PHVO 8 ⎭⎬ 78% Corl et al. (2001)
Pasture 16 > 91% Kay et al. (2004)
Rumen Output3
Grass silage/concentrate 8–11 > 80% Lock & Garnsworthy
plus various plant oils (2002)
Corn silage/alfalfa hay TMR 5–7 > 93% Piperova et al. (2002)
Grass silage/concentrate 19 > 74% Shingfield et al. (2003)
plus fish oil
Modeling of isotope kinetics4
Alfalfa silage and 4.8 80 Mosely et al. (2006)
hay/concentrate
1
Diet abbreviations are TMR (total mixed ration) and PHVO (partially hydrogenated vegetable oil).
2
Direct estimates of endogenous synthesis determined by use of sterculic oil as a source of cyclopropoic
fatty acids to inhibit Δ9-desaturase.
3
Indirect estimates determined by measuring rumen outflow of cis-9,trans-11 CLA and comparing this
to the quantity of cis-9,trans-11 CLA secreted in milk fat.
4
Based on modeling of isotope kinetics.
vaccenic acid into the abomasum of dairy cows and observed a 31% increase in
the concentration of rumenic acid in milk fat. This investigation clearly
demonstrated the potential for endogenous synthesis, but additional studies
were needed to determine its actual importance. To address this, three ap-
proaches were used (see Table 3). One approach was to inhibit Δ9-desaturase
directly, and this involved the use of sterculic oil which contains two
cyclopropene fatty acids, sterculic acid and malvalic acid, which are specific
inhibitors of Δ9-desaturase (Jeffcoat & Pollard, 1977). Griinari et al. (2000),
using this approach, estimated that 64% of the rumenic acid in milk fat was of
endogenous origin in cows fed a diet based on alfalfa hay and corn grain. This
represented the first direct demonstration that endogenous synthesis was the
major source of rumenic acid in milk fat. Subsequent investigations using the
same approach extended results to other dietary situations (total mixed diets
with or without plant oils and pasture) and in all cases endogenous synthesis
was the predominant source of the rumenic acid in milk fat (Corl et al., 2001;
Kay et al., 2004). Results from grazing cows are of special note because
pasture is high in linolenic acid and endogenous synthesis accounted for > 91%
of the total rumenic acid in milk fat (Kay et al., 2004).
The second approach to quantify the contribution of endogenous synthesis to

milk fat rumenic acid was indirect and involved a comparison of ruminal
outflow with secretion in milk; rumen output of rumenic acid would represent
the maximum proportion of rumenic acid secreted in milk fat and the remainder
would have to be derived from endogenous synthesis (Table 3). For this
approach, representative samples of digesta were obtained and data for rumenic
acid content were combined with marker-derived estimates of flow rates of
digesta. Lock & Garnsworthy (2002) estimated rumen output of rumenic acid
in non-lactating cows and then extrapolated results to lactating cows on the
basis of feed intake. Their estimates indicated that endogenous synthesis
accounted for over 80% of the rumenic acid in milk fat in cows fed a grass
silage/concentrate diet supplemented with various plant oils. Comparable
results were obtained in similar studies using this approach with lactating cows
fed a range of diets (Piperova et al., 2002; Shingfield et al., 2003).
The third approach to examine the importance of endogenous synthesis in
lactating dairy cows involves the use of isotope kinetics (Table 3). Mosley
et al. (2006) used radiolabelled vaccenic acid and modelled the data for
enrichment of rumenic acid in milk fat; they concluded that 80% of the rumenic
acid in milk fat originated from endogenous synthesis from vaccenic acid.
Overall, investigators using different diets and experimental approaches
have found similar results; the major source of rumenic acid in milk fat is
endogenous synthesis. Thus, endogenous synthesis is the basis for rumenic
acid being the predominant CLA isomer in milk fat, and the relatively constant
ratio between vaccenic acid and rumenic acid observed in milk fat reflects the
substrate:product relationship for Δ9-desaturase.
2. The Δ9-desaturase enzyme system

The predominance of endogenous synthesis as the source of rumenic acid in
milk fat highlights the critical role of Δ9-desaturase in the biology of rumenic
acid. This enzyme is also known as stearoyl-CoA desaturase (EC 1.14.19.1)
in biochemistry texts because stearic acid is its most common substrate. The
CoA ester of vaccenic acid is the substrate for the formation of rumenic acid,
but the preferred substrates for Δ 9-desaturase are stearoyl-CoA and
palmitoyl-CoA, which are converted to oleoyl-CoA and palmitoleoyl-CoA,
respectively (Ntambi, 1999; Ntambi & Miyazaki, 2004). For ruminants, a
substantial activity of Δ9-desaturase has been reported in mammary tissue
(e.g. Bickerstaffe & Annison, 1970; Kinsella, 1972), adipose tissue (e.g.
Chang et al., 1992; Barber et al., 2000) and in intestinal epithelium
(Bickerstaffe & Annison, 1969). Genes coding for the Δ9-desaturase enzyme
have been identified and cloned for both bovine (Chung et al., 2000) and
ovine (Ward et al., 1998) animals. In the cow and sheep two genes for Δ9-
desaturase have been elucidated, whereas four isoforms of Δ9-desaturase
have been identified in mice, two in rats, and three in hamsters (Ntambi &
Miyazaki, 2004; Lengi & Corl, 2008).
Our understanding of the regulation of Δ9-desaturase in ruminants is limited,
with current knowledge coming mainly from investigations on rodents. Δ9-
desaturase has no known allosteric or feedback inhibition involving its substrates
or products; rather, it is regulated by dietary factors such as glucose and long-
chain PUFA, and by hormones such as insulin and glucagon (Ntambi &
Miyazaki, 2004). The enzyme protein has a relatively short half-life (~4 h) and,
thus, gene transcription is its major point of regulation (Ozols, 1997). Both,
PUFA and trans-10,cis-12 CLA, down-regulate its gene expression but rumenic
acid has no effect (Ntambi & Miyazaki, 2004). Interestingly, the cyclopropene
fatty acids in sterculic oil do not affect Δ9-desaturase gene expression or protein
synthesis, but they directly inhibit the activity of the enzyme (Gomez et al.,
2003).
The relationship between substrate and product for Δ9-desaturase is reflected
by the desaturase index, defined as the ratio [product of desaturase ÷ (product
+ substrate for desaturase)] (Kelsey et al., 2003). The desaturase index in milk
fat represents a proxy for Δ9-desaturase and a several-fold range is observed
among individuals providing a strong indication that there are genetic differ-
ences among individuals with respect to this enzyme. This is discussed in more
detail in Section B.5.
3. Origin of other CLA isomers in milk fat

Yurawecz et al. (1998) were the first to identify the presence of trans-7, cis-9
CLA in milk fat and across studies, the level of trans-7, cis-9 CLA in milk fat
has generally been about 10% of rumenic acid and several-fold greater than that
of any of the other CLA isomers (see review by (Lock & Bauman, 2004). Early
investigations with Δ9-desaturase from rat liver had established that trans-7
18:1 could serve as a substrate for this enzyme (Mahfouz et al., 1980; Pollard
et al., 1980). Furthermore, trans-7 18:1 is present in rumen outflow, albeit at
low concentrations (Table 2), being produced as an intermediate in the
biohydrogenation of C18 PUFA as discussed earlier. Based on this, when
Yurawecz et al. (1998) initially discovered trans-7, cis-9 CLA in milk fat, they
speculated that it might originate from endogenous synthesis. Using the same
approaches as discussed for rumenic acid, researchers have consistently re-
ported that the source of trans-7, cis-9 CLA in milk fat was endogenous
synthesis via Δ9-desaturase from ruminally produced trans-7 18:1 (Palmquist
et al., 2005).
In contrast to cis-9, trans-11 and trans-7, cis-9, the other isomers of CLA
found in the milk fat of ruminants appear to originate exclusively from rumen
output. This conclusion is based, in large part, on the fact that these minor cis-
trans, trans-cis, cis-cis, and trans-trans isomers are detected in rumen fluid
(Palmquist et al., 2005; Bauman & Lock, 2006), and estimates of digesta flow
indicate that rumen output is more than adequate to account for the trace
amounts secreted in milk fat (Piperova et al., 2002; Shingfield et al., 2003).
Furthermore, there has been no demonstration that other mammalian desaturases
act in a manner analogous to Δ9-desaturase to synthesize additional CLA
isomers endogenously from other trans-18:1 fatty acids. Thus, these CLA
isomers found at trace levels in milk fat are of rumen origin and represent
intermediates formed in the biohydrogenation of PUFA.
Information on the effect of diet on the production of minor isomers of CLA
in the rumen and alterations in their content in milk fat is limited. Best
described are diet-induced changes in trans-10,cis-12 CLA, and its biological
effects in the dairy cow and impact on milk fat synthesis (see reviews by
Bauman & Lock, 2006; Shingfield & Griinari, 2007; Bauman et al., 2008). The
ability to better characterize rumen biohydrogenation pathways and establish
their relationship to specific rumen bacteria and diets are important areas for
future research.
4. Endogenous synthesis of rumenic acid in humans and other species

In addition to ruminants, the conversion of dietary vaccenic acid to rumenic acid
occurs in other mammalian species including humans. Parodi (1994) was the first
to propose that vaccenic acid could be converted to rumenic acid in non-
ruminants based on the observation that vaccenic acid is desaturated to rumenic
acid by the Δ9-desaturase enzyme from rat liver microsomes (Mahfouz et al.,
1980; Pollard et al., 1980). Santora et al. (2000) first reported and quantified the
desaturation of vaccenic acid to rumenic acid in mice. Subsequent studies in
different animal models have provided evidence for a dose-dependent endog-
enous synthesis of rumenic acid by Δ9-desaturase and its accumulation in a wide
range of animals and body tissues (Glaser et al., 2002; Lock et al., 2004; Kraft
et al., 2006; Shingfield et al., 2007). Furthermore, several studies in humans
have reported that rumenic acid, detected in blood serum or erythrocyte mem-
branes, was in part derived from the bioconversion of dietary vaccenic acid (e.g.
Adlof et al., 2000; Turpeinen et al., 2002; Kuhnt et al., 2006); the estimated
average conversion rate of vaccenic acid into rumenic acid was 19–25%. There
is a wide range in the conversion rate within subjects suggesting that environment
and/or gene polymorphisms of the Δ9-desaturase may alter gene expression,
enzyme activity or substrate specificity. In addition, gender-specific differences
in vaccenic acid desaturation rate were observed, which could be partly related to
hormones or to body fat mass (Kuhnt et al., 2006). Overall, the conversion of
vaccenic acid into rumenic acid should be taken into account in future studies
when determining the CLA supply. As mentioned earlier, Parodi (2003) has sug-
gested that rumenic acid intake multiplied by 1.4 would provide an estimate of the
effective physiological dose of rumenic acid derived from ruminant products.
5. Factors affecting rumenic acid content in milk fat
The discovery of health benefits of rumenic acid and a recognition of its

potential as a bioactive food component in dairy products has stimulated
research to identify factors that affect the rumenic acid content of milk fat.
These efforts have focused on enhancing the CLA content per unit of fat and
centred on rumenic acid as the predominant CLA isomer. As highlighted in the
preceding discussion on the origin of rumenic acid, there are four possible areas
to consider when attempting to increase the rumenic acid content of milk fat
(Lock & Bauman, 2004): (i) increase the 18-carbon PUFA precursors in the
diet (linoleic and linolenic acid); (ii) maintain rumen biohydrogenation path-
ways that result in the production of vaccenic acid as an intermediate; (iii)
inhibit the final step in the biohydrogenation of C18 PUFA so that vaccenic acid
accumulates; and (iv) increase Δ9-desaturase activity and accordingly the
desaturation of vaccenic acid to rumenic acid in the mammary gland (Figure 2).
Numerous studies have shown that diet is the most significant factor affect-
ing the rumenic acid content of milk fat, and its concentration can be increased
several-fold by dietary means (see reviews by Chilliard et al., 2001; Stanton
et al., 2003; Lock & Bauman, 2004). As cited above, one key to increasing milk
rumenic acid is to increase the dietary intake of C18 PUFA, thereby providing
more substrate for rumen biohydrogenation. The dietary supply of linoleic and
linolenic acid is most easily increased by the addition of plant oils rich in these
fatty acids. A range of plant oils have been investigated and shown to be
effective in increasing the level of rumenic acid in milk fat. In addition to the
feeding of ‘free’ oils, a range of oilseeds containing both linoleic and linolenic
acid have been shown to be effective in increasing the rumenic acid content of
milk fat. In general, oil seeds rich in C18 PUFA and processed so that the oil is
accessible to the bacteria involved in biohydrogenation, result in greater
increases in milk rumenic acid compared with whole oilseeds, but are not as
efficient as using the pure oil.
The amount of 18-carbon PUFA that can be added to the diets of dairy cows
is limited due to the adverse effects these PUFA can have on the metabolism of
rumen bacteria, thereby impairing rumen fermentation and animal perform-
ance (Palmquist et al., 2005). Thus, dairy cattle diets are generally restricted to
less than 7% total lipid, and this provides an upper limit in the use of lipid
supplements. When a high level of oil is added, up to 10-fold increases in the
rumenic acid content of milk fat are observed. As a result, however, of negative
effects on rumen bacteria, these levels are often transient and decline within a
few weeks to stabilize at ~4-fold to 5-fold increases (e.g. Bauman et al., 2000).
In addition, there is often a fine line between supplying additional lipid
supplements to increase milk fat rumenic acid content and causing changes in
the rumen environment; referred to as the ‘isomerase shift’, under some
conditions the rumen environment may be modified to produce more trans-10
18:1 and trans-10,cis-12 CLA as intermediates and this results in a dramatic

reduction in milk fat synthesis (Palmquist et al., 2005; Bauman & Lock, 2006;
Shingfield & Griinari, 2007). Thus, it is important to prevent the isomerase
shift and rather maintain the trans-11 C18 pathway of rumen biohydrogenation
(Figure 2).
Another means through which dietary and nutritional factors can increase
the CLA content of milk fat is by inhibiting the terminal step in biohydrogenation
(Figure 2). This typically occurs either directly or indirectly via changes in the
rumen environment; the net result is an accumulation of vaccenic acid, thereby
increasing the rumen outflow of this precursor for the endogenous synthesis of
CLA. A limited number of bacterial species have been shown to carry out the
final biohydrogenation step and, presumably, changes in the rumen environ-
ment lead to a reduction in these species and/or a reduction in their capacity to
reduce vaccenic acid to stearic acid. Several dietary situations also have these
effects and they include alterations in the forage:concentrate ratio, dietary
supplements of fish oil and the use of ionophores. The most consistently
effective of these is the use of fish oils. Fish oils themselves provide very little
C18 PUFA precursors to allow for increased rumen vaccenic acid output,
indicating that this increase occurs through an inhibition of the biohydrogenation
of vaccenic acid. Indeed, 22:6n–3 (docosahexaenoic acid, DHA), a major n–3
fatty acid in fish oil, has been shown to promote accumulation of vaccenic acid
in mixed ruminal cultures when incubated with linoleic acid (AbuGhazaleh &
Jenkins, 2004). Both linoleic and linolenic acid are plentiful in forages and
concentrates which provide sufficient C18 PUFA precursors. A range of fish
and marine oils have been used with success, and similar to the supply of C18
PUFA discussed above, both lipid supplements and fish by-products (fish
meal) have been shown to be effective (e.g., Offer et al., 1999; Abu-Ghazaleh
et al., 2001; Shingfield et al., 2003). Marine algae also contain long-chain
PUFA and have also been efficacious (Franklin et al., 1999).
Furthermore, supplementing lactating dairy cow diets with the ionophore
‘Rumensin’ (Monensin sodium, Elanco Animal Health, Greenfield, Indiana,
USA) has been shown to increase the rumen outflow of trans-18:1
biohydrogenation intermediates, a response associated with changes in bacte-
rial species present in the rumen. When Rumensin is added, the bacteria
responsible for the complete biohydrogenation of trans-18:1 intermediates to
stearic acid are inhibited, thus increasing the rumen outflow of biohydrogenation
intermediates (Fellner et al., 1997; Jenkins et al., 2003). In some studies this
has resulted in increases in the rumenic acid content of milk fat via increases in
the rumen outflow of vaccenic acid and increased endogenous synthesis of
rumenic acid in the mammary gland (e.g. Bell et al., 2006). As discussed
previously, however, changes in the rumen environment that result in an
isomerase shift and increased formation of biohydrogenation intermediates
other than vaccenic acid will not result in increases in the rumenic acid content
of milk fat. These changes are often associated with increased rumen outflow
of trans-10 18:1 and trans-10,cis-12 CLA which results in dramatic reductions
in milk fat synthesis (Bauman & Lock, 2006).
The most effective dietary treatments for increasing the rumenic acid content
of milk fat are those that increase both the supply of C18 PUFA and modify the
rumen environment. The most widely studied of these is the use of fresh
pasture, with numerous studies indicating that fresh pasture results in a 2-fold
to 3-fold increase in the rumenic acid content of milk fat (e.g., Stanton et al.,
1997; Dhiman et al., 1999). The degree of response, however, decreases as the
pasture matures and the proportion in the diet decreases. Correspondingly,
seasonal effects on milk rumenic acid content have been reported, with the
trend that content is greatest when fresh pasture is plentiful, and decreases
throughout the growing season (Auldist et al., 2002; Lock & Garnsworthy,
2003). These results cannot be explained fully in terms of the fatty acid
composition and supply of PUFA that grass provides; therefore, there must be
additional factors or components of grass that promote the production of
vaccenic acid in the rumen, and these lessen in effect as the pasture matures
(Lock & Bauman, 2004). Presumably, these factors inhibit the conversion of
vaccenic acid to stearic acid, as discussed previously. The effect of different
farming systems has also been investigated, with systems differentiated by the
amount and type of forages typically fed to cows. In general, production
systems with the greatest proportion of fresh forage in the diet often result in the
highest level of rumenic acid in milk fat.
Although the use of fresh pasture has striking effects on enhancing the
rumenic acid content of milk fat, a similar or even greater increase is possible
using standard dietary ingredients such as plant oils/oilseeds and fish oil/fish
meal supplements. Further, there is some indication that dietary regimes
involving a combination of supplements can have an additive effect on
increasing the level of rumenic acid in milk; for example Whitlock et al. (2002)
observed higher levels with a combination of plant oil and fish oil than when
either was fed alone. In all of the dietary situations designed to enhance the
level of rumenic acid in milk fat, it is vital that the normal vaccenic acid
pathway of biohydrogenation is maintained. If shifts in biohydrogenation
occur, then the pattern of trans-18:1 fatty acids is altered resulting in a
reduction in the rumen output of vaccenic acid, and as a consequence a
reduction in the level of rumenic acid in milk fat.
The final method to enhance the level of rumenic acid in milk is to increase
endogenous synthesis and this must be the basis for the variation among cows
in a herd fed the same diet. Undoubtedly, the variation in Δ9-desaturase among
individuals has a genetic basis but this has not been examined directly.
However, an indirect evaluation is possible because milk fat contains four
major fatty acid pairs that represent a product/substrate relationship for Δ9-
desaturase, myristoleic/myristic acid, palmitoleic/palmitic acid, oleic/stearic
acid and rumenic acid/vaccenic acid. As mentioned earlier, ratios for these
pairs of fatty acids, referred to as a desaturase index, represent a proxy for Δ9-
desaturase activity. In the largest study to examine this, Kelsey et al. (2003)
reported a 3-fold difference in both milk fat content of rumenic acid and
desaturase index among individuals consuming the same diet. Other investiga-
tions have also observed a 2-fold to 3-fold range in desaturase index among
cows in the same herd (e.g. Lock & Garnsworthy, 2002; Peterson et al., 2002).
Peterson et al. (2002) also demonstrated a consistency in the individual hierar-
chy in desaturase index over time when cows were fed the same diet and a
consistency in the individual hierarchy when cows were switched between
diets. Furthermore, the same studies also observed a similar 2-fold to 3-fold
range in the milk fat content of rumenic acid among individual animals (Lock
& Garnsworthy, 2002; Peterson et al., 2002).
In a structured genetic study involving 2400 cows, the heritability in the
desaturase index (using myristoleic/myristic acid ratio) was approximately 0.3
(Garnsworthy & Royal, 2005). Presumably this variation reflects individual
differences in the activity of Δ9-desaturase involving the regulation of gene
expression, primary or tertiary structure of the enzyme due to gene
polymorphisms, post-translational modifications, or other factors affecting the
interaction between the enzyme and the substrate or product. Furthermore, it
was recently demonstrated that the myristoleic/myristic acid ratio correlates
well with concentration of Δ9-desaturase mRNA in somatic cells extracted
from milk, and that between-cow variability seems to account for more
variation in Δ9-desaturase mRNA than does stage of lactation or diet (Feng
et al., 2007). Thus, cows could potentially be selected for high mammary
expression of Δ9-desaturase. Whether this would have a significant impact on
milk fatty acid composition, however, remains to be established.
Several specific physiological factors have been examined for effects on the
level of CLA in milk fat, but because of the large impact of diet and the wide
range among individuals, it is important that these comparisons involve a
reasonable number of cows fed a common diet. Studies meeting these condi-
tions found that the rumenic acid content of milk fat and the desaturase index
had no relationship to parity or stage of lactation (days in milk) (Kelsey et al.,
2003; Lock et al., 2005a). Likewise, they observed that milk fat content of
rumenic acid and desaturase index also had no relationship to milk yield, milk
fat percent or yield of milk fat (Kelsey et al., 2003; Lock et al., 2005a). The
investigation by Kelsey et al. (2003) involved over 200 cows fed the same diet
and found no difference between Holstein and Brown Swiss breeds. In contrast,
several studies have reported breed differences in the rumenic acid content of
milk fat (e.g. Lawless et al., 1999; Whitlock et al., 2002), which may reflect
differences in desaturase index among breeds. However, these studies often
involved very few animals or were confounded by diet, or both. Using a larger
data set, DePeters et al. (1995) reported breed differences in the desaturase
index in milk fat of dairy cows, consistent with the suggestion that the activity
of Δ9-desaturase is higher in Holstein than in Jersey mammary tissue (Beaulieu
& Palmquist, 1995). If breed differences exist they would, however, appear to
be minor compared with the magnitude of dietary effects and variation among
cows in terms of both the rumenic acid content of milk fat and desaturase index.
6. Manufacturing and product quality considerations related to rumenic

acid enriched milk fat
Central to marketing and consumer acceptance of rumenic acid enriched foods
is a consideration of the effects of processing and storage, and the final sensory
characteristics of rumenic acid enriched products. Many dairy products un-
dergo a microbial fermentation during processing and the effects of this on the
rumenic acid content have been of special interest. Several studies have
investigated this and found that food processing and manufacturing have little
or no effect on rumenic acid content (e.g. Werner et al., 1992; Shantha et al.,
1995). As emphasized in the review by Parodi (2003), any changes in the
rumenic acid content related to processing or to storage are minimal when
compared to the variations associated with on-farm differences in diet formu-
lations and variation among individual cows. Thus, the final concentration of
rumenic acid in dairy products is, in large part, related to the rumenic acid
concentration in the raw milk fat and the fat content of the final product.
Consumer acceptability of rumenic acid enriched dairy products is also
dependent on their taste and organoleptic properties. Off-flavours due to fatty
acid oxidation are of prime concern because diet formulation methods used to
naturally enhance milk fat with respect to rumenic acid generally cause an
increase in the proportion of unsaturated fatty acids in the milk fat (Lock &
Bauman, 2004). Reports on sensory characteristics and quality of naturally-
enriched dairy products, typically having a 2-fold to 3-fold increase in milk fat
rumenic acid, have generally indicated no differences from un-enriched dairy
products (e.g. Baer et al., 2001; Ramaswamy et al., 2001a; Ramaswamy et al.,
2001b; Gonzalez et al., 2003). For example, Lynch et al. (2005) compared the
flavour, organoleptic and storage characteristics of standard 2%-fat milk with
2%-fat milk that had an approximately 10-fold higher level of rumenic acid.
The naturally-enhanced milk (rumenic acid and vaccenic acid equalled 47 and
121 mg/g fatty acids, respectively) was produced through individual selection
and nutritional management of the cows. Initial evaluation of the milk and
evaluation over a 14 day post-pasteurization period indicated no flavour
differences as determined by triangle taste tests. Similarly, sensory results
indicated no differences in susceptibility to the development of oxidized off-
flavours between the control and rumenic acid enhanced milks, even when milk
was stored under light (Lynch et al., 2005). Thus, flavour and consumer
acceptability were maintained in a dairy product with substantially enhanced
levels of rumenic acid and vaccenic acid. Two studies have reported adverse
effect of rumenic acid enrichment on sensory and storage properties of milk;
however, one of these studies used extremely high levels of fish oil in the cows’
diet (Lacasse et al., 2002) and the other study involved a post-harvest fortifica-
tion of milk fat with synthetic CLA during the manufacturing process (Campbell
et al., 2003). Overall, results to date indicate that manufacturing and quality
characteristics were similar for dairy products naturally-enriched with CLA
compared with regular dairy products, and consumer acceptability was compa-
rable to un-enriched dairy products.
C. Biological effects of rumenic acid: cancer

There are a number of reviews available providing a broad overview of the
biological effects of CLA isomers. Thus, the emphasis in this and in the
following section will be on rumenic acid, particularly its biological effects
when supplied as a natural component of the diet. Although other CLA isomers
are present in milk fat, they appear to be present at concentrations much too low
to have a significant biological effect as functional food components. Rumenic
acid appears to play no role in the anti-obesity and anti-diabetogenic effects
that have been observed with mixed CLA isomers. Rather, it is the trans-10,cis-
12 CLA isomer that is responsible for these biological effects. Furthermore,
recent investigations have reported that under some situations, CLA supple-
ments may cause undesirable effects related to hyperinsulinaemia and insulin
resistance; these have also been specifically attributed to trans-10,cis-12 CLA
and have not been observed with rumenic acid. A detailed discussion of
specific biological effects of CLA isomers can be found elsewhere (e.g. Belury,
2002; Bauman l et al., 2006; Parodi, 2006).
Since the original finding of the anti-mutagenic properties of CLA (Pariza
et al., 1979; Ha et al., 1987), its anti-carcinogenic effects have received
widespread interest. Although there are 56 CLA isomers possible, and so far 20
isomers have been detected in foods and biological samples (e.g. Delmonte
et al., 2005), the anti-carcinogenic effects are primarily ascribed to the action
of rumenic acid, the most prevalent CLA isomer in the human food chain, and
trans-10,cis-12 a minor CLA isomer in ruminant-derived foods. There are
biomedical models for most types of cancer and many of these have been used
to investigate the role of CLA (and specifically rumenic acid) as an anti-
carcinogen (see reviews by Belury, 2002; Banni et al., 2003; Bauman et al.,
2006). These include the use of human cancer cell lines, transplanted cell lines
and in situ organ site carcinogenesis models. The latter are of particular value
in cancer investigations and dietary supplements of CLA have been shown to
be effective in inhibiting chemically-induced skin papillomas, fore-stomach
neoplasia, preneoplastic lesions and tumours in the colon and mammary gland
(Parodi, 2004; Bauman et al., 2006). Table 4 highlights a partial list of invest-
Table 4. Partial list of investigations examining the anti-carcinogenic effects of

conjugated linoleic acid (CLA) isomers in cancer models.
Cell cultures In vivo Animal models
• Breast • Skin papillomas

• Ovarian • Forestomach neoplasia
• Prostate • Colon aberrant crypt foci
• Colon • Mammary premaligant lesions
• Liver • Mammary adenocarcinomas
• Lung
• Melanoma Tumor transplants
• Mesothelioma • Breast cancer cells
• Glioblastoma • Prostrate cancer cells
• Leukemia
igations examining the anti-carcinogenic effects of CLA isomers in various

cancer models.
1. Cell culture studies.

Despite numerous cell culture experiments, mainly conducted with CLA
mixtures (rumenic acid and trans-10,cis-12 in a 50:50 ratio), the effect of
rumenic acid on different cancer cells is an area of study that is still progressing.
One of the fields most researched is the inhibitory effect of CLA on breast
cancer, where rumenic acid is reported to be an active isomer. In various
studies, rumenic acid was shown to inhibit the proliferation of the oestrogen
receptor positive (ER+) human breast cancer cell line MCF-7 (e.g., Chujo
et al., 2003; Miller et al., 2003b). Interestingly, rumenic acid exhibited an
inhibitory effect only on the proliferation of the ER+ MCF-7cell line, but not on
an oestrogen receptor negative breast cancer cell line (MDA-MB-231), even at
the highest concentration of 200 µmol/l (Tanmahasamut et al., 2004). Rumenic
acid showed the least growth inhibitory effects (~20%) in MCF-7 cells when
compared to other CLA isomers investigated (most to least potent: 9c,11c >
10t,12c > 9t,11t > 11c,13t > rumenic acid). In contrast, Fite et al. (2007) did not
detect any growth inhibition of rumenic acid in either cell type. In more applied
studies, it was reported that milk fat triacylglycerol-bound CLA, consisting
primarily of rumenic acid, suppressed growth of MCF-7 cells (O’Shea et al.,
2000; Miller et al., 2003b). The same group indicated that vaccenic acid elicits
anti-proliferative effects in MCF-7 cells most likely as a result of its conversion
into rumenic acid (Miller et al., 2003a).
In studies with gastric (SW480, SGC-7901), colorectal (Caco-2, DLD-1,
HT-29, MIP-101), and prostate (PC-3, DU-145) cancer cell lines, rumenic acid
inhibited cell growth and proliferation depending on the concentration added to
the cell culture (Miller et al., 2002; Palombo et al., 2002; Ochoa et al., 2004;
De la Torre et al., 2005; Beppu et al., 2006; Kim et al., 2006). However, in
most of those studies, with the exception of those conducted in colon cancer
cell lines, the trans-10,cis-12 CLA isomer was much more efficacious than
rumenic acid. Yet, other studies showed that rumenic acid did not repress cell
proliferation in the colon cancer cells Caco-2 and HT-29 (Kim et al., 2002; Cho
et al., 2003; Lee et al., 2006). On the contrary, when human colon cancer cells
(SW480) were treated with milk fat triacylglycerol-bound CLA (consisting
primarily of rumenic acid) cell growth was significantly suppressed (Miller
et al., 2003b).
One study evaluated the effectiveness of rumenic acid on RLh-84 (rat
hepatoma cell line) where rumenic acid was found to be a growth promoter
rather than an inhibitor (Yamasaki et al., 2002). The data suggest that rumenic
acid may elicit different effects in various cell lines.
Taken together, the inhibitory potency of rumenic acid varies considerably
among different cell lines, depending also on experimental conditions, such as
concentration of rumenic acid and duration of the treatment. Clearly, in vitro
data with cell cultures demonstrate that the pure single isomer rumenic acid was
only moderately effective against treatment of cancer especially when com-
pared with other pure CLA isomers. However, it is conceivable that rumenic
acid derived from milk is more proficient in suppressing tumour cell growth
than the synthetic single isomer rumenic acid, indicating that its precursor
vaccenic acid (or other active components in milk) are involved in the modu-
lation of tumour cell growth.
2. Animal studies
Because of the difficulty in obtaining pure rumenic acid for animal studies,
previous studies on animal models in cancer research focused on CLA mix-
tures, mainly rumenic acid and trans-10,cis-12 in a ratio of 1:1. Thus, there
have been only a few animal studies documenting effects of rumenic acid in
decreasing the risk of various cancer types, many of which utilized rumenic
acid enriched milk fat.
Diets, containing either rumenic acid enriched butter fat (0.8% rumenic acid)
or 1% purified rumenic acid, fed to rats 30 days prior to N-nitroso-N-methyl
nitrosourea (MNU) treatment, were found to be equally effective in inhibiting
mammary tumorigenesis (reduced mammary epithelial mass, decreased size of
the terminal end bud population, suppressed proliferation of terminal end bud
population cells, and inhibited mammary tumour yield (Table 5) (Ip et al.,
1999). In a subsequent study, the same group verified that dietary rumenic acid
at a dose level of 0.5% fed to rats for 6 or 24 weeks after MNU administration
was also effective in inhibiting the development of premalignant lesions (33–
36%) and the total number of carcinomas (35–40%) in the mammary gland (Ip
et al., 2002). In a similar study, 1% rumenic acid fed to rats for 6 or 20 weeks
Table 5. Bioassay of mammary cancer prevention in rats fed different sources of

conjugated linoleic acids (CLA)1. Adapted from Ip et al. (1999).
CLA content (µg/mg lipid) Mammary tumours

Dietary treatment Total CLA Plasma Mammary fat Incidence Total No.
in diet (%)
Control butter 0.1 5.4a 7.2a 28/30a 92a

(93%)
High VA/CLA butter 0.8 23.3c 36.5c 15/30b 43b
(50%)
Control butter & 0.8 18.4c 26.2b 16/30b 46b
synthetic CLA (53%)
1
Dietary treatments were initiated at weaning and continued for 30 days. All animals were then injected
with methylnitrosourea (MNU) to induce mammary tumours and switched to a 5% corn oil diet with no
CLA. They remained on this diet for 24 weeks and were then sacrificed for tissue analysis. Values with
unlike superscripts in the same column (a, b, c) differ (P < 0.05). VA, vaccenic acid.
after MNU treatment resulted in 45% decrease in tumour mass. However,

tumour multiplicity and latency remained unaffected (Lavillonniere et al.,
2003). Feeding a diet containing 2% vaccenic acid for 6 weeks after MNU
treatment inhibited premalignant lesions by 50% in the mammary gland to the
same extent as 1% rumenic acid (Banni et al., 2001). In this study, vaccenic
acid was metabolized to rumenic acid at a conversion rate of ~50%. A study in
a specific mouse model of mammary cancer that used transplanted mammary
tumour cells showed that 0.1% and 0.25% rumenic acid in the diet decreased
pulmonary tumour burden after spontaneous metastasis but did not affect
primary tumour growth or latency (Hubbard et al., 2003). In a recent study,
0.5% rumenic acid fed to transgenic mice over-expressing wild-type erbB2 in
the mammary epithelium (a protein giving higher aggressiveness in breast
cancers) either prepubertal or postpubertal, respectively, did not alter tumori-
genesis or metastasis (Ip et al., 2007).
Using a benzo[a]pyrene-induced forestomach neoplasia model, the addition
of 0.5% of rumenic acid (75% or 98% purity) in the diet fed over a period of
7 weeks to postpubertal mice, reduced significantly the tumour size, but not
average tumour number (Chen et al., 2003). Moreover, 98% pure rumenic acid
induced stronger inhibition of carcinogenesis than 78% pure rumenic acid.
Another group showed that 1% rumenic acid, fed over a time course of
30 weeks, reduced the incidence rate of 1,2-dimethylhydrazine-induced colon
tumours in rats (Park et al., 2006). In contrast, rumenic acid at 1% of the diet,
for 8 weeks, had no effect on tumour number and size in the large and small
intestine in Min mice, an animal model for intestinal carcinogenesis (Rajakangas
et al., 2003).
Since rumenic acid is found predominately in the dairy fat component of
human diets, a series of studies have used the rat prepubertal mammary cancer
model to investigate the anti-carcinogenic potential of rumenic acid when

supplied as a naturally-enriched butter that was produced using dietary regimes
described previously. The use of a functional food approach would have many
advantages as a strategy to prevent cancer. As described earlier (Table 3), the
initial investigation established that rumenic acid was an effective anti-car-
cinogen when it was supplied as a dietary food in a natural form (esterified in
triacylglycerols; Ip et al., 1999). Importantly, tissue concentrations of rumenic
acid were greater in rats fed the butter enriched in vaccenic acid and rumenic
acid than in rats fed a comparable amount of chemically-synthesized rumenic
acid, suggesting the possibility of endogenous synthesis from vaccenic acid.
Subsequent investigations established that dietary vaccenic acid derived from
butter enriched in vaccenic and rumenic acid also resulted in a dose-dependent
increase in the accumulation of rumenic acid in the mammary fat pad, which
was accompanied by a parallel decrease in tumour incidence and tumour
number (Corl et al., 2003), and that the anti-carcinogenic effects of vaccenic
acid were predominately, perhaps exclusively, mediated through its conver-
sion to rumenic acid via Δ9-desaturase (Lock et al., 2004). Therefore, vaccenic
acid and rumenic acid derived from milk fat are both anti-carcinogenic and this
series of pre-clinical investigations clearly demonstrates the feasibility of a
functional food approach using dairy products enriched in vaccenic acid and
rumenic acid in the prevention of mammary cancer.
In summary, the majority, but not all, of the studies described above indicate
a protective effect of rumenic acid at all stages of the tumorigenic process from
preventing the occurrence to modifying the spread of tumours. Studies using
the natural form of rumenic acid (milk fat) have proven to be particularly
effective. Finally, it is noteworthy that rumenic acid may be more effective in
preventing the development of tumours, a role consistent with its potential
value as a functional food component, rather than having an impact as a cancer
therapy (Bauman et al., 2006).
A number of issues remain to be resolved before any definitive conclusions
can be drawn concerning the anti-cancer effects of rumenic acid. These
include: the optimal (lowest) dose and form of rumenic acid (free fatty acid, or
in its original triacylglycerol form in milk lipids); interactions among CLA
isomers and other functional active milk fat components such as butyric acid,
branched-chain fatty acids, or sphingolipids; the role of rumenic acid uptake
and disposition, organ and tissue specificity; the appropriate animal model, and
the problem of extrapolation from rodent models to human populations.
3. Human studies
Human studies of rumenic acid and cancer have been scarce. Evaluating the
specific role of rumenic acid in health maintenance and the prevention of
cancer in humans is difficult, since cancer takes many years to develop. Thus,
documenting that dietary rumenic acid is beneficial in health maintenance and

the prevention of this disease is a major challenge (Bauman et al., 2006). The
only data available on the effects of rumenic acid on cancer in humans come
from epidemiological studies which, not surprisingly, offer conflicting results.
There have been just three epidemiologic studies which correlated the risk of
breast cancer with the intake of rumenic acid. In a Finnish case control study
with premenopausal and postmenopausal women, a low intake of rumenic acid
was associated with increased risk of breast cancer in postmenopausal, but not
in premenopausal, women (Aro et al., 2000). A relatively large cohort study,
carried out in the Netherlands, found a very weak, but positive correlation
between rumenic acid intake and breast cancer incidence in postmenopausal
women (Voorrips et al., 2002). Interestingly, there was no association between
food items containing rumenic acid and breast cancer risk. A case-control
study, conducted in the USA, found that risk of ER-negative breast cancer
among premenopausal women was reduced in the highest quartile of rumenic
acid intake. No association, however, was observed between intakes of rumenic
acid and overall risk of premenopausal or postmenopausal breast cancer
(McCann et al., 2004).
Taken together, although in vitro and in vivo evidence of the role of rumenic
acid in cancer prevention is persuasive, this role remains to be proven because
data obtained from epidemiological studies are not conclusive and cannot be
taken as definitive. The application processes would be hastened by using
biomarkers as potential surrogate endpoints to predict reduced cancer risk,
rather than measures of morbidity, but to date there are no consensus biomarkers
for most types of cancer. Conflicting results among these studies likely stem
from several sources, but there appear to be three over-riding limitations. First,
simply knowing the intake of dairy products is not sufficient; rumenic acid is
associated with the fat in dairy products which varies widely among particular
dairy foods. Even when the rumenic acid content is expressed on a fat basis, a
several-fold variation in CLA within a dairy product line is still observed due
to the effects of dietary practices and individual cow differences in dairy
production as discussed earlier. Furthermore, butter is often added to other
foods during manufacturing and is an ingredient in many recipes. Endogenous
synthesis is also a factor, and the kinetic study by Turpeinen et al. (2002)
demonstrated a substantial variation among individuals in the extent of vaccenic
acid conversion to rumenic acid. Second, dietary measurements and laboratory
methods used often have poor reliability and the adequacy of the analysis of
rumenic acid in the existing databases for fatty acid composition of foods is
questionable. Third, the necessarily statistical nature of these studies makes
data generation and interpretation inherently susceptible to confounding fac-
tors. Of particular concern when attempting to isolate the effects of rumenic
acid through dietary surveys is the wide array of other compounds, that are
found in the same food sources as rumenic acid and may also have roles in the
development of cancer. These issues are discussed further in Bauman et al.

(2006). Clearly, assessing the role of dietary rumenic acid (particularly as a
natural component of our diet) for the prevention of cancer presents some
unusual difficulties, and thus many of the traditional approaches to evaluate
human health effects have substantial limitations.
Finally, in humans it is unclear if the amounts of rumenic acid consumed via
a normal diet are functionally significant. This fact has become especially
problematic because current guidelines suggest that the amount of fat from
ruminant meat and dairy products, which is the major source of rumenic acid,
should be reduced. Nonetheless, taking into account the endogenous synthesis
of rumenic acid from ruminant-derived vaccenic acid, and the increases that
can be achieved in the rumenic acid and vaccenic acid content of dairy products
as highlighted earlier, it appears likely that the potential protective benefits of
dietary rumenic acid intake may be achievable using functional food products
that have been naturally enhanced in vaccenic acid and rumenic acid (Bauman
et al., 2006).
4. Potential mechanisms of action

Mechanisms of action of rumenic acid have not, as yet, been fully elucidated.
Over the past five years, however, substantial progress has been made to
explain the molecular basis for the anti-carcinogenic activity of rumenic acid.
Once taken up into the cell, rumenic acid may impact the formation and
metabolism of eicosanoids as well as cellular events that regulate signal
transduction and gene expression. The possible mechanisms of rumenic acid as
an anti-tumour nutrient include alteration of tissue membrane fatty acid com-
position and subsequently alteration of eicosanoid metabolism; modulation of
cell proliferation and apoptosis via regulation of gene expression; and inhibi-
tion of angiogenesis. Taken together, the multiple mechanisms attributed to
rumenic acid, and/or its metabolites, suggest a high degree of tissue specificity.
In addition, it is conceivable that the cancer type/site, species/strain, age, type
of carcinogen, stage of carcinogenesis, etc may play an important role in the
mechanisms of action as well. Finally, it is important to note that, with regard
to rumenic acid, biological effects are mostly related to prevention rather than
treatment of cancers; it is inherently more difficult to ascertain mechanisms of
action for the prevention of a chronic disease compared with mechanisms
related to treating or curing such a disease.
D. Biological effects of rumenic acid: atherosclerosis

Atherosclerosis is characterized by the accumulation of cholesterol, lipids and
lipid peroxidation product deposits in large-sized and medium-sized arteries;
inflammatory and oxidative mechanisms, coupled with dyslipidaemia, initiate
the atheroma formation. This leads to a proliferation of certain cell types within
the arterial wall that gradually impinge on the vessel lumen and impede blood
flow. Atherosclerosis is the primary cause of CHD around the world, and in
North America, is the underlying cause of nearly half of all deaths (Lusis,
2000).
1. Animal studies
Changes in both total plasma cholesterol and individual lipoprotein cholesterol
concentrations have been implicated as major determinants of atherosclerosis
disease risk. This has led to a number of studies specifically investigating the
effects of rumenic acid on cholesterol and lipoprotein metabolism. The ability
of rumenic acid to alter blood lipid profiles has been tested in various animal
models, such as the apoE knock-out mouse, the Golden Syrian hamster, and the
New Zealand White rabbit. In apoE–/– mice, diets rich in rumenic acid de-
creased triacylglycerol and triacylglycerol-rich lipoproteins and raised
high-density lipoprotein cholesterol (HDL-C) in diabetic and non-diabetic
mice, respectively (Nestel et al., 2006). In Golden Syrian hamsters, rumenic
acid has been demonstrated to have neutral and/or beneficial effects on blood
lipids. Mitchell et al. (2005) showed that when supplemented in amounts as
high as 1% of diet, rumenic acid had no effect on blood lipid profile. However,
Ledoux et al. (2007) determined that adding 1% rumenic acid to the diet
decreased total cholesterol, low-density lipoprotein cholesterol (LDL-C), and
HDL-C, when compared to animals fed control diets. Differences in the control
diets used in these studies should be noted. In the study in which rumenic acid
had neutral effects, it was compared to a diet high in non-conjugated linoleic
acid isomers. In the study in which rumenic acid had beneficial effects, it was
compared to a diet rich in oleic acid. It is noteworthy that both oleic and linoleic
acid have been shown to reduce total cholesterol and LDL-C (Mensink et al.,
2003). These results indicate that consuming rumenic acid may improve
cholesterol levels, and certainly not adversely affect them. While both of the
above studies were conducted with pure rumenic acid sources, other studies
have been conducted using whole foods to deliver the experimental rumenic
acid. Valeille et al. (2005) supplemented butter with ~1% rumenic acid which
decreased the non-HDL-C to HDL-C ratio in Syrian Golden hamsters com-
pared to hamsters fed regular butter. In another study, when butter naturally
enriched with rumenic acid (~1%) and vaccenic acid was fed to Golden Syrian
hamsters, total cholesterol, LDL-C, and non-HDL-C to HDL-C ratio were
lower than in those fed a control diet (Valeille et al., 2006). This is in agreement
with Lock et al. (2005b) who also used the Golden Syrian hamster to examine
the potential of rumenic acid when fed as a component of a functional food
(butter enriched in vaccenic acid and rumenic acid) as part of a diet that was
high in cholesterol (0.2%) and fat (20%). Compared with the control animals,
12 a
Control
Cholesterol (mmol/litre) 10 VA/RA-enriched butter
b
8
4
a
2 b
a
b
0
Total LDL HDL LDL/HDL
Plasma Cholesterol
Figure 3. Effect of butter enriched in vaccenic and rumenic acids (VA/RA) on plasma cholesterol
levels in the cholesterol-fed Golden Syrian hamster. Adapted from Lock et al. (2005b). Standard error is
indicated by error bars over each column. P < 0.01 for treatment effects with significance differences
indicated by letter differences (a, b) over the columns.
those fed the enriched butter showed a number of beneficial effects, including
reduced total plasma cholesterol, and very-low-density lipoprotein cholesterol
(VLDL) and LDL, suggesting that CLA may modify the production of athero-
genic lipoproteins by the liver (Figure 3). In addition, the enriched butter
produced a less atherogenic profile than an equivalent diet in which the
enriched butter was replaced with trans fatty acids from partially hydrogenated
vegetable oil (Lock et al., 2005b). A similar study in New Zealand White
rabbits determined that rumenic acid enriched butter had a neutral effect on
plasma lipoprotein cholesterol (Bauchart et al., 2007). Finally, rumenic acid
was reported to have neutral effects on plasma and hepatic cholesterol, as well
as hepatic esterified cholesterol, in rabbits, when compared to animals fed a
cholesterol-free diet (Kritchevsky et al., 2004).
Although rumenic acid has been widely recognised for its ability to improve
plasma cholesterol and lipoprotein profiles, perhaps more impressive is evi-
dence demonstrating its ability to reduce the size of atherosclerotic lesions, and
halt the progression of such lesions in animal models. Supplementation of a
high-fat, high-cholesterol diet with 1% rumenic acid resulted in fewer fatty
streak lesions in the aortas of Golden Syrian hamsters when compared with
those fed isomers of non-conjugated linoleic acid (Mitchell et al., 2005).
Although findings were not statistically significant, there were nearly 20%
fewer fatty lesions present than in control animals. New Zealand White rabbits
fed a 50:50 mixture of rumenic acid and trans-10,cis-12 CLA isomer mixture
at 0.1% of diet exhibited a lower plaque diameter in the abdominal aorta than
animals fed a CLA-free diet (Kritchevsky et al., 2000). When the isomer blend
was increased to 0.5% and 1.0%, plaque thickness was decreased in both the
thoracic aorta and aortic arch (Kritchevsky et al., 2000), indicating a decrease
in the progression of atherosclerosis. Rumenic acid, alone, significantly re-
duced the size of lesions in both the thoracic aorta and aortic arch of New
Zealand White rabbits fed an atherogenic diet (Kritchevsky et al., 2004). In a
recent study in which butter was naturally-enriched either with vaccenic acid
and rumenic acid or with trans-10 18:1, the trans-10 18:1-enriched butter was
found to have detrimental effects as evidenced by increases in plasma concen-
trations of total cholesterol, LDL-C, VLDL-C, and non-HDL-C:HDL-C ratio
(Roy et al., 2007). In contrast, vaccenic/rumenic enriched butter had neutral
effects on blood lipid profiles and tended to reduce aortic lipid deposition.
It is important to note that an elevated and/or altered plasma lipid level is
only one of a wide range of risk factors that contribute to the clinical manifes-
tations of atherosclerosis in humans (Lusis, 2000). Consequently, in some
studies, the reduced incidence of atherosclerosis in animals fed CLA isomers
was not accompanied by an improvement in the plasma lipid profile during the
CLA-feeding phase (e.g. Wilson et al., 2000). Reasons for these effects are not
fully understood. Atherosclerosis, however, is a chronic inflammatory disease
(Libby, 2002) and several important anti-inflammatory effects have been
associated with rumenic acid.
As previously stated, when Golden Syrian hamsters were fed butter supple-
mented with rumenic acid, they showed improved cholesterol profiles and
lower total fatty streak development in the aorta (Valeille et al., 2005). These
hamsters also had lower inflammatory serum amyloid A levels and improved
anti-oxidized LDL paraoxonase activity; as well as down-regulated expression
of proinflammatory cytokines such as tumour necrosis factor-α (TNF-α) and
interleukin-1β (IL-1β); decreased cyclooxygenase 2 (COX-2) activation of the
transcription factor peroxisome proliferator-activating receptor (PPAR)/liver X
receptor (LXR)-α signalling cascade; and increased ATP-binding cassette
subfamily A1 expression in the aorta (Valeille et al., 2005). Similar results
were demonstrated in a study in which hamsters were fed a diet supplemented
with up to 9 g/kg of rumenic acid as rumenic acid enriched milk fat. Hamsters
fed a high-fat diet (67 g/100 g saturated fatty acids), supplemented with 2.6%
of lipids from rumenic acid, had increased ATP-binding cassette subfamily A-
1 gene expression in the aorta, decreased LDL-peroxidability index, and
down-regulation of the proinflammatory IL-1β gene in the aorta when com-
pared to hamsters fed the control diet without rumenic acid supplementation
(Valeille et al., 2006).
In apo E–/– mice, a high-fat, high-cholesterol diet containing 7% rumenic acid
increased five post-translationally modified forms of heat shock protein 70kD
(HSP 70kD) (de Roos et al., 2005). In humans, elevated levels of this protein
are associated with low coronary artery disease risk, perhaps due to its inverse
association with the activity of COX-2, a proinflammatory cytokine (de Roos
et al., 2005). HSP 70kD attenuates NF-kB activation, an important transcrip-
tion factor in the expression of proinflammatory genes (de Roos et al., 2005),
which lends insight into the mechanism behind the ability of rumenic acid to
reduce inflammation.
2. Human studies
Since results from studies with biomedical models indicate potential beneficial
effects of rumenic acid, there is obvious interest in the effects of rumenic acid
consumption in foods on the risk of atherosclerosis in humans. The use of
surrogate biomarkers for disease risk is more readily achievable for atheroscle-
rosis than for cancer in humans and a number of genetic and environmental risk
factors have been identified, with the relative abundance of the different
lipoproteins being of primary importance (Lusis, 2000). To date, there have been
no epidemiological studies that have examined the relationship between intake
of rumenic acid derived from foods and risk of atherosclerosis. Several human
intervention studies involving dietary supplements of rumenic acid in the form of
capsules have reported plasma lipid variables as secondary observations, but
most used mixed isomers of CLA and found inconsistent results (see Bauman
et al., 2006). However, recent studies examined the specific effects of rumenic
acid on blood lipids in healthy subjects. In normal healthy men and women, a
capsule containing a mixture of rumenic acid and trans-10,cis-12 CLA, admin-
istered three times daily with food, resulted in reduced triacylglycerol levels
when administered in a 50:50 ratio, and decreased VLDL-C concentrations when
administered at an 80:20 ratio, respectively (Noone et al., 2002). When rumenic
acid and trans-10,cis-12 CLA where examined against each other in a dose-
response study in healthy men, rumenic acid supplementation consistently
resulted in lower total cholesterol, LDL-C, and LDL-C:HDL-C, with a dose-
response observed for total cholesterol and LDL-C (Tricon et al., 2004). In an
attempt to examine the effects of rumenic acid on lipoprotein profiles in the
context of the human experience, foods naturally enriched with rumenic acid
were fed to healthy men. At the conclusion of 6 weeks of eating naturally-
enriched rumenic acid milk, butter, and cheese alongside a normal diet, there
were no significant changes in total cholesterol, LDL-C, or HDL-C (Tricon et al.,
2006), although it is worth noting that subjects consumed twice the level of dairy
fat compared with baseline values. The enriched dairy foods changed LDL-C
fatty acid composition, but had no significant effect on LDL particle size or
susceptibility to oxidation (Tricon et al., 2006). These studies indicated that
human consumption of naturally-occurring rumenic acid, such as from dairy
products, had a neutral effect on lipoprotein profiles, with the potential for a
protective effect, dependent on dose.
In humans, the effect of rumenic acid on inflammation is difficult to discern.

Most studies have used isomer blends, making it difficult to determine the fatty
acids responsible for the observed effects. When fed a 50:50 rumenic acid to
trans-10,cis-12 CLA isomer blend for 12 weeks, both men and women showed
increases in C-reactive protein and interleukin-6 (Steck et al., 2007). However,
these markers, usually indicators of inflammation, were still well within the
normal range. Supplementation with an 80:20 mix of the same isomers,
respectively, showed no effect on markers of inflammation, including inter-
leukin-2 or TNF-α (Nugent et al., 2005), indicating a neutral effect of the
blend, which was predominant in rumenic acid. In regards to these findings,
further investigation is warranted into the effects of rumenic acid on inflamma-
tion in humans.
Overall, results indicate that rumenic acid has beneficial effects in reducing
risk factors for atherosclerosis. In particular, the available data indicate that
consumption of dairy foods rich in rumenic acid, although naturally high in
saturated fatty acids, may be protective against the development of atheroscle-
rosis.
E. Summary
Research in animal science has traditionally focused on the productivity and
well-being of food-producing animals. However, there is also a growing
recognition of the consumer desire for healthier and more nutritious foods. The
contributions of animal-derived foods in supplying essential nutrients have
been recognized for some time, but consumers are increasingly aware that
foods also contain components that can have positive effects on health mainte-
nance and the prevention of chronic diseases. Referred to as ‘bioactive food
components’ or ‘functional foods’, CLA in milk fat is such a component. Milk
fat contains many isomers of CLA, but rumenic acid predominantly, and this is
the CLA isomer with bioactive potential in relation to cancer and atherosclero-
sis. The uniqueness of rumenic acid in ruminant-derived foods is related to
rumen biohydrogenation of PUFA. Research over the last decade has estab-
lished that rumenic acid in milk fat originates mainly from endogenous
synthesis by mammary Δ9-desaturase from vaccenic acid, a biohydrogenation
intermediate produced in the rumen. Thus, an understanding of dietary lipid
supply, rumen fermentation and mammary synthesis of fat is required and
strategies to increase milk fat rumenic acid centre on enhancing rumen output
of vaccenic acid and increasing tissue activity of Δ9-desaturase.
The anti-carcinogenic activity of rumenic acid has been demonstrated with
animal models and in vitro studies for a wide range of cancer types. The anti-
cancer effects of rumenic acid have been particularly impressive in biomedical
models of breast cancer. By comparison, investigations of rumenic acid effects
on atherosclerosis are more limited. A number of studies using animal models
have reported that dietary supplementation with rumenic acid can improve the
plasma profile of lipids and lipoproteins, reduce the development of atheroscle-
rotic lesions, and indeed even induce the regression of pre-existing lesions. The
rumenic acid used for most studies of cancer and atherosclerosis has been
synthetically produced from vegetable oils. Of special importance are recent
results showing that beneficial effects were also achieved when the dietary
supply of rumenic acid was provided by a natural food. These studies are
among the first to demonstrate the feasibility of a functional food approach
using the natural form of rumenic acid (esterified) in a naturally-enriched food
(dairy products). Furthermore, these investigations also established that the
vaccenic acid in milk fat is a functional food component with beneficial effects
on cancer and atherosclerosis because it can be used for endogenous synthesis
of rumenic acid in humans.
Extending the results from biomedical models to implications of rumenic
acid and vaccenic acid as functional food components in humans is limited and
problematic. These evaluations are challenging because the development of
cancer and atherosclerosis has long latency periods and there is a lack of
consensus biomarkers, especially for cancer. Further, there have been signifi-
cant advances in the analysis of rumenic acid and specific trans fatty acids, and
these techniques have not yet been applied to updating values in food databases.
Finally, successful application of bioactive food components found in milk fat
includes the education of the public that not all fatty acids have the same
biological effects. This is of special importance with the introduction of food
labelling of trans fatty acid content as undesirable and the fact that both
vaccenic acid and rumenic acid are trans fatty acids. Overall, studies summa-
rized in this review demonstrate the exciting potential of rumenic acid and
vaccenic acid in dairy products as bioactive food components that may provide
benefits in health maintenance and disease prevention as well as improve the
public perception of milk and dairy products.
References
Abu-Ghazaleh AA, Schingoethe DJ and Hippen AR (2001) Conjugated linoleic acid and
other beneficial fatty acids in milk from cows fed soybean meal, fish meal, or both. J.
Dairy Sci., 84, 1845–1850.
Abu-Ghazaleh AA and Jenkins TC (2004) Short communication. Docosahexaenoic acid
promotes vaccenic acid accumulation in mixed ruminal cultures when incubated with
linoleic acid. J. Dairy Sci., 87, 1047–1050.
Adlof RO, Duval S and Emken EA (2000) Biosynthesis of conjugated linoleic acid in humans.
Lipids, 35, 131–135.
Aro A, Mannisto S, Salminen I, Ovaskainen ML, Kataja V and Uusitupa M (2000) Inverse
association between dietary and serum conjugated linoleic acid and risk of breast cancer
in postmenopausal women. Nutr. Cancer, 38, 151–157.
Auldist MJ, Kay JK, Thomson NA, Napper AR and Kolver ES (2002) Brief communication.
Concentrations of conjugated linoleic acid in milk from cows grazing pasture or fed a total
mixed ration for an entire lactation. Proc. New Zealand Soc. Anim. Prod., 62, 240–247.
Baer RJ, Ryali J, Schingoethe DJ, Kasperson KM, Donovan DC, Hippen AR and Franklin ST
(2001) Composition and properties of milk and butter from cows fed fish oil. J. Dairy Sci.,
84, 345–353.
Banni S, Angioni E, Murru E, Carta G, Melis MP, Bauman DE, Dong Y and Ip C (2001)
Vaccenic acid feeding increases tissue levels of conjugated linoleic acid and suppresses
development of premalignant lesions in rat mammary gland. Nutr. Cancer, 41, 91–97.
Banni S, Heys SD and Wahle KWJ (2003) Conjugated linoleic acids as anticancer nutrients:
studies in vivo and cellular mechanisms. In: Advances in Conjugated Linoleic Acid
Research, Volume 2 (JL Sebedio, WW Christie and RO Adlof, eds) AOCS Press,
Champaign, Illinois, USA.
Barber MC, Ward RJ, Richards SE, Salter AM, Buttery PJ, Vernon RG and Travers MT
(2000) Ovine adipose tissue monounsaturated fat content is correlated to depot-specific
expression of the stearoyl-CoA desaturase gene. J. Anim. Sci., 78, 62–68.
Bauchart D, Roy A, Lorenz S, Chardigny JM, Ferlay A, Gruffat D, Sebedio JL, Chilliard Y
and Durand D (2007) Butters varying in trans 18:1 and cis-9,trans-11 conjugated linoleic
acid modify plasma lipoproteins in the hypercholesterolemic rabbit. Lipids, 42, 123–133.
Bauman DE, Barbano DM, Dwyer DA and Griinari JM (2000) Technical note. Production
Bauman DE and Lock AL (2006) Conjugated linoleic acid: biosynthesis and nutritional
significance. In: Advanced Dairy Chemistry, Volume 2: Lipids, 3rd Edition, (PF Fox and
PLH McSweeney, eds) Kluwer Academic/Plenum Publishers, New York, USA, pp.93–
136.
Bauman DE and Lock AL (2008) Modifying animal fat to enhance animal and human health.
Proc. 29th Canadian Western Nutrition Conference, pp.265–277.
Bauman DE, Lock AL, Corl BA, Salter AM, Ip C and Parodi PW (2006) Milk fatty acids and
human health potential role of conjugated linoleic acid and trans fatty acids. In: Ruminant
Physiology: Digestion, Metabolism and Impact of Nutrition on Gene Expression, Immu-
nology and Stress (K Sejrsen, T Hvelplund and MO Nielsen, eds) Wageningen Academic
Publishers, Wageningen, The Netherlands, pp.523–555.
Bauman DE, Perfield JW, 2nd, Harvatine KJ and Baumgard LH (2008) Regulation of fat
synthesis by conjugated linoleic acid: lactation and the ruminant model. J. Nutr., 138,
403–409.
Beaulieu AD and Palmquist DL (1995) Differential-effects of high-fat diets on fatty-acid
composition in milk of Jersey and Holstein cows. J. Dairy Sci., 78, 1336–1344.
Bell JA, Griinari JM and Kennelly JJ (2006) Effect of safflower oil, flaxseed oil, monensin,
and vitamin E on concentration of conjugated linoleic acid in bovine milk fat. J. Dairy
Sci., 89, 733–748.
Belury MA (2002) Dietary conjugated linoleic acid in health: physiological effects and
mechanisms of action. Annu. Rev. Nutr., 22, 505–531.
Beppu F, Hosokawa M, Tanaka L, Kohno H, Tanaka T and Miyashita K (2006) Potent
inhibitory effect of trans9,trans11 isomer of conjugated linoleic acid on the growth of
human colon cancer cells. J. Nutr. Biochem., 17, 830–836.
Bickerstaffe R and Annison EF (1969) Glycerokinase and desaturase activity in pig, chicken
and sheep intestinal epithelium. Comp. Biochem. Physiol., 31, 47–54.
Bickerstaffe R and Annison EF (1970) The desaturase activity of goat and sow mammary
tissue. Comp. Biochem. Physiol., 35, 653–665.
Booth RG, Dann WJ, Kon SK and Moore T (1933) A new variable factor in butter fat. Chem.
Ind., 52, 270.
Campbell W, Drake MA and Larick DK (2003) The impact of fortification with conjugated
linoleic acid (CLA) on the quality of fluid milk. J. Dairy Sci., 86, 43–51.
Chang JHP, Lunt DK and Smith SB (1992) Fatty acid composition and fatty acid elongase
and stearoyl-CoA desaturase activities in tissues of steers fed high oleate sunflower seed.
J. Nutr., 122, 2074–2080.
Chen BQ, Yang YM, Gao YH, Liu JR, Xue YB, Wang XL, Zheng YM, Zhang JS and Liu
RH (2003) Inhibitory effects of c9,t11-conjugated linoleic acid on invasion of human
gastric carcinoma cell line SGC-7901. World J. Gastroenterol., 9, 1909–1914.
Chilliard Y, Ferlay A and Doreau M (2001) Effect of different types of forages, animal fat
or marine oils in cow’s diet on milk fat secretion and composition, especially conjugated
linoleic acid (CLA) and polyunsaturated fatty acids. Livest. Prod. Sci., 70, 31–48.
Cho HJ, Lee HS, Chung CK, Kang YH, Ha YL, Park HS and Park JH (2003) Trans-10, cis-
12 conjugated linoleic acid reduces insulin-like growth factor-II secretion in HT-29
human colon cancer cells. J. Med. Food, 6, 193–199.
Chujo H, Yamasaki M, Nou S, Koyanagi N, Tachibana H and Yamada K (2003) Effect of
conjugated linoleic acid isomers on growth factor-induced proliferation of human breast
cancer cells. Cancer Lett., 202, 81–87.
Chung M, Ha S, Jeong S, Bok J, Cho K, Baik M and Choi Y (2000) Cloning and
characterization of bovine stearoyl CoA desaturase 1 cDNA from adipose tissues. Biosci.
Biotechnol. Biochem., 64, 1526–1530.
Corl BA, Barbano DM, Bauman DE and Ip C (2003) cis-9,trans-11 CLA derived endogenously
from trans-11 18:1 reduces cancer risk in rats. J. Nutr., 133, 2893–2900.
Corl BA, Baumgard LH, Dwyer DA, Griinari JM, Phillips BS and Bauman DE (2001) The
role of delta(9)-desaturase in the production of cis-9,trans-11 CLA. J. Nutr. Biochem., 12,
622–630.
De la Torre A, Debiton E, Durand D, Chardigny JM, Berdeaux O, Loreau O, Barthomeuf C,
Bauchart D and Gruffat D (2005) Conjugated linoleic acid isomers and their conjugated
derivatives inhibit growth of human cancer cell lines. Anticancer Res., 25, 3943–3949.
de Roos B, Rucklidge G, Reid M, Ross K, Duncan G, Navarro MA, Arbones-Mainar JM,
Guzman-Garcia MA, Osada J, Browne J, Loscher CE and Roche HM (2005) Divergent
mechanisms of cis-9,trans-11- and trans-10, cis-12-conjugated linoleic acid affecting
insulin resistance and inflammation in apolipoprotein E knockout mice: a proteomics
approach. FASEB Journal, 19, 1746–1748. Epub.
Delmonte P, Kataoka A, Corl BA, Bauman DE and Yurawecz MP (2005) Relative retention
order of all isomers of cis/trans conjugated linoleic acid FAME from the 6,8 to 13,15
positions using silver ion HPLC with 2 different elution system. Lipids, 40, 509–514.
DePeters EJ, Medrano JF and Reed BA (1995) Fatty acid composition of milk fat from three
breeds of dairy cattle. Can. J. Anim. Sci., 75, 264–269.
Dhiman TR, Anand GR, Satter LD and Pariza MW (1999) Conjugated linoleic acid content
of milk from cows fed different diets. J. Dairy Sci., 82, 2146–2156.
Fellner V, Sauer FD and Kramer JKG (1997) Effect of nigericin, monensin, and tetronasin
on biohydrogenation in continuous flow-through ruminal fermenters. J. Dairy Sci., 80,
921–928.
Feng S, Salter AM, Parr T and Garnsworthy PC (2007) Extraction and quantitative analysis
of stearoyl-coenzyme A desaturase mRNA from dairy cow milk somatic cells. J. Dairy
Sci., 90, 4128–4136.
Fite A, Goua M, Wahle KW, Schofield AC, Hutcheon AW and Heys SD (2007) Potentiation
of the anti-tumour effect of docetaxel by conjugated linoleic acids (CLAs) in breast
cancer cells in vitro. Prostaglandins Leukot. Essent. Fatty Acids, 77, 87–96.
Franklin ST, Martin KR, Baer RJ, Schingoethe DJ and Hippen AR (1999) Dietary marine
algae (Schizochytrium sp.) increases concentrations of conjugated linoleic, docosa-
hexaenoic and trans vaccenic acids in milk of dairy cows. J. Nutr., 129, 2048–2054.
Garnsworthy PC and Royal MD (2005) Heritability of delta-9 desaturase enzyme activity in
dairy cows. Proc. 26th World Congress of the International Society for Fat Research,
Prague, Czech Republic, pp.55.
Glaser KR, Wenk C and Scheeder MRL (2002) Effects of feeding pigs increasing levels of
C18:1 trans fatty acids on fatty acid composition of backfat and intramuscular fat as well
as backfat firmness. Arch. Anim. Nutr., 56, 117–130.
Gomez FE, Bauman DE, Ntambi JM and Fox BG (2003) Effects of sterculic acid on stearoyl-
CoA desaturase in differentiating 3T3-L1 adipocytes. Biochem. Biophys. Res. Commun.,
300, 316–326.
Gonzalez S, Duncan SE, O’Keefe SF, Sumner SS and Herbein JH (2003) Oxidation and
textural characteristics of butter and ice cream with modified fatty acid profiles. J. Dairy
Sci., 86, 70–77.
Griinari JM and Bauman DE (1999) Biosynthesis of conjugated linoleic acid and its
incorporation into meat and milk in ruminants. In: Advances in Conjugated Linoleic Acid
Research (MP Yurawecz, MM Mossoba, JKG Kramer, MW Pariza and G Nelson, eds)
AOCS Press, Champaign, Illinois, USA, pp.180–200.
Griinari JM, Corl BA, Lacy SH, Chouinard PY, Nurmela KVV and Bauman DE (2000)
Conjugated linoleic acid is synthesized endogenously in lactating dairy cows by delta(9)-
desaturase. J. Nutr., 130, 2285–2291.
Ha YL, Grimm NK and Pariza MW (1987) Anticarcinogens from ground beef: heat altered
derivatives of linoleic acid. Carcinogenesis, 8, 1881–1887.
Harfoot CG and Hazlewood GP (1997) Lipid metabolism in the rumen. In: The Rumen
Microbial Ecosystem (Second Edition) (PN Hobson and DS Stewart, editors) Chapman
& Hall, London, UK, pp.382–426.
Hubbard NE, Lim D and Erickson KL (2003) Effect of separate conjugated linoleic acid
isomers on murine mammary tumorigenesis. Cancer Lett., 190, 13–19.
Ip C, Banni S, Angioni E, Carta G, McGinley J, Thompson HJ, Barbano D and Bauman DE
(1999) Conjugated linoleic acid-enriched butter fat alters mammary gland morphogen-
esis and reduces cancer risk in rats. J. Nutr., 129, 2135–2142.
Ip C, Dong Y, Ip MM, Banni S, Carta G, Angioni E, Murru E, Spada S, Melis MP and Saebo
A (2002) Conjugated linoleic acid isomers and mammary cancer prevention. Nutr.
Cancer, 43, 52–58.
Ip MM, McGee SO, Masso-Welch PA, Ip C, Meng X, Ou L and Shoemaker S (2007) The
t10,c12 isomer of conjugated linoleic acid stimulates mammary tumorigenesis in
transgenic mice overexpressing erbB2 in the mammary epithelium. Carcinogenesis, 28,
1269–1276.
Jahreis G, Fritsche J and Steinhart H (1997) Conjugated linoleic acid in milk fat: high
variation depending on production system. Nutr. Res., 17, 1479–1484.
Jeffcoat R and Pollard MR (1977) Studies on the inhibition of the desaturases by cyclopropenoid
fatty acids. Lipids, 12, 480–485.
Jenkins TC, Fellner V and McGuffey RK (2003) Monensin by fat interactions on trans fatty
acids in cultures of mixed ruminal microorganisms grown in continuous fermentors fed
corn or barley. J. Dairy Sci., 86, 324–330.
Jiang J, Bjoerck L, Fonden R and Emanuelson M (1996) Occurrence of conjugated cis-
9,trans-11-octadecadienoic acid in bovine milk: effects of feed and dietary regimen. J.
Dairy Sci., 79, 438–445.
Kay JK, Mackle TR, Auldist MJ, Thomson NA and Bauman DE (2004) Endogenous
synthesis of cis-9,trans-11 conjugated linoleic acid in dairy cows fed fresh pasture. J.
Dairy Sci., 87, 369–378.
Kelly ML, Berry JR, Dwyer DA, Griinari JM, Chouinard PY, Van Amburgh ME and Bauman
DE (1998) Dietary fatty acid sources affect conjugated linoleic acid concentrations in
milk from lactating dairy cows. J. Nutr., 128, 881–885.
Kelsey JA, Corl BA, Collier RJ and Bauman DE (2003) The effect of breed, parity, and stage
of lactation on conjugated linoleic acid (CLA) in milk fat from dairy cows. J. Dairy Sci.,
86, 2588–2597.
Kim EJ, Holthuizen PE, Park HS, Ha YL, Jung KC and Park JH (2002) Trans-10, cis-12-
conjugated linoleic acid inhibits Caco-2 colon cancer cell growth. Am. J. Physiol.
Gastrointest. Liver Physiol., 283, G357–367.
Kim EJ, Shin HK, Cho JS, Lee SK, Won MH, Kim JW and Park JH (2006) Trans-10,cis-12
conjugated linoleic acid inhibits the G1-S cell cycle progression in DU145 human
prostate carcinoma cells. J. Med. Food, 9, 293–299.
Kinsella JE (1972) Stearyl CoA as a precursor of oleic acid and glycerolipids in mammary
microsomes from lactating bovine: possible regulatory step in milk triglyceride synthe-
sis. Lipids, 7, 349–355.
Kraft J, Collomb M, Mockel P, Sieber R and Jahreis G (2003) Differences in CLA isomer
Kraft J, Hanske L, Mockel P, Zimmermann S, Hartl A, Kramer JKG and Jahreis G (2006) The
conversion efficiency of trans-11 and trans-12 18:1 by Δ9-desaturation differs in rats.
J. Nutr., 136, 1209–1214.
Kramer JKG, Parodi PW, Jensen RG, Mossoba MM, Yurawecz MP and Adlof RO (1998)
Kritchevsky D, Tepper SA, Wright S, Czarnecki SK, Wilson TA and Nicolosi RJ (2004)
Conjugated linoleic acid isomer effects in atherosclerosis: growth and regression of
lesions. Lipids, 39, 611–616.
Kritchevsky D, Tepper SA, Wright S, Tso P and Czarnecki SK (2000) Influence of conjugated
linoleic acid (CLA) on establishment and progression of atherosclerosis in rabbits. J. Am.
Coll. Nutr., 19, 472S–477S.
Kuhnt K, Kraft J, Moeckel P and Jahreis G (2006) Trans-11-18:1 is effectively Δ9-
desaturated compared with trans-12-18:1 in humans. Br. J. Nutr., 95, 752–761.
Lacasse P, Kennelly JJ, Delbecchi L and Ahnadi CE (2002) Addition of protected and
unprotected fish oil to diets for dairy cows. I. Effects on the yield, composition and taste
of milk. J. Dairy Res., 69, 511–520.
Lavillonniere F, Chajes V, Martin JC, Sebedio JL, Lhuillery C and Bougnoux P (2003)
Dietary purified cis-9,trans-11 conjugated linoleic acid isomer has anticarcinogenic
properties in chemically induced mammary tumors in rats. Nutr. Cancer, 45, 190–194.
Lawless F, Stanton C, L’Escop P, Devery R, Dillon P and Murphy JJ (1999) Influence of
breed on bovine milk cis-9,trans-11-conjugated linoleic acid content. Livest. Prod. Sci.,
62, 43–49.
Ledoux M, Laloux L, Fontaine JJ, Carpentier YA, Chardigny JM and Sebedio JL (2007)
Rumenic acid significantly reduces plasma levels of LDL and small dense LDL
cholesterol in hamsters fed a cholesterol- and lipid-enriched semi-purified diet. Lipids,
42, 135–141.
Lee SH, Yamaguchi K, Kim JS, Eling TE, Safe S, Park Y and Baek SJ (2006) Conjugated
linoleic acid stimulates an anti-tumorigenic protein NAG-1 in an isomer specific manner.
Carcinogenesis, 27, 972–981.
Lengi AJ and Corl BA (2008) Comparison of pig, sheep and chicken SCD5 homologs:
evidence for an early gene duplication event. Comp. Biochem. Physiol. B. Biochem. Mol.
Biol., 150, 440–446.
Libby P (2002) Inflammation in atherosclerosis. Nature, 420, 868–874.
Lock AL and Bauman DE (2004) Modifying milk fat composition of dairy cows to enhance
fatty acids beneficial to human health. Lipids, 39, 1197–1206.
Lock AL, Bauman DE and Garnsworthy PC (2005a) Effect of production variables on the cis-
9,trans-11 conjugated linoleic acid content of cows’ milk. J. Dairy Sci., 88, 2714–2717.
Lock AL, Corl BA, Barbano DM, Bauman DE and Ip C (2004) The anticarcinogenic effect
of trans-11 18:1 is dependent on its conversion to cis-9,trans-11 CLA by Δ9-desaturase
in rats. J. Nutr., 134, 2698–2704.
Lock AL and Garnsworthy PC (2002) Independent effects of dietary linoleic and linolenic
fatty acids on the conjugated linoleic acid content of cows’ milk. Anim. Sci., 74, 163–176.
Lock AL and Garnsworthy PC (2003) Seasonal variation in milk conjugated linoleic acid and
delta(9)-desaturase activity in dairy cows. Livest. Prod. Sci., 79, 47–59.
Lock AL, Horne CAM, Bauman DE and Salter AM (2005b) Butter naturally enriched in
conjugated linoleic acid and vaccenic acid alters tissue fatty acids and improves the
plasma lipoprotein profile in cholesterol-fed hamsters. J. Nutr., 135, 1934–1939.
Lusis AJ (2000) Atherosclerosis. Nature, 407, 233–241.
Lynch JM, Lock AL, Dwyer DA, Norbaksh R, Barbano DM and Bauman DE (2005) Flavor
and stability of pasteurized milk with elevated levels of conjugated linoleic acid and
vaccenic acid. J. Dairy Sci., 88, 489–498.
Mahfouz MM, Valicenti AJ and Holman RT (1980) Desaturation of isomeric trans-
octadecenoic acids by rat liver microsomes. Biochim. Biophys. Acta, 618, 1–12.
McCann SE, Ip C, Ip MM, McGuire MK, Muti P, Edge SB, Trevisan M and Freudenheim JL
(2004) Dietary intake of conjugated linoleic acids and risk of premenopausal and
postmenopausal breast cancer, Western New York Exposures and Breast Cancer Study
(WEB study). Cancer Epidemiol. Biomarkers Prev., 13, 1480–1484.
Mensink RP, Zock PL, Kester ADM and Katan MB (2003) Effects of dietary fatty acids and
carbohydrates on the ratio of serum total to HDL cholesterol and on serum lipids and
apolipoproteins: a meta-analysis of 60 controlled trials. Am. J.Clin. Nutr., 77, 1146–1155.
Miller A, McGrath E, Stanton C and Devery R (2003a) Vaccenic acid (t11-18:1) is converted
to c9,t11-CLA in MCF-7 and SW480 cancer cells. Lipids, 38, 623–632.
Miller A, Stanton C, Murphy J and Devery R (2003b) Conjugated linoleic acid (CLA)-
enriched milk fat inhibits growth and modulates CLA-responsive biomarkers in MCF-7
and SW480 human cancer cell lines. Br. J. Nutr., 90, 877–885.
Miller A, Stanton C and Devery R (2002) Cis 9,trans 11- and trans 10,cis 12-conjugated
linoleic acid isomers induce apoptosis in cultured SW480 cells. Anticancer Res., 22,
3879–3887.
Mitchell PL, Langille MA, Currie DL and McLeod RS (2005) Effect of conjugated linoleic
acid isomers on lipoproteins and atherosclerosis in the Syrian Golden hamster. Biochim.
Biophys. Acta, 1734, 269–276.
Moore T (1939) Spectroscopic changes in fatty acids VI: General. Biochem. J., 33, 1635–
1638.
Mosley EE, Shafii B, Moate PJ and McGuire MA (2006) Cis-9,trans-11 conjugated linoleic
acid is synthesized directly from vaccenic acid in lactating dairy cattle. J. Nutr., 136, 570–
575.
National Research Council (1996) Carcinogens and anticarcinogens in the human diet.
National Academy Press, Washington, DC, USA.
Nestel P, Fujii A and Allen T (2006) The cis-9,trans-11 isomer of conjugated linoleic acid
(CLA) lowers plasma triglyceride and raises HDL cholesterol concentrations but does not
suppress aortic atherosclerosis in diabetic apoE-deficient mice. Atherosclerosis, 189,
282–287.
Noone EJ, Roche HM, Nugent AP and Gibney MJ (2002) The effect of dietary supplemen-
tation using isomeric blends of conjugated linoleic acid on lipid metabolism in healthy
human subjects. Br. J. Nutr., 88, 243–251.
Ntambi JM (1999) Regulation of stearoyl-CoA desaturase by polyunsaturated fatty acids and
cholesterol. J. Lipid Res., 40, 1549–1558.
Ntambi JM and Miyazaki M (2004) Regulation of stearoyl-CoA desaturases and role in
metabolism. Prog. Lipid Res., 43, 91–104.
Nugent AP, Roche HM, Noone EJ, Long A, Kelleher DK and Gibney MJ (2005) The effects
of conjugated linoleic acid supplementation on immune function in healthy volunteers.
Eur. J. Clin. Nutr., 59, 742–750.
Ochoa JJ, Farquharson AJ, Grant I, Moffat LE, Heys SD and Wahle KW (2004) Conjugated
linoleic acids (CLAs) decrease prostate cancer cell proliferation: different molecular
mechanisms for cis-9,trans-11 and trans-10,cis-12 isomers. Carcinogenesis, 25, 1185–
1191.
O’Shea M, Devery R, Lawless F, Murphy J and Stanton C (2000) Milk fat conjugated linoleic
acid (CLA) inhibits growth of human mammary MCF-7 cancer cells. Anticancer. Res.,
20, 3591–3601.
Offer NW, Marsden M, Dixon J, Speake BK and Thacker FE (1999) Effect of dietary fat
supplements on levels of n–3 poly-unsaturated fatty acids, trans acids and conjugated
linoleic acid in bovine milk. Anim. Sci., 69, 613–625.
Ozols J (1997) Degradation of hepatic stearyl CoA delta9-desaturase. Mol. Biol. Cell, 8,
2281–2290.
Palmquist DL, Lock AL, Shingfield KJ and Bauman DE (2005) Biosynthesis of conjugated
linoleic acid in ruminants and humans. In: Advances in Food and Nutrition Research (SL
Taylor, ed.) Elsevier Inc., San Diego, California, USA, pp.179–217.
Palombo JD, Ganguly A, Bistrian BR and Menard MP (2002) The antiproliferative effects
of biologically active isomers of conjugated linoleic acid on human colorectal and
prostatic cancer cells. Cancer Lett., 177, 163–172.
Pariza MW, Ashoor SH, Chu FS and Lund DB (1979) Effects of temperature and time on
mutagen formation in pan-fried hamburger. Cancer Lett., 7, 63–69.
Pariza MW (1999) The biological activities of conjugated linoleic acid. In: Advances in Con-
jugated Linoleic Acid Research, Volume 1 (MP Yurawecz, MM Mossoba, JKG Kramer,
MW Pariza and G Nelson, eds) AOCS Press, Champaign, Illinois, USA, pp.12–20.
Park HS, Chun CS, Kim S, Ha YL and Park JH (2006) Dietary trans-10,cis-12 and cis-9,trans-
11 conjugated linoleic acids induce apoptosis in the colonic mucosa of rats treated with
1,2-dimethylhydrazine. J. Med. Food, 9, 22–27.
Parodi PW (1977) Conjugated octadecienoic acids of milk fat. J. Dairy Sci., 60, 1550–1553.
Parodi PW (1994) Conjugated linoleic acid — an anticarcinogenic fatty acid present in milk
fat. Aust. J. Dairy Tech., 49, 93–97.
Parodi PW (2003) Conjugated linoleic acid in food. In: Advances in Conjugated Linoleic Acid
Champaign, Ilinois, USA, pp.101–122.
Parodi PW (2004) Milk fat in human nutrition. Aust. J. Dairy Tech., 59, 3–59.
Parodi PW (2006) Nutritional significance of milk lipids. In: Advanced Dairy Chemistry,
Volume 2: Lipids, 3rd Edition (PF Fox and PLH McSweeney, eds) Kluwer Academic/
Plenum Publishers, New York, USA, pp.601–639.
Peterson DG, Kelsey JA and Bauman DE (2002) Analysis of variation in cis-9,trans-11
conjugated linoleic acid (CLA) in milk fat of dairy cows. J. Dairy Sci., 85, 2164–2172.
Piperova LS, Sampugna J, Teter BB, Kalscheur KF, Yurawecz MP, Ku Y, Morehouse KM
and Erdman RA (2002) Duodenal and milk trans octadecenoic acid and conjugated
linoleic acid (CLA) isomers indicate that postabsorptive synthesis is the predominant
source of cis-9-containing CLA in lactating dairy cows. J. Nutr., 132, 1235–1241.
Pollard MR, Gunstone FD, James AT and Morris LJ (1980) Desaturation of positional and
geometric isomers of monoenoic fatty acids by microsomal preparations from rat liver.
Lipids, 15, 306–314.
Rajakangas J, Basu S, Salminen I and Mutanen M (2003) Adenoma growth stimulation by
the trans-10,cis-12 isomer of conjugated linoleic acid (CLA) is associated with changes
in mucosal NF-kappaB and cyclin D1 protein levels in the Min mouse. J. Nutr., 133,
1943–1948.
Ramaswamy N, Baer RJ, Schingoethe DJ, Hippen AR, Kasperson KM and Whitlock LA
(2001a) Composition and flavor of milk and butter from cows fed fish oil, extruded
soybeans, or their combination. J. Dairy Sci., 84, 2144–2151.
Ramaswamy N, Baer RJ, Schingoethe DJ, Hippen AR, Kasperson KM and Whitlock LA
(2001b) Short communication. Consumer evaluation of milk high in conjugated linoleic
acid. J. Dairy Sci., 84, 1607–1609.
Ritzenthaler KL, McGuire MK, Falen R, Shultz TD, Dasgupta N and McGuire MA (2001)
Estimation of conjugated linoleic acid intake by written dietary assessment methodolo-
gies underestimates actual intake evaluated by food duplicate methodology. J. Nutr., 131,
1548–1554.
Roy A, Chardigny JM, Bauchart D, Ferlay A, Lorenz S, Durand D, Gruffat D, Faulconnier
Y, Sebedio JL and Chilliard Y (2007) Butters rich either in trans-10-C18:1 or in trans-
11-C18:1 plus cis-9,trans-11 CLA differentially affect plasma lipids and aortic fatty
streak in experimental atherosclerosis in rabbits. Animal, 1, 467–476.
Santora JE, Palmquist DL and Roehrig KL (2000) Trans-vaccenic acid is desaturated to
conjugated linoleic acid in mice. J. Nutr., 130, 208–215.
Scimeca JA (1999) Cancer inhibition in animals. In: Advances in Conjugated Linoleic Acid
Research, Volume I (MP Yurawecz, MM Mossoba, JKG Kramer, MW Pariza and G
Nelson, eds) AOCS Press, Champaign, Illinois, USA, pp.420–443.
Shantha NC, Ram LN, O’Leary J, Hicks CL and Decker EA (1995) Conjugated linoleic acid
concentrations in dairy products as affected by processing and storage. J. Food Sci., 60,
695–697.
Shingfield KJ, Ahvenjarvi S, Toivonen V, Arola A, Nurmela KVV, Huhtanen P and Griinari
JM (2003) Effect of dietary fish oil on biohydrogenation of fatty acids and milk fatty acid
content in cows. Anim. Sci., 77, 165–179.
Shingfield KJ, Ahvenjarvi S, Toivonen V, Vanhatalo A and Huhtanen P (2007) Transfer of
absorbed cis-9,trans-11 conjugated linoleic acid into milk is biologically more efficient
than endogenous synthesis from absorbed vaccenic acid in lactating cows. J. Nutr., 137,
1154–1160.
Shingfield KJ and Griinari JM (2007) Role of biohydrogenation intermediates in milk fat
depression. Eur. J. Lipid Sci. Tech., 109, 799–816.
Stanton C, Lawless F, Kjellmer G, Harrington D, Devery R, Connolly JF and Murphy J (1997)
Dietary influences on bovine milk cis-9,trans-11-conjugated linoleic acid content. J.
Food Sci., 62, 1083–1086.
Stanton C, Murphy J, McGrath E and Devery R (2003) Animal feeding strategies for
conjugated linoleic acid enrichment of milk. In: Advances in Conjugated Linoleic Acid
Champaign, Illinois, USA, pp.123–145.
Steck SE, Chalecki AM, Miller P, Conway J, Austin GL, Hardin JW, Albright CD and
Thuillier P (2007) Conjugated linoleic acid supplementation for twelve weeks increases
lean body mass in obese humans. J. Nutr., 137, 1188–1193.
Tanmahasamut P, Liu J, Hendry LB and Sidell N (2004) Conjugated linoleic acid blocks
estrogen signaling in human breast cancer cells. J. Nutr., 134, 674–680.
Tricon S, Burdge GC, Jones EL, Russell JJ, El-Khazen S, Moretti E, Hall WL, Gerry AB,
Leake DS, Grimble RF, Williams CM, Calder PC and Yaqoob P (2006) Effects of dairy
products naturally enriched with cis-9,trans-11 conjugated linoleic acid on the blood lipid
profile in healthy middle-aged men. Am. J. Clin. Nutr., 83 744–753.
Tricon S, Burdge GC, Kew S, Banerjee T, Russell JJ, Jones EL, Grimble RF, Williams CM,
Yaqoob P and Calder PC (2004) Opposing effects of cis-9,trans-11 and trans-10,cis-12
conjugated linoleic acid on blood lipids in healthy humans. Am. J. Clin. Nutr., 80, 614–
620.
Turpeinen AM, Mutanen M, Aro A, Salminen I, Basu S, Palmquist DL and Griinari JM (2002)
Bioconversion of vaccenic acid to conjugated linoleic acid in humans. Am. J. Clin. Nutr.,
76, 504–510.
Valeille K, Férézou J, Amsler G, Quignard-Boulang‚ A, Parquet M, Gripois D, Dorovska-
Taran V and Martin JC (2005) A cis-9,trans-11 conjugated linoleic acid-rich oil reduces
the outcome of atherogenic process in hyperlipidemic hamster. Am. J. Physiol. Heart
Circ. Physiol., 289, H652–H659.
Valeille K, Ferezou J, Parquet M, Amsler G, Gripois D, Quignard-Boulange A and Martin
JC (2006) The natural concentration of the conjugated linoleic acid, cis-9,trans-11 in milk
fat has antiatherogenic effects in hyperlipidemic hamsters. J. Nutr., 136, 1305–1310.
Voorrips LE, Brants HAM, Kardinaal AFM, Hiddink GJ, van den Brandt PA and Goldbohm
RA (2002) Intake of conjugated linoleic acid, fat, and other fatty acids in relation to
postmenopausal breast cancer: the Netherlands cohort study on diet and cancer. Am. J.
Clin. Nutr., 76, 873–882.
Ward RJ, Travers MT, Richards SE, Vernon RG, Salter AM, Buttery PJ and Barber MC
(1998) Stearoyl-CoA desaturase mRNA is transcribed from a single gene in the ovine
genome. Biochim. Biophys. Acta, 1391, 145–156.
Werner SA, Luedecke LO and Shultz TD (1992) Determination of conjugated linoleic acid
content and isomer distribution in three cheddar-type cheeses: effects of cheese cultures,
processing and aging. J. Agric. Food Chem., 40, 1817–1821.
Whitlock LA, Schingoethe DJ, Hippen AR, Kalscheur KF, Baer RJ, Ramaswamy N and
Kasperson KM (2002) Fish oil and extruded soybeans fed in combination increase
conjugated linoleic acids in milk of dairy cows more than when fed separately. J. Dairy
Sci., 85, 234–243.
Wilson TA, Nicolosi RJ, Chrysam M and Kritchevsky D (2000) Conjugated linoleic acid
reduces early aortic atherosclerosis greater than linoleic acid in hypercholesterolemic
hamsters. Nutr. Res., 20, 1795–1805.
Yamasaki M, Chujo H, Koga Y, Oishi A, Rikimaru T, Shimada M, Sugimachi K, Tachibana
H and Yamada K (2002) Potent cytotoxic effect of the trans10,cis12 isomer of conjugated
linoleic acid on rat hepatoma dRLh-84 cells. Cancer Lett., 188, 171–180.
Yurawecz MP, Roach JAG, Sehat N, Mossoba MM, Kramer JKG, Fritsche J, Steinhart H and
Ku Y (1998) A new conjugated linoleic acid isomer, 7 trans,9 cis-octadecadienoic acid,
in cow milk, cheese, beef and human milk and adipose tissue. Lipids, 33, 803–809.
CHAPTER 9
Biosynthesis, synthesis and biological activity
of trans-10,cis-12 conjugated linoleic acid
(CLA) isomer
DELPHINE TISSOT-FAVRE1 AND MARK WALDRON2
1
Nestlé Purina PetCare, St Louis, Missouri, USA
2
A. Introduction
Conjugated linoleic acids (CLA) consist of several positional and geometric
isomers. Although there are 28 known isomers, CLA is not one molecule, but
rather is a mixture of primarily cis-9,trans-11 18:2 and trans-10,cis-12 18:2
CLA isomers. In fact it is these two isomers that are responsible for all known
biological effects of CLA. The cis-9,trans-11 isomer is the most abundant form
found in food, the trans-10,cis-12 isomer being the second most abundant form
(Fritsche, 1999). Until recently, investigations of the effects of CLA on bio-
logical processes used a mixture of these two isomers and thus it was impossible
to ascertain which of the primary isomers was eliciting the biological effect.
Numerous biological effects and health benefits have been attributed to CLA
including weight loss, improved body composition, enhanced immune func-
tion, anti-carcinogenic and anti-atherosclerotic effects, improved glucose control
and insulin sensitivity, and others. There are several outstanding reviews of the
literature reviewing the health benefits of CLA (Bhattacharya et al. 2006;
Wahle et al., 2004; Wang & Jones, 2004; Pariza et al.; and Sébédio, Christie &
Adlof, 2003). However, nearly all reviews of CLA include reports in which
both isomers were fed or tested in vitro. Recent advances in technology have
led to the availability of the purified isomers and thus a great deal of studies
have been designed and conducted investigating the biological effects of the
individual isomers. However, until recently scientists were unable to assess
which isomer was eliciting these reported biological effects. There is a growing
body of literature indicating that the trans-10,cis-12 isomer may be primarily
responsible for several of these observed effects. A complete review of CLA
and its biological effects are beyond the scope of this chapter. Rather this
231
chapter will focus on the biosynthesis, synthesis and biological effects of the
trans-10,cis-12 CLA isomer and focus on the newer and more relevant findings
with respect to the biological activity of the trans-10,cis-12 CLA isomer.
B. Biosynthesis
1. Biosynthesis in mammals
The investigations on CLA isomers showed that two main isomers, cis-9,trans-
11 and trans-10,cis-12 isomers occur naturally in foods derived from ruminants
(Chin et al., 1992; Kraft et al., 2003; Kramer et al., 1998; Kramer et al., 2004;
Parodi, 1977; Parodi, 1999). Two routes for CLA synthesis in ruminants were
described and occur in two different parts of the ruminant animal (Bauman
et al., 1999; Griinari & Bauman, 1999). The first route is a ruminal biohydro-
genation of dietary linoleic acid performed by the ruminal bacterial population.
The second route, occurring in the animal’s tissue, is the desaturation of
vaccenic acid by endogenous Δ9-desaturase (Griinari & Bauman, 1999; Mosley
et al., 2006b), which is also active in producing cis-9,trans-11 isomer in
mammary glands of other mammals such as human and rat (Lock et al., 2004;
Mosley et al., 2006a).
In humans, no conversion of trans-10 18:1 into trans-10,cis-12 18:2 CLA by
the endogenous Δ12-desaturase was detected (Adlof et al., 2000). The predomi-
nant isomer in ruminant and milk fat is cis-9,trans-11 18:2 representing 80 to
90% of the total milk CLA (Chin, et al., 1992; Parodi, 1977) and more than
90% of total CLA in the subcutaneous and intra-muscular fat of steers (Fritsche
& Fritsche, 1998). However, the content of trans-10,cis-12 CLA isomer can be
increased in lean beef or milk fat by specific dietary conditions (Griinari &
Bauman, 1999; Peterson et al., 2003, Xu et al., 2006). Some studies have
demonstrated that the rumen fermentation pathway was altered by feeding a
low-fibre diet resulting in an increase in trans-10 octadecenoic acid content in
the milk fat. Griinari and Bauman (1999) proposed that this new route,
alternative to the biohydrogenation process and induced by the diet, would
involve a specific bacterial cis-9,trans-10 isomerase that would form a trans-
10,cis-12 conjugated double bond structure as a first reaction.
Sieber and co-workers (2004) showed that the trans-10,cis-12 CLA isomer
is biosynthesized by some bacteria in the rumen only from linoleic acid.
Furthermore, protecting unsaturated fatty acids from bacterial hydrogenation
by thermal treatment of the diet increased the production of trans-10,cis-12
isomer in lean beef and the level of this CLA isomer in the intra-muscular fat
(Xu et al., 2006). In addition, in vitro studies showed that acidic pH conditions
(pH from 6 to 7) as well as aerobic conditions influence drastically the
production of trans-10,cis-12 CLA isomer by ruminal bacteria (Choi et al.,
2007; Kawahara et al., 2007).
BIOSYNTHESIS, SYNTHESIS AND BIOACTIVITY OF TRANS-10,CIS-12 CLA 233
Table 1. CLA production by microorganisms.*
Microorganisms producing CLA Cis-9,trans-11 Trans-10,cis-12 Productivity

(mg/l)
Lactobacillus acidophilus 67–85 10 131–4900

delbrueckii ssp.
lactis
delbrueckii ssp.
bulgaricus
brevis 550
casei 85 12 111
paracasei ssp. 70–200
paracasei
pentosus 80–130
plantarum 21–38 2700–40000
fermentum
rhamnosus 64 36 1410–4400
reuteri 59 300
Lactococcus casei
lactis ssp. cremoris
lactis ssp. lactis
lactis
Streptococcus thermophilus
Bifidobacterium adolescentis 46 34 3.5
angulatum 50 50 1.2
bifidum 100 1.0
breve 91 398
dentium 78 1 160
infantis 74 7 24.6
lactis 72 2 170
pseudocatenulatum 72 9 23.3
Propionibacterium shermanii 110
freudenreichii ssp. 93 265
freudenreichii
freudenreichii ssp. 85–95 490–1600
shermanii
acidipropionici
technicum
acnes 85
Pediococcus acidilactici 1400
Enterococcus faecium 100
Butyrivibrio fibrisolvens 95 220
Megasphaera elsdenii 15 85 7 µg/mg
protein
Delacroixia coronata 98 10.5 mg/ml
in 7 days
* Adapted from Alonso et al., 2003; Ando et al., 2003; Ando et al., 2004; Coakley et al., 2003; Jiang et
al., 1998; Kim et al., 2002; Kim, 2003; Kishino et al. 2002; Lee et al., 2006; Lee et al., 2003; Ogawa et
al., 2001; Ogawa et al., 2005 and Sieber et al., 2004. Results obtained using linoleic acid, ricinoleic acid
or castor oil as substrates.
2. Production by bacteria, fungi and yeasts
Apart from the ruminal bacteria, an extensive literature describes different

families of bacteria able to produce CLA isomers such as Lactobacilli,
Lactococci, Bifidobacteria and Propionibacteria (Table 1) (Alonso et al., 2003;
Ando et al., 2003; Ando et al., 2004; Coakley et al., 2003; Jiang et al., 1998;
Kim et al., 2002; Kim, 2003; Kishino et al., 2002; Lee et al., 2003; Ogawa
et al., 2001; Ogawa et al., 2005; Sieber et al., 2004). Depending on the
conditions of growth; the use of washed or immobilized cells; specific substrates
such as linoleic or ricinoleic acids; lactate or fructo-oligosaccharides and the
type of reaction, the level of CLA production can be increased (Ando et al.,
2003; Ando et al., 2004; Coakley et al., 2003; Fukuda et al., 2005; Lee et al.,
2003; Lee et al., 2003; Oh et al., 2003; Vahvaselka et al., 2004).
However, the CLA isomer predominantly produced by bacteria is cis-
9,trans-11, except for a few species which were shown to synthesize
trans-10,cis-12 from linoleic acid. Among those species, Propionibacterium
acnes and Megasphaera elsdenii were identified as good producers of trans-
10,cis-12 CLA (Kim et al., 2002; Liavonchanka et al., 2006; Wallace et al.,
2007). Some Bifidobacteria, isolated from human intestine, were reported to
produce trace amount of trans-10,cis-12 CLA (Coakley et al., 2003; Devillard
et al., 2007) as well as Lactobacillus acidophilus and Lactobacillus casei from
human origin (Alonso et al., 2003). Recently, a strain of Lactobacillus
plantarum, isolated from baby’s faeces and significantly producing trans-
10,cis-12 CLA, was shown to cause a specific reduction of white adipose tissue
in diet-induced obese mice (Lee et al., 2006). The study was an 8-week oral
treatment on mice fed with 1 × 107 or 1 × 109 cfu of washed CLA-producing
Lactobacillus plantarum. The mice were colonized with Lactobacillus
plantarum and up to 1.906 µg/ml of trans-10,cis-12 CLA was detected in the
sera after 3 weeks. The treatment induced a decrease of body weight gain and
of liver steatosis. However, no relationship was seen between a high-dose and
low-dose of CLA-producing Lactobacillus plantarum (Lee et al., 2006).
Finally, few studies showed production of CLA by some other microorgan-
isms. For instance, the filamentous fungus Delacroixia coronata transformed
vaccenic (trans-11 18:2) acid into 98% of cis-9,trans-11 CLA isomer (Ogawa
et al., 2004)
3. Production of trans-10,cis-12 CLA isomer by genetically-modified

organisms and microorganisms
As previously described, Propionibacterium acnes isolated from mouse cae-
cum or human skin (Verhulst 1987) and Megasphaera elsdenii from rumen
origin (Kim et al., 2000; Kim et al., 2002) are the main known bacterial species
that can isomerize linoleic acid to produce trans-10,cis-12 CLA isomer in
significant amounts. For instance, (Verhulst, 1987) showed that Propionibac-

terium acnes is able to convert approximately 85% of linoleic acid in
trans-10,cis-12 CLA isomer in 24 h via the isomerization of the cis-9 double
bond of linoleic acid into trans-10, implying the involvement of flavin adenine
dinucleotide (FAD) cofactor (Liavonchanka et al., 2006, Verhulst, 1987).
In recent times, the gene encoding the linoleate isomerase, that converts
linoleic acid into trans-10,cis-12 CLA, has been cloned from P. acnes (Rosson
et al., 2001). Since then, this P. acnes linoleic acid isomerase (PAI) enzyme
has been successfully cloned and over-expressed in bacteria, yeast and transgenic
plants (Hornung et al., 2005; Kohno-Murase et al., 2006; Rosberg-Cody et al.,
2007). Thus, Hornung and co-workers (2005) transformed Saccharomyces
cerevisiae and tobacco plants with the PAI sequence, altered for optimal codon
usage, and showed production of trans-10,cis-12 CLA in yeast (up to 5.7% of
total free fatty acids) and tobacco seeds (up to 0.3% in esterified acids, and to
15% in free fatty acids). Furthermore, using seed-specific promoters from
oleosin and globulin genes, Kohno-Murase et al. (2006) constructed a transgenic
rice plant containing PAI and showed its ability to produce CLA trans-10,cis-
12. Finally, PAI was cloned and over-expressed in Escherichia coli and
Lactococcus lactis resulting in the conversion of 30 to 50% of linoleic acid to
trans-10,cis-12 CLA (Rosberg-Cody et al., 2007). In this latter study, the
authors showed that the fatty acids produced partially keep their anti-prolifera-
tive activity which is useful for cancer prevention (Rosberg-Cody et al., 2007).
C. Chemical synthesis of trans-10,cis-12 CLA isomer

Trans-10,cis-12 CLA isomer can be found in many foods; however, principle
dietary sources are dairy products, such as milk and processed cheese, and
other foods derived from ruminants. In contrast to cis-9,trans-11 CLA, which
represents up to 80% of the CLA isomers naturally detected in food products,
the amounts of trans-10,cis-12 CLA are much lower than those shown to have
a significant biological effect. Different methods have been used to produce
CLA isomers mixtures or pure isomers.
1. Alkaline isomerization of linoleic acid

The commercial process used to produce CLA has been performed tradition-
ally by alkaline isomerization of linoleic acid. This reaction, known as diene
conjugation since 1951 (Nichols et al., 1951), is carried out under strong alkali
conditions and synthesizes predominantly a mixture of the cis-9,trans-11 (43–
45%) and trans-10,cis-12 (43–45%) isomers (Chipault & Hawkins, 1959).
These two major isomers are accompanied by other CLA isomers (with double
bonds in 8,10 or 11,13 positions) as well as geometrical isomers (Christie et al.,
1997). Depending on the severity of alkaline isomerization conditions such as
reaction time, the CLA isomer composition varies considerably (Ackman,

1998). Thus, more than 99% of cis-9,trans-11 and trans-10,cis-12 isomers can
be produced in synthetic conditions by limiting to 70% the conversion level of
linoleic acid into CLA (Reaney et al., 1999). Natural vegetable oils containing
linoleic acid (e.g. sunflower or safflower oils) cannot be used directly as
substrate but have to be converted into free fatty acids at high temperatures.
Different isomer compositions can also be obtained using diverse solvents
(such as ethylene glycol, glycerol, propylene glycol, tert-butanol, water,
dimethyl sulphoxide, dimethylformamide), catalysts (hydroxide of lithium,
sodium or potassium) and reaction vessels. Nevertheless, to avoid undesired
toxic residues when the final product is intended for human food applications,
propylene glycol, glycerol or ethanol/water are preferably utilized. For cost
reasons, sodium hydroxide is the favoured catalyst (Reaney et al., 1999).
To enrich in beneficial isomers, some alternative methods were developed.
For instance, alkali-isomerization of methyl linoleate, followed by a fractional
crystallization in acetone resulted in the formation of the two major CLA
isomers cis-9,trans-11 and trans-10,cis-12 with an isomeric purity of 90–97%
(Berdeaux et al., 1998).
2. Photocatalytic synthesis of CLA isomers

More recently, a ‘greener’ technology was developed to produce a mixture of
CLA isomers by ultra-violet (UV) photo-isomerization of linoleic acid acid
containing oil at 22ºC and using only an iodine catalyst (Jain & Proctor, 2006,
2007a). After 144 h of irradiation, a total CLA yield of 24% (w/w) total oil was
obtained using 0.15% (w/w) iodine. However, approximately 75% of the total
CLA content was found to be all trans isomers (mainly trans-8,trans-10, trans-
9,trans-11, and trans-10,trans-12 isomers). Kinetic study of the isomers showed
that, at the beginning of the reaction, in the presence of high content of linoleic
acid, there is a production of the cis,trans and trans,cis isomers which slows
down with the disappearance of linoleic acid (Jain & Proctor, 2007a). In the
latest study, photo-isomerized soybean oil was used to fry potato chips after
removal of the residual iodine. A 1 ounce serving of those chips was shown to
contain approximately 2.4 g of CLA (Jain & Proctor 2007b). Unfortunately,
the chips still contained predominantly all-trans isomers (68% of the total CLA
content) and only 0.2 g of cis-9,trans-11 and trans-10,cis-12 (8.4%) CLA
isomers.
5. Purification of CLA isomers

CLA solutions, available commercially, are mixtures of cis-9,trans-11 and
trans-10,cis-12 isomers usually produced by alkaline isomerization. As each
isomer was reported to have different physiological activity, fractionation of
the active isomers is required. The purification of these isomers was performed
successfully using an enzymatic method with lipases from either Candida
rugosa or Geotrichum candidum (McNeill et al., 1999; Nagao et al., 2002,
2003). The process comprises lipase-catalysed selective esterification with
lauryl alcohol, molecular distillation, and urea adduct fractionation under
strictly controlled conditions in ethanol. As a result, purification of 1.0 kg of
the CLA mixture yielded pure cis-9,trans-11 (93.1%) and trans-10,cis-12
(95.3%) CLA isomers.
Recently, a novel and efficient method by crystallization in acetone of the
two CLA isomers in the presence of medium-chain fatty acid additives was
reported (Uehara et al., 2006). Crystals containing mainly trans-10,cis-12
isomer were obtained by addition of lauric and decanoic acids, whereas
octanoic acid yielded crystals containing predominantly cis-9,trans-11 isomer
(Uehara et al., 2006).
D. Biological activity of trans-10,cis-12 CLA isomer
1. Effects of trans-10,cis-12 CLA isomer on body composition and energy

metabolism
Some of the earliest reports of the nutritional benefits of CLA were its effect as
an anti-obesity agent, specifically promoting weight loss and improving body
composition. Indeed the effects of CLA as an anti-obesity agent are well
documented in animals (Park et al., 1997; West et al., 1998, 2000). However
most, if not all, of the early studies were conducted using a mixture of isomers
containing cis-9,trans-11 and trans-10,cis-12. The question remained as to
which isomer is responsible for the observed enhancement of body composi-
tion and weight loss? And can the observations in animal models be extrapolated
to humans? There are numerous reports of the ability of CLA to reduce body fat
and enhance lean body mass in several animal models (Park et al., 1997;
DeLaney and West, 2000; Gavino et al., 2000; Ostrowaska et al., 1999).
However, as indicated previously, most of these studies report the effects of
feeding or supplementing with both CLA isomers. Thus it remains unclear
which of the primary isomers is responsible for the weight loss and improved
body composition?
The availability of purified isomers has led to many studies designed and
conducted to elucidate which isomer, cis-9,trans-11 or trans-10,cis-12, pro-
motes weight loss and enhances body composition. Collectively, these studies
have demonstrated that it is the trans-10,cis-12 isomer that is primarily, if not
solely, responsible for promoting a reduction of fat mass and enhancing lean
body composition (Park et al., 1999; Pariza et al., 2001). Although, the support
for the trans-10,cis-12 isomer in reducing fat mass is overwhelming, there are
exceptions in the literature. For example, reports of the efficacy of the trans-
10,cis-12 isomer on body composition in hamsters are mixed. An often cited

study in support of the trans-10,cis-12 isomer as the one promoting enhanced
body composition did not test the isomer individually but rather tested the cis-
9,trans-11 isomer alone or in conjunction with the trans-10,cis-12 isomer
compared to linoleic acid (Gavino et al., 2000). The authors concluded that cis-
9,trans-11 isomer may need to act synergistically with other isomers to exert its
biological effects (Gavino et al., 2000). Simon et al. (2005) reported that
supplementation with 0.5% trans-10,cis-12 CLA resulted in a significant
reduction in adipose tissue mass in hamsters; they suggested that a reduction in
preadipocyte differentiation into mature adipocytes may account for the fat-
lowering effect. Conversely, Bissonauth et al. (2006) reported no effect of the
trans-10,cis-12 isomer on bodyweight gain in hamsters. In mice, the trans-
10,cis-12 isomer was reported to induce body composition changes (Park et al.,
1999). In another study, mice were fed a modified AIN (American Institute of
Nutrition) diet containing 0.5% trans-10,cis-12 CLA or soybean oil for 14 days
(LaRosa et al. 2006). Collectively the mice lost 85% of white adipose tissue,
95% of adipocyte lipid droplet volume and 47% of the number of adipocytes
(LaRosa et al., 2006). This recent study provides further evidence that it is the
trans-10,cis-12 CLA isomer that elicits the beneficial effect with weight loss
and body fat reduction. However, the profound effect of supplementing with
trans-10,cis-12 CLA isomer may be limited to rodents.
Few studies testing the isomers independently in humans have been re-
ported. Several clinical trials in humans have been conducted and they reported
either weight loss or enhancement in body composition (Blankson et al., 2000;
Steck et al., 2007). But in these cases as well as others reported, both isomers
were provided as part of the treatment and, thus, it is impossible to determine
which isomer is inducing the effect in humans. There is at best limited evidence
in human trials investigating the effects of individual isomers on weight loss
and body composition. However, Malpuech-Brugere et al. (2004) reported that
neither the cis-9,trans-11 or the trans-10,cis-12 isomer had an effect on body
composition in middle-aged overweight men and women. Indeed, it appears
that the effect of CLA on bodyweight and body composition may be limited to
other animals. But why is this the case? Several explanations are plausible and
include experimental design, dose, genetic predisposition, age, and others.
Nonetheless, given the breadth of evidence in animal trials it is clear that the
trans-10,cis-12 isomer of CLA is the causative agent producing weight loss
and enhancing body composition in rodent models.
2. Anti-obesity mechanisms of trans-10,cis-12 CLA isomer

The ability of CLA and specifically the trans-10,cis-12 isomer to modulate
obesity and weight gain in animals is well documented. However, the mecha-
nisms by which the trans-10,cis-12 isomer exerts this effect have not been fully
elucidated. Proposed mechanisms by which weight loss and enhanced body

composition are achieved include; increased energy expenditure, increased fat
oxidation, decreased adipocyte size, decreased energy intake, and changes in
enzymatic activity in pathways involving fatty acid metabolism and lipo-
genesis (Park et al., 1999; West et al., 1998; Azain et al., 2000; Poulos et al.,
2001). How does CLA elicit its anti-obesity effects? And which specific
mechanisms are involved?
Numerous investigators have focused on the mechanisms by which CLA,
and specifically the trans-10,cis-12 isomer, imparts its beneficial effects with
respect to decreasing obesity and reducing fat mass. Mechanistic studies have
revealed that the trans-10,cis-12 isomer contributes to fat reduction and weight
loss via several mechanisms. Evans et al. (2000) using pure isomers demon-
strated that it is the trans-10,cis-12 isomer and not the cis-9, trans-11 isomer
that contributes to fat reduction. Specifically, supplementation with the trans-
10,cis-12 isomer resulted in a reduction in adipose size which occurred via the
induction of apoptosis. Reports of other mechanisms exist as well. Martin and
co-workers (2000) demonstrated that the trans-10,cis-12 isomer results in
increased fat oxidation in male rats. Several groups have shown that the trans-
10,cis-12 isomer inhibits key lipogenic enzymes in rodent models (Choi et al.,
2000; Park and Pariza, 2001). In 3T3-L1 preadipocyte cells, the trans-10,cis-
12 CLA isomer significantly reduced metalloproteinase-2 secretion and
triacylglycerol content (Moon et al., 2007), thus demonstrating the role that
trans-10,cis-12 CLA plays in the regulation of obesity onset. Collectively,
these studies clearly demonstrate that the trans-10,cis-12 isomer contributes to
a reduction in fat mass, through enhanced energy expenditure, a reduction in
adipocyte size and formation, and by regulation of lipogenic enzymes.
A recent series of excellent publications has further demonstrated the effect
of the trans-10,cis-12 isomer in reducing fat mass at the level of regulating gene
expression. LaRosa et al. (2007) investigated the effect of the individual
isomers on uncoupling protein (UCP) expression in white and brown adipose
tissue. Three groups of animals (hamsters) were supplemented with one of
three different amounts of trans-10,cis-12 CLA isomer: a control group (no
CLA) and two groups with high and low trans-10,cis-12 CLA isomer diets (0.5
or 1 g/100g diet) respectively. Supplementation with the trans-10,cis-12 iso-
mer reduced adipose depot weights but did not have an overall effect on
bodyweight. Additionally, leptin mRNA expression was reduced. The authors
concluded that induction of uncoupling proteins (1 and 2) in white and brown
adipose tissue by trans-10,cis-12 CLA is responsible for the reduction in
adiposity often reported with CLA supplementation (LaRosa et al., 2006).
After further work, the same group observed and concluded that it is the trans-
10,cis-12 and not the cis-9,trans-11 isomer which is responsible for affecting
gene expression in preadipocytes. Using microarrray technology, the authors
demonstrated that the trans-10,cis-12 isomer effect is mediated by the inte-
grated stress response pathway in preadipocytes, and may be the cause of

proinflammatory cytokines observed in adipocytes treated with trans-10,cis-
12 CLA isomer (LaRosa et al., 2007). The authors concluded that it is the
proinflammatory cytokines in early stage adipocytes that may indeed cause the
loss of lipids in mature adipocytes (LaRosa et al., 2007).
3. Effects of trans-10,cis-12 CLA isomer on gene expression

An excellent study was recently published which reported the effect of trans-
10,cis-12 CLA isomer supplementation on white adipose tissue gene expression
using microarray analysis (LaRosa et al., 2006). The study revealed that
administration of the trans-10,cis-12 isomer profoundly affects many bio-
chemical pathways. In particular, the authors reported that feeding the
trans-10,cis-12 CLA isomer resulted in 2600 to 4200 transcripts changing by
more than twofold (LaRosa et al., 2006). Among those genes significantly
affected are genes involved in glucose and fatty acid import or biosynthesis
which were significantly reduced. Furthermore, consistent with earlier reports,
mRNA for uncoupling protein 1 (UCP1) and carnitine palmitoyltransferase 1
(CPT-1) were increased. The authors suggested that these increases may
contribute to increased thermogenesis and a decrease in adipocyte formation
(La Rosa et al., 2006). The authors also observed that key adipocyte regulatory
factors were significantly affected. Perhaps of greater interest was the profound
increase in the genes related to the inflammatory response. These include
marked increases in IL-6, IL-8, and TNF ligands (LaRosa et al., 2006). The
authors conclude that the increased macrophage density, the altered transport
of lipids, and increased energy expenditure may all contribute to the cumula-
tive weight loss that was observed in white adipose tissue (LaRosa et al., 2006).
The trans-10,cis-12 isomer has profound effects on gene expression. Several
recent publications have investigated specifically the role of either the trans-
10,cis-12 or the cis-9,trans-11 isomer and their effect on gene expression in
commonly used cell culture models. In particular, a recent study revealed that
the trans-10,cis-12 isomer and not the cis-9,trans-11 isomer has profound
effects on gene expression in Caco-2 cells (Murphy et al., 2007). The results
indicate that gene expression, at least in this cell model, is highly affected by
trans-10,cis-12 and not the cis-9,trans-11 isomer and in large measure these
gene changes appear to be correlated with the cell cycle, carcinogenesis, cell
proliferation, and DNA metabolism. Using microarray techniques, Caco-2
cells were incubated with either the trans-10,cis-12 and cis-9, trans-11 CLA
isomers or linoleic acid and assayed for changes in gene expression. In total,
expression of 918 genes was significantly affected by incubation with trans-
10,cis-12 isomer. The most profound changes were linked to the following
biological processes: lipid metabolism, cell cycle, DNA metabolism, and cell
proliferation. Specifically, the cell cycle pathway was significantly down
regulated by trans-10,cis-12. Additionally, genes involved in apoptosis were

unregulated. These findings are consistent with other similar reports on the role
of CLA and apoptosis (Kim et al., 2002; Cho et al., 2005). The authors suggest
that many of the observed changes in gene expression give credence to
previous observations that CLA positively affect vital intestinal processes,
specifically carcinogenesis and calcium transport (Murphy et al., 2007). These
recent studies using microarray technology further our understanding of the
mechanisms in which the trans-10,cis-12 CLA isomer elicit their biological
effects. Of particular interest is the observation that the trans-10,cis-12 CLA
isomer profoundly affects gene expression and that the cis-9,trans-11 isomer
was reported to have no effect.
4. Effect of trans-10,cis-12 CLA isomer in cancer

One of the prominent beneficial effects attributed to CLA is its role as an anti-
cancer agent. Investigation of CLA and the two primary isomers and their effect
as an anti-cancer agent come primarily, if not solely, from in vitro and animal
studies. The earliest reports of the beneficial effects of CLA and cancer were
reported by the Pariza group (Pariza, 1998, Pariza, et al., 2000, 2003). Subse-
quently other studies have also shown CLA to inhibit cancer formation in murine
models (Ha et al., 1987, 1990; Ip et al., 2002; Kim et al., 2005). However, until
recently few studies have been undertaken investigating the effect of the
individual isomers on cancer formation and progression. As with other observed
health benefits of CLA, the question raised is which isomer is responsible for
these effects? And by which mechanism are they achieved? Collectively, the
literature indicates that the trans-10,cis-12 isomer is responsible for anti-cancer
effects attributed to CLA supplementation. However, the literature is not
conclusive and indeed in some studies each isomer appears to be effective in
reducing tumorigenesis. Indeed one early report indicated that the cis-9,trans-11
isomer had the greater anti-proliferative effects on colorectal and prostate
cancer cells in vitro (Palombo et al. 2002). Ip and co-workers (2002) investi-
gated the effect of each isomer, cis-9,trans-11 or trans-10,cis-12, to inhibit
premalignant lesions and carcinomas in rats and reported that each isomer was
equally effective in reducing premalignant lesions and mammary carcinomas. In
mice supplemented with either isomer, the trans-10,cis-12 group had fewer
tumours, but there was no difference in tumour size between the groups (Chen
et al., 2003). Although the literature is mixed with respect to the effects of the
individual isomers on cancer proliferation (i.e. reports indicate that both isomers
are effective in reducing cancer cell proliferation) there is a growing body of
evidence indicating that it is the trans-10,cis-12 isomer that is more effective as
an anti-cancer agent (Miller et al., 2002). However, as many of these reports are
from chemically-induced tumours in non-human models great care must be
taken when extrapolating to human populations.
The literature is expanding rapidly with respect to the mechanisms by which

CLA, and specifically the trans-10,cis-12 isomer, elicits its anti-cancer effects.
The trans-10,cis-12 isomer has anti-proliferative effects in several cancer cell
lines. Recently Cho et al. (2005, 2006) demonstrated that the trans-10,cis-12
isomer decreases Erb83 expression and inhibits G1-S progression in HT-29
human colon cells. In HT-29 colon cancer cell lines, the trans-10,cis-12 isomer
and not the cis-9,trans-11 isomer arrested the cell cycle at the G1 phase (Lim-
do et al., 2005, Cho et al., 2006). Others have demonstrated that a reduction in
cellular proliferation in HT-29 colon cancer cells was mediated by inhibition of
insulin like growth factor-II expression (Cho et al., 2003; Kim et al., 2003).
Kemp and co-workers (2003) demonstrated that the trans-10,cis-12 isomer was
more effective that the cis-9,trans-11 isomer in inhibiting cell proliferation in
colon and breast cancer cell lines. The same group recently reported that the
trans-10,cis-12 isomer was more effective than linoleic acid and the cis-
9,trans-11 isomer in producing anti-proliferative effects in breast cancer cell
lines via the inhibition of COX-2 transcription (Degner et al., 2006). Fite and
co-workers (2007) reported that CLA, specifically the trans-10,cis-12 isomer,
had an additive effect in conjunction with docetexel in producing greater IC50
and IC70 in breast cancer cells. Another study also reports the efficacy of CLA
supplementation, specifically the trans-10,cis-12 isomer, in breast cancer.
Wang and co-workers (2006) found that both CLA isomers increased the
PTPgamma mRNA expression levels in primary cultured normal breast cells,
normal breast stromal cells, and breast cancer epithelial cells. They concluded
that dietary CLA might serve as a chemo-preventive agent in human breast
cancer via upregulation of PTPgamma, a reported tumour suppressor. The
trans-10,cis-12 isomer and not the cis-9,trans-11 isomer is also effective in
inhibiting Caco-2 colon cancer cells (Kim et al., 2002).
Thus, there is a growing body of evidence supporting the notion that the
trans-10,cis-12 isomer is responsible for the anti-cancer effects observed with
CLA supplementation. However, the literature is not conclusive. Ip and co-
workers (2007) reported that long-term supplementation with trans-10,cis-12
isomer increased mammary tumorigenesis in transgenic mice. Transgenic mice
over-expressing erbB2 in mammary epithelium were fed either a control diet or
one containing one of the two isomers at 0.5% of the diet from weaning. As the
authors indicate, unexpectedly, the trans-10,cis-12 CLA isomer increased
mammary epithelium hyperplasia, and increased tumour development. Addi-
tionally, lung metastasis was also significantly greater in the trans-10,cis-12
group. In a subsequent study, they showed that this observation was not limited
to erbB2 transgenic mice, but was also observed in wild type mice as well (Ip
et al., 2007). The authors conclude that it may be advisable to avoid consuming
trans-10,cis-12 CLA whereas the cis-9,trans-11 is safe. Subsequent studies
need to be conducted to confirm these observations.
The literature supports the hypothesis that it is the trans-10,cis-12 CLA
isomer that protects against cancer and has anti-cancer properties. However,
this is not conclusive and in some cases either isomer is effective or, as in one
report, long-term supplementation of the trans-10,cis-12 isomer actually in-
creased tumour formation. The mechanisms by which the trans-10,cis-12
isomer elicits these anti-cancer effects include inhibition of cellular prolifera-
tion, increased apoptosis, and changes in COX-2 expression as well as other
proposed mechanisms. However, given the recent report by Ip and co-workers
(2007), more coordinated long-term research is needed to ensure the safety and
efficacy of the trans-10,cis-12 isomer as an agent that protects against cancer.
5. Effect of trans-10,cis-12 CLA isomer on immune function

Enhanced immune function is among the biological effects associated with
CLA supplementation and there are many reports of the efficacy of CLA to
attenuate the immune response or function in animals. However, reports of the
effect of CLA on immune function in animals are mixed. In humans, CLA
supplementation has only shown modest or no effect on immune function.
There are even fewer studies investigating the effect of individual isomers on
immune parameters. Indeed each of the isomers may have independent effects.
Lymphocytes isolated from mice supplemented with the trans-10,cis-12 iso-
mer produced more immunoglobulin A (IgA) and and IgM but not IgG
compared with mice fed the cis-9,trans-11 isomer (Yamasaki et al., 2003).
Yamasaki and co-workers also showed that the cis-9,trans-11 isomer increased
CD8+ T cells indicating that each isomer may exert independent effects. Tricon
et al. (2004) reported that both isomers decreased lymphocyte activation, but
had no effect on lymphocyte subpopulations, ex vivo cytokine production or C-
reactive protein. Eder et al. (2003) reported that treatment of aortic endothelial
cells with either pure cis-9,trans-11 or trans-10,cis-12 CLA isomers resulted in
less eicosanoid and nitric oxide production. The authors concluded that this
was not favourable for endothelial function (Eder et al., 2003). Albers et al.
(2003) reported that CLA supplementation may enhance the immune system’s
ability to mount a response to new challenge. But as with most studies, the
supplementation was with either a 50:50 mix of the two isomers or 80:20 (80%
cis-9,trans-11: 20% trans-10,cis-12). Furthermore, the observed effect on
hepatitis B vaccination was observed in the 50:50 group and not in the 80:20
group. Additionally, other aspects of immune function were not affected, such
as DTH response, NK cell activity, production of TNF-α, IL-1β, IL-6, and
PGE2 (Albers et al., 2003). As with weight control and cancer, only recently
have studies been conducted elucidating which isomer exerted its biological
effects and what parameters are affected. It is clear that CLA supplementation
can and does affect both the adaptive and innate immune function and further-
more it appears that each isomer has independent effects. We will focus on the
effects attributed to the trans-10,cis-12 isomer.
The trans-10,cis-12 isomer modulates and enhances immune cell function

(Pariza, 2001). Kang and et al. (2007) reported that in pigs the addition of
trans-10,cis-12 isomer resulted in enhanced phagocytosis mediated by TNF-α
by activating PPAR-gamma. Although limited, there are reports of enhanced
immune function in humans via the action of trans-10,cis-12 CLA. Stachowska
et al. (2007) reported recently that the trans-10,cis-12 CLA improved phago-
cytosis in human monocytes/macrophages. Isolated human monocytes were
incubated with the trans-10,cis-12 CLA for 7 days and assayed for phagocyto-
sis as well as COX-2 expression and NF-κB activation. Both isomers increased
the percentage of cells participating in phagocytosis and both isomers had
decreased NF-κB activation compared to control. The trans-10,cis-12 isomer
had the greatest reduction of COX-2 expression and both isomers had signifi-
cantly greater PPAR gamma expression (Stachowska et al., 2007). The authors
concluded that the CLA isomers increased macrophage phagocytosis via
inhibition of PGE2 production. The decrease in PGE2 production was a result of
decreased COX-2 production. CLA supplementation reduces inflammation in
vitro and in vivo. Li et al. (2005) investigated the effects of the individual
isomers on COX-2 protein and mRNA expression as well as the cellular
mechanism by which these are affected. Murine macrophages, RAW cells,
were incubated at varying concentrations with either cis-9,trans-11 CLA or
trans-10,cis-12 CLA isomers. Treatment with the trans-10,cis-12 isomer re-
sulted in an 80% reduction of COX-2 expression. Accordingly, PGE2 production
was significantly reduced compared to the cis-9,trans-11 treated cells (Li et al.,
2005). In the same study, in vivo investigations were carried out. Similar
findings were observed for the mice fed the trans-10,cis-12 isomer but not for
the cis-9,trans-11 fed mice. The authors concluded that the decrease in COX-
2 mRNA expression was due to inhibition of the NF-κB pathway by the
trans-10,cis-12 isomer (Li et al., 2005).
The literature is insufficient and certainly inconclusive regarding which
isomer elicits which effect on immune function, whether adaptive or innate.
Clearly, further research needs to be conducted in this area to clarify the
mechanism by which CLA enhances immune function. As O’Shea et al. (2004)
concludes, such work may lead to the development of nutritional-based
approaches to addressing a myriad of immune-related disorders, i.e. protection
against infection and inflammatory conditions.
6. Effects of trans-10,cis-12 CLA isomer on cardiovascular disease

CLA supplementation is also reported to have anti-atherosclerosis benefits. In
large measure, these are studies in which a mixture of the isomers was fed or
used in in vitro experiments. Furthermore, it is clear that the anti-atherosclero-
sis effects of CLA can be attributed largely if not solely to the cis-9,trans-11
isomer. Few studies have investigated the effects of the trans-10,cis-12 isomer
on factors associated with cardiovascular disease. The trans-10,cis-12 CLA

isomer does not have anti-atherogenic properties. In a cross-over study, healthy
men were supplemented with a CLA-enriched preparation containing either the
cis-9,trans-11 or trans-10,cis-12 CLA isomer. The two isomers had opposing
effects, the cis-9,trans-11 isomer group having significantly improved blood
lipids (Tricon et al., 2004). It is clear from this study, as well as reviews cited
earlier, that the improvement in parameters associated with cardiovascular
health is not attributable to the trans-10,cis-12 CLA isomer and as such will not
be discussed further.
7. Effect of trans-10,cis-12 CLA isomer on glucose tolerance and insulin

sensitivity
The literature is mixed regarding the effect of the trans-10,cis-12 isomer on
insulin resistance/insulin sensitivity. In two studies the authors concluded that
the trans-10,cis-12 isomer improves insulin sensitivity. Insulin sensitivity was
improved in ZDF rats fed a mixture of the two isomers (Ryder et al., 2001).
Although, the trans-10,cis-12 isomer was not fed alone, the diet rich in the cis-
9,trans-11 (91%) isomer had no effect. The authors suggested that the
trans-10,cis-12 isomer may have beneficial effects with respect to insulin
resistance. Insulin-resistant obese Zucker rats supplemented with the trans-
10,cis-12 isomers had improved glucose tolerance and insulin response
(Henriksen et al., 2003). The rats were fed a mix of the two isomers (50:50), or
a diet enriched in either the cis-9,trans-11 (76%) or trans-10,cis-12 isomer
(90%). The trans-10,cis-12 isomer group had improved glucose and insulin
response. The authors concluded that the observed improvement in insulin
response is attributed to the trans-10,cis-12 isomer and that the cis-9,trans-11
isomer is metabolically neutral. In another study, Sprague-Dawley rats were
fed a high-fat diet, including a mixture or one of the two isomers. Both CLA
isomer groups had significantly greater acyl coenzyme A oxidase mRNA in the
muscle and all CLA groups had improved glucose tolerance and improved
insulin resistance (Choi et al., 2004). The authors concluded that Sprague-
Dawley rats fed a high-fat diet have improved glucose response with CLA
supplementation and the possible mechanisms are increased fat oxidation and
energy expenditure. However, there is some controversy around this conclu-
sion. Pariza argues that the insulin resistance observed in these mouse studies
is a by-product of the experimental conditions (Pariza, 2004).
Conversely, several studies indicate that CLA supplementation results in
hyperinsulinaemia and insulin resistance in mouse models (Tsuboyama-Kasaoka
et al., 2000, Roche et al., 2002, Clement et al., 2002). These studies demon-
strated that observed insulin resistance in mice was concomitant with the
absence of leptin. However as with most studies, the individual isomers were
not investigated. A study employing ob/ob mice reported that supplementation
with the trans-10,cis-12 isomer resulted in increased glucose and contributed

to insulin resistance (Roche et al., 2002). Collectively, findings from rodent
studies indicate that potential beneficial effects of CLA and specifically the
trans-10,cis-12 isomer may be attributable to the total fat content of the diet. In
human clinical trials the results are less favourable. In two separate studies,
both isomers were shown to decrease insulin sensitivity in humans with
metabolic syndrome (Riserus et al., 2001, 2002). Subjects were supplemented
with 3.4 g/day of CLA mixture or trans-10,cis-12 isomer for 12 weeks. Supple-
mentation with the trans-10,cis-12 isomer resulted in increased insulin resistance
and impaired glucose control (Riserus et al., 2002). Another study in obese
men concluded that the trans-10,cis-12 isomer impaired insulin sensitivity.
The authors concluded that supplementation with the trans-10,cis-12 isomer
induces hyperproinsulinaemia in obese individuals (Riserus et al., 2004).
Results from studies investigating the effect of CLA or one of the isomers on
insulin resistance and glucose control are mixed. Additionally, they vary in
both duration of the study, and often the metabolic state of the animal (normal,
obese, or metabolic syndrome). More studies need to be conducted to accu-
rately answer the question of the effect of CLA or the isomers and their
subsequent effect on insulin resistance.
E. Conclusion
Conjugated linoleic acid (CLA) is the common name for a group of positional
and geometric isomers of linoleic acid. The two primary isomers are the cis-9,
trans-11 and trans-10,cis-12 linoleic acid. The earliest health observations of
CLA go back to 1992 (Chin et al., 1992). Since those early observations
considerable research has been conducted on the biological and health effects
of CLA. The reported health benefits attributed to CLA consumption include:
anti-obesity, anti-atherosclerotic/anti-atherogenic, anti-cancer, and enhanced
immune function. As CLA preparations are most commonly a mixture of the
two isomers, the question remains as to which isomer elicited which biological
effect? And by what mechanisms are these effects elicited? Techniques have
been developed for the synthesis and purification of the two isomers allowing
for studies addressing the question of which isomer elicited the observed
biological effects. There is a vast amount of literature using in vitro cell culture
systems and in non-human animal models supporting the efficacy of CLA
supplementation for a variety of health benefits. However, conclusive evidence
of the specific benefits of CLA supplementation for humans is lacking. With
respect to which biological effect is attributable to which isomer, recent studies
have clarified to a great degree which isomer is responsible for which biologi-
cal effect as well as elucidating the mechanisms by which these biological
effects are achieved.
In this regard, it is clear that supplementation with the trans-10,cis-12 isomer
profoundly decreases fat mass and improves body composition in rodents. The
effect of CLA supplementation on reducing fat mass and improving body
composition is less clear in humans as clinical trials have failed to demonstrate
uniform results. Why are the effects observed in rodents conclusive yet not in
humans? As indicated previously, the reasons may range from experimental
design, dose, and duration of the study, age of the study population, gender, or
genetic predisposition. Furthermore, some reports have indicated that the
trans-10,cis-12 isomer may contribute to insulin resistance and hyper-
insulinaemia.
Thus, future studies are necessary to confirm these observations and en-
sure that long term supplementation with the trans-10,cis-12 isomer is safe.
It should be noted that Pariza et al. (2004) suggest that the lack of uniformity
amongst these observations is due to experimental conditions. CLA supple-
mentation modulates the immune response and indeed each isomer appears
to have differential effects, yet as with weight loss the literature is inconclu-
sive with respect to modulation of immune function in humans. Thus, further
research needs to be conducted to confirm both the effect of CLA supple-
mentation as well as which isomer elicits which effect. For example, in
rodents supplementation with the trans-10,cis-12 results in greater IgA and
IgM production compared to the cis-9,trans-11 CLA isomer. Another study
testing both the mixture of the isomers and the isomers individually observed
a decrease in proinflammatory eicosanoids (leukotriene B 4 and
prostaglandin E2). Thus, the potential exists for CLA, a mixture of the iso-
mers or alone as an anti-inflammatory application, to address other specific
needs with respect to immune function.
The trans-10,cis-12 isomer also appears to have potent anti-cancer activity
in vitro and in animal trials. To our knowledge, there are no clinical trials
demonstrating the ability of CLA to reduce tumorigenesis or tumour growth
in humans. And although the trans-10,cis-12 CLA isomer appears to have
potent anti-cancer activity in vitro and in vivo, some studies indicate that the
cis-9,trans-11 CLA isomer also has potent anti-cancer potential. Further
research in a more controlled manner needs to be conducted to verify the
effects and mechanisms by which the independent isomers elicit their bio-
logical effects. As suggested by others, such research could lead to the
development of therapeutic applications to address obesity, inflammatory
conditions, and improve the host response to infection, as well as other
potential applications.
Of course safety is of utmost importance. A recent report indicates that
during long term supplementation the trans-10,cis-12 isomer may actually be
deleterious. Thus, careful trials must also be conducted to confirm both the
safety and efficacy of supplementation with the trans-10,cis-12 isomer.
References
Ackman, RG (1998) Laboratory preparation of conjugated linoleic acids. J. Am. Oil Chem.
Soc., 75, 1227–1469.
Adlof, RO, Duval, S and Emken, EA (2000) Biosynthesis of conjugated linoleic acid in
humans. Lipids, 35, 131–135.
Albers, R, van der Wielen, RPJ, Brink, EJ, Hendriks, HFJ, Dorovska-Taran, VN and Mohede,
ICM (2003) Effects of cis-9, trans-11 and trans-10, cis-12 conjugated linoleic acid (CLA)
on immune function in healthy men. Eur. J. Clin. Nutr., 57, 595–603.
Alonso, L, Cuesta, EP and Gilliland, SE (2003) Production of free conjugated linoleic acid
by Lactobacillus acidophilus and Lactobacillus casei of human intestinal origin. J. Dairy
Sci., 86, 1941–1946.
Ando, A, Ogawa, J, Kishino, S and Shimizu, S (2003) CLA production from ricinoleic acid
by lactic acid bacteria, J. Am. Oil Chem. Soc., 80, 889–894.
Ando, A, Ogawa, J, Kishino, S and Shimizu, S (2004) Conjugated linoleic acid production
from castor oil by Lactobacillus plantarum JCM 1551. Enzyme and Microbial Technol.,
35, 40–45.
Azain, MJ, Hausamann, DB, Sisk, MB, Flatt, WP and Jewell, DE (2000) Dietary conjugated
linoleic acid reduces rat adipose tissue size rather than cell number. J. Nutr., 130, 1548–
1554.
Bauman, DE, Baumgard, LH and Griinari, JM (1999) Biosynthesis of conjugated linoleic
acid in ruminants, Proceedings of the American Society of Animal Science. ASAS
Annual meeting, Conjugated Linoleic Acid: Effects on Metabolism and Health
Minisymposium (Indianapolis, Indiana, USA) pp.1–15 (www.asas.org/jas/symposia/
proceedings/0937.pdf)
Battacharya, A, Banu, J, Rahman, M, Causey, J and Fernandes, G (2006) Biological effects
of conjugated linoleic acids in health and disease. J. Nutr. Biochem., 17, 789–810.
Bissonauth, V, Chouinard,Y, Marin, J, Leblanc, N, Richard, D and Jacques, H (2006) The
effects of t10,c12 CLA isomer compared with c9,t11 CLA isomer on lipid metabolism
and body composition in hamsters. J. Nutr. Biochem., 17, 597–603.
Blankson, H, Stakkestad, JY, Fagertun, H, Thorn,E, Wadstein, J and Gudmundsen, O (2000)
Conjugated linoleic acid reduces body fat mass in overweight and obese humans. J. Nutr.,
130, 2943–2948.
Chin, SF, Liu, W, Storkson, J, Ha, YL and Pariza, M (1992) Dietary sources of conjugated
dienoic isomers of linoleic acid, a newly recognized class of anticarcinogens. J. Food
Comp.Anal., 5, 185–197.
Chipault, JR and Hawkins, JM (1959) The determination of conjugated cis-trans and trans-
trans methyl octadecadienoates by infrared spectrometry, J. Am. Oil Chem. Soc., 36,
535–539.
Cho, HJ, Kim, WK, Jung, JI, Kim, EJ, Lim, SS, Kwon, DY and Park, JH (2005) Trans-10,12-
cis, not cis-9,trans-11, conjugated linoleic acid decreases ErbB3 expression in HT-29
human cancer cells. World J. Gasteroenterol., 11, 5142–5150.
Cho, HJ, Lee, HS, Chung, CK, Kang, YH, Ha, YL, Park, HS and Park, JH (2003) Trans-
10,cis-12 conjugated linoleic acid reduces insulin-like growth factor-II in HT-29 human
cancer cells. J. Med. Food, 6, 193–199.
Choi, Y, Kim, YC, Han, YB, Park, Y, Pariza, MW and Ntambi, JM (2000) The trans-10,cis-
12 isomer of conjugated linoleic acid downregulates stearoyl-CoA desaturase 1 gene
expression in 3T3-L1 adipocytes. J. Nutr, 130, 1920–1924.
Choi, JS, Jung, MH, Park, HS and Song, J (2004) Effect of conjugated linoleic acid isomers
on insulin resistance and mRNA levels of genes regulating energy metabolism in high-
fat-fed rats. Nutrition, 20, 1008–10017.
Choi, N-J, Imm, JY, Oh, S, Kim, B-C, Hwang, H-J and Kim, YJ (2007) Effect of pH and
oxygen on conjugated linoleic acid production by mixed rumen bacteria from cows fed
high concentrate and high forage diets, Animal Feed Science and Technology, 123–124,
643–653.
Christie, WW, Dobson, G and Gunstone, FD (1997) Isomers in commercial samples of
conjugated linoleic acid. Lipids, 32, p.1231.
Clement, L, Poirier, H, Niot, I, Bocher, V, Guerre-Millo, M, Krier, S, Staels, B and Besnard,
P (2002) Dietary trans-10,cis-12 conjugated linoleic acid induces hyperinsulinemia and
fatty liver in the mouse. J. Lipid Research, 43, 1400–1409.
Coakley, M, Ross, RP, Nordgren, M, Fitzgerald, G, Devery, R and Stanton, C (2003)
Conjugated linoleic acid biosynthesis by human-derived Bifidobacterium species. J. Appl.
Microbiol., 94, 138–145.
Degner, SC, Kemp, MQ, Bowden, T and Romangnolo, DF (2006) Conjugated linoleic acid
attenuates cyclooxygenase-2 transcriptional activity via an anti-Ap-1 mechanism in
MCF-7 breast cancer cells. J. Nutr, 136, 421–427.
DeLaney, JP and West, DB (2000) Changes in body composition wit conjugated linoleic acid.
J. Am. Coll. Nutr., 19, 487S–493S.
Devillard, E, McIntosh, FM, Duncan, SH and Wallace, RJ (2007) Metabolism of linoleic acid
by human gut bacteria: different routes for biosynthesis of conjugated linoleic acid.
J. Bacteriol., 189, 2566–2570.
Eder, K, Schlesser, S, Becker, K and Korting, R (2003) Conjugated linoleic acids lower the
release of eicosanoids and nitric oxide from human aortic endothelial cells. J. Nutr., 133,
4083–4089.
Evans, M, Geigermann, C, Cook, J, Curits, L, Kuebler, B and MacIntosh, M (2000)
conjugated linoleic acid suppresses triglyceride accumulation and induces apoptosis in
3T3-L1 preadipocytes. Lipids, 35, 134–138.
Fite, A, Goua, M, Wahle, KWJ, Schofield, AC, Hutcheson, AW and Heys, SD (2007)
Potentiation of the anti-tumor effect of docetexel by conjugated linoleic acids (CLAs) in
breast cancer cells in vitro. Pros. Leuko. Essent. Fatty Acids, 77, 87–96.
Fritsche, S and Fritsche, J (1998), Occurrence of conjugated linoleic acid isomers in beef.
J. Am. Oil Chem. Soc., 75, 1449–1451.
Fritsche, JRR and Steinhard, H (1999) Formation, contents and estimation of daily intake of
conjugated linoleic acid isomers and trans-fatty acids in food. Advances in Comjugated
Linoleic Acid Research, 1, 378–396.
Fukuda, S, Furuya, H, Suzuki, Y, Asanuma, N and Hino, T (2005) A new strain of
Butyrivibrio fibrisolvens that has high ability to isomerize linoleic acid to conjugated
linoleic acid. J. Gen. Appl. Microbiol., 51, 105–113.
Gavino, VC, Gavio, G, Leblanc, M-J and Tuchweber, B (2000) An isomeric mixture of
conjugated linoleic acid but not pure cis-9,trans-11-octadecadienoic acid affects body
weight gain and plasma lipids in hamsters. J. Nutr., 130, 27–29.
Griinari, JM and Bauman, DE (1999) Biosynthesis of conjugated linoleic acid and its
incorporation into meat and milk in ruminants. In: Advances in Conjugated Linoleic Acid
Research, First edition (Yurawecz, MP, Mossaba, MM, Kramer, JKG, Pariza, MW,
Nelson, GJ, eds) AOCS Press, Champaign, Illinois, USA, pp.180–200.
Ha, YI, Grimm, NK and Pariza, MW (1987) Anticarcinogens from fried ground beef: heat-
altered derivatives of linoleic acid. Carcinogenesis, 8, 1881–1887.
Ha, YI, Storkson, J and Pariza, MW (1990) Inhibition of benzo(a)pyrene-induced mouse
forestomach neoplasia by conjugated dienoic derivatives of linoleic acid. Cancer Res.,
15, 1097–1101.
Houseknecht, KL, Vanden, Heuvel, JP, Moya-Camarena, SY (1998) Dietary conjugated
linoleic acid normalizes impaired glucose tolerance in Zucker diabetic fatty fa/fa rat.
Biochem. Biophys. Res. Commun., 244, 678–682.
Hornung, E, Krueger, C, Pernstich, C, Gipmans, M, Porzel, A and Feussner, I (2005)
Production of (10E,12Z)-conjugated linoleic acid in yeast and tobacco seeds. Biochim.

Biophys. Acta, 1738, 105–114.
Ip, C, Dong,Y, Ip, MM, Banni, S, Carta, G, Angioni, E, Murru, E, Spada, S, Melis,MP and
Saebo, A (2002) Conjugated linoleic acid isomers and mammary gland. Nutr. Cancer, 41,
911–97.
Ip, MM, McGee, SO, Masso-Welch, PA, Ip, C, Meng, X, Ou, L and Shoemaker, SF (2007)
The t10,c12 isomer of conjugated linoleic acid stimulates mammary tumorigenesis in
transgenic mice over-expressing erbB2 in mammary epithelium. Carcionogenesis, 28,
1269–1276.
Jain, VP and Proctor, A (2006) Photocatalytic production and processing of conjugated
linoleic acid-rich soy oil. J. Agric. Food Chem., 54, 5590–5596.
Jain, VP and Proctor, A (2007a) Kinetics of photoirradiation-induced synthesis of soy oil-
conjugated linoleic acid isomers. J. Agric. Food Chem., 55, 889–894.
Jain, VP and Proctor, A (2007b) Production of conjugated linoleic acid-rich potato chips,
J. Food Sci., 72, pp.S75–S78.
Jiang, J, Bjorck, L and Fonden, R (1998) Production of conjugated linoleic acid by dairy
starter cultures. J. Appl. Microbiol., 85, 95–102.
Kang, J-H, Lee, G-S, Jeung, E-B and Yang, M-P (2007) Trans-10,cis-12 conjugated linoleic
acid increased phagocytosis of porcine peripheral blood polymorphonuclear cell in vitro.
Brit. J. Nutr., 97, 117–125.
Kawahara, S, Sakka, K, Niimi, M, Kawamura, O and Mugumura, M (2007) In-vitro synthesis
of conjugated linoleic acids by rumen bacteria. Miyazaki Daigaku Nogakubu Kenkyu
Hokoku, 53, 101–106.
Kemp, MQ, Brandon, DJ and Romangnolo, DF (2003) Conjugated linoleic acid inhibits cell
proliferation through a p53-dependant mechanism: effect on the expression of G1-
restriction points in breast and colon cancer cells. J. Nutr., 133, 3670–3677.
Kim, EJ, Holthuizen, PE, Park, HS, Ha, YL, Jung, KC and Park, JH (2002) Trans-10,12-cis
conjugated linoleic acid inhibits Caco-2 colon cancer cell growth. Am. J. Physiol. Liver
Physiol., 283, G357–367.
Kim, YJ (2003) Partial inhibition of biohydrogenation of linoleic acid can increase the
conjugated linoleic acid production of Butyrivibrio fibrisolvens A38. J. Agric. Food
Chem., 51, 4258–4262.
Kim, YJ, Liu, RH, Bond, DR and Russell, JB (2000) Effect of linoleic acid concentration on
conjugated linoleic acid production by Butyrivibrio fibrisolvens A38. Appl. Environ.
Microbiol., 66, 5226–5230.
Kim, YJ, Liu, RH, Rychlik, JL and Russell, JB (2002) The enrichment of a ruminal bacterium
(Megasphaera elsdenii YJ-4) that produces the trans-10,cis-12 isomer of conjugated
linoleic acid, J. Appl. Microbiol., 92, 976–982.
Kim, JH, Hubbard, NE, Ziboh, V and Erickson, KL (2005) Conjugated linoleic acid reduction
of murine mammary tumor cell growth through 5-hydroxyeicosatetraenoic acid. Biochim.
Biophys. Acta, 1687, 103–109.
Kishino, S, Ogawa, J, Omura, Y, Matsumura, K and Shimizu, S (2002) Conjugated linoleic
acid production from linoleic acid by lactic acid bacteria. J. Am. Oil Chem. Soc., 79, 159–
163.
Kohno-Murase, J, Iwabuchi, M, Endo-Kasahara, S, Sugita, K, Ebinuma, H and Imamura, J
(2006) Production of trans-10,cis-12 conjugated linoleic acid in rice. Transgenic Res.,
15, 95–100.
Kraft, J, Collomb, M, Mockel, P, Sieber, R and Jahreis, G (2003) Differences in CLA isomer
Kramer, JKG, Cruz-Hernandez, C, Deng, ZY, Zhou, JQ, Jahreis, G and Dugan, MER (2004)
Analysis of conjugated linoleic acid and trans 18:1 isomers in synthetic and animal
products. Am. J. Clin. Nutr., 79, 1137S–1145S.
Kramer, JKG, Parodi, PW, Jensen, RG, Mossoba, MM, Yurawecz, MP and Adlof, RO (1998)
LaRosa, CP, Miner, J, Xia, Y, Zhou, Y, Kachman, S and Fromm, ME (2006) Trans-10,cis-
12 conjugated linoleic acid causes inflammation and delipidation of white adipose tissue
in mice: a microarray and histologic analysis. Physiol. Genomics, 27, 282–294.
LaRosa, PC, Riethoven, JJ, Chem H, Xia, Y, Chen, M, Miner, JL and Fromm, ME (2007)
Trans-10,cis-12 conjugated linoleic acid activates the integrated stress response pathway
in adipocytes. Physiol. Genomics, 31, 544–553.
Lee, HY, Park, JH, Seok, SH, Baek, MW, Kim, DJ, Lee, KE, Paek, KS, Lee, Y and Park, JH
(2006) Human originated bacteria, Lactobacillus rhamnosus PL60, produce conjugated
linoleic acid and show anti-obesity effects in diet-induced obese mice, Biochim. Biophys.
Acta, 1761, 736–744.
Lee, SO, Hong, GW and Oh, DK (2003) Bioconversion of linoleic acid into conjugated
linoleic acid by immobilized Lactobacillus reuteri. Biotechnology Progress, 19, 1081–
1084.
Lee, SO, Kim, CS, Cho, SK, Choi, HJ, Ji, GE and Oh, DK (2003) Bioconversion of linoleic
acid into conjugated linoleic acid during fermentation and by washed cells of Lactoba-
cillus reuteri. Biotechnology Letters, 25, 935–938.
Li, G, Barnes, D, Butz, D, Bjorling, D and Cook, ME (2005) 10t,12c-conjugated linoleic acid
inhibits lipopolysaccharide-induced cyclooxygenase expression in vitro and in vivo.
J. Lipid Res., 46, 2134–2142.
Liavonchanka, A, Hornung, E, Feussner, I and Rudolph, MG (2006) Structure and mecha-
nism of the Propionibacterium acnes polyunsaturated fatty acid isomerase. Proc. Natl.
Acad. Sci. USA, 103, pp.2576–2581.
Lim-do, Y, Tyner, AL, Park, JB, Lee, Jy, Choi, YH and Park, JH (2005) Inhibition of colon
cancer cell proliferation by the dietary compound conjugated linoleic acid is mediated by
the CDK inhibitor p21CIP1/WAF1. J. Cell Physiol., 205, 107–113.
Lock, AL, Corl, BA, Barbano, DM, Bauman, DE and Ip, C (2004) The anticarcinogenic effect
of trans-11 18:1 is dependent on its conversion to cis-9, trans-11 CLA by delta9-
desaturase in rats. J. Nutr., 134, 2698–2704.
Malpuech-Brugére, C, Verboeket-van de Venne, W PHG, Mensink, RP, Arnal, M-A, Morio,
B, Brandolini, M, Saebo, A, Lassel, TS, Chardigny, JM, Sébédio, JL and Beaufrere
(2003) Effects of two conjugated linoleic acid isomers on body fat mass in overweight
humans. u Obes. Res., 12, 591–598.
Martin, JC, Gregoire,S and Siess, MH (2000) Effects of conjugated linoleic acid isomers on
lipid-metabolizing enzymes in male rats. Lipids, 35, 91–98.
McNeill, GP, Rawlins, C and Peilow, AC (1999) Enzymatic enrichment of conjugated
linoleic acid isomers and incorporation into triglycerides. J. Am. Oil Chem. Soc., 76,
1265–1268.
Miller, A, Snanton, C and Devenery, R (2002) Cis 9, trans 11- and trans 10, cis 12-conjugated
linoleic acid isomers induce apoptosis in cultured SW480 cells. Anticancer Res., 22,
3879–3887.
Moon, HS, Lee, HG, Seo, JH, Chung, CS, Guo, DD, Kim, TG, Choi, YJ and Cho, CS (2007)
Leptin-induced matrix matelloproteinase-2 secretion is suppressed by trans-10,cis-12
conjugated linoleic acid. Biochem. Biophys. Res. Commun., 356, 955–960.
Mosley, EE, McGuire, MK, Williams, JE and McGuire, MA (2006a) Cis-9,trans-11
conjugated linoleic acid is synthesized from vaccenic acid in lactating women. J. Nutr.,
136, 2297–2301.
Mosley, EE, Shafii, DB, Moate, PJ and McGuire, MA (2006b) Cis-9,trans-11 conjugated
linoleic acid is synthesized directly from vaccenic acid in lactating dairy cattle. J. Nutr.,
136, 570–575.
Murphy, EF, Hooivled, GJ, Muller, M, Calogero, RA and Cashman, KD (2007) Conjugated
linoleic acid alters global gene expression in human intestinal-like Caco-2 cells in an
isomer-specific manner. J. Nutr, 137, 2359–2365.
Nagao, T, Shimada, Y, Yamauchi-Sato, Y, Yamamoto, T, Kasai, M, Tsutsumi, K, Sugihara,
A and Tominaga, Y (2002) Fractionation and enrichment of CLA isomers by selective
esterification with Candida rugosa lipase. J. Am. Oil Chem. Soc., 79, 303–308.
Nagao, T, Yamauchi-Sato, Y, Sugihara, A, Iwata, T, Nagao, K, Yanagita, T, Adachi, S and
Shimada, Y (2003) Purification of conjugated linoleic acid isomers through a process
including lipase-catalyzed selective esterification. Biosci. Biotechnol. Biochem., 67,
1429–1433.
Nichols, PL, Herb, SF and Riemenschneider, RW (1951) Isomers of conjugated fatty acids.
I. Alkaline-isomerized linoleic acid. J. Chem. Soc., 73, 247–252.
Ogawa, J, Kishino, S ando, A, Sugimoto, S, Mihara, K and Shimizu, S (2005) Production of
conjugated fatty acids by lactic acid bacteria, J. Biosci. Bioeng., 100, 355–364.
Ogawa, J, Matsumura, K, Kishino, S, Omura, Y and Shimizu, S (2001) Conjugated linoleic
acid accumulation via 10-hydroxy-12 octadecaenoic acid during microaerobic transfor-
mation of linoleic acid by Lactobacillus acidophilus. Appl. Environ. Microbiol., 67,
1246–1252.
Ogawa, J, Sugimoto, S, Ito, T, Kishino, S, Ando, A and Shirasaka, N (2004) Screening and
analysis of microbial reactions useful for CLA production. Abstracts of AOCS 95th
Annual Meeting & Expo, 9–12 May 2004, Cincinnati, Ohio, USA.
Oh, DK, Hong, GH, Lee, Y, Min, SG, Sin, HS and Cho, SK (2003) Production of conjugated
linoleic acid by isolated Bifidobacterium strains. World Journal of Microbiology &
Biotechnology, 19, 907–912.
O’Shea, M, Bassaganya-Riera, J and Mohede, ICM (2004) Immunomodulatory properties of
conjugated linoleic acid. Am. J. Clin. Nutr., 79(S), 1199S–2006S.
Ostrowska, E, Muralitharan, M, Cross, RD, Bauman, DE and Dunshea, FR (1999) Dietary
conjugated linoleic acids increase lean tissue and decrease fat deposition in growing pigs.
J. Nutr., 129, 2037–2042.
Palombo, JD, Ganguly, A, Bistrian, BR and Menard, MP (2002) The antiproliferative effects
of biologically active isomers of conjugated linoleic acid on human colorectal and
prostate cancer cells. Cancer Lett., 28, 163–172.
Pariza, MW (1988) Dietary fat and cancer risk; evidence and research needed. Ann. Rev.
Nutr., 167–183.
Pariza, MW (2004) Perspective on the safety and effectiveness of conjugated linoleic acid.
Am. J. Clin. Nutr., 79 (suppl), 1132S–1136S.
Pariza, MW, Park, Y, Cook, ME, (2000) Mechanisms of action of conjugated linoleic acid:
evidence and speculation. Proc Soc Exptl Biol Med, 223, 8–13.
Pariza, MW, Park, Y and Cook, ME (2001) The biologically active isomers of conjugated
linoleic acid. Prog. Lipid Res., 40, 283–298.
Pariza, MW, Park, Y, Xu, X, Ntambi, J and Kang, KC (2003) Speculation on the mechanisms
of action of conjugated linoleic acid. In: Advances in Conjugated Linoleic Acid Research,
Volume 2 (Sébédio, JL, Christie, WW and Adolf, R, eds), AOCS Press, Champaign,
Illinois, USA, pp.251–258.
Park, Y, Albright, KJ, Liu, W, Storkson, JM, Cook, ME,and Pariza, MW (1997) Effect of
conjugated linoleic acid on body composition in mice. Lipids, 32, 853–858.
Park,Y, Storkson, JM, Albright, KJ, Liu, W and Pariza, MW (1999) Evidence that the trans-
10,cis-12 isomer of conjugated linoleic acid induces body composition changes in mice.
Lipids, 34, 235–241.
Park, Y and Pariza, MW (2001) Lipoxygenase inhibitors inhibit heparin releasable lipopro-
tein lipase activity in 3T3-L1 adipocytes and enhance bodyweight reduction mice by
conjugated linoleic acid. Biochim. Biophys. Acta, 1534, 27–33.
Park, Y and Pariza, MW (2001) The effects of dietary conjugated nonadecadienoic acid on
body composition in mice. Biochim. Biophys. Acta, 1533, 171–174.
Parodi, PW. 1977, Conjugated octadecadienoic acids of milk fat. J. Dairy Sci., 60, 1550–
1553.
Parodi, PW. 1999, Conjugated linoleic acid and other anticarcinogenic agents of bovine milk
fat. J. Dairy Sci., 82, 1339–1349.
Plourde, M, Jew, S, Cunnane, SC and Jones PJH (2008) Conjugated linoleic acids: why the
discrepancy between animal and human studies? Nutr. Rev., 66, 415–421.
Poulos, SP, Sisk, M, Hausmann, DB, Azain, MJ and Hausamann, GJ, (2001) Pre- and
postnatal dietary conjugated linoleic acid alters adipose development, body weight, gain
and body composition in Sprague-Dwley rats. J. Nutr., 131, 2722–2733.
Peterson, DG, Matitashvili, EA and Bauman, DE (2003) Diet-induced milk fat depression in
dairy cows results in increased trans-10,cis-12 CLA in milk fat and coordinate suppres-
sion of mRNA abundance for mammary enzymes involved in milk fat synthesis. J. Nutr.,
133, 3098–3102.
Reaney, MJT, Liu, YD and Westcott, ND (1999) Commercial production of conjugated
linoleic acid. In: Advances in Conjugated Linoleic Acid Research, First edition (Yurawecz,
MP, Mossoba, MM, Kramer, JKG, Pariza, MW, and Nelson, GJ, eds) AOCS Press,
Champaign, Illinois, USA, pp.39–54.
Riserus, U, Berglund, L and Vessby, B (2001) Conjugated linoleic acid (CLA) reduced
abdominal adipose tissue in obese middle-aged men with signs of metabolic syndrome:
a randomized control trial. J. Obes. Relat. Metab. Disord., 25, 1129–1135.
Riserus, U, Amer, P, Brismar, K and Vessby, B (2002) Treatment with dietary trans-10,cis-
12 conjugated linoleic acid isomers causes isomer-specific insulin resistance in obese
men with the metabolic syndrome. Diabetes Care, 25, 1516–1521.
Riserus, U, Vessby, B, Amer, P and Zethelius, B. (2004) Supplementation with trans-10,cis-
12 conjugated linoleic acid induces hyperproinsulinemia in obese men: close association
with impaired insulin sensitivity. Diabetologia, 47, 1016–1019.
Roche, HM, Noon, E, Sewter, C, Savage, D, Gibney, MJ, O’Rahilly, S and Vidal-Puig, AJ
(2002) Isomer-independent metabolic effects of conjugated linoleic acid: insights from
molecular markers and sterol regulatory element-binding protein-1c and LXR alpha.
Diabetes, 51, 2037–2044.
Rosberg-Cody, E, Johnson, MC, Fitzgerald, GF, Ross, PR and Stanton, C (2007) Heterolo-
gous expression of linoleic acid isomerase from Propionibacterium acnes and
anti-proliferative activity of recombinant trans-10,cis-12 conjugated linoleic acid. Micro-
biology, 153, 2483–2490.
Rosson, RA, Deng, MD, Grund, AD and Peng, SS (2001) Linoleate isomerase. World Patent
WO 01/00846 1-118.
Ryder, JW, Portocarrero, CP, Song XM, Cui, L, Yu, M, Combatsiaris, T, Galuska, D,
Bauman, DE, Barbano, DM, Charron, MJ, Zierath, JR and Housenecht, KL (2001)
Isomer-specific antidiabetic properties of conjugated linoleic acid. Improved glucose
tolerance, skeletal muscle insulin action and UCP-2 gene expression. Diabetes, 50, 1149–
1157.
Sébédio, J-L, Christie, WW and Adlof, R (eds) (2003) Advances in Conjugated Linoleic Acid
Research, Volume 2. AOCS Press, Champaign, Illinois, USA.
Sieber, R, Collomb, M, Aeschlimann, A, Jelen, P and Eyer, H (2004) Impact of microbial
cultures on conjugated linoleic acid in dairy products — a review, International Dairy
Journal, 14, 1–15.
Simon, E, Macarulla, MT, Fernandez-Quinteal, A, Rodriguez,VM and Portillo, MP (2005)
Body fat-lowering effect of conjugated linoleic acid is not due to increased lipolysis.
J. Phys. Biochem, 61, 363–370.
Stachowska, E, Baskiewicz-Masiuk, M, Dziedziejko, V, Adler, G, Bober, J, Machalinski, B
and Chlubek, D (2007) Conjugated linoleic acids can change phagocytosis of human
monocytes/macrophages by reduction of Cox-2 expression. Lipids, 42, 707–716.
Steck, SE, Chalecki, AM, Miller, P, Conway, J, Austin, GL, Hardin, JW, Albright, CD and
Thuillier, P (2007) Conjugated linoleic acid supplementation for twelve weeks increases
lean body mass in obese humans. J. Nutr., 137, 1188–1193.
Tricon, S, Burdge, GC, Kew, S, Banerjeee, T, Russel, JJ, Grimble, RF, Williams, CM, Calder,
PC and Yaqoob, P (2004) Effects of cis-9,trans-11 and trans-10,cis-12 conjugated
linoleic acid on immune cell function in healthy humans. Am. J. Clin. Nutr., 80, 1626–
1633.
Tricon, S, Burdge, GC, Kew, S, Banerjeee, T, Russel, JJ, Jones, EL, Grimble, RF, Williams,
CM, Yaqoob, P and Calder, PC (2004) Opposing effects of cis-9, trans-11 and trans-
10,cis-12 conjugated linoleic acid on blood lipids in healthy humans. Am. J. Clin. Nutr.,
80, 614–620.
Tsuboyama-Kasaoka N, Takahashi, M, Tanemura, K, Kim, HJ, Tange, T, Okuyama, H,
Kasai, M, Ikemoto, S and Ezaki, O (2000) Conjugated linoleic acid supplementation
reduces adipose tissue by apoptosis and develops lipodystrophy in mice. Diabetes, 49,
1534–1542.
Uehara, H, Suganuma, T, Negishi, S, Satoru, U and Sato, K (2006) A novel method for solvent
fractionation of two CLA isomers., J. Am. Oil Chem. Soc., 83, 261–267.
Vahvaselka, M, Lehtinen, P, Sippola, S and Laakso, S (2004) Enrichment of conjugated
linoleic acid in oats (Avena sativa L.) by microbial isomerization, J. Agric. Food Chem.,
52, 1749–1752.
Verhulst, A (1987) Isomerization of polyunsaturated long chain fatty acids by Propionibac-
teria. Syst. Appl. Microbiol., 9, 12–15.
Wahle, KWJ, Heys, SD and Rotondo, D (2004) Conjugated linoleic acids; are they beneficial
or detrimental to health? Prog. Lipid Res., 43, 553–587.
Wallace, RJ, McKain, N, Shingfield, KJ and Devillard, E (2007) Isomers of conjugated
linoleic acids are synthesized via different mechanisms in ruminal digesta and bacteria.
J. Lipid Res., 48, pp.2247–2254.
Wang, LS, Huang, YW, Sugimoto, Y, Liu, S, Chang, HL, Ye, W, Shu, S and Lin, YC (2006)
Conjugated linoleic acid (CLA) up-regulates the estrogen-regulated cancer suppressor
gene, protein tyrosine phosphatase gamma (PTPgamma), in human breast cells. Antican-
cer Res., 26, 27–34.
Wang, YW and Jones, PJH (2004) Conjugated linoleic acid and obesity control: efficacy and
mechanisms. Intl J. Obesity, 28, 941–955.
West, DB, DeLaney, JP, Camet, PM, Blohm, FY, Truett, AA and Sciemaca, J (1998) Effects
of conjugated linoleic acid on body fat and energy metabolism in the mouse. Am. J.
Physiol., 275, R667–R672.
West, DB, Blohm, FY, Truett, AA and DeLaney, JP (2000) Conjugated linoleic acid
persistently increases total energy expenditure in AKR/J mice without increasing
uncoupling protein gene expression. J. Nutr., 130, 2471–2477.
Yamasaki, M, Chujo, H, Hiarao, A, Koyanagi, N, Okamoto, T, Tojo, N, Oishi, A, Iwata, T,
Yamauchi-Sato, Y, Yamamoto, T, Tsutsumi, K, Tachibana, H and Yamada, K (2003)
Immunoglobulin and cytokine production from spleen lymphocytes is modulated in
C57BL/6J mice by dietary cis-9,trans-11 and trans-10,cis-12 conjugated linoleic acid.
J. Nutr., 133, 784–788.
Xu, C, Lee, H, Lee, B, Wang, J, Hong, Z, Kim, T, Kang, S, Choi, N, Roh, S and Choi, Y (2006)
Production of lean beef containing a high content of trans-10,cis-12 conjugated linoleic
acid by feeding a high-temperature-micro-time-treated diet with extruded soybean.
Biosci. Biotechnol. Biochem., 70, 2589–2597.
CHAPTER 10
Observational epidemiological studies on
intake of trans fatty acids and risk of ischaemic
heart disease
MARIANNE UHRE JAKOBSEN AND KIM OVERVAD
Department of Cardiology and Department of Clinical Epidemiology, Aalborg

Hospital, Aarhus University Hospital, Aalborg, Denmark
A. Background
1. Sources and intake

Milk fat and meat from ruminant animals contain trans fatty acids (TFA) as a
result of bacterial hydrogenation of unsaturated fatty acids in the rumen
(ruminant TFA). TFA are also present in margarines, shortenings and frying
fats as a result of industrial partial hydrogenation of unsaturated vegetable oils
(industrial TFA). The predominant products are trans 18:1 isomers, whether
from bacterial hydrogenation in the rumen or industrial partial hydrogenation
of unsaturated vegetable oils (Precht & Molkentin, 1995). However, the
isomeric distribution of trans 18:1 in milk fat differs considerably from the
distribution in margarines and shortenings (Precht & Molkentin, 1995, 1996);
in milk fat the predominant TFA is vaccenic acid (trans 18:1n–7) whereas
in margarines and shortenings the predominant TFA is elaidic acid (trans
18:1n–9). Besides positional isomers of trans 18:1, milk fat contains trans
14:1, trans 16:1 and trans 18:2 (Jakobsen et al., 2006; Precht & Molkentin,
1995). Margarines contain trans 18:2 isomers in addition to positional isomers
of trans 18:1, but TFA of other chain length (trans 14:1 and trans 16:1) only
occur in traces (Precht & Molkentin, 1995).
In North America, mean daily intakes of TFA have been estimated to be 3–4 g
per person (Craig-Schmidt, 2006). Diets in northern Europe traditionally have
contained more TFA than in Mediterranean countries where olive oil is used.
According to the TRANSFAIR study, intakes of TFA in Europe have been
estimated to range from minimal values in Italy, Portugal, Greece, and Spain
(mean daily intakes: 0.6–0.7% of energy intake, 1.4–2.1 g/person) to
255
greater values in Germany, Finland, Denmark, Sweden, France, UK, Belgium,

Norway, the Netherlands, and Iceland (mean daily intakes: 0.9–2.0% of energy
intake, 2.1–5.4 g/person) (Hulshof et al., 1999; van Poppel, 1998). Among these
countries, 28–79% of total TFA intake was obtained from ruminant sources
(Hulshof et al., 1999). Decreases in dietary TFA have been observed due to in-
dustrial modification and changes in dietary habits. Two studies have indicated
that the intake of TFA in the USA may have decreased (Harnack et al., 2003; Oh
et al., 2005). Harnack et al. (2003) reported that mean daily TFA intakes among
women had decreased from 2.8% of energy intake (5.4 g/person) in 1980–82 to
2.2% of energy intake (4.7 g/person) in 1995–97 and among men from 3.0% of
energy intake (8.4 g/person) in 1980–82 to 2.2% of energy intake (6.4 g/person)
in 1995–97. Oh et al. (2005) reported that mean daily TFA intakes had decreased
from 2.2% of energy intake in 1980 to 1.6% of energy intake in 1998 among
women. In Denmark, the content of industrial TFA in margarines has been re-
duced and consequently the mean daily intakes of industrial TFA from margarines
and shortenings has decreased from 2.2 g/person in 1992 to 0.35 g/person in
1999 (Leth et al., 2003). In contrast, the intake of ruminant TFA may have been
constant, as the amount of milk fat and ruminant meat available for con-sumption
has been rather constant (Fagt & Trolle, 2001). In Denmark, the daily median
intake of ruminant TFA has been estimated to be 0.5% of energy intake (1.8 g/
person) among women and men aged 30–80 years (Jakobsen et al., 2006). In the
Netherlands, the daily mean TFA intake decreased from 4.3% of energy intake
(10.9 g/person) in 1995 to 1.9% of energy intake (4.4 g/person) in 1995 among
men (Oomen et al., 2001). Intake of ruminant TFA remained constant during this
time period at around 0.7% of energy intake; the decrease in intake of TFA was
thus due to decreased intake of industrial TFA (Oomen et al., 2001).
2. Biological effects of TFA and mechanisms

The ratio of plasma total cholesterol concentration to high-density lipoprotein
(HDL) cholesterol concentration has a strong positive association with the risk of
ischaemic heart disease (IHD) (Shai et al., 2004). A meta-analysis of human
dietary trials shows that intake of TFA increases the ratio of total to HDL-
cholesterol more than intake of other fatty acids (Mensink et al., 2003). Intake of
TFA increases serum activity of cholesterylester transfer protein (van Toll et al.,
1995); the enzyme which transfers cholesterol esters from HDL-cholesterol to
lipoproteins of lower density. This increased activity may partly explain the
unfavourable ratio of total to HDL-cholesterol seen with intake of TFA.
However, all the dietary intervention studies contributing to this analysis used
TFA derived from industrial sources. An interesting question is whether TFA
obtained from ruminant sources have different effects compared with those
obtained from industrial sources. Two randomized dietary trials have compared
the effects on plasma lipids of TFA derived from ruminant sources and from
EPIDEMIOLOGICAL STUDIES ON INTAKE OF TRANS FATTY ACIDS 257
industrial sources (Chardigny et al., 2008; Motard-Belanger et al., 2008).

Chardigny et al. (2008) compared two experimental diets, either high in TFA
derived from ruminant sources or high in TFA derived from industrial sources (in
each case, around 5% of energy intake, 11–12 g/d). The interpretation of that
study is difficult as only cholesterol concentrations at the end of the intervention
periods are provided. Among women, total cholesterol, HDL- and low-density
lipoprotein (LDL) cholesterol concentrations were higher after intake of TFA
obtained from ruminant sources than after intake of TFA obtained from industrial
sources (Chardigny et al., 2008). Among men, only minor differences were
observed (Chardigny et al., 2008). The wash-out period between the two
interventions was only one week and, as cholesterol concentrations at the start of
the second intervention period were not provided, carry-over effects may have
affected this study. Motard-Belanger et al. (2008) compared four experimental
diets among men; high in TFA derived from ruminant sources (3.7% of energy
intake, 10.2 g/d), high in TFA derived from industrial sources (3.7% of energy
intake, 10.2 g/d), moderate in TFA derived from ruminant sources (1.5% of
energy intake, 4.2 g/d), and low in TFA from any source (0.8% of energy intake,
2.2 g/d) (control diet). Also, the cholesterol concentrations before the different
interventions were not provided in that study. However, the wash-out periods
were from 3–13 weeks making carry-over effects less likely. The high TFA diets,
whether ruminant TFA or industrial TFA, increased plasma LDL-cholesterol
concentrations, marginally decreased plasma HDL-cholesterol concentrations,
and increased the ratio of total cholesterol to HDL-cholesterol (Motard-Belanger
et al., 2008). The moderate ruminant TFA diet was associated with the most
favourable cholesterol concentrations, although, compared with the control diet,
none of these differences reached significance (Motard-Belanger et al., 2008).
Other biological effects of intake of TFA may also contribute to the risk of
IHD. Intake of TFA may induce inflammation, endothelial dysfunction, adi-
posity, and insulin resistance (Mozaffarian et al., 2006); the mechanisms
behind these effects are not well established.
3. Aim
The aim of this chapter is to summarize and evaluate the results from observa-
tional epidemiological studies on the intake of TFA and the risk of IHD.
B. Methods
The selection criteria for the observational epidemiological studies were that
they included data on TFA intake and development of IHD in the general
population. Measures of TFA intake included both direct measures of dietary
TFA intakes and proportions of TFA in adipose tissue, whole blood, plasma, or
erythrocytes which are considered to provide biomarkers of TFA intake
(Baylin et al., 2002; Baylin et al., 2005; Sun et al., 2007a).
A PubMed search (MEDLINE citations) of papers published from January

1966 onwards was performed in May 2008 using the following keywords:
trans fatty acids AND (acute coronary syndrome OR cardiovascular disease
OR coronary artery disease OR coronary heart disease OR ischaemic heart
disease OR sudden cardiac death OR sudden death). The emerged papers were
searched for additional relevant references. Only papers published in English
were considered.
C. Results
1. Included studies
In total, 31 potential studies on the association between TFA intake and risk of
IHD emerged from the PubMed search (Aro et al., 1995; Ascherio et al., 1994;
Ascherio et al., 1996; Baylin et al., 2003; Bolton-Smith et al., 1996; Clifton
et al., 2004; Colon-Ramos et al., 2006; Dlouhy et al., 2003; Fritsche et al.,
1998; Ghahremanpour et al., 2008; Harris et al., 2007; Heckers et al., 1977;
Hodgson et al., 1996; Jakobsen et al., 2008; Kromhout et al., 1995; Lemaitre
et al., 2002; Lemaitre et al., 2006; Lopes et al., 2007; Mozaffarian et al., 2007;
Oomen et al., 2001; Pedersen et al., 2000; Pietinen et al., 1997; Roberts et al.,
1995; Siguel & Lerman, 1993; Singh et al., 1996; Sun et al., 2007b; Sun et al.,
2007c; Thomas et al., 1983a; Thomas et al., 1983b; van de Vijver et al., 1996;
Willett et al., 1993). Seven studies were excluded as the selection criteria were
not fulfilled (Fritsche et al., 1998; Heckers et al., 1977; Hodgson et al., 1996;
Mozaffarian et al., 2007; Singh et al., 1996; Thomas et al., 1983b; van de
Vijver et al., 1996). Four studies were included after search in references (Hu
et al., 1997; Hu et al., 1999; Oh et al., 2005; Thomas & Scott, 1981). In total,
28 studies met the selection criteria.
2. Ecologic and cross-sectional studies

In the Seven Countries Study, men aged 40–59 years were enrolled and 16
cohorts were established in the seven countries (Kromhout et al., 1995). Risk
factors of IHD were measured and dietary information was collected from
small random samples of the cohort participants (Kromhout et al., 1995). A
positive (ecologic) correlation between the average percentage of energy
intake obtained from trans 18:1n–9 and age-adjusted mortality rate from IHD
was found (r = 0.78, p value < 0.001) (Kromhout et al., 1995).
In the cross-sectional study by Bolton-Smith et al. (1996), intakes of TFA
were statistically not significantly associated with IHD after control for dietary
and non-dietary IHD risk factors among women who had undiagnosed IHD at
baseline; for total TFA the odds ratio (OR) for the highest compared with the
lowest quintile was 1.36 (95% confidence interval, CI: 0.94–1.98, p value for
trend = 0.11); for ruminant TFA the OR for highest compared with lowest
quintile was 0.96 (95% CI: 0.58–1.59, p value for trend = 0.78); and for
industrial TFA the OR for highest compared with lowest quintile was 1.26
(95% CI: 0.92–1.72, p value for trend = 0.11). Intake of ruminant TFA was
negatively associated with IHD after control for dietary and non-dietary IHD
risk factors among men who had undiagnosed IHD at baseline (OR for highest
compared with lowest quintile, OR = 0.65 (95% CI: 0.41–1.04, p value for
trend = 0.02) (Bolton-Smith et al., 1996). Intakes of total and industrial TFA
were not associated with IHD; for total TFA the OR for highest compared with
lowest quintile was 0.87 (95% CI: 0.60–1.24, p value for trend = 0.48); for
industrial TFA the OR for highest compared with lowest quintile was 1.08
(95% CI: 0.79–1.48, p value for trend = 0.99) (Bolton-Smith et al., 1996).
3. Case-control studies
In total, 18 case-control studies met the selection criteria; one study using direct
measures of dietary TFA intakes (Ascherio et al., 1994), 15 studies using
biomarkers of TFA intakes (Aro et al., 1995; Baylin et al., 2003; Colon-Ramos
et al., 2006; Dlouhy et al., 2003; Ghahremanpour et al., 2008; Harris et al.,
2007; Lemaitre et al., 2002; Lemaitre et al., 2006; Pedersen et al., 2000; Roberts
et al., 1995; Siguel & Lerman, 1993; Sun et al., 2007b; Sun et al., 2007c; Thomas
& Scott, 1981; Thomas et al., 1983a), and two studies using both direct measures
of dietary TFA intakes and biomarkers of TFA intakes (Clifton et al., 2004;
Lopes et al., 2007). Three of the case-control studies were nested case-control
studies (Lemaitre et al., 2006; Sun et al., 2007b; Sun et al., 2007c). Two of these
studies (Sun et al., 2007b; Sun et al., 2007c) were nested in the Nurses’ Health
Study and therefore these two studies cannot be considered independent of the
follow-up studies by Willett et al. (1993), Hu et al. (1997), Hu et al. (1999) and
Oh et al. (2005) also based on data from the Nurses’ Health Study (described
below). The studies by Baylin et al. (2003) and Colon-Ramos et al. (2006) were
partly based on the same participants and therefore these studies cannot be
considered independent. In the 2003 study (Baylin et al., 2003), the investigators
reported results on adipose tissue total TFA; trans 16:1n–7; total trans 18:1; and
total trans 18:2 before a reduction of TFA in foods. In the 2006 study (Colon-
Ramos et al., 2006), the investigators reported results on adipose tissue total
TFA, total trans 18:1, and total trans 18:2 before and after a reduction of TFA in
foods. For this review, the 2003 study (Baylin et al., 2003) was selected for
results on adipose tissue trans 16:1n–7 and the 2006 study (Colon-Ramos et al.,
2006) was selected for results on adipose tissue total TFA, total trans 18:1, and
total trans 18:2 before and after a reduction of TFA in foods. Characteristics of the
case-control studies included are shown in Tables 1 and 2. Cases and controls
were sampled in two independent time periods before and after a reduction of
TFA in foods. The results from the two different time periods can be considered
independent and are referred to as two separate studies.
Table 1. Case-control studies investigating the association between intake of trans fatty acids and the risk of ischaemic heart disease.
260
Characteristics of participants Exposure Results

First author Number Sex Age, y Mea- Expres- Variation Description Schematic
(year) (mean sure sed (mean±SD)
±SD)
OR (95% CI) Variabes adjusted Total R-TFA I-TFA
for in analysis TFA
Ascherio 239 cases F&M 58±10 FFQ g/d Total TFA: Total TFA: Sex, age, BMI, family ↑‡ →‡ ↑
(1994) of first AMI 4.7±2.6 Q1(median)=1.7: history of IHD,
(no history OR=1.00 (reference) physical activity,
of hyper- Q2(median)=2.5: smoking status,
choleste- OR=0.63(0.31–1.30) alcohol consumption,
rolaemia, Q3(median)=3.4: intakes of SFA, MUFA,
diabetes, or OR=1.03 (0.53–2.00 linoleic acid, and chole-
angina Q4(median)=4.5: sterol, and hypertension;
pectoris) OR=1.35 (0.69–2.63) further adjustment for
Q5(median)=6.5: education, occupation,
OR=2.02 (1.03–3.93) marital status, personal-
p value for trend=0.002 ity type, use of multi-
vitamin supplement,
R-TFA: R-TFA:
use of aspirin, and
NA Q1(median)=0.5:
intakes of fibre, caro-
OR=1.00 (reference)
tene, vitamin E, and
Q2(median)=0.7:
vitamin C did not
OR=0.64 (0.33–1.23)
change the results of
Q3(median)=1.0:
intake of total TFA
OR=0.87 (0.46–1.66)
Q4(median)=1.2:
OR=0.86 (0.43–1.71)
Q5(median)=1.8:
OR=1.23 (0.60–2.50)
p value for trend=0.09
I-TFA: I-TFA:
NA Q1(median)=0.8:
OR=1.00 (reference)
Q2(median)=1.6:
OR=1.46 (0.74–2.88)
Q3(median)=2.3:
OR=1.16 (0.57–2.34)
Q4(median)=3.3:
OR=1.26 (0.61–2.61)
Q5(median)=5.0:
OR=3.05 (1.48–6.31)
282 controls 57±10 Total TFA:

matched with 3.8±2.0
the cases for
sex and age R-TFA:
(no history NA
of hyper-
choleste-
rolaemia, I-TFA:
diabetes, NA
angina pect-
oris, or AMI)
Cases g/d NA Total TFA: ↑ → ↑
(energy- Q1(median)=NA:
adjusted) OR=1.00 (reference)
Q2(median)=NA:
EPIDEMIOLOGICAL STUDIES ON INTAKE OF TRANS FATTY ACIDS
OR=0.81 (0.42–1.57)
Q3(median)=NA:
OR=0.40 (0.19–0.83)
Q4(median)=NA:
261
OR=0.72 (0.36–1.48)
Table 1. continued
262

±SD)
for in analysis TFA
Q5(median)=NA:
OR=2.03 (0.98–4.22)
R-TFA:
Q1(median)=NA:
OR=1.00 (reference)
Q2(median)=NA:
OR=1.17 (0.59–2.34)
Q3(median)=NA:
OR=1.12 (0.53–2.34)
Q4(median)=NA:
OR=1.00 (0.45–2.21)
Q5(median)=NA:
OR=1.02 (0.43–2.41)
I-TFA:
Q1(median)=NA:
OR=1.00 (reference)
Q2(median)=NA:
OR=0.74 (0.38–1.46)
Q3(median)=NA:
OR=0.43 (0.21–0.90)
Q4(median)=NA:
OR=0.63 (0.30–1.32)
Q5(median)=NA:
OR=1.94 (0.93–4.04)
Controls NA
Clifton 79 cases F&M 56±13 FFQ g/d NA Total TFA: Intakes of energy and SFA →
(2004) of first IHD Q1(median)=1.6:
(no history OR=1.00 (reference)
of hyper Q2(median)=2.4:
triglyce- OR=1.37 (95% CI NA)
ridaemia, Q3(median)=3.0:
hyper- OR=1.51 (95% CI NA)
cholesterol- Q4(median)=3.7:
aemia, or OR=0.81 (95% CI NA)
diabetes) Q5(median)=5.5:
OR=0.98 (95% CI NA)
p value for trend=NS
174 controls 56±11 NA
matched with
the cases for
sex, age, and
postal code
(no history of
hypertriglyce-
ridaemia,
hyperchole-
sterolaemia,
diabetes, or
IHD)
263
Table 1. continued.
264

±SD)
for in analysis TFA
Lopes (2007) 297 cases of M 57±11 FFQ g/d NA Total TFA: Age, BMI, family →
first AMI Q1(median)=0.3: history of AMI,
OR=1.00 (reference) education, physical
Q2(median)=0.6: activity, smoking
OR=0.79 (0.47–1.33) status, and intake of
Q3(median)=1.2: energy; further
OR=0.66 (0.38–1.14) adjustment for
Q4(median)=1.4: dyslipidaemia, hyper-
OR=0.81 (0.48–1.37) tension, diabetes,
p value for trend=0.34 and angina pectoris
did not change the results
(no history
of AMI)
AMI, acute myocardial infarction; BMI, body mass index; CI, confidence interval; F, Female; FFQ, food frequency questionnaire; I-TFA, industrial trans fatty acids;
M, Male; MUFA, monounsaturated fatty acids; NA, not available; NS, not statistically significant; OR, odds ratio; Q, quintile/quartile; R-TFA, ruminant trans fatty
acids; SFA, saturated fatty acids; SD, standard deviation; TFA, trans fatty acids.
†↑ Statistically significant positive association.
‡→ No statistically significant association.
Table 2. Case-control studies investigating the association between biomarkers of trans fatty acids and the risk of ischaemic heart disease.
Characteristics of participant Exposure Results

First author Sex Age, y Mea- Expre- Variation Description Schematic
(year), number (mean sure ssed (mean±SD)
±SD) OR (95% CI) Variables adjusted Total
for in analysis TFA Specific TFA isomers
Total Total Total
trans trans trans
16:1† 18:1 18:2
Thomas (1981)
136 cases of IHD M 53±NA Subcu- % of FA NA OR NA; total TFA was not →
death taneous statistically different between
AT (abdo- cases and controls
minal)
95 controls of death 52±NA NA
from unrelated causes
matched with the cases
for sex, age, and area
Thomas(1983a)
136 cases of M 53±NA Subcu- % of FA NA OR NA; total trans 16:1 was ↑‡ →§
IHD death taneous significantly higher in cases
AT (abdo- than in controls (p value=
minal) < 0.005); total trans 18:1
was not statistically different
between cases and controls
95 controls of death 52±NA NA
from unrelated causes
matched with the cases
for sex, age, and area

Siguel (1993)
40 cases of IHD F&M 31–72 Plasma % of FA Trans OR NA; trans 16:1, was Plasma SFA and PUFA ↑
16:1, n–7: positively associated with
265
0.40±0.14 IHD
Table 2. continued
266

Total Total Total
trans trans trans
16:1† 18:1 18:2
54 controls 25–65 Trans

16:1, n–7:
0.31±0.07
Cases mg/dl NA OR NA; trans 16:1, n–7 was Plasma SFA and PUFA ↑
positively associated with IHD
Controls NA
Aro (1995)
671 cases of first M 55±9 Subcu- % of FA NA Total trans 18:1: Age, BMI, smoking →
AMI taneous AT Q1(median)=0.5: status, and centre;
(buttock) OR=1.00 (reference) further adjustment for
Q2(median)=1.3: AT oleic acid, linoleic
OR=0.68 (0.41–1.13) acid, and arachidonic
Q3(median)=1.8: acid did not change
OR=1.05 (0.63–1.75) the results

Q4(median)=2.5:
OR=0.97 (0.56–1.67)
p value for trend=NS
Total trans 18:1(after exclusion
of participants from Spanish
centres):
Q1(median)=1.1:
OR=1.00 (reference)
Q2(median)=1.6:
OR=1.16 (0.79–1.71)
Q3(median)=2.0:
OR=1.53 (1.02–2.28)
Q4(median)=2.6:
OR=1.44 (0.94–2.20)
717 controls matched 53±9 NA
with the cases for age
(no history of AMI)
Roberts (1995)
66 cases of SCD M 55±8 Subcu- % of FA Total TFA: Total TFA: Age, smoking status, → → →
(no history of IHD) taneous 2.7±0.6 Q1=≤2.3: AT oleic acid and
AT (abdo- OR=1.00 (reference) linoleic acid, hyper-
minal) Q2=2.3–2.7: tension, and diabetes
OR=1.35 (0.60, 3.06)
Q3=2.7–3.0:
OR=1.25 (0.53, 2.95)
Q4=3.0–3.4:
OR=0.85 (0.32, 2.21)
Q5=≥3.4:
OR=0.63 (0.22, 1.84)
p value for trend=NA
Total Total trans 18:1:
trans 18:1: Q1=<=1.8:
2.1±0.6 OR=1.00 (reference)
Q2=1.8–2.1:
OR=0.96 (0.42, 2.21)
Q3=2.1–2.4:
OR=1.14 (0.48, 2.70)
Q4=2.4–2.7:
OR=0.61 (0.24, 1.56)
Q5=≥2.8:
OR=0.59 (0.19, 1.83)

trans 18:2: Q1=≤0.5:
267
0.6±0.2 OR=1.00 (reference)

Table 2. continued
268

Total Total Total
trans trans trans
16:1† 18:1 18:2
Q2=0.5–0.5:
OR=1.26 (0.49–3.21)
Q3=0.6–0.6:
OR=1.05 (0.40–2.78)
Q4=0.6–0.7:
OR=1.37 (0.53–3.56)
Q5=≥0.7:
OR=0.99 (0.35–2.84)
286 controls matched 57±7 Total TFA:
with the cases for sex 2.9±0.7
and age Total trans18:1:
2.3±0.7
Total trans 18:2:
0.6±0.2
Pedersen (2000)
100 cases of first F&M 62±8 Subcu- % of FA NA Total TFA: Sex, age, waist to hip →
AMI (no history taneous AT Q1(cut-off)=<3.4: ratio, family history of
of diabetes) (buttock) OR=1.00 (reference) IHD, smoking status,
Q2(cut-off)=>3.4: and AT linoleic acid;
OR=1.04 (0.32–3.37) further adjustment
Q3(cut-off)=>3.8: for AT α-linolenic
OR=1.07 (0.33–3.48) acid did not alter
Q4(cut-off)=>4.4: the associations
OR=0.54 (0.14–2.07)
Q5(cut-off)=>4.8:
OR=1.49 (0.47–4.69)
with the cases for sex,
age, and geographical
location (no history
of diabetes or AMI)
Lemaitre (2002)
179 cases of SCD F&M 60±10 Erythro % of FA Total TFA: Total TFA: Age, weight, height, ↑ → ↑
(nohistory of -cytes 2.1±0.6 IQ range=0.9: family history of AMI
IHD OR=1.47 (1.01–2.13) or sudden death,
Total Total trans 18:1: education, physical
trans 18:1: IQ range=0.8: activity, smoking
1.7±0.5 OR=0.77 (0.48–1.24) status, intake of SFA,
Total Total trans 18:2: hypertension, and
trans 18:2: IQ range=0.1: diabetes; trans 18:1
0.2±0.1 OR=3.05 (1.71–5.44) and trans 18:2 were
also adjusted for erythro-
cyte long-chain n-3
PUFA and were assessed
simultaneously
with the cases for 2.0±0.5
sex and age (no Total trans 18:1:
history of IHD) 1.6±0.5
Total trans 18:2:
0.2±0.1
269
Table 2. continued
270

Total Total Total
trans trans trans
16:1† 18:1 18:2
Baylin (2003)#
482 cases of first F&M 57±10 Subcuta- % of FA NA Trans 16:1, n–7: Income, years living ↑
AMI (no history neous AT Q1(median)=0.0: in the house, physical
of CVD) (buttock) OR=1.00 (reference) activity, smoking status,
Q2(median)=0.1: alcohol consumption,
OR=1.57 (0.83–2.98) AT α-linolenic acid,
Q3(median)=0.1: intakes of energy, SFA,
OR=1.39 (0.73–2.66) and vitamin E, hyper-
Q4(median)=0.1: tension, and diabetes
OR=1.34 (0.65–2.79) (conditioned on the
Q5(median)=0.1: matching factors: sex,
OR=2.58 (1.22–5.43) age, and area of residence);
p value for trend=0.025 further adjustment for
BMI, waist to hip ratio,
use of multivitamin
supplement, and intakes

of folate and fibre, did not
change the results
with the cases for
sex, age, and area of
residence (no history
of AMI)
Dlouhy (2003)
34 cases of IHD F&M 64±7 Subcuta- % of FA Total TFA: OR NA; total TFA (p value → →
neous AT 2.9±1.2 =NS) and total trans 18:1
(site NA) Total trans 18:1 (p value=0.05) were not
2.3±1.1 statistically different between
cases and controls
46 controls (no 61± 9 Total TFA:
history of IHD) 2.6±0.9
Total trans 18:1:
2.0±0.8
Clifton (2004)
79 cases of first F&M 57±12 Subcuta- % of FA Total OR NA; in multiple analyses Employment, intakes →
IHD (no history of neous AT trans 16:1: total trans 16:1 was not an of energy and SFA, hyper-
hypertriglyce- (abdo- 0.1±0.1 independent predictor of tension, TG, cholesterol,
ridaemia, hyper- minal) AMI and date of biopsy; all
cholesterolaemia, ATFA were assessed
or diabetes) simultaneously
167 controls 57±11 Total trans 16:1
matched with the 0.1±0.1
cases for sex, age,
and postal code (no
history of hypertri-
glyceridaemia, hyper-
cholesterolaemia,
diabetes, or IHD)
Colon-Ramos
(2006) 1994–99#
477 cases of first F&M 57±10 Subcuta- % of FA NA Total TFA: Income, physical ↑ → ↑
AMI (no history of neous AT Q1(median)=1.9: activity, smoking status,
CVD) (buttock) OR=1.00 (reference) alcohol consumption, AT
Q2(median)=2.5: α-linolenic acid, intakes
OR=1.37 (0.80–2.35) of energy, SFA, and

Q3(median)=3.0: vitamin E, hypertension,
OR=1.91 (1.08–3.37) and diabetes (conditioned
Q4(median)=3.6: on the matching factors:
271
OR=1.86 (1.04–3.32) sex, age, and area of

Table 2. continued
272

Total Total Total
trans trans trans
16:1† 18:1 18:2
Q5(median)=4.4: residence); further adjust-

OR=3.28 (1.68–6.42) ment for weight, height,
p value for trend=<0.001 waist to hip ratio,
Total trans18:1: education, other AT FA,
Q1(median)=0.9: and intakes of protein,
OR=1.00 (reference) cholesterol, fibre, folate,
Q2(median)=1.3: margarine, fish, and
OR=1.84 (1.09–3.11) dairy products did not
Q3(median)=1.6: change the results
OR=1.54 (0.87–2.71)
Q4(median)=2.0:
OR=2.13 (1.21–3.74)
Q5(median)=2.5:
OR=1.75 (0.97–3.15)
Total trans18:2:
Q1(median)=0.8:
OR=1.00 (reference)
Q2(median)=1.0:
OR=1.04 (0.60–1.80)
Q3(median)=1.2:
OR=2.16 (1.18–3.96)
Q4(median)=1.5:
OR=2.86 (1.50–5.46)
Q5(median)=2.0:
OR=4.76 (2.24–10.11)
p value for trend=≤0.001
matched with the
cases for sex, age,
and area of residence
(no history of AMI)
Colon-Ramos
(2006) 2000–03#
1,320 cases of first F&M 59±11 Subcuta- % of FA NA Total TFA: Income, physical → → →
AMI (no history of neous AT Q1(median)=1.8: activity, smoking status,
of CVD) (buttock) OR=1.00 (reference) alcohol consumption,
Q2(median)=2.3: hypertension, AT
OR=0.78 (0.58,1.04) α-linolenic acid, intakes
Q3(median)=2.6: of energy, SFA, and
OR=1.03 (0.80,1.43) vitamin E, and diabetes
Q4(median)=2.9: (conditioned on the
OR=0.88 (0.65,1.18) matching factors: sex,
Q5(median)=3.4: age, and area of resi-
OR=1.03 (0.75,1.42) dence); further adjustment
p value for trend=0.65 for weight, height, waist
Total trans 18:1: to hip ratio, education,
Q1(median)=0.9: other AT FA, and intakes
OR=1.00 (reference) of protein, cholesterol,
Q2(median)=1.1: fibre, folate, margarine,
OR=0.94 (0.71,1.24) fish, and dairy products
Q3(median)=1.3: did not change the results
OR=0.92 (0.69,1.23)
Q4(median)=1.5:
OR=0.97 (0.72,1.30)
Q5(median)=1.9:
OR=1.02 (0.75,1.37)
Total trans 18:2:

Q1(median)=0.7:
OR=1.00 (reference)
Q2(median)=0.9:
273
OR=1.09 (0.81–1.46)
Table 2. continued
274

Total Total Total
trans trans trans
16:1† 18:1 18:2
Q3(median)=1.1:
OR=1.10 (0.81–1.50)
Q4(median)=1.2:
OR=1.05 (0.75–1.46)
Q5(median)=1.4:
OR=1.15 (0.80–1.64)
1,320 controls 59±11 NA
matched with the
cases for sex, age,
and area of residence
(no history of AMI)
Lemaitre (2006)
214 cases of fatal F&M 77±6 Plasma % of FA Total TFA: Total TFA: Education, smoking → → ↓** ↑
IHD NA IQ range =1.4:OR=0.94 status, plasma long-chain
(0.65–1.34) n–3 PUFA, diabetes,

Total Total trans 16:1: congestive heart failure,
trans 16:1: IQ range=0.1:OR=0.95 and stroke (conditioned
0.3±0.1 (0.64–1.42) on the matching factors:
sex, age, CVD, date of
Total Total trans 18:1: blood draw, and clinic
trans 18:1: Proportions above compared site); trans 18:1 and
2.0±0.7 with proportions below the trans 18:2 were assessed
20th percentile:OR=0.34 simultaneously
(0.18–0.63)
trans 18:2: IQ range=0.1:
0.3±0.1 OR=1.68 (1.21–2.33)
matched with the NA
cases for sex, age, Total trans 16:1:
history of CVD, date 0.3±0.1
of blood draw, clinic Total trans 18:1:
site, entry cohort, 2.0±0.8
and follow-up Total trans 18:2:
duration ≥ that of 0.3±0.1
the case person
Harris (2007)
94 cases of ACS F&M 46±5 Whole % of FA Total TFA: Total TFA: Sex, race, age, BMI, → → →
blood 2.1±0.7 OR(1 SD increase)=0.87 education, smoking
(0.55–1.37) status, alcohol con-
Total Total trans 18:1: sumption, TG, LDL-
trans 18:1: OR(1 SD increase)=0.80 cholesterol, HDL-
1.7±0.6 (0.51–1.25) cholesterol, diabetes,
Total Total trans 18:2: and PCE
trans 18:2: OR(1 SD increase)=1.41
0.4±0.2 (0.84–2.39)
cases for sex, race, Total trans 18:1:
and age 1.6±0.8
Total trans 18:2:
0.4±0.2
Lopes (2007)
49 cases of first M 55±9 Subcuta- % of FA Total TFA: Total TFA: Age, BMI, family history ↓
AMI neous AT 0.6±0.1 T1:OR=1.00 (reference) of AMI, education,
(buttock) T2:OR=0.25 (0.06–0.96) physical activity, and

T3:OR=0.04 (0.01–0.32) intake of energy; further
p value for trend=0.001 adjustment for smoking
status did not change the
275
results
Table 2. continued
276

Total Total Total
trans trans trans
16:1† 18:1 18:2
49 controls (no 61±11 Total TFA:

history of AMI) 0.8±0.4
Sun (2007b)
166 cases of IHD F 61± 6 Erythro- % of FA Total TFA: Total TFA: BMI, menopausal ↑ ↑ ↑
(no history of CVD) cytes 1.8± 0.4 Q1(mean)=1.2: status, postmenopausal
OR=1.0 (reference) hormone use, family
Q2(mean)=1.5: history of AMI, physical
OR=1.6 (0.7–3.6) activity, alcohol con-
Q3(mean)=1.7: sumption, erythrocyte
OR=1.6 (0.7–3.4) n–6 PUFA and long-
Q4(mean)=2.2: chain n–3 PUFA, hyper-
OR=3.3 (1.5–7.2) tension, hypercholesterol-
p value for trend=<0.01 aemia, and diabetes
Total Total trans 18:1: (conditioned on the
trans 18:1: Q1(mean)=0.8: matching factors: age,
1.3±0.3 OR=1.0 (reference) smoking status, fasting

Q2(mean)=1.0: status at blood draw,
OR=1.1 (0.5–2.4) and date of blood draw);
Q3(mean)=1.2: further adjustment for
OR=1.3 (0.6–2.7) use of vitamin E
Q4(mean)=1.6: supplement, erythrocyte
OR=3.1 (1.5–6.7) SFA, MUFA, and
p value for trend=<0.01 α-linolenic acid, and
Total Total trans 18:2: intakes of fibre, folate,
trans 18:2: Q1(mean)=0.3: fruits, and vegetables did
0.4±0.1 OR=1.0 (reference) not change the results
Q2(mean)=0.3:
OR=1.5 (0.7–3.4)
Q3(mean)=0.4:
OR=2.5 (1.1–5.7)
Q4(mean)=0.5:
OR=2.8 (1.2–6.3)
p value for trend=<0.01
cases for age, smoking Total trans 18:1:
status, fasting status 1.2±0.4
at blood draw, and Total trans 18:2:
date of blood draw 0.4±0.1
(no history of CVD)
Sun (2007c)
166 cases of IHD (no F 61±6 Plasma % of FA Trans Trans 16:1, n–7: BMI, menopausal status, →
history of CVD) 16:1 n–7: T1(mean)=0.1: postmenopausal hormone
0.1±0.0 OR=1.00 (reference) use, family history of
T2(mean)=0.2: AMI, physical activity,
OR=0.79 (0.46–1.36) alcohol consumption, use
T3(mean)=0.2: of aspirin, plasma linoleic
OR=0.79 (0.44–1.42) acid and total TFA,
p value for trend=NS hypertension, hyper-
cholesterolaemia, and
diabetes (conditioned on
the matching factors: age,
smoking status,fasting
status at blood draw, and
date of blood draw);

further adjustment for
intakes of fibre, calcium,
vitamin D, folate, fruits,
and vegetables did not
277
change the results

Table 2. continued
278

Total Total Total
trans trans trans
16:1† 18:1 18:2
327 controls 60±6 Trans 16:1 n–7

cases for age, smoking
status, fasting status at
blood draw, and date
of blood draw (no
history of CVD)
Cases Erythro- Trans Trans 16:1, n–7: BMI, menopausal status, →
cytes 16:1 n–7: T1(mean)=0.1: postmenopausal hormone
0.1±0.0 OR=1.00 (reference) use, family history of
T2(mean)=0.1: AMI, physical activity,
OR=0.79 (0.44–1.43) alcohol consumption,
T3(mean)=0.2: use of aspirin, erythrocyte
OR=0.98 (0.53–1.83) linoleic acid and total
p value for trend=NS TFA, hypertension, hyper-
cholesterolaemia, and
diabetes (conditioned on
the matching factors: age,
smoking status, fasting
status at blood draw,
and date of blood draw);
further adjustment for
intakes of fibre, calcium,
vitamin D, folate,
fruits, and vegetables
did not change the results
Controls Trans 16:1 n–7:
0.1±0.0
Ghahremanpour (2008)
105 cases of IHD F&M 54±9 Subcuta- % of FA Total TFA: Total TFA: Smoking status, AT ↑ → ↑ →
(no history of hyper- neous AT 8.9±2.5 IQ range=13.7: PUFA, hypertension,
cholesterolaemi, (buttock) OR=1.41 (1.0–1.8) and TG
diabetes, or CVD) Total Total trans 16:1:
0.5±0.3 OR=0.80 (0.3–1.9)
6.6±2.1 OR=1.32 (1.0–1.8)
Total Total trans18:2:
1.8±0.8 OR=1.85 (0.6–4.8)
with the cases for 8.0±2.2
sex and age (no Total trans 16:1:
history of hyper- 0.5±0.3
cholesterolaemi, Total trans 18:1:
diabetes, or CVD) 6.0±1.8
Total trans 18:2:
1.5±0.7
ACS, acute coronary syndrome; AMI, acute myocardial infarction; AT, adipose tissue; BMI, body mass index; CE, cholesterol esters; CI, confidence interval; CVD,
cardiovascular disease; F, Female; FA, fatty acids; FFQ, food frequency questionnaire; HDL, high-density lipoprotein; IQ, interquintile; LDL, low-density lipoprotein; M,
Male; MUFA, monounsaturated fatty acids; NA, not available; NS, not statistically significant; OR, odds ratio; PCE, percutaneous coronary intervention; PUFA,
polyunsaturated fatty acids; Q, quintile/quartile; SCD, sudden cardiac death; SFA, saturated fatty acids; SD, standard deviation; T, tertile; TG, triacylglycerol; TFA, trans fatty
acids.
† Trans 16:1n–7 or the sum of trans 16:1n–7 and 16:1n–9.
‡ ↑ Statistically significant positive association or statistically significant higher in cases than in controls.
§ → No statistically significant association or no statistically significant difference between cases and controls.
# The studies by Baylin et al. (2003) and Colon-Ramos et al. (2006) were partly based on some of the same participants and therefore these studies cannot be considered
independent. In the study by Baylin et al. (2003), the investigators reported results on AT total TFA, trans 16:1n–7, total trans 18:1, and total trans 18:2 before a reduction
of TFA in foods. In the study by Colon-Ramos et al. (2006), the investigators reported results on AT total TFA, total trans 18:1, and total trans 18:2 before and after a reduction
of TFA in foods. For this review, the study by Baylin et al. (2003) was selected for results on trans 16:1n–7 and the study by Colon-Ramos et al. (2006) was selected for total
TFA, total trans 18:1, and total trans 18:2 before and after a reduction of TFA in foods. Cases and controls were sampled in two independent time periods before and after
a reduction of TFA in foods. The results from the two different time periods can be considered independent and are referred to as two separate studies (Colon-Ramos (2006)
279
1994–99 and Colon-Ramos (2006) 2000–03).

**↓ Statistically significant negative association or statistically significant lower in cases than in controls.
a. Intake of TFA and risk of IHD

Three studies investigated associations between the intake of total TFA and the
risk of IHD (Ascherio et al., 1994; Clifton et al., 2004; Lopes et al., 2007)
(Table 1). One study showed a significant positive association between intake of
total TFA and risk of acute myocardial infarction (Ascherio et al., 1994) and two
studies showed no significant associations between intake of total TFA and risk
of IHD (Clifton et al., 2004; Lopes et al., 2007) (Table 1). One study investi-
gated the associations between the intakes of TFA obtained from ruminant and
industrial sources and the risk of acute myocardial infarction (Ascherio et al.,
1994) (Table 1). In that study, the energy-adjusted intake of ruminant TFA was
not significantly associated with the risk of acute myocardial infarction whereas
the energy-adjusted intake of industrial TFA was significantly associated with
the risk of myocardial infarction (Table 1). However, the study indicated a trend
for a positive association between energy-adjusted intake of ruminant TFA and
risk of acute myocardial infarction (Table 1).
Intake of saturated fatty acids is a risk factor of IHD at least partially
mediated by their effect on LDL-cholesterol concentration in plasma. The
ruminant TFA intake is probably directly correlated to saturated fatty acids
intake in most studies. Thus, by controlling for saturated fatty acids intake in
analyses of ruminant TFA intake and risk of IHD, the models address the effect
of ruminant TFA intake independent of this factor. In the studies by Ascherio
et al. (1994) and Clifton et al. (1994), analyses were controlled for intake of
saturated fatty acids (Table 1).
Whether intake of TFA should be expressed as crude intake (absolute intake)
or relative to energy intake (energy-adjusted intake) depends on the mecha-
nisms behind a potential association between TFA intake and IHD. As indicated,
the mechanisms are not fully explored. In the study by Ascherio et al. (1994),
both approaches were used (Table 1).
b. Adipose tissue TFA and risk of IHD

In total, 12 case-control studies investigated the association between adipose
tissue TFA and risk of IHD (Aro et al., 1995; Baylin et al., 2003; Clifton et al.,
2004; Colon-Ramos et al., 2006; Dlouhy et al., 2003; Ghahremanpour et al.,
2008; Lopes et al., 2007; Pedersen et al., 2000; Roberts et al., 1995; Thomas &
Scott, 1981; Thomas et al., 1983a) (Table 2). Two studies showed significant
positive associations between adipose tissue total TFA and risk of IHD (Colon-
Ramos et al., 2006; Ghahremanpour et al., 2008), one study showed a significant
negative association (Lopes et al., 2007), and five studies showed no signifi-
cant associations (Colon-Ramos et al., 2006; Dlouhy et al., 2003; Pedersen
et al., 2000; Roberts et al., 1995; Thomas & Scott, 1981) (Table 2).
Regarding the specific TFA, two studies showed significant positive asso-
ciations between adipose tissue total trans 16:1 and risk of IHD (Baylin et al.,
2003; Thomas et al., 1983a) and two studies showed no significant associations
(Clifton et al., 2004; Ghahremanpour et al., 2008) (Table 2). Among the two
studies which showed no significant associations, one study, however, indi-
cated a trend for a negative association between adipose tissue trans 16:1 and
risk of IHD (Ghahremanpour et al., 2008) (Table 2). One study showed a
significant positive association between adipose tissue total trans 18:1 and risk
of IHD (Ghahremanpour et al., 2008) and six studies showed no significant
associations (Aro et al., 1995; Colon-Ramos et al., 2006; Dlouhy et al., 2003;
Roberts et al., 1995; Thomas et al., 1983a) (Table 2). However, among the six
studies which showed no significant associations, three studies indicated
trends for positive associations between adipose tissue trans 18:1 and risk of
IHD (Aro et al., 1995; Colon-Ramos et al., 2006; Dlouhy et al., 2003) and one
study indicated a trend for a negative association (Roberts et al., 1995)
(Table 2). In the study by Aro et al. (1995), the trend for a positive association
between adipose tissue total trans 18:1 appeared after exclusion of participants
from the two Spanish centres for whom proportions of adipose tissue TFA were
very low compared with participants from other countries. One study showed
a significant positive association between adipose tissue total trans 18:2 and
risk of acute myocardial infarction (Colon-Ramos et al., 2006) and three
studies showed no significant associations between adipose tissue total trans
18:2 and risk of IHD (Colon-Ramos et al., 2006; Ghahremanpour et al., 2008;
Roberts et al., 1995) (Table 2). Among the three studies which showed no
significant associations, two studies, however, indicated trends for positive
associations between adipose tissue trans 18:2 and risk of IHD (Colon-Ramos
et al., 2006; Ghahremanpour et al., 2008) (Table 2). In the study by Clifton
et al. (2004), adipose tissue trans 18:1n–7 was positively associated with risk
of IHD after control for dietary IHD risk factors, whereas other positional
isomers within trans 18:1 (18:1n–8 and 18:1n–9) were not significantly
associated with risk of IHD.
By controlling for other adipose tissue fatty acids in analyses of adipose
tissue TFA and risk of IHD, the models address the effects of the latter
independently of the former. One study reported results from multiple analyses
without and with control for other adipose tissue fatty acids (Pedersen et al.,
2000). In that study, investigating the association between total adipose tissue
TFA and risk of acute myocardial infarction the significant positive association
became non-significant after additional control for adipose tissue linoleic acid
(cis 18:2n–6) (Pedersen et al., 2000).
Five studies controlled for dietary variables (energy intake, saturated fatty
acids, or vitamin E) in multiple analyses of adipose tissue TFA and risk of IHD
(Baylin et al., 2003; Clifton et al., 2004; Colon-Ramos et al., 2006; Lopes
et al., 2007) (Table 2). By controlling for these dietary variables, the models
address the effect of adipose tissue TFA independent of these dietary variables.
However, as the proportion of adipose tissue TFA is expressed as percentage of
total adipose tissue fatty acids the proportion of TFA also depends indirectly on
for, example, the saturated fatty acids intake (i.e. because adipose tissue TFA
is expressed as percentage of total fatty acids, any increase or decrease in other
fatty acids will result in reciprocal changes in the TFA percentage).
In five studies reporting results from multiple analyses, the analyses in-
cluded control for plasma total cholesterol or diabetes mellitus which may be
considered as potential metabolic effects of intake of TFA (Baylin et al., 2003;
Clifton et al., 2004; Colon-Ramos et al., 2006; Roberts et al., 1995) (Table 2).
c. IHD risk and TFA in whole blood, plasma and erythrocytes

Six case-control studies investigated the association between IHD risk and the
TFA levels of whole blood, plasma, or erythrocytes (Harris et al., 2007;
Lemaitre et al., 2002; Lemaitre et al., 2006; Siguel & Lerman, 1993; Sun et al.,
2007b; Sun et al., 2007c) (Table 2). Two studies showed significant positive
associations between erythrocyte total TFA and risk of IHD (Lemaitre et al.,
2002; Sun et al., 2007b) and two studies showed no significant associations
between whole blood or plasma total TFA and risk of IHD (Harris et al., 2007;
Lemaitre et al., 2006) (Table 2).
Regarding the specific TFA, one study showed a significant positive associa-
tion between plasma trans 16:1n–7 and risk of IHD (Siguel & Lerman, 1993)
and two studies showed no significant associations between plasma or erythro-
cyte total trans 16:1 and risk of IHD (Lemaitre et al., 2006; Sun et al., 2007c)
(Table 2). Among the two studies which showed no significant associations,
one indicated a trend for a negative association between plasma or erythrocyte
trans 16:1n–7 and risk of IHD (Sun et al., 2007c) (Table 2). One study showed
a significant positive association between erythrocyte total trans 18:1 and risk
of IHD (Sun et al., 2007b), one study showed a significant negative association
between plasma total trans 18:1 and risk of fatal IHD (Lemaitre et al., 2006),
and two studies showed no significant associations between whole blood or
erythrocyte total trans 18:1 and risk of IHD (Harris et al., 2007; Lemaitre et al.,
2002) (Table 2). However, the two studies which showed no significant
associations indicated trends for negative associations between whole blood or
erythrocyte trans 18:1 and risk of IHD (Harris et al., 2007; Lemaitre et al.,
2002) (Table 2). Three studies showed significant positive associations be-
tween plasma or erythrocyte total trans 18:2 and risk of IHD (Lemaitre et al.,
2002; Lemaitre et al., 2006; Sun et al., 2007b) and one study showed no
significant association between whole blood total trans 18:2 and risk of acute
coronary syndrome (Harris et al., 2007) (Table 2). However, the study by
Harris et al. (2007) indicated a trend for a positive association between whole
blood trans 18:2 and risk of acute coronary syndrome (Table 2). In the study by
Lemaitre et al. (2002), specific erythrocyte positional isomers within trans
18:1 (18:1n–7, 18:1n–8 and 18:1n–9) were not significantly associated with
risk of sudden cardiac death after control for dietary and non-dietary IHD risk
factors. In the study by Sun et al. (2007b), specific erythrocyte position isomers
within trans 18:1 (18:1n–7, 18:1n–9 and 18:1n–12) and within trans 18:2
(18:2n–6tt, 18:2n–6ct and 18:2 n–6tc) were significantly positively associated
with risk of IHD after control for dietary and non-dietary IHD risk factors.
One study reported results from multiple analyses without and with control for
other fatty acids (Sun et al., 2007b). In that study, investigating the association
between erythrocyte total TFA, total trans 18:1, and total trans 18:2 and risk of
IHD, the significant positive associations became stronger after additional
control for erythrocyte n–6 PUFA and long-chain n–3 PUFA (Sun et al., 2007b).
In two studies, total trans 18:1 and total trans 18:2 were assessed simultaneously
(Lemaitre et al., 2002, 2006) (Table 2). In the study by Lemaitre et al. (2002),
erythrocyte total trans 18:1 was non-significantly positively associated with risk
of sudden cardiac death in analysis without control for erythrocyte total trans
18:2; however, when assessed simultaneously, erythrocyte total trans 18:1
became non-significantly negatively associated with risk of sudden cardiac
death. Erythrocyte total trans 18:2 was significantly positively associated with
risk of sudden cardiac death in analyses without and with control for erythrocyte
total trans 18:1 (Lemaitre et al., 2002). In the study by Lemaitre et al. (2006),
plasma total trans 18:1 was significantly negatively associated with risk of fatal
IHD in analyses without and with control for plasma total trans 18:2. Plasma total
trans 18:2 was non-significantly positively associated with risk of fatal IHD in
analysis without control for plasma total trans 18:1; however, when assessed
simultaneously, plasma total trans 18:2 became significantly positively associ-
ated with risk of fatal IHD (Lemaitre et al., 2006).
One study controlled for intake of saturated fatty acids in multiple analyses
of erythrocyte total TFA, total trans 18:1, and total trans 18:2 and risk of
sudden cardiac death (Lemaitre et al., 2002) (Table 2). By controlling for
intake of saturated fatty acids, the models address the effect of erythrocyte TFA
independent of saturated fatty acids intake. However, as the proportion of
erythrocyte TFA is expressed as percentage of total erythrocyte fatty acids the
proportion of TFA also depends indirectly on the saturated fatty acids intake.
In five studies reporting results from multiple analyses, the analyses in-
cluded control for LDL-cholesterol, HDL-cholesterol, hypercholesterolaemia
or diabetes mellitus which may be considered as potential metabolic effects of
intake of TFA (Harris et al., 2007; Lemaitre et al., 2002; Lemaitre et al., 2006;
Sun et al., 2007b; Sun et al., 2007c) (Table 2).
d. Effect modification by sex and age

Effect modification is an interaction among multiple possible cause-and-effect
relationships, where the estimate of the effect of one factor on a disease process
depends on other factors in the study. Two case-control studies on intake of
TFA and risk of IHD were carried out both among women and men (Ascherio
et al., 1994; Clifton et al., 2004) and one study was carried out among men only
(Lopes et al., 2007) (Table 1). In the study by Ascherio et al. (1994), no effect
Table 3. Follow-up studies investigating the association between the intake of trans fatty acids and the risk of ischaemic heart disease.
284
Characteristics Exposure Event Follow- Results

of participants up, y
First author Sex and Measure, Measure Number Description Schematic
(year), number Age, y Expressed,
(mean±SD) Variation RR† (95% CI) Variables adjusted Total R-TFA I-TFA
for in analysis TFA
Willett (1993) F, NA FFQ, g/d Non-fatal 356 8 R-TFA: Age, BMI, family →‡ ↑§
69,181 (no (energy- AMI or Q1:RR=1.00 history of AMI, meno-
history of adjusted), fatal IHD (reference) pausal status, post-
hypercholes- NA Q2:RR=0.76 menopausal hormone
terolaemia, (0.51–1.12) use, smoking status,
diabetes, stroke, Q3:RR=0.69 alcohol consumption,
angina pectoris, (0.43–1.10) use of multivitamin
or AMI) who Q4:RR=0.55 supplement, intakes
reported no (0.31–0.96) of energy, SFA,
change in margar- Q5:RR=0.59 MUFA, and linoleic
ine intake in (0.30–1.17) acid, and hypertension
previous 10 y p value for trend=0.23
I-TFA:
Q1:RR=1.00
(reference)
Q2:RR=1.43
(1.00–2.04)
Q3:RR=1.11
(0.74–1.66)
Q4:RR=1.39
(0.91–2.13)
Q5:RR=1.78
(1.12–2.83)
Ascherio (1996) M, 40–75 FFQ, g/d Non-fatal 734 6 Total TFA: Age, BMI, family →
43,757 (no (energy- AMI or Q1(median)=1.5: history of AMI,
history of diabetes, adjusted), fatal IHD RR=1.00 (reference) profession, physical
peripheral arterial NA Q2(median)=2.2: activity, smoking
disease, stroke, RR=1.12 (0.86–1.44) status, alcohol con-
angina pectoris, Q3(median)=2.7: sumption, intake of
transient ischaemic RR=1.12 (0.87–1.46) fibre, hypertension,
attack, coronary Q4(median)=3.3: and hypercholeste-
artery surgery, RR=1.12 (0.86–1.46) rolaemia
or AMI) Q5(median)=4.3:
RR=1.21 (0.93–1.58)
E%, NA Total TFA: Age, BMI, family →
RR(2 unit)=1.13 history of AMI,
(0.81–1.58) profession, physical
activity, smoking
status, alcohol con-
sumption, intakes of
energy, fat, and fibre,
hypertension, and
hypercholeste-
rolaemia
g/d Fatal 229 Total TFA: →
(energy- IHD Q1(median)=1.5:
adjusted), RR=1.00 (reference)
NA Q2(median)=2.2:
RR=1.63 (1.01–2.62)
Q3(median)=2.7:
RR=1.18 (0.71–1.96)
Q4(median)=3.3:
RR=1.59 (0.98–2.60)
Q5(median)=4.3:
285
Table 3. continued
286

for in analysis TFA
RR=1.41 (0.86–2.32)
E%, NA Total TFA: →
RR(2 unit)=0.93
(0.52–1.69)
Pietinen (1997) M, 50–69 FFQ, g/d Non-fatal 1,399 6 Total TFA: Age, BMI, education, →
21,930 smokers (energy- AMI or Q1(median)=1.3: physical activity,
(no history of adjusted), fatal IHD RR=1.00 (reference) smoking status, alcohol
diabetes, stroke, NA Q2(median)=1.7: consumption, intakes
exercise-related RR=1.10 (0.93–1.31) of energy and fibre,
chest pain, angina Q3(median)=2.0: blood pressure, and
pectoris, or AMI) RR=0.97 (0.82–1.16) treatment group;
Q4(median)=2.7: further adjustment
RR=1.07 (0.90–1.28) for intake of SFA,
Q5(median)=6.2: linoleic acid, and
RR=1.14 (0.96–1.35) cholesterol slightly

p value for trend=0.16 increased the RR
Fatal IHD 635 Total TFA: ↑ ↓# ↑
Q1(median)=1.8:
RR=1.00 (reference)
Q2(median)=2.4:
RR=1.05 (0.81–1.36)
Q3(median)=2.9:
RR=1.12 (0.87–1.45)
Q4(median)=3.5:
RR=0.90 (0.69–1.18)
Q5(median)=6.2:
RR=1.39 (1.09–1.78)
R-TFA:
Q1(median)=0.6:
RR=1.00 (reference)
Q2(median)=1.1:
RR=0.97 (0.75–1.25)
Q3(median)=1.5:
RR=0.91 (0.70–1.19)
Q4(median)=1.9:
RR=0.90 (0.69–1.17)
Q5(median)=2.5:
RR=0.83 (0.62–1.11)
I-TFA:
Q1(median)=0.1:
RR=1.00 (reference)
Q2(median)=0.4:
RR=0.87 (0.68–1.11)
Q3(median)=0.8:
RR=0.77 (0.60–1.01)
Q4(median)=1.6:
RR=0.94 (0.73–1.20)
Q5(median)=5.1:
RR=1.23 (0.97–1.55)
287
Table 3. continued
288

for in analysis TFA
Oomen (2001) M, 71±5 Dietary Non-fatal 98 10 Total TFA: Age, BMI smoking ↑ → →
667 (no history history, E%, AMI or RR(2 unit)=1.28 status, alcohol con-
of angina pectoris Total TFA: fatal IHD (1.01–1.61) sumption, use of
or AMI) 4.3±2.2 R-TFA: vitamin supplement,
R-TFA: RR(0.5 unit)=1.17 intakes of energy,
0.7± 0.2 (0.69–1.98) SFA, MUFA, PUFA,
I-TFA I-TFA (trans 18:1): cholesterol, and fibre
(trans 18:1): RR(0.5unit)=1.05
2.1±1.2 (0.94–1.17)
I-TFA I-TFA (other TFA):
(other TFA): RR(0.5unit)=1.07
1.6±1.4 (0.99–1.15)
Fatal IHD 49 Total TFA: →
RR(2 unit)=1.33
(0.96–1.86)
Oh (2005) F, NA FFQ, E%. Non-fatal 1,766 20 Total TFA: Age, BMI, family ↑
78,778 (no NA AMI or Q1(median)=1.3: history of AMI, meno-
history of hyper- fatal IHD RR=1.00 (reference) pausal status, post-
cholesterolaemia, Q2(median)=1.6: menopausal hormone
diabetes, CVD, RR=1.08 (0.92–1.26) use, physical activity,
or cancer) Q3(median)=1.9: smoking status,
RR=1.29 (1.09–1.53) alcohol consumption,
Q4(median)=2.2: use of multivitamin
RR=1.19 (0.99–1.44) supplement, use of
Q5(median)=2.8: vitamin E supplement,
RR=1.33 (1.07–1.66) use of aspirin, intakes
p value for trend=0.01 of energy, protein,
SFA, MUFA, PUFA,
α-linolenic, long-chain
n–3 PUFA, chole-
sterol , fibre, fruits,
and vegetables, and
hypertension
Jakobsen (2008) F&M, Food record, Non-fatal 374 18 R-TFA: Sex, age, family →
3,686 (no history 30–71 (80% g/d, AMI or RR(0.5 unit)=0.97 history of AMI,
of diabetes or central) R-TFA: fatal IHD (0.91–1.04) education, smoking
IHD) 90% central status, alcohol con-
range:0.6–3.6 sumption, intakes of
protein, SFA, MUFA,
PUFA, cholesterol,
fibre, and weighted
intake of foods con-
taining I-TFA, blood
pressure, and cohort
identification**
g/d(energy- R-TFA: Sex, age, BMI, family →
adjusted), RR(0.5 unit)=1.05 history of AMI,
NA (0.92–1.19) education, physical
activity, smoking
status, alcohol con-
sumption, intakes of
energy, protein, SFA,
MUFA, PUFA, chole-
sterol, fibre, and
289
290
Table 3. continued

for in analysis TFA
weighted intake of
foods containing
I-TFA, blood
pressure, and cohort
identification**
AMI, acute myocardial infarction; BMI, body mass index; CI, confidence interval; E%, percentage of energy intake; F, Female; FFQ, food frequency questionnaire;
I-TFA, industrial trans fatty acids; M, Male; MUFA, monounsaturated fatty acids; NA, not available; PUFA, polyunsaturated fatty acids; Q, quintile; R-TFA,
ruminant trans fatty acids; RR, relative risk; SD, standard deviation; SFA, saturated fatty acids; TFA, trans fatty acids.
† Relative association; in most studies hazard ratio.
‡ → No statistically significant association.
§ ↓ Statistically significant positive association.
# ↑ Statistically significant negative association.
** A variable was included if it changed the beta coefficient for the R-TFA variable 10% or more.
modification by sex was found. In total, 11 studies on biomarkers of intake of

TFA and risk of IHD were carried out both among women and men (Baylin
et al., 2003; Clifton et al., 2004; Colon-Ramos et al., 2006; Dlouhy et al.,
2003; Ghahremanpour et al., 2008; Harris et al., 2007; Lemaitre et al., 2002;
Lemaitre et al., 2006; Pedersen et al., 2000; Siguel & Lerman, 1993), two
studies were carried out among women only (Sun et al., 2007b; Sun et al.,
2007c) and five studies were carried out among men only (Aro et al., 1995;
Lopes et al., 2007; Roberts et al., 1995; Thomas & Scott, 1981; Thomas et al.,
1983a) (Table 2). Sex-specific results did not suggest sex differences in the
association between biomarkers of intake of TFA and risk of IHD (Table 2). In
the studies by Lemaitre et al. (2002) and Siguel & Lerman (1993), no effect
modification by sex was found.
Only one study investigated potential effect modification by age (Lemaitre
et al., 2002). In that study, investigating the association between erythrocyte
total TFA, total trans 18:1, and total trans 18:2 and risk of sudden cardiac death
no effect modification by age was found (Lemaitre et al., 2002).
4. Follow-up studies
Eight follow-up studies met the selection criteria (Ascherio et al., 1996; Hu
et al., 1997; Hu et al., 1999; Jakobsen et al., 2008; Oh et al., 2005; Oomen
et al., 2001; Pietinen et al., 1997; Willett et al., 1993). The studies by Willett
et al. (1993), Hu et al. (1997), Hu et al. (1999), and Oh et al. (2005) were based
on the same participants and therefore these studies cannot be considered
independent. In the 1993 study by Willett et al., the investigators reported
results on intakes of total TFA and TFA obtained from ruminant and industrial
sources after 8 years follow-up. In the Hu et al. 1997 and 1999, and Oh et al.
2005 studies, the investigators reported results on intake of total TFA after 14,
14, and 20 years follow-up, respectively. For this review, the 1993 paper by
Willett et al. was selected for results on intakes of TFA obtained from ruminant
and industrial sources, and the 2005 Oh et al. paper was selected for results on
intake of total TFA due to the long follow-up. Characteristics of the follow-up
studies included are shown in Table 3.

Four studies investigated the association between the intake of total TFA and the
risk of IHD (Ascherio et al., 1996; Oh et al., 2005; Oomen et al., 2001; Pietinen
et al., 1997) (Table 3). Two studies showed significant positive associations
between energy-adjusted intake of total TFA and risk of non-fatal and fatal IHD
(Oh et al., 2005; Oomen et al., 2001) and two studies showed no significant
associations between energy-adjusted intake of total TFA and risk of non-fatal
and fatal IHD (Ascherio et al., 1996; Pietinen et al., 1997) (Table 3). One study
showed a significant positive association between energy-adjusted intake of total
TFA and risk of fatal IHD (Pietinen et al., 1997) and two studies showed no
significant associations between energy-adjusted intake of total TFA and risk of
fatal IHD (Ascherio et al., 1996; Oomen et al., 2001) (Table 3).
Four studies investigated the associations between intake of TFA obtained
from ruminant sources and risk of IHD (Jakobsen et al., 2008; Oomen et al.,
2001; Pietinen et al., 1997; Willett et al., 1993) (Table 3). One study showed a
significant negative association between energy-adjusted intake of ruminant
TFA and risk of fatal IHD (Pietinen et al., 1997) and three studies showed no
significant associations between energy-adjusted intake of ruminant TFA and
risk of IHD (Jakobsen et al., 2008; Oomen et al., 2001; Willett et al., 1993)
(Table 3). Among the three studies which showed no significant associations,
one study, however, indicated a trend for a negative association (Willett et al.,
1993) and one study indicated a trend for a positive association (Oomen et al.,
2001) (Table 3). Three studies investigated the associations between intake of
TFA obtained from industrial sources and risk of IHD (Oomen et al., 2001;
Pietinen et al., 1997; Willett et al., 1993) (Table 3). Two studies showed
significant positive associations between energy-adjusted intake of industrial
TFA and risk of IHD (Pietineno et al., 1997; Willett et al., 1993) and one study
showed no significant association (Oomen et al., 2001) (Table 3). In the study
by Oomen et al. (2001), however, the results indicated a trend for a positive
association (Table 3).
In the study by Pietinen et al. (1997) energy-adjusted intake of trans 18:1n–9
was significantly positively associated with risk of fatal IHD after control for
dietary and non-dietary IHD risk factors (relative risk, RR, for highest compared
with lowest quintile = 1.37, 95% CI: 1.07–1.75, p value for trend = 0.002).
Saturated fatty acids intake is a potential confounder in analyses of TFA
intake and risk of IHD. Of the six studies (Ascherio et al., 1996; Jakobsen et al.,
2008; Oh et al., 2005; Oomen et al., 2001; Pietinen et al., 1997; Willett et al.,
1993) on TFA intake and risk of IHD, all but that of Ascherio et al. were
controlled for saturated fatty acids intake in multiple analyses (Table 3).
In one study reporting results from multiple analysis, the analysis included
control for hypercholesterolaemia which may be considered as a potential
metabolic effect of intake of TFA (Ascherio et al., 1996) (Table 3).
Whether intake of TFA should be expressed as absolute intake or energy-
adjusted intake depends on the mechanisms behind a potential association
between TFA intake and IHD. In the study by Jakobsen et al. (2008), both
approaches were used (Table 3).
b. Effect modification by sex and age

Two studies were carried out among women only (Oh et al., 2005; Willett et al.,
1993) and three studies were carried out among men only (Ascherio et al., 1996;
Oomen et al., 2001; Pietinen et al., 1997) (Table 3). Sex-specific results did not
suggest sex differences in the association between intake of TFA and risk of IHD
(Table 3). One study was carried out both among women and men (Jakobsen
et al., 2008). In that study there were no overall associations between energy-
adjusted ruminant TFA intake and risk of IHD (Table 3). Among women,
however, an indication of a negative association between energy-adjusted
ruminant TFA intake and risk of IHD was found (RR per 0.5 g intake = 0.77, 95%
CI: 0.55–1.09); no association was found among men (RR per 0.5 g intake = 1.05,
95% CI: 0.94–1.17). The p value for effect modification by sex was 0.52.
In two studies that investigated potential effect modification by age (Jakobsen
et al., 2008; Oh et al., 2005) the associations between TFA intake and risk of
IHD differed by age category. In the study by Oh et al. (2005), a positive
significant association between the percentage of energy intake obtained from
total TFA intake and risk of IHD was found among women less than age 65 years
(RR for highest compared with lowest quintile = 1.50, 95% CI: 1.13–2.00,
p value for trend = 0.01) but not among women aged 65 years or more (RR for
highest compared with lowest quintile = 1.15, 95% CI: 0.80–1.66, p value for
trend = 0.49, p value for effect modification by age = 0.03). In the study by
Jakobsen et al. (2008), an indication of a negative association between energy-
adjusted ruminant TFA intake and risk of IHD was found among women less
than age 60 years (RR per 0.5 g intake, intake, = 0.59, 95% CI: 0.35–1.01);
among women aged 60 years or more the RR per 0.5 g intake was 0.81 (95% CI:
0.57–1.15). The p value for effect modification by age was 0.22. Among men,
less than age 60 years the RR per 0.5 g intake was 0.94 (95% CI: 0.78–1.14) and
among men aged 60 years or more the RR per 0.5 g intake was 1.09 (95% CI:
0.97–1.22). The p value for effect modification by age was 0.17.

The results of the case-control and follow-up studies investigating the associa-
tion between the intake of TFA and the risk of IHD are summarized in Table 4.
Among those relating to ruminant TFA, one study showed a significant
negative association (Pietinen et al., 1997) and four studies showed no signifi-
cant associations (Ascherio et al., 1994; Jakobsen et al., 2008; Oomen et al.,
2001; Willett et al., 1993) (Table 4). However, among the four studies which
showed no significant associations, two indicated trends for positive associa-
tions (Ascherio et al., 1994; Oomen et al., 2001) and one indicated a trend for
a negative association (Willett et al., 1993) (Table 4). Among the studies
relating to industrial TFA, three studies showed significant positive associa-
tions (Ascherio et al., 1994; Pietinen et al., 1997; Willett et al., 1993) and one
study showed no significant association (Oomen et al., 2001) (Table 4). How-
ever, the results of the Oomen study indicated a trend for a positive association
(Table 4). In the cross-sectional study by Bolton-Smith et al. (1996), intakes of
294
Table 4. Summary of the results of the case-control and follow-up studies investigating the association between intake of trans fatty acids
and risk of ischaemic heart disease.
Positive association Negative association No association

No. of studies Reference(s)* No. of studies Reference(s) No. of studies Reference
Ruminant TFA 2 (Ascherio (1994)); 2 (Willett (1993)); 1 Jakobsen (2008)

(Oomen (2001)) Pietinen (1997)
Industrial TFA 4 Willett (1993); 0 – 0 –
Ascherio (1994);
Pietienen (1997);
(Oomen (2001))
* A bracket indicates that the association was not statistically significant. TFA, trans fatty acid.
total TFA, ruminant TFA, and industrial TFA were not significantly associated
with IHD among women who had undiagnosed IHD at baseline. Among men
who had undiagnosed IHD at baseline, there was a significant negative
association between intake of ruminant TFA and IHD; intakes of total and
industrial TFA were not significantly associated with IHD (Bolton-Smith
et al., 1996).
In summary, the studies using direct measure of intake of TFA suggest that
intake of TFA derived from industrial sources is associated with a higher risk
of IHD. There is no observational epidemiological evidence of a harmful effect
on IHD risk of intake of TFA obtained from ruminant sources.
b. Biomarkers of intake of TFA and risk of IHD

The results of the case-control studies investigating the association between
biomarkers of intake of TFA and risk of IHD are summarized in Table 5.
Among the studies investigating the association between tissue total trans
16:1 and risk of IHD, three studies showed significant positive associations
(Baylin et al., 2003; Siguel & Lerman, 1993; Thomas et al., 1983a) and four
studies showed no significant associations (Clifton et al., 2004;
Ghahremanpour et al., 2008; Lemaitre et al., 2006; Sun et al., 2007c) (Ta-
ble 5). However, among the four studies which showed no significant
associations, those of Ghahremanpour et al. and Sun et al. indicated trends
for negative associations. Among the studies investigating the association
between tissue total trans 18:1 and risk of IHD, two studies showed signifi-
cant positive associations (Ghahremanpour et al., 2008; Sun et al., 2007b),
one study showed a significant negative association (Lemaitre et al., 2006),
and eight studies showed no significant associations (Aro et al., 1995; Co-
lon-Ramos et al., 2006; Dlouhy et al., 2003; Harris et al., 2007; Lemaitre
et al., 2002; Roberts et al., 1995; Thomas et al., 1983a) (Table 5). However,
among the eight studies which showed no significant associations, three
(Aro et al., Colon-Ramos et al., Dlouhy et al.) indicated trends for positive
associations and three (Harris et al., Lemaitre et al., Roberts et al.) indicated
trends for negative associations (Table 5). Among the studies investigating
the association between tissue total trans 18:2 and risk of IHD, four studies
showed significant positive associations (Colon-Ramos et al., 2006;
Lemaitre et al., 2002; Lemaitre et al., 2006; Sun et al., 2007b) and four
studies showed no significant associations (Colon-Ramos et al., 2006;
Ghahremanpour et al., 2008; Harris et al., 2007; Roberts et al., 1995) (Ta-
ble 5). However, among the four studies which showed no significant
associations, three (Colon-Ramos et al., Ghahremanpour et al., Harris et al.)
indicated trends for positive associations (Table 5).
In summary, the studies using biomarkers of intakes of TFA suggest that
intake of total trans 18:2 is associated with a higher risk of IHD whereas results
on total trans 16:1 and total trans 18:1 are not consistent.
Table 5. Summary of the results of the case-control studies investigating the association between biomarkers of trans fatty acids and risk of
296
ischaemic heart disease.
Positive association Negative association No association

No. of Reference(s)* No. of Reference(s) No. of Reference(s)
studies studies studies
Total trans 16:1† 3 Thomas (1983a); 2 (Sun (2007c)); 2 Clifton (2004);
Siguel (1993); (Ghahremanpour (2008)) Lemaitre (2006)
Baylin (2003)‡
Total trans 18:1 5 (Aro (1995)); 4 (Roberts (1995)); 2 Thomas (1983a);
(Dlouhy (2003)); (Lemaitre (2002)); Colon-Ramos (2006) 2000–03‡
(Colon-Ramos (2006) 1994–99‡); Lemaitre (2006);
Sun (2007b); (Harris (2007))
Ghahremanpour (2008)
Total trans 18:2 7 Lemaitre (2002); 0 – 1 Roberts (1995)
Colon-Ramos (2006) 1994–99‡;
(Colon-Ramos (2006) 2000–03‡);
Lemaitre (2006);
(Harris (2007));
Sun (2007b);
(Ghahremanpour (2008))
* A bracket indicates that the association was not statistically significant.

† Trans 16:1n–7 or the sum of trans 16:1n–7 and 16:1n–9.
‡ The studies by Baylin et al. (2003) and Colon-Ramos et al. (2006) were partly based on some of the same participants and therefore these studies cannot be considered
independent. In the study by Baylin et al. (2003), the investigators reported results on adipose tissue total trans fatty acids (TFA), trans 16:1n–7, total trans 18:1, and
total trans 18:2 before a reduction of TFA in foods. In the study by Colon-Ramos et al. (2006), the investigators reported results on adipose tissue total TFA, total
trans 18:1, and total trans 18:2 before and after a reduction of TFA in foods. For this review, the study by Baylin et al. (2003) was selected for results on trans 16:1n–
7 and the study by Colon-Ramos et al. (2006) was selected for total TFA, total trans 18:1, and total trans 18:2 before and after a reduction of TFA in foods. Cases
and controls were sampled in two independent time periods before and after a reduction of TFA in foods. The results from the two different time periods can be considered
independent and are referred to as two separate studies (Colon-Ramos (2006) 1994–99 and Colon-Ramos (2006) 2000–03).
D. Discussion
One ecologic study (Kromhout et al., 1995), one cross-sectional study (Bolton-
Smith et al., 1996), 19 case-control studies (Aro et al., 1995; Ascherio et al.,
1994; Baylin et al., 2003; Clifton et al., 2004; Colon-Ramos et al., 2006;
Dlouhy et al., 2003; Ghahremanpour et al., 2008; Harris et al., 2007; Lemaitre
et al., 2002; Lemaitre et al., 2006; Lopes et al., 2007; Pedersen et al., 2000;
Roberts et al., 1995; Siguel & Lerman, 1993; Sun et al., 2007b; Sun et al.,
2007c; Thomas & Scott, 1981; Thomas et al., 1983a), and six follow-up studies
(Ascherio et al., 1996; Jakobsen et al., 2008; Oh et al., 2005; Oomen et al.,
2001; Pietinen et al., 1997; Willett et al., 1993) met the selection criteria and
were included in this review of observational epidemiological studies on the
intake of TFA and the risk of IHD. The studies using direct measure of intake
of TFA suggest that intake of TFA obtained from industrial sources is associ-
ated with a higher risk of IHD. There is no observational epidemiological
evidence of a harmful effect on IHD risk of intake of TFA obtained from
ruminant sources. The studies using biomarkers of intakes of TFA suggest that
intake of trans 18:2 is associated with a higher risk of IHD whereas results on
total trans 16:1 and total trans 18:1 are not consistent.
2. Quality of the studies

In this review the studies were ordered according to study design.
a. Selection problems
The study populations in this review included both men and women and a broad
range of age. However, the measures of exposure (intake of TFA and propor-
tions of TFA in tissue) were at different levels and also the variation in
measures of exposure differed. Both aspects may have implications for the
external validity of the studies.
In most case-control studies, only surviving cases with IHD were included.
If TFA affects the disease prognosis, selection bias may be of concern. The
known mechanisms behind the effects of TFA do not lend support to this
concern. In the follow-up studies, censoring due to death from other causes
associated with both intake of TFA and risk of IHD seemed limited and
selection bias is thus unlikely to have affected the study results.
b. Information problems
Measures of TFA intake included both direct measures of dietary TFA intakes
and proportions of TFA in adipose tissue, whole blood, plasma, or erythrocytes
as biomarkers of TFA intakes.
The primary method used for quantification of dietary TFA intakes was food
frequency questionnaires. A potential source of random measurement error
arises from dietary self-reporting methods. Generally, random measurement

error leads to under-estimation of the true risk and to the loss of statistical power
for testing associations. Moreover, lack of repeated assessment of dietary intake
excludes possible analytic approaches to reduce measurement error due to
changes in the participants’ habitual diets over time, as well as changes in food
composition. The latter is a particular issue with respect to investigations of
associations between TFA intake and risk of IHD, because of the efforts made by
the food industry to decrease TFA in margarines and shortenings which took
place in many countries during the 1990s. In Denmark, the content of industrial
TFA in margarines has been reduced and consequently the mean daily intakes of
industrial TFA from margarines and shortenings has decreased from 2.2 g/person
in 1992 to 0.35 g/person in 1999 (Leth et al., 2003). In contrast, the intake of
ruminant TFA may have been constant, as the amount of milk fat and ruminant
meat available for consumption has been rather constant (Fagt & Trolle, 2001).
Only one of the included studies used repeated measurements of diet in analyses
of the association between intake of TFA and risk of IHD (Oh et al., 2005). In the
studies using adipose tissue, whole blood, plasma or erythrocytes as biomarkers
of TFA intakes, fatty acids were analysed by gas chromatography; this method
accurately quantifies fatty acids (Delmonte & Rader, 2007). Measurement of
proportions of TFA in tissue partly reduces the measurement error associated
with collecting dietary information by dietary self-reporting methods. For fatty
acids which cannot be synthesized de novo in the body, such as linoleic acid,
α-linolenic acid (cis 18:3n–3) and TFA, the proportions in adipose tissue, whole
blood, plasma, and erythrocytes are considered to be accurate measures of dietary
exposure (Baylin et al., 2002 and 2005; Sun et al., 2007a). The proportion of
TFA in plasma, erythrocytes, and adipose tissue, may be used as indices of
integrated exposure over the previous weeks, months, and year, respectively
(Craig-Schmidt, 2006). It is not possible to quantify dietary TFA intakes from
proportions of tissue TFA. Therefore, the findings from the studies using
biomarkers of intake of TFA cannot be directly translated into public health
recommendations.
The proportion of TFA in tissue is expressed as percentage of total fatty acids
in the tissue. Thus, the proportion of TFA in tissue depends on the intake of
TFA but also on the intake, oxidation and biosynthesis of other fatty acids.
Consequently, this way of expressing the value does not reflect individual
differences in TFA intake that are unrelated to differences in dietary composi-
tions (dietary patterns) between persons (e.g. differences in intakes of amounts
of foods – total energy intake – due to differences in physical activity level).
Whether this variation is extraneous depends on the mechanisms behind the
associations between TFA intake and risk of IHD. In studies using direct
measures of dietary TFA intakes, this variation is reflected if intake is expressed
as absolute intake (a measure of dietary composition and total amount of food,
i.e. total energy intake) whereas this variation is not reflected if intake is
expressed as energy-adjusted intake (a measure of dietary composition).

However, at present we have limited insight into the mechanisms behind the
associations. Therefore, at present, the optimal approach for obtaining further
insight into the association between intake of TFA and risk of IHD may be to
evaluate both the absolute and the energy-adjusted intake.
The Danish population may be one of the best populations in which to study
the association between ruminant TFA intake and risk of IHD due to the wide
range of ruminant TFA intake (Jakobsen et al., 2006 and 2008). In the study by
Jakobsen et al. (2008), no evidence of a higher risk associated with intake of
ruminant TFA was found within the wide range of intakes among both women
(90% central range: 0.5–3.1 g/day, 0.3–1.4% of energy intake) and men (90%
central range: 0.6– 4.1 g/day, 0.3–1.4% of energy intake). The findings from
epidemiological studies are in agreement with the findings from the dietary
trial by Motard-Belanger et al. (2008) which showed that intake of 4.2 g/d
(1.5% of energy intake), does not affect risk factors of IHD (the upper limit of
actual amounts of ruminant TFA consumed in diets). However, these findings
do not exclude the possibility that ruminant TFA may have adverse effects if
consumed in the generally higher amounts typical of industrial TFA (see
Table 1 and 3 for amounts). In the study by Motard-Belanger (2008), the high
TFA diets (10.2 g/d), whether ruminant or industrial TFA, significantly in-
creased the ratio of total cholesterol to HDL-cholesterol compared with the
moderate ruminant TFA diet (4.2 g/d). Moreover, in the case-control study by
Colon-Ramos et al. (2006), the association between adipose tissue total TFA,
total trans 18:1 and total trans 18:2 and risk of acute myocardial infarction
were investigated before and after reduction of TFA content in foods. Before
industrial modification, total TFA was significantly positively associated with
the risk of acute myocardial infarction (OR for highest compared with lowest
quintile = 3.28, 95% CI: 1.68–6.42, p value for trend = < 0.001) primarily
accounted for by total trans 18:2 (Table 2). In contrast, after industrial modifi-
cation, the association with acute myocardial infarction was no longer significant
(OR for highest compared with lowest quintile = 1.03, 95% CI: 0.75–1.42,
p value for trend = 0.65) (Table 2). Before industrial modification the medians
within quintiles were 1.9, 2.5, 3.0, 3.6 and 4.4 and afterwards 1.8, 2.3, 2.6, 2.9
and 3.4 (Table 2). In future studies, it is recommended to use appropriate
methods (such as restricted cubic spline models) to study potential threshold
effects.
In most case-control studies, intake of TFA was assessed using biomarkers.
Early symptoms appearing before diagnosis may have affected the dietary
habits of the cases. Information bias may thus have affected the study results
most likely over-estimating the strength of the association between TFA intake
and risk of IHD. In the follow-up studies, information on development of IHD
was obtained and classified independent of exposure to TFA intake. Informa-
tion bias is thus an unlikely explanation of the study results.
c. Effect modification by sex and age

The rationale behind the evaluation of effect modification by sex and age was
that the association between intake of TFA and risk of IHD may be modified by
sex and age due to differences in the underlying risk functions.
Sex-specific results did not suggest sex differences in the association be-
tween intakes of TFA and risk of IHD or between biomarkers of intake of TFA
and risk of IHD. However, in the cross-sectional study by Bolton-Smith (1996)
no significant association was found among women who had undiagnosed IHD
at baseline whereas intake of ruminant TFA was significantly negatively assoc-
iated with IHD among men who had undiagnosed IHD at baseline. In contrast,
the study by Jakobsen et al. (2008) indicated a negative association between
intake of ruminant TFA and risk of IHD among women but not among men.
In two studies, the associations between TFA intake and risk of IHD differed
by age category (Jakobsen et al., 2008; Oh et al., 2005). These studies indi-
cated stronger associations between intake of TFA and risk of IHD among
younger women compared with older women.
d. Confounding and modeling considerations

In studies on TFA intake, other dietary components may potentially confound
the results. Intake of saturated fatty acids is one of the potential confounders
because saturated fatty acids intake is a risk factor for IHD and because
ruminant TFA intake is correlated to saturated fatty acids intake in most
studies, partly because of shared food sources. By controlling for saturated
fatty acids intake, the models address the effects of ruminant TFA intake
independent of saturated fatty acids intake. The variation in food sources
contributing to variation in ruminant TFA intake, however, is then restricted to
variation in food sources not directly overlapping food sources contributing to
variation in saturated fatty acids intake. In the study by Jakobsen et al. (2008),
however, linear regression analyses were used to identify the most important
food groups contributing to variation in ruminant TFA intake in analyses
without and with control for saturated fatty acids intake. The results indicated
that the underlying dietary patterns might be similar without and with control
for saturated fatty acids intake implying that the interpretation of results from
analyses not controlled for saturated fatty acids do not differ from the interpre-
tation of results from analyses controlled for saturated fatty acids intake.
In studies on tissue TFA, other tissue fatty acids and diet may be of concern.
By controlling for other tissue fatty acids in analyses of tissue TFA, the models
address the effects of the latter independently of the former. However, due to
the percentage expression, the interpretation of results for one tissue fatty acid
controlled for other tissue fatty acids is not straightforward. By controlling for
saturated fatty acids intake, the models address the effect of tissue TFA
independent of saturated fatty acids intake. However, as the proportion of
tissue TFA is expressed as percentage of total tissue fatty acids the proportion
of TFA also depends indirectly on the saturated fatty acids intake. Moreover,
by controlling for saturated fatty acids intake, the variation in food sources
contributing to variation in tissue TFA is then restricted to variation in food
sources not directly overlapping food sources contributing to variation in
saturated fatty acids intake which are mostly the same foods. Therefore,
inclusion of fatty acids intake in multiple analyses should be considered
carefully to avoid models for which the results have no clear interpretation.
In a number of the included studies, the analyses included control for
potential metabolic effects of intake of TFA. As these factors (plasma cho-
lesterol concentration and history of diabetes mellitus) may represent steps
in the causal chain between intake of TFA and risk of IHD (i.e. the effect of
intake of TFA on risk of IHD is mediated through the effect of these interme-
diate factors) they should not be treated simply as potential confounding
factors, but instead analysed in more detail taking into account their inter-
mediate nature.
3. Specific TFA and risk of IHD

Observational epidemiological studies on the association between the intake of
TFA and the risk of IHD suggest that intake of TFA obtained from industrial
sources is associated with an increased risk of IHD whereas TFA obtained from
ruminant sources is not associated with an increased risk of IHD. Potentially
different effects of TFA derived from ruminant versus industrial sources on
IHD risk may relate to lower amounts of ruminant TFA consumed in diets or to
different biological effects of specific TFA. Studies using biomarkers of
intakes of TFA suggest that intake of trans 18:2 is associated with a higher risk
of IHD. The adverse effects of industrial compared with ruminant TFA may
relate to trans 18:2 isomers. However, small amounts of trans 18:2 isomers are
also present in milk fat and meat from ruminant animals (Precht & Molkentin,
1995). The trans 18:1 isomers are the predominant TFA in the diet (Hulshof
et al., 1999). In milk fat from ruminant animals, the predominant TFA is
vaccenic acid (trans 18:1n–7) whereas in margarines and shortenings the
predominant TFA are elaidic acid (trans 18:1n–9) and trans 18:1n–8 (Precht &
Molkentin, 1995 and 1996). Thus, the observed differences in risk associated
with intake of TFA from ruminant sources versus industrial sources could also
relate to different biological effects of specific positional isomers within trans
18:1. Only a few of the included studies reported results on positional isomers
within trans 18:1 and results were not consistent. Further insight into the
association between intake of TFA and risk of IHD would, therefore, probably
be obtained from studies of the relationship of the risk to specific TFA rather
than total TFA.
."
4. Conclusions and perspectives

The studies using direct measure of intake suggest that intake of TFA obtained
from industrial sources is associated with an increased risk of IHD. There is no
observational epidemiological evidence of a harmful effect on IHD risk of
intake of TFA obtained from ruminant sources. The studies using biomarkers
of intakes of TFA suggest that intake of trans 18:2 is associated with an
increased risk of IHD, whereas results for total trans 16:1 and total trans 18:1
are not consistent. Further insight into the association between intake of TFA
and risk of IHD would probably be obtained from studies of the relationship of
the risk to specific TFA rather than total TFA.
We have limited insight into the mechanisms behind the associations between
intake of TFA and risk of IHD and further studies are warranted. Further
insight into the association between intake of TFA and risk of IHD would
probably be obtained from studies of the relationship of the risk to specific
TFA rather than total TFA. In analyses of tissue TFA (which are considered to
provide biomarkers of TFA intake) and risk of IHD, inclusion of fatty acid
intake in multiple analyses should be considered carefully to avoid models for
which the results have no clear interpretation. Moreover, we recommend the
use of appropriate methods (such as restricted cubic spline models) to study
potential threshold effects. Finally, we recommend that in future studies
potential effect modification by sex and age should be considered in order to
obtain further insight into the association between intake of TFA and risk of
IHD.
Acknowledgements
The authors collaborate within the project ‘Hepatic and adipose tissue and
functions in the metabolic syndrome’ (HEPADIP, www.hepadip.org), which is
supported by the European Commission as an Integrated Project under the 6th
Framework Program (Contract LSHM-CT-2005-018734) and within the
research programme of the ‘Danish obesity research centre’ (DanORC,
(www.danorc.dk), which is supported by the Danish Council for Strategic
Research (Contract 2101-06-0005).
References
Aro, A, Kardinaal, AF, Salminen, I, Kark, JD, Riemersma, RA, Delgado-Rodriguez, M,
Gomez-Aracena, J, Huttunen, JK, Kohlmeier, L, Martin, BC, Martin-Moreno, JM,
Mazaev, VP, Ringstad, J, Thamm, M, van’t Veer, P and Kok, FJ (1995) Adipose tissue
isomeric trans fatty acids and risk of myocardial infarction in nine countries: the
EURAMIC study. Lancet, 345, 273–278.
Ascherio, A, Hennekens, CH, Buring, JE, Master, C, Stampfer, MJ and Willett, WC (1994)
Trans-fatty acids intake and risk of myocardial infarction. Circulation, 89, 94–101.
Ascherio, A, Rimm, EB, Giovannucci, EL, Spiegelman, D, Stampfer, M and Willett, WC
(1996) Dietary fat and risk of coronary heart disease in men: cohort follow up study in
the United States. Br. Med. J., 313, 84–90.
Baylin, A, Kabagambe, EK, Ascherio, A, Spiegelman, D and Campos, H (2003) High 18:2
trans-fatty acids in adipose tissue are associated with increased risk of nonfatal acute
myocardial infarction in Costa Rican adults. J. Nutr., 133, 1186–1191.
Baylin, A, Kim, MK, Donovan-Palmer, A, Siles, X, Dougherty, L, Tocco, P and Campos,
H (2005) Fasting whole blood as a biomarker of essential fatty acid intake in epide-
miologic studies: comparison with adipose tissue and plasma. Am. J. Epidemiol., 162,
373–381.
Bolton-Smith, C, Woodward, M, Fenton, S and Brown, CA (1996) Does dietary trans fatty
acid intake relate to the prevalence of coronary heart disease in Scotland? Eur. Heart J.,
17, 837–845.
Chardigny, JM, Destaillats, F, Malpuech-Brugere, C, Moulin, J, Bauman, DE, Lock, AL,
Barbano, DM, Mensink, RP, Bezelgues, JB, Chaumont, P, Combe, N, Cristiani, I, Joffre,
F, German, JB, Dionisi, F, Boirie, Y and Sebedio, JL (2008) Do trans fatty acids from
industrially produced sources and from natural sources have the same effect on cardio-
vascular disease risk factors in healthy subjects? Results of the Trans Fatty Acids
Collaboration (TRANSFACT) study. Am. J. Clin. Nutr., 87, 558–566.
Clifton, PM, Keogh, JB and Noakes, M (2004) Trans fatty acids in adipose tissue and the food
supply are associated with myocardial infarction. J. Nutr., 134, 874–879.
Colon-Ramos, U, Baylin, A and Campos, H (2006) The relation between trans fatty acid
levels and increased risk of myocardial infarction does not hold at lower levels of trans
fatty acids in the Costa Rican food supply. J. Nutr., 136, 2887–2892.
Craig-Schmidt, MC (2006) World-wide consumption of trans fatty acids. Atheroscler.
Suppl., 7, 1–4.
Delmonte, P and Rader, JI (2007) Evaluation of gas chromatographic methods for the
determination of trans fat. Analyt. Bioanalyt. Chem., 389, 77–85.
Dlouhy, P, Tvrzicka, E, Stankova, B, Vecka, M, Zak, A, Straka, Z, Fanta, J, Pachl, J,
Kubisova, D, Rambouskova, J, Bilkova, D and Andel, M (2003) Higher content of 18:1
trans fatty acids in subcutaneous fat of persons with coronarographically documented
atherosclerosis of the coronary arteries. Ann. Nutr. Metab, 47, 302–305.
Fagt, S and Trolle, E (2001) 1. Forsyningen af fødevarer 1955–1999. Udviklingen i
danskernes kost — forbrug, indkøb og vaner (Food statistics 1955–1999. Development
in dietary intake in the Danish population — consumption, purchase and habits). (In
Danish). National Food Agency, Søborg, Denmark.
Fritsche, J, Steinhart, H, Kardalinos, V and Klose, G (1998) Contents of trans-fatty acids in
human substernal adipose tissue and plasma lipids: relation to angiographically docu-
mented coronary heart disease. Eur. J. Med. Res., 3, 401–406.
Ghahremanpour, F, Firoozrai, M, Darabi, M, Zavarei, A and Mohebbi, A (2008) Adipose
tissue trans fatty acids and risk of coronary artery disease: a case-control study. Ann. Nutr.
Metab., 52, 24–28.
Harnack, L, Lee, S, Schakel, SF, Duval, S, Luepker, RV and Arnett, DK (2003) Trends in the
trans-fatty acid composition of the diet in a metropolitan area: the Minnesota Heart
Survey. J. Am. Diet. Assoc., 103, 1160–1166.
Harris, WS, Reid, KJ, Sands, SA and Spertus, JA (2007) Blood omega-3 and trans fatty acids
in middle-aged acute coronary syndrome patients. Am. J. Cardiol., 99, 154–158.
Heckers, H, Korner, M, Tuschen, TW and Melcher, FW (1977) Occurrence of individual
trans-isomeric fatty acids in human myocardium, jejunum and aorta in relation to
different degrees of atherosclerosis. Atherosclerosis, 28, 389–398.
Hodgson, JM, Wahlqvist, ML, Boxall, JA and Balazs, ND (1996) Platelet trans fatty acids
in relation to angiographically assessed coronary artery disease. Atherosclerosis, 120,

147–154.
Hu, FB, Stampfer, MJ, Manson, JE, Rimm, E, Colditz, GA, Rosner, BA, Hennekens, CH and
Willett, WC (1997) Dietary fat intake and the risk of coronary heart disease in women.
N. Engl. J. Med., 337, 1491–1499.
Hu, FB, Stampfer, MJ, Rimm, E, Ascherio, A, Rosner, BA, Spiegelman, D and Willett, WC
(1999) Dietary fat and coronary heart disease: a comparison of approaches for adjusting
for total energy intake and modeling repeated dietary measurements. Am. J. Epidemiol.,
149, 531–540.
Hulshof, KF, Erp-Baart, MA, Anttolainen, M, Becker, W, Church, SM, Couet, C, Hermann-
Kunz, E, Kesteloot, H, Leth, T, Martins, I, Moreiras, O, Moschandreas, J, Pizzoferrato,
L, Rimestad, AH, Thorgeirsdottir, H, van Amelsvoort, JM, Aro, A, Kafatos, AG,
Lanzmann-Petithory, D and van Poppel, G (1999) Intake of fatty acids in western Europe
with emphasis on trans fatty acids: the TRANSFAIR Study. Eur. J. Clin. Nutr., 53, 143–
157.
Jakobsen, MU, Bysted, A, Andersen, NL, Heitmann, BL, Hartkopp, HB, Leth, T, Overvad,
K and Dyerberg, J (2006) Intake of ruminant trans fatty acids in the Danish population
aged 1–80 years. Eur. J. Clin. Nutr., 60, 312–318.
Jakobsen, MU, Overvad, K, Dyerberg, J and Heitmann, BL (2008) Intake of ruminant trans
fatty acids and risk of coronary heart disease. Int. J. Epidemiol., 37, 173–182.
Kromhout, D, Menotti, A, Bloemberg, B, Aravanis, C, Blackburn, H, Buzina, R, Dontas, AS,
Fidanza, F, Giampaoli, S, Jansen, A, Karvonen, M, Katan, M, Nissinen, A, Nedeljkovic,
S, Pekkanen, J, Pekkarinen, M, Punsar, S, Räsänen, L, Simic, B and Toshima, H (1995)
Dietary saturated and trans fatty acids and cholesterol and 25-year mortality from
coronary heart disease: the Seven Countries Study. Prev. Med., 24, 308–315.
Lemaitre, RN, King, IB, Mozaffarian, D, Sotoodehnia, N, Rea, TD, Kuller, LH, Tracy, RP
and Siscovick, DS (2006) Plasma phospholipid trans fatty acids, fatal ischemic heart
disease, and sudden cardiac death in older adults: the cardiovascular health study.
Circulation, 114, 209–215.
Lemaitre, RN, King, IB, Raghunathan, TE, Pearce, RM, Weinmann, S, Knopp, RH, Copass,
MK, Cobb, LA and Siscovick, DS (2002) Cell membrane trans-fatty acids and the risk
of primary cardiac arrest. Circulation, 105, 697–701.
Leth, T, Bysted, A, Hansen, K and Ovesen, L (2003) Trans FA content in Danish margarines
and shortenings. J. Am. Oil Chem. Soc., 80, 475–478.
Lopes, C, Aro, A, Azevedo, A, Ramos, E and Barros, H (2007) Intake and adipose tissue
composition of fatty acids and risk of myocardial infarction in a male Portuguese
community sample. J. Am. Diet. Assoc., 107, 276–286.
Mensink, RP, Zock, PL, Kester, AD and Katan, MB (2003) Effects of dietary fatty acids and
apolipoproteins: a meta-analysis of 60 controlled trials. Am. J. Clin. Nutr., 77, 1146–
1155.
Motard-Belanger, A, Charest, A, Grenier, G, Paquin, P, Chouinard, Y, Lemieux, S, Couture,
P and Lamarche, B (2008) Study of the effect of trans fatty acids from ruminants on blood
lipids and other risk factors for cardiovascular disease. Am. J. Clin. Nutr., 87, 593–599.
Mozaffarian, D, Abdollahi, M, Campos, H, Houshiarrad, A and Willett, WC (2007)
Consumption of trans fats and estimated effects on coronary heart disease in Iran. Eur.
J. Clin. Nutr., 61, 1004–1010.
Mozaffarian, D, Katan, MB, Ascherio, A, Stampfer, MJ and Willett, WC (2006) Trans fatty
acids and cardiovascular disease. N. Engl. J. Med., 354, 1601–1613.
Oh, K, Hu, FB, Manson, JE, Stampfer, MJ and Willett, WC (2005) Dietary fat intake and risk
of coronary heart disease in women: 20 years of follow-up of the Nurses’ Health Study.
Am. J. Epidemiol., 161, 672–679.
Oomen, CM, Ocke, MC, Feskens, EJM, van Erp-Baart, MJ, Kok, FJ and Kromhout, D
(2001) Association between trans fatty acid intake and 10-year risk of coronary heart
disease in the Zutphen Elderly Study: a prospective population-based study. Lancet,
357, 746–751.
Pedersen, JI, Ringstad, J, Almendingen, K, Haugen, TS, Stensvold, I and Thelle, DS (2000)
Adipose tissue fatty acids and risk of myocardial infarction – a case-control study. Eur.
J. Clin. Nutr., 54, 618–625.
Pietinen, P, Ascherio, A, Korhonen, P, Hartman, AM, Willett, WC, Albanes, D and Virtamo,
J (1997) Intake of fatty acids and risk of coronary heart disease in a cohort of Finnish men.
The Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study. Am. J. Epidemiol., 145,
876–887.
Precht, D and Molkentin, J (1995) Trans fatty acids: implications for health, analytical
methods, incidence in edible fats and intake (a review). Nahrung, 39, 343–374.
Precht, D and Molkentin, J (1996) Rapid analysis of the isomers of trans-octadecenoic acid
in milk fat. Int. Dairy J., 6, 791–809.
Roberts, TL, Wood, DA, Riemersma, RA, Gallagher, PJ and Lampe, FC (1995) Trans
isomers of oleic and linoleic acids in adipose tissue and sudden cardiac death. Lancet, 345,
278–282.
Shai, I, Rimm, EB, Hankinson, SE, Curhan, G, Manson, JE, Rifai, N, Stampfer, MJ and Ma,
J (2004) Multivariate assessment of lipid parameters as predictors of coronary heart
disease among postmenopausal women: potential implications for clinical guidelines.
Circulation, 110, 2824–2830.
Siguel, EN and Lerman, RH (1993) Trans-fatty acid patterns in patients with angiographically
documented coronary artery disease. Am. J. Cardiol., 71, 916–920.
Singh, RB, Niaz, MA, Ghosh, S, Beegom, R, Rastogi, V, Sharma, JP and Dube, GK (1996)
Association of trans fatty acids (vegetable ghee) and clarified butter (Indian ghee) intake
with higher risk of coronary artery disease in rural and urban populations with low fat
consumption. Int. J. Cardiol., 56, 289–298.
Sun, Q, Ma, J, Campos, H, Hankinson, SE and Hu, FB (2007a) Comparison between plasma
and erythrocyte fatty acid content as biomarkers of fatty acid intake in US women. Am.
J. Clin. Nutr., 86, 74–81.
Sun, Q, Ma, J, Campos, H, Hankinson, SE, Manson, JE, Stampfer, MJ, Rexrode, KM, Willett,
WC and Hu, FB (2007b) A prospective study of trans fatty acids in erythrocytes and risk
of coronary heart disease. Circulation, 115, 1858–1865.
Sun, Q, Ma, J, Campos, H and Hu, FB (2007c) Plasma and erythrocyte biomarkers of dairy
fat intake and risk of ischemic heart disease. Am. J. Clin. Nutr., 86, 929–937.
Thomas, LH and Scott, RG (1981) Ischaemic heart disease and the proportions of hydrogen-
ated fat and ruminant-animal fat in adipose tissue at post-mortem examination: a
case-control study. J. Epidemiol. Community Health, 35, 251–255.
Thomas, LH, Winter, JA and Scott, RG (1983a) Concentration of 18:1 and 16:1 transunsaturated
fatty acids in the adipose body tissue of decedents dying of ischaemic heart disease
compared with controls: analysis by gas liquid chromatography. J. Epidemiol. Commu-
nity Health, 37, 16–21.
Thomas, LH, Winter, JA and Scott, RG (1983b) Concentration of transunsaturated fatty acids
in the adipose body tissue of decedents dying of ischaemic heart disease compared with
controls. J. Epidemiol. Community Health, 37, 22–24.
van de Vijver, LP, van Poppel, G, van Houwelingen, A, Kruyssen, DA and Hornstra, G (1996)
Trans unsaturated fatty acids in plasma phospholipids and coronary heart disease: a case-
control study. Atherosclerosis, 126, 155–161.
van Poppel, G (1998) Intake of trans fatty acids in western Europe: the TRANSFAIR study.
Lancet, 351, 1099.
van Toll, A, Zock, PL, van Gent, T, Scheek, LM and Katan, MB (1995) Dietary trans fatty
acids increase serum cholesterylester transfer protein activity in man. Atherosclerosis,
115, 129–134.
Willett, WC, Stampfer, MJ, Manson, JE, Colditz, GA, Speizer, FE, Rosner, BA, Sampson,
LA and Hennekens, CH (1993) Intake of trans fatty acids and risk of coronary heart
disease among women. Lancet, 341, 581–585.
CHAPTER 11
Dietary Trans Fatty Acids and Cardiovascular
Disease Risk
CORINNE MALPUECH-BRUGÈRE1,2, BÉATRICE MORIO1,2 AND RONALD

P. MENSINK3
1
Clermont Université, UFR Médecine, UMR 1019 Nutrition Humaine,
Clermont-Ferrand, France
2
INRA, UMR 1019 Nutrition Humaine, Saint Genès Champanelle, France
3
Maastricht University, Nutrition and Toxicology Research Institute Maastricht,
Department of Human Biology, Maastricht, The Netherlands
In this chapter, the clinical evidence linking the consumption of the different
trans fatty acids (TFA) to cardiovascular disease risk factors is discussed. The
main results from epidemiological studies on the relationship between TFA
intake and cardiovascular risk are summarized in the first section (for detailed
information see Chapter 10).
A. Evidence from epidemiological studies

This part summarizes the results from prospective epidemiological studies on
the relationship between TFA intake and cardiovascular risk. This topic is
discussed in further detail in Chapter 10. The association between the intake of
TFA and the risk for cardiovascular disease (CVD) has been examined in four
prospective cohort studies (Ascherio et al., 1996; Pietinen et al., 1997; Oomen
et al., 2001; Oh et al., 2005). In general, the intake of TFA was positively
related to CVD risk. Except for one study (Ascherio et al., 1996), the ‘p for
trend’ was statistically significant, indicating that the relative risk increased,
the higher the intake of total TFA.
Ascherio and co-workers (1996) examined the association between TFA
intake and the incidence of coronary heart disease (CHD) in 43,757 health
professionals from the USA, who were at the start of the study aged 40 to 75
years, and free of diagnosed cardiovascular disease or diabetes. During the
follow-up period, 734 coronary events were documented. After multivariate
analyses, the relative risks for men in the highest versus the lowest quintile of
TFA intake (median intakes of 4.3 versus 1.5 g/d) were 1.40 for myocardial
infarction and 1.78 for fatal CHD. The relationships were attenuated after
further adjustment for fibre intake.
In Finland, Pietinen and co-workers (1997) examined the relationships
307
between the intakes of TFA and the risk of CHD in a cohort of 21,930 smoking
men aged 50 to 69 years. At the start of the study, subjects were free of
diagnosed CVD. After 6.1 years of follow-up, 1399 major coronary events and
635 coronary deaths were reported. After adjustment for several potential
confounding variables, a significant positive relationship was observed be-
tween the intake of TFA and the risk of coronary death. For men in the top
quintile of TFA intake (median = 6.2 g/d), the relative risk of coronary death
was 1.39 as compared with men in the lowest quintile of intake (median = 1.8 g/
d). For all major coronary events, associations were in the same direction, but
in general somewhat less pronounced.
In the Zutphen Elderly Study from the Netherlands, 667 men (aged 64–84
years) and free of coronary heart disease at baseline, were studied prospec-
tively. After 10 years of follow-up, there were 98 cases of fatal or non-fatal
CHD. After multivariate analyses, a positive relationship was found between
TFA intake at baseline with the 10-year risk of CHD. It was estimated that a
difference of 2% of energy in TFA intake at baseline increased the risk for CHD
by 28% (Oomen et al., 2001).
Oh and co-workers (Oh et al., 2005) examined the relationship between TFA
intake with the risk of CHD among 78,778 US women, who were initially free
of cardiovascular disease and diabetes. During 20 years of follow-up, 1766
incident CHD cases were reported. TFA intake was positively associated with
CHD risk. The multivariate relative risk for the highest vs the lowest quintile of
intake was 1.33. Median intakes of TFA intakes in these two quintiles were
respectively 2.8 and 1.3% of energy. This association was particularly evident
among women younger than age 65 years.
No such associations have been found between the intakes of TFA from
ruminant sources with CVD risk (Oomen et al., 2001; Willett et al., 1993). In
one study, even a negative relationship was reported (Pietinen et al., 1997).
Also, a recent study (Jakobsen et al., 2008) did not suggest that the intake of
ruminant TFA is associated with a higher risk of CVD. For women, even a
negative relationship was reported. It should be noted however that the intake
of TFA from animal origin is low, which may have masked any relationship. In
fact, based on the data of prospective cohort studies, Weggemans and co-
workers (2004) suggested that in prospective cohort studies no differences in
the risk of CHD is evident between total, ruminant and industrial TFA when the
intake is below 2.5 g/d.
B. Evidence from metabolic, randomized studies
1. Effect of TFA on lipid and lipoprotein metabolism

When compared with an isocaloric amount of carbohydrates, TFA from
partially hydrogenated oils increase serum total and LDL-cholesterol concen-
DIETARY TRANS FATTY ACIDS AND CARDIOVASCULAR DISEASE RISK 309
trations. In fact, the LDL-cholesterol increasing effect of TFA is comparable to

that of palmitic acid. Effects on HDL-cholesterol are comparable to those of
carbohydrates. As a result, TFA from hydrogenated sources have, in compari-
son with carbohydrates and other fatty acids, the most unfavourable effect on
the ratio of serum total cholesterol to HDL-cholesterol. Effects on apoB and
apo-A1 concentration followed those of LDL- and HDL-cholesterol. Finally,
exchanging carbohydrates for TFA does not affect serum triacylglycerol levels
(Mensink et al., 2003).
Trans isomers of monounsaturated fatty acids (MUFA) from partially hydro-
genated oils have been studied most intensively. In only one study, the focus
was on the effects of trans isomers of α-linolenic acid. In that multi-centre
study, it was found that these trans isomers might increase the plasma ratios of
total cholesterol to HDL-cholesterol, as compared to α-linolenic (Vermunt et
al., 2001). In another study, it was demonstrated that partially hydrogenated
fish oils also unfavourably affected lipid risk indicators for coronary heart
disease (Almendingen et al., 1995).
Like LDL, lipoprotein(a) [Lp(a)] is also related to an increased risk of CVD.
In several intervention studies, it has been demonstrated that TFA from
partially hydrogenated oils increased Lp(a) concentrations, as compared with
saturated fatty acids or cis-unsaturated fatty acids (Mensink et al., 1992; Nestel
et al., 1992).
Until recently, hardly anything was known on the effects of TFA from
ruminants on the serum lipoprotein profile. Lately, however, Chardigny and
co-workers (2008) have compared the effects on lipoprotein metabolism of 5%
of energy of TFA from hydrogenated sources with the same amount of TFA
from ruminant sources. In women, but not in men, TFA from ruminant sources
beneficially increased HDL-cholesterol concentrations. Increases in LDL-
cholesterol and triacylglycerol concentrations were also observed. For both
genders, the serum total to HDL-cholesterol ratio did not change. No effects on
Lp(a) were observed. In a second study (Motard-Bélanger et al., 2008) in which
only healthy men participated, the effects of TFA from ruminant sources were
studied at two different levels of intake (1.5% and 3.7% of energy). In addition,
TFA from hydrogenated sources were fed to the volunteers at an intake of 3.6%
of energy. It was found that at extreme high intakes, TFA (3.7% of energy) from
both ruminant and hydrogenated sources adversely affected cholesterol me-
tabolism. At intakes of 1.5% of energy, no effects of TFA from ruminant
sources on the serum lipoprotein profile were observed.
2. Effect of TFA on haemostatic function

Few human intervention studies have examined the effects of TFA on markers
for platelet aggregation, coagulation and fibrinolysis, three main determinants
of haemostatic function. Almendingen and co-workers (1996) reported that a
diet enriched in partially hydrogenated soybean oil unfavourably changed

concentrations of plasminogen activator inhibitor type 1 (PAI-1) antigen and
PAI-1 activity as compared with a diet rich in partially hydrogenated fish oil or
butter fat. Fibrinogen concentrations were adversely affected on butter com-
pared with partially hydrogenated fish oil. However, Mutanen and Aro (1997)
concluded that stearic (18:0) acid and TFA from a partially hydrogenated
vegetable oil had similar effects on markers of coagulation and fibrinolysis. In
contrast, collagen-induced platelet aggregation was positively changed on the
diet rich in TFA. On the other hand, in vitro thromboxane B2 production,
adenosine diphosphate (ADP)-induced platelet aggregation, and production of
endothelial prostacyclin (PGI2) were not different between the two diets.
Comparable effects of cis MUFA and trans MUFA on coagulation factors have
been reported by Louheranta and co-workers (1999). Also, trans isomers of
α-linolenic acid did not change haemostatic function when compared with
α-linolenic acid (Tholstrup et al., 2003). Sanders and co-workers (2000) did
not find any indications that a single meal rich in TFA affected postprandial
activation of factor VII differently when compared with meals enriched with
other fatty acids. Tholstrup and co-workers (2003) on the other hand found that
TFA, like other unsaturated fatty acids, increased postprandial FVII activation
and decreased that of tissue plasminogen activator concentrations, as compared
with saturated fatty acids.
3. Effect of TFA on inflammatory markers and endothelial cell function

The effects of TFA on blood lipids are now well characterized, in particular
effects on the LDL- to HDL-cholesterol ratio or the total to HDL-cholesterol
ratio. Also, the relationship between the total to HDL-cholesterol ratio and risk
of CVD is clearly established. Few recent studies suggest that the intake of TFA
can also affect other markers of the risk of CVD such as inflammation and
endothelial dysfunction. This is important as a great number of arguments
support a central role of inflammation in all phases of atherosclerosis (Ross,
1999; Basu et al., 2006) (Table 1).
Interleukin-6 (IL-6) and C-reactive protein (CRP, an acute-phase protein)
have been shown to be independent markers for prediction of cardiovascular
events (Libby, 2002; Ridker, 2007). CRP has been considered as a potential
contributor to inflammatory diseases including atherosclerosis as well as a
marker of cardiovascular risk. The plasma CRP concentration has been vali-
dated as a sensitive, independent biomarker and predictor of CVD (Yu and
Rifai, 2000). Recently, the development of a highly sensitive method to
improve the high-sensitivity C-reactive protein (hs-CRP) enables the detection
of subclinical inflammation. Clinically, levels of hs-CRP greater than 3 mg/l
indicate elevated risk for myocardial infarction and stroke, even among appar-
ently healthy individuals with low-to-normal lipid levels (Ndumele et al.,
Table 1. Relevant biomarkers of inflammation and their possible role in atherosclerosis (adapted from Basu et al., 2006).
Biomarkers of inflammation Potential sources Possible role in atherosclerosis
CRP, SAA, fibrinogen (acute phase proteins) Liver, adipose tissue, macrophages, smooth CRP associated with production of
muscle cells, endothelial cells inflammatory cytokines, chemokines, tissue
factor expression, chemotaxis of monocytes.
Downregulation of eNOS and
prostacyclin. Predisposition to a state of
hypercoagulability (increased PAI-1, and
decreased tPA).
Cytokines (IL-1β, IL-6, TNF, IL-7, TNF-α, Endothelial cells, macrophages, adipose tissue Pro-atherogenic and augment monocyte-
IL-18 etc) endothelial adhesion. IL-7 activates
monocytes and enhances production
of inflammatory cytokines. IL-18, a strong
independent predictor of mortality in patients
with coronary artery disease.
Chemokines (MCP-1, IL-8) Endothelial cells, macrophages Stimulate chemotaxis.
Adhesion molecules (ICAM, VCAM, Endothelial cells Promote monocyte–endothelial adhesion.
E-selectin, P-selectin)
PAI-1 (inhibitor of fibrinolysis) Endothelial cells, adipose tissue, macrophages Reduced plasma fibrinolysis.
Promotes atherothrombosis.
ICAM, intercellular adhesion molecule; MCP-1, monocyte chemoattractant protein-1; PAI-1, plasminogen activator inhibitor-1; SAA, serum amyloid A; VCAM,
vascular cell adhesion molecule.
DIETARY TRANS FATTY ACIDS AND CARDIOVASCULAR DISEASE RISK
311
2006). Otherwise, higher circulating levels of IL-6 or TNF-α as well as the

soluble TNF-α receptors 1 and 2 (sTNF-R1 and sTNF-R2) have been shown to
be predictive of heart failure mortality (Rauchhaus et al., 2000).
Several observational studies have reported a positive correlation between
diets with a high content of trans fatty acids and biomarkers of inflammation in
healthy persons. Mozaffarian and co-workers (2004) investigated the associa-
tions between TFA intake and concentrations of CRP, IL-6, sTNF-R1 and
sTNF-R2 systemic inflammatory markers in 823 generally healthy women
enrolled in the Nurses’ Health Study (NHS) and Nurses’ Health Study II
(NHSII). They found an interaction between TFA intake and CRP and/or IL-6
only in women with higher body mass index (BMI) (P for interaction = 0.03 for
each). The higher susceptibility effects of TFA intake on CRP and IL-6
production in women with higher BMI could be related to a response of the
adipose tissue: this endocrine tissue is indeed able to secrete proinflammatory
cytokines (TNF-α and IL-6). Otherwise, these authors found, after adjustment
for age, that TFA intake was positively associated with sTNF-R1 and sTNF-R2
(P for trend < 0.001 for each). Women in the highest quintile of dietary trans fat
intake had 10% higher circulating levels of sTNF-R1 and 12% higher levels of
sTNF-R2 levels than women in the lowest quintile of trans fat intake, which
reflect the activity of the TNF system. The authors proposed that activation of
the TNF system may represent a mediating step between TFA consumption and
risks of coronary artery disease and diabetes.
Lennie and co-workers (2005) studied the effect of consumption of trans
fatty acids on circulating levels of IL-6, TNFα and associated soluble receptors
in patients with heart failure (Lopez-Garcia et al., 2005). Indeed, it was
reported that the consumption of TFA increases the proinflammatory cytokines
in healthy persons (Mozaffarian et al., 2004) and that the heart failure pathol-
ogy induces a proinflammatory state and the increased proinflammatory cytokine
levels are associated with decreased survival of patients with heart failure
(Rauchhaus et al., 2000; Lennie et al., 2005). They found that trans fats were
related to TNFα levels, but not to soluble receptor levels: TNFα levels were
significantly elevated in patients with diets rich in trans fats, i.e. patients who
consumed more than 3 g/day of TFA. Circulating levels of IL-6, and soluble
TNF receptor levels, were not related to the TFA. The authors argue it was
possible that the inflammatory stimuli from heart failure pathology were
sufficient to mask the effects of dietary trans fats on soluble receptor levels
(Lennie et al., 2005).
Moreover, analysis of food frequency questionnaires in 1986 and 1990 from
the Nurses’ Health Study I cohort revealed that CRP levels were 73% higher
among those in the highest quintile of trans fat intake (i.e. 3.7 g/d), compared
with the lowest quintile (1.5 g/d); IL-6 levels were 17% higher and sTNFR-2
levels 5% higher. TFA intake was positively related to plasma concentration
of CRP (P = 0.009), sTNFR-2 (P = 0.002) (Lopez-Garcia et al., 2005).
Controlled dietary intervention studies demonstrate the modulation of mark-

ers of inflammation by specific fatty acids, especially. One of the first studies
which report an effect of the consumption of trans fatty acid on inflammatory
parameters was the study of Han and co-workers (2002). They found that the
32-day consumption of diets containing hydrogenated fats rich in trans fatty
acids increases the production of the inflammatory cytokines TNFα and IL-6 in
men and women with moderately elevated LDL cholesterol levels, which is
alone a factor for development of atherosclerosis. These two inflammatory
cytokines are implicated in the development of atherosclerosis. They affect
lipid metabolism by inhibiting lipoprotein lipase and stimulating lipolysis by
modifying lipid homeostasis at the gene expression and metabolite levels.
These include reduction of HDL-cholesterol, increase of LDL-cholesterol, and
elevated expression of cholesterogenic genes (Popa et al., 2007). Moreover,
TNF may induce and initiate vascular injury leading to endothelial dysfunction
(Madge and Pober, 2001).
In a study with 39% of energy from fat in the diet, the replacement of 8% of
the fat with TFA for 5 weeks led to an increase of the CRP and E-selectin
plasma concentration in healthy men (Baer et al., 2004). Motard-Belanger and
co-workers (2008) found that the circulating level of CRP was not statistically
different at the end of the 4-week experimental nutritional intervention for 38
healthy men between the following four experimental isoenergetic (2500 kcal)
diets: high in TFA from ruminants, 10.2 g; moderate in TFA from ruminants,
4.2 g; high in TFA from industrial sources, 10.2 g; and low in TFA from any
source, 2.2 g (control diet) (Motard-Bélanger et al., 2008; Mozaffarian et al.,
2004). The supplementation with food items containing TFA (11–12 g/d,
representing 5% of daily energy) from ruminants and from industrial sources
for 4 weeks, in a randomized, double blind, controlled, cross-over design has
no differential effect on the plasma hs-CRP concentration of 46 healthy
subjects (22 men and 24 women) (unpublished data, source Chardigny and co-
workers, 2008).
4. Effect of TFA on endothelial cell function

Increasing evidence indicates the important role of endothelial dysfunction in
the development of cardiovascular disease through induction by inflammatory
mediators (Ross, 1999; Basu et al., 2006). Moreover, dietary factors, especially
fatty acids, might influence cardiovascular risk. In order to assess the dietary
impact of trans fatty acids on the endothelium activation, Lopez-Garcia and co-
workers (2005) conducted a cross sectional study of 730 apparently healthy
women from the Nurses’ Health Study, aged between 43 and 69 years, and
measured the plasmatic concentrations of markers of endothelial activation
such as E-selectin, soluble intercellular adhesion molecule (sICAM)-1, and
soluble vascular cell adhesion molecule (sVCAM)-1. E-selectin is an adhesion
molecule that mediates the initial rolling of inflammatory cells along endothe-
lial cells and which is preferentially expressed on the endothelium overlaying
the atherosclerotic plaques. Intracellular adhesion molecules-1 (ICAM-1) and
vascular adhesion molecule-1 (VCAM-1) are thought to regulate attachment
and transendothelial migration of leukocytes (Ross, 1999, Basu et al., 2006).
Both macrophages and endothelial cells produce ICAM-1 in response to
inflammatory cytokines such as interleukin-1 (IL-1), and TNF-α (Steffens &
Mach, 2004). Biomarkers of endothelial dysfunction found to be higher in the
quintile of subjects with the highest consumption of trans fatty acids (3.7 g/d)
compared to those in the lowest quintile (1.5 g/d) were: E-selectin 20%, and
soluble cell adhesion molecules (sICAM-1 and sVCAM-1) both 10%. E-selectin,
sICAM-1 and sVCAM-1 are adhesion proteins whose concentrations are
increased in heart diseases and the early stage of atherosclerosis. TFA intake
was positively related to E-selectin (P = 0.003), sICAM-1 (P = 0.007) and
sVCAM-1 (P = 0.001) in linear regression models after controlling for age,
BMI, physical activity, smoking status, alcohol consumption, intake of mono-
, poly- and saturated fatty acids, and postmenopausal hormone therapy. They
concluded that trans fatty acids adversely affect endothelial function, which
might explain in part the association of these fatty acids with risk of cardiovas-
cular disease and a higher intake of trans fatty acids could favour inflammation
and adversely affect endothelial function (Lopez-Garcia et al., 2005).
5. Relation of cardiovascular disease risk to risk of type 2 diabetes

Impairment of insulin action (insulin resistance) is an important feature of
several diet-related chronic diseases such as obesity, hypertension and type 2
diabetes mellitus. Occurrence of insulin resistance precedes by several decades
the appearance of type 2 diabetes.
In vitro studies have clearly demonstrated that saturated fatty acids can alter
insulin sensitivity (Chavez et al., 2003), suggesting that excess dietary intake
of these fats may enhance the risk of developing type 2 diabetes. In that respect,
high intake of saturated fatty acids has been shown to induce insulin resistance
(Parker et al., 1993). Furthermore, study of the Normative Aging cohort
showed that saturated fatty acids remained a significant predictor of fasting and
postprandial insulin concentrations, even after adjustment for total body adi-
posity and body-fat distribution (Parker et al., 1993). Regarding TFA, reports
on insulin sensitivity are more limited and remain controversial. Among the
first studies published on that topic, that of Christiansen and co-workers (1997)
showed in obese and insulin resistant subjects that a diet high in industrial trans
MUFA induced an increase of postprandial insulinaemia, suggesting an in-
creased insulin resistance. An important body of evidence relating TFA intake
to risk of type 2 diabetes was first published by Salmeron and co-workers
(2001), and updated by Hu and co-workers (2001). These arguments came from
the Nurses Health Study which, over 14 years, prospectively followed 84 204
healthy women aged 34–59 years in 1980. The data suggested that total lipid
intake, compared with equivalent energy intake from carbohydrates, was not
associated with risk of type 2 diabetes in women (Salmeron et al., 2001). By
contrast, TFA increased this risk whereas PUFA reduced it. Indeed, for a 5%
increase in energy from PUFA, the relative risk was 0.63 (0.53, 0.76; P < 0.0001)
and for a 2% increase in energy from TFA the relative risk was 1.39 (1.15, 1.67;
P < 0.001). In addition, substituting non-hydrogenated PUFA for TFA sub-
stantially reduced the risk of type 2 diabetes. The authors estimated that
replacing 2% of energy from TFA isoenergetically with PUFA would lead to a
40% lower risk. Subgroup analyses revealed that effects were primarily ob-
served in obese women, possibly due to the fact that these women were already
more insulin resistant compared with non-obese women. Regarding the other
fatty acids, no relationships were observed between incidence of type 2 diabe-
tes and intakes of saturated fatty acids or MUFA (Salmeron et al., 2001). The
Iowa Women’s Health Study showed that diabetes incidence was negatively
associated with dietary PUFA, vegetable fat, and trans fatty acids and posi-
tively associated with omega-3 fatty acids (Meyer et al., 2001). Relative risks
for diabetes among quintiles of intake of TFA were 1.00, 1.01, 0.93, 1.09, 0.86
and 0.88 (P = 0.03) (Meyer et al., 2001). In contrast, the intakes of oleic acid,
trans fatty acids, long-chain n–3 fat, and linolenic acid were not associated with
risk of diabetes after multivariate adjustment in men included in the Health
Professionals Follow-up Study (van Dam et al., 2002). The authors explained
that the differences between these three studies came from the different
amounts of TFA consumed by the subjects. Indeed the median intake at
baseline was 1.2% of energy in their study vs 2.2% in the Nurses Health Study,
suggesting that adverse effects of TFA on insulin have been observed at high
levels of TFA (Christiansen et al., 1997). In addition, the sources of TFA were
not examined, so no conclusions could be brought regarding the respective
effects of dairy and hydrogenated fats. However, differences in the relative
abundance of each trans MUFA isomer in food products could probably lead
to different metabolic effects.
Short term intervention studies have not been very conclusive either. We
reported in Wistar rats that an 8 week diet enriched (4% of energy intake) in
either dairy, industrial, or control MUFA did not alter insulin and glucose
responses to an intraperitoneal glucose tolerance test (1 g/kg). Furthermore, we
found that in C2C12 myotubes, vaccenic (dairy TFA) and elaidic (industrial
TFA) acids did not modify insulin sensitivity compared with oleic acid. In
addition, four to five week dietary challenges have shown that dietary trans
MUFA from both sources (industrial vs ruminant) do not impair insulin
sensitivity in healthy subjects, even if there is an alteration of the lipoprotein
metabolism (Louheranta et al., 1999; Lovejoy et al., 2001; Tholstrup et al.,
2006). Therefore, increasing evidence suggests that dietary trans MUFA of
dairy or industrial origin at nutritional levels (< 4% energy intake) may not
impair insulin sensitivity in healthy subjects. However, this question remains to
be elucidated in subjects at risk for diabetes, because Christiansen and co-
workers (1997) observed the deleterious effect of a diet high in industrial trans
MUFA in obese and insulin resistant subjects.
C. Conclusion
To conclude, while trans MUFA from hydrogenated oils have been clearly
implicated in the pathogenesis of cardiovascular diseases, those from dairy
products may be less deleterious. Furthermore, dietary intakes of trans MUFA
from dairy origin are far from being high enough to constitute a serious threat
for public health. Therefore, regular moderate consumption of dairy fats could
be tolerated with respect to cardiovascular risks. Regarding the role of trans
MUFA from both origins on the incidence of type 2 diabetes, studies carried
out on the topic are still controversial. Mechanistic and intervention studies
showed no significant effect of trans MUFA on muscle insulin resistance,
suggesting that these fatty acids may not increase the risk for diabetes.
However, further intervention studies are required to examine the impact of
trans MUFA in individuals at risk for metabolic syndrome and diabetes.
References
Almendingen, K, Jordal, O, Kierulf, P, Sandstad, B and Pedersen, JI (1995) Effects of
partially hydrogenated fish oil, partially hydrogenated soybean oil, and butter on serum
lipoproteins and Lp[a] in men. J. Lipid Res., 36, 1370–1384.
Almendingen, K, Seljeflot, I, Sandstad, B and Pedersen, JI (1996) Effects of partially
hydrogenated fish oil, partially hydrogenated soybean oil, and butter on hemostatic
variables in men. Arterioscler. Thromb. Vasc. Biol., 16, 375–380.
Armstrong, RA, Chardigny, JM, Beaufrere, B, Bretillon, L, Vermunt, SH, Mensink, RP,
Macvean, A, Elton, RA, Sebedio, JL and Riemersma, RA (2000) No effect of dietary
trans isomers of alpha-linolenic acid on platelet aggregation and haemostatic factors in
european healthy men. The TRANSLinE study. Thromb. Res., 100, 133–141.
Ascherio, A, Rimm, EB, Giovannucci, EL, Spiegelman, D, Stampfer, M and Willett, WC
(1996) Dietary fat and risk of coronary heart disease in men: cohort follow up study in
the United States. Brit. J. Med., 313, 84–90.
Baer, DJ, Judd, JT, Clevidence, BA and Tracy, RP (2004) Dietary fatty acids affect plasma
markers of inflammation in healthy men fed controlled diets: a randomized crossover
study. Am. J. Clin. Nutr., 79, 969–973.
Basu, A, Devaraj, S and Jialal, I (2006) Dietary factors that promote or retard inflammation.
Arterioscler. Thromb. Vasc. Biol., 26, 995–1001.
Barbano, DM, Mensink, RP, Bezelgues, JB, Chaumont, P, Combe, N, Cristiani, I, Joffre,
F, German, JB, Dionisi, F, Boirie, Y and Sebedio, JL (2008) Do trans fatty acids from
vascular disease risk factors in healthy subjects? Results of the Trans Fatty Acids
Chavez, JA, Knotts, TAe , Wang, LP, Li, G, Dobrowsky, RT, Florant, GL and Summers,
SA (2003) A role for ceramide, but not diacylglycerol, in the antagonism of insulin signal
transduction by saturated fatty acids. J. Biol. Chem., 278, 10297–10303.
Christiansen, E, Schnider, S, Palmvig, B, Tauber-Lassen, E and Pedersen, O (1997) Intake
of a diet high in trans monounsaturated fatty acids or saturated fatty acids. Effects on
postprandial insulinemia and glycemia in obese patients with NIDDM. Diabetes Care,
20, 881–887.
Han, SN, Leka, LS, Lichtenstein, AH, Ausman, LM, Schaefer, EJ and Meydani, SN (2002)
Effect of hydrogenated and saturated, relative to polyunsaturated, fat on immune and
inflammatory responses of adults with moderate hypercholesterolemia. J. Lipid Res., 43,
445–452.
Hu, FB, Manson, JE, Stampfer, MJ, Colditz, G, Liu, S, Solomon, CG and Willett, WC (2001)
Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N. Engl. J. Med., 345,
790–797.
fatty acids and risk of coronary heart disease. Int. J. Epidemiol., 37, 173–182.
Lennie, TA, Chung, ML, Habash, DL and Moser, DK (2005) Dietary fat intake and
proinflammatory cytokine levels in patients with heart failure. J. Card. Fail., 11, 613–
618.
Libby, P (2002) Inflammation in atherosclerosis. Nature, 420, 868–874.
Lopez-Garcia, E, Schulze, MB, Meigs, JB, Manson, JE, Rifai, N, Stampfer, MJ, Willett, WC
and Hu, FB (2005) Consumption of trans fatty acids is related to plasma biomarkers of
inflammation and endothelial dysfunction. J. Nutr., 135, 562–566.
Louheranta, AM, Turpeinen, AK, Vidgren, HM, Schwab, US and Uusitupa, MI (1999) A
high-trans fatty acid diet and insulin sensitivity in young healthy women. Metabolism,
48, 870–875.
Lovejoy, JC, Champagne, CM, Smith, SR, DeLany, JP, Bray, GA, Lefevre, M, Denkins, YM
and Rood, JC (2001) Relationship of dietary fat and serum cholesterol ester and
phospholipid fatty acids to markers of insulin resistance in men and women with a range
of glucose tolerance. Metabolism, 50, 86–92.
Madge, LA and Pober, JS (2001) TNF signaling in vascular endothelial cells. Exp. Mol.
Pathol., 70, 317–325.
Mensink, RP, Zock, PL, Katan, MB and Hornstra, G (1992) Effect of dietary cis and trans
fatty acids on serum lipoprotein[a] levels in humans. J. Lipid Res., 33, 1493–1501.
Mensink, RP, Zock, PL, Kester, AD and Katan, MB (2003) Effects of dietary fatty acids and
apolipo-proteins: a meta-analysis of 60 controlled trials. Am. J. Clin. Nutr., 77, 1146–1155.
Meyer, KA, Kushi, LH, Jacobs, DR, Jr and Folsom, AR (2001) Dietary fat and incidence of
type 2 diabetes in older Iowa women. Diabetes Care, 24, 1528–1535.
P and Lamarche, B (2008) Study of the effect of trans fatty acids from ruminants on blood
Mozaffarian, D, Pischon, T, Hankinson, SE, Rifai, N, Joshipura, K, Willett, WC and Rimm,
EB (2004) Dietary intake of trans fatty acids and systemic inflammation in women. Am.
J. Clin. Nutr., 79, 606–612.
Mutanen, M and Aro, A (1997) Coagulation and fibrinolysis factors in healthy subjects
consuming high stearic or trans fatty acid diets. Thromb. Haemost., 77, 99–104.
Ndumele, CE, Pradhan, AD and Ridker, PM (2006) Interrelationships between inflammation,
C-reactive protein, and insulin resistance. J. Cardiometab. Syndr., 1, 190–196.
Nestel, P, Noakes, M, Belling, B, McArthur, R, Clifton, P, Janus, E and Abbey, M (1992)
Plasma lipoprotein lipid and Lp[a] changes with substitution of elaidic acid for oleic acid
in the diet. J. Lipid Res., 33, 1029–1036.
Oh, K, Hu, FB, Manson, JE, Stampfer, MJ and Willett, WC (2005) Dietary fat intake and risk
of coronary heart disease in women: 20 years of follow-up of the nurses’ health study. Am
J Epidemiol, 161, 672–679.
Oomen, CM, Ocke, MC, Feskens, EJ, van Erp-Baart, MA, Kok, FJ and Kromhout, D (2001)
Association between trans fatty acid intake and 10-year risk of coronary heart disease in
the Zutphen Elderly Study: a prospective population-based study. Lancet 357, 746–751.
Parker, DR, Weiss, ST, Troisi, R, Cassano, PA, Vokonas, PS and Landsberg, L (1993)
Relationship of dietary saturated fatty acids and body habitus to serum insulin concen-
trations: the Normative Aging Study. Am. J. Clin. Nutr. 58, 129–136.
Pietinen, P, Ascherio, A, Korhonen, P, Hartman, AM, Willett, WC, Albanes, D and Virtamo,
J (1997) Intake of fatty acids and risk of coronary heart disease in a cohort of Finnish men.
The Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study. Am. J. Epidemiol., 145,
876–887.
Popa, C, Netea, MG, van Riel, PL, van der Meer, JW and Stalenhoef, AF (2007) The role of
TNF-alpha in chronic inflammatory conditions, intermediary metabolism, and cardio-
vascular risk. J. Lipid Res., 48, 751–762.
Rauchhaus, M, Doehner, W, Francis, DP, Davos, C, Kemp, M, Liebenthal, C, Niebauer, J,
Hooper, J, Volk, HD, Coats, AJS, Stefan, DM, Anker, D (2000) Plasma cytokine
parameters and mortality in patients with chronic heart failure. Circulation, 102, 3060–
3067.
Ross, R (1999) Atherosclerosis is an inflammatory disease. Am. Heart J., 138, S419–420.
Ridker, PM (2007) C-reactive protein and the prediction of cardiovascular events among
those at intermediate risk: moving an inflammatory hypothesis toward consensus. J. Am.
Coll. Cardiol., 49, 2129–2138.
Salmeron, J, Hu, FB, Manson, JE, Stampfer, MJ, Colditz, GA, Rimm, EB and Willett, WC
(2001) Dietary fat intake and risk of type 2 diabetes in women. Am. J. Clin. Nutr., 73,
1019–1026.
Sanders, TA, de Grassi, T, Miller, GJ and Morrissey, JH (2000) Influence of fatty acid chain
length and cis/trans isomerization on postprandial lipemia and factor VII in healthy
subjects (postprandial lipids and factor VII). Atherosclerosis, 149, 413–420.
Steffens, S and Mach, F (2004) Inflammation and atherosclerosis. Herz, 29, 741–748.
Tholstrup, T, Miller, GJ, Bysted, A and Sandstrom, B (2003) Effect of individual dietary fatty
acids on postprandial activation of blood coagulation factor VII and fibrinolysis in
healthy young men. Am. J. Clin. Nutr., 77, 1125–1132.
Tholstrup, T, Raff, M, Basu, S, Nonboe, P, Sejrsen, K and Straarup, EM (2006) Effects of
butter high in ruminant trans and monounsaturated fatty acids on lipoproteins, incorpo-
ration of fatty acids into lipid classes, plasma C-reactive protein, oxidative stress,
hemostatic variables, and insulin in healthy young men. Am. J. Clin. Nutr., 83, 237–243.
van Dam, RM, Willett, WC, Rimm, EB, Stampfer, MJ and Hu, FB (2002) Dietary fat and meat
intake in relation to risk of type 2 diabetes in men. Diabetes Care, 25, 417–424.
Vermunt, SH, Beaufrere, B, Riemersma, RA, Sebedio, JL, Chardigny, JM, Mensink, RP and
TransLinE Investigators (2001) Dietary trans alpha-linolenic acid from deodorised
rapeseed oil and plasma lipids and lipoproteins in healthy men: the TransLinE Study. Br.
J. Nutr., 85, 387–392.
Weggemans, RM, Rudrum, M and Trautwein, EA (2004) Intake of ruminant versus industrial
trans fatty acids and risk of coronary heart disease — what is the evidence? Eur. J. Lipid
Sci. Technol.,106, 390–397.
Willett, WC, Stampfer, MJ, Manson, JE, Colditz, GA, Speizer, FE, Rosner, BA, Sampson,
LA and Hennekens, CH (1993) Intake of trans fatty acids and risk of coronary heart
disease among women. Lancet, 341, 581–585.
Yu, H and Rifai, N (2000) High-sensitivity C-reactive protein and atherosclerosis: from
theory to therapy. Clin. Biochem., 33, 601–610.
CHAPTER 12
Dietary trans fatty acids: from the mother’s
diet to the infant
JEAN-MICHEL CHARDIGNY1 AND NICOLE COMBE2
1
UMR 1019 INRA Université Clermont I, Clermont-Ferrand, France ; CRNH
Auvergne, Clermont-Ferrand, France.
2
ITERG, Département de Nutrition, Bordeaux, France.
A. Introduction
The occurrence of trans fatty acids (TFA) in human tissues depends on dietary
intake, as humans do not synthesize these fatty acid isomers. It is now well known
that TFA are provided by three different dietary sources : the ruminant fats (milk
and dairy products, meat, tallow) (Wolff et al., 1998) partially hydrogenated
vegetable oils (margarines produced with partially hydrogenated vegetable oils,
shortenings, bakeries and chips) (Craig-Schmidt, 1992) and to a lesser extent,
refined vegetable oils (Wolff, 1993). In the latter, TFA are mainly PUFA isomers,
whereas the other sources contain mainly trans monoenes.
Several in vivo animal studies (Anderson et al., 1975; Privett et al., 1967;
Berdeaux et al., 1998) and in vitro experiments in animal tissues (Mahfouz
et al., 1984; Cook and Emken, 1990; Berdeaux et al., 1998; Blond et al., 1995),
as well as human fibroblasts studies (Rosenthal and Doloresco, 1984), have
demonstrated that TFA impair the microsomal desaturation and chain elonga-
tion of linoleic acid and α-linolenic acid to long-chain polyunsaturated fatty
acids. Among all trans isomers, the 9t12t-18:2 isomer exerts the strongest
impairment of Δ6 desaturase activity (Anderson et al., 1975; Shimp et al.,
1982; Hwang and Kinsella, 1979; Brenner and Peluffo, 1969). In adults, the
inhibitory effect of TFA does not cause serious undesirable health hazards
when dietary linoleic acid is sufficient (Zevenbergen et al., 1988; Zevenbergen
and Haddeman, 1989). However, the infant and the foetus are sensitive to
factors that can interfere with the synthesis of long-chain polyunsaturated fatty
acids, due to an extensive need for arachidonic and docosahexaenoic acids for
the development of the neurological structures of the brain and retina (Innis
et al., 1999; Clandinin, 1999; Carlson and Neuringer, 1999) and for the
synthesis of eicosanoids (Sellmayer & Koletzko, 1999).
319
Table 1. Concentration of trans 18:1 acid isomers (% of total fatty acids) in human milk
in various countries (adapted from Mosley et al., 2005).
Authors Year Country TFA content (%)

Koletzko et al. 1988 Germany 3.1
Koletzko et al. 1991 Nigeria 1.2*
Boatella et al. 1993 Spain 0.2–4.5
Chardigny et al. 1995 France 1.9
Chen et al. 1995 Canada 7.2
Chen et al. 1997 China 0.22
Chen et al. 1997 Hong Kong 0.88
Precht & Molkentin 1999 Germany 2.5
Dlouhy et al. 2002 Czech Rep. 3.6
Mojska et al. 2003 Poland ≈ 2.4
Mosley et al. 2005 USA 5.1
Friesen & Innis 2006 Canada 4.6
Luna et al. 2007 Spain < 1.3
Szabo et al. 2007 Germany 0.86 (German, born in Germany)
1.16 (German, born elsewhere)
0.64 (Turkish)
0.99 (Others)
Tinoco et al. 2008 Brazil 2.2–2.3
* total TFA (i.e. not only trans-18:1 acid isomers).
B. TFA in infant nutrition
1. TFA in human milk

Since the late 1980s, the profile of fatty acids including TFA found in human
milk has been reported in the literature (Table 1). However, some data must
now to be considered with caution, because the TFA content of dietary fat has
decreased in the last decade as information on their deleterious effects has been
published (see other chapters).
The highest TFA content, 7.2% of fatty acids, was found in Canadian milk
samples (Chen et al., 1995). On the other hand, the content in French samples
collected during the same period was only 1.9% (Chardigny et al., 1995).
Similar results were reported for samples from Germany (Koletzko et al., 1988;
Precht & Molkentin, 1999) and Poland (Mojska et al., 2003), but samples from
Spain contained less TFA (Jensen, 1999, Luna et al., 2007). For Brazilian milk
samples, the TFA content was reported recently to be 2.34% in colostrum and
2.19% in mature milk (Tinoco et al., 2007). Mosley and co-workers (2005)
reported that in the USA human milk samples (mean of 81 samples) contain
about 5.1% of trans 18:1, a content close to those reported for Canada.
Friesen and Innis (2006) reported more recently that the TFA content of
Canadian breast milk decreased in the last decade, from more than 7% of total
fatty acids in 1998 to less than 5% in 2006.
DIETARY TFA: FROM THE MOTHER’S DIET TO THE INFANT 321
Table 2. Profile of trans-18:1 isomers (% of total trans-18 acid isomers) in breast milk
analysed in the Aquitaine study (adapted from Combe et al., 2005). SD, standard
deviation.
Δ Breast milk (n = 11)
Trans positional Mean SD Minimum Maximum

18:1 isomers
Δ6 + Δ8 6.08 4.30 0.80 13.73
Δ9 18.06 8.34 6.54 30.84
Δ10 17.22 6.39 9.31 31.73
Δ11 37.96 7.03 24.94 51.24
Δ12 7.33 2.36 3.82 11.52
Δ13 + Δ14 8.46 5.60 1.61 21.36
Δ15 2.66 1.68 0.53 5.80
Δ16 2.24 2.69 0.00 9.27
Clearly, it can be stated that TFA content in human milk depends on dietary
habits, with a higher content found in countries were consumption of TFA is
greatest. The trans 18:1 isomeric profile could also differ according to dietary
habits, such as the major source of TFA (partially hydrogenated vegetable oils
versus ruminant fat) as observed in the Aquitaine study (Combe et al., 1995).
In 1997, the mean trans 18:1 content of 27 mature breast milk samples was
1.78% of total fatty acids; dietary records of correspondent mothers showed
that ruminant fat was the main source of trans 18:1 (66%) compared with
(33%). Complete analysis of the trans positional 18:1 isomer profile of 11 milk
samples indicated a high proportion of Δ11 trans 18:1, i.e. 38% of total trans
18:1 isomers, the most prevalent isomer derived from ruminant fat (Table 2)
(Combe et al., 1995).
Recent data have shown that the TFA content of French human milk is
significantly lower in 2007 than in 1997 (Vaysse et al., 2008). In 2007, the
average trans 18:1 content of 142 mature milk samples collected from eight
human milk banks was 0.96 +/– 0.30% of total fatty acids versus 1.78 +/–
0.86% in 1997. This decreasing level of TFA observed in 2007 is probably
explained by a parallel decrease in level of TFA in margarine composition
(< 1% TFA) during this last decade.
2. TFA in infant formulas

Infant formulas are alternatives to human milk feeding and their TFA content
depends on the fat material used in processing. The TFA content of infant
formulas was reported as early as 1990, when Permanyer et al. analysed 28
samples from the Spanish market. The percentage of trans isomers ranged from
0.6 to 4.5% when determined by GC analysis, with trans 18:1 as the major
component. In the mid-1990s, we assessed the TFA content of infant formulas
available in France and in Canada. The range of trans 18:1 content was 0.6 to
3.1% in Canadian formulas, with a higher content in liquid items when
compared to powdered ones, and 0.2 to 4.3% in French formulas (Chardigny
et al., 1996). However, current data are lacking and the TFA content has
probably decreased since the 1990s, as a result of the abundant literature now
available describing adverse effects.
C. TFA: from the mother to foetus and infant

It has been clearly demonstrated that TFA cross the placenta and are transferred
to the developing human foetus
1. Placental transfer
In the early 1980s, it was demonstrated that [1-14C]-elaidic acid crosses the
placental barrier in rats (Moore & Dhopeshwarkar, 1980). However, contradic-
tory data were obtained in piglets from mothers fed a high trans diet (Pettersen
& Opstvedt, 1989). In humans, Johnston and co-workers (Johnston et al.,
1958) first reported that the transfer was limited. More recently, Koletzko and
Müller (1990) reported that cord blood in term infants presented the same TFA
content than maternal plasma lipids. Similar data were obtained in the Nether-
lands, with a direct correlation between trans 18:1 in maternal plasma and
foetal tissues (Van Houwelingen & Hornstra, 1994). The ILSI expert panel
further concluded that: “TFA are transferred by the placenta to the foetus and
incorporated into foetal tissues” (Carlson et al., 1997). TFA have been reported
to represent 1–2% of total fatty acids in blood lipid fractions in preterm and
term newborns as well as 1–12 month infants (Koletzko, 1992; Decsi &
Koletzko, 1994). The transfer is probably limited as illustrated by the low
content compared to TFA level in the maternal diet (Pettersen & Opstvedt,
1989). Comparison of the TFA content in maternal and umbilical plasma from
French pregnant women (Boue et al., 2001) was in accordance with this point
of view. TFA levels were significantly higher (p = 0.001) in maternal (0.9% of
total fatty acids) than in umbilical plasma total lipids (0.6%). Moreover, the
trans isomer pattern in cord plasma lipids was different from that of the mother
(Figure 1). The trans 18:1 acid level in maternal plasma (0.64% of total fatty
acids) exceeded significantly (p < 0.001) that found in umbilical plasma
(0.24%), while trans 18:2 acid percentage was significantly higher (p < 0.001)
in umbilical (0.30%) than in maternal plasma (0.16%). More precisely, the
9t12c-18:2 isomer and the sum 9c13t+9t12t-18:2 were significantly higher
( p < 0.001) in umbilical (0.15% and 0.08% of total fatty acids respectively)
than in maternal plasma (0.05%), whereas no difference was observed between
these two compartments with respect to 9c12t-18:2 isomer level (0.06% of total
fatty acids in maternal plasma and 0.07% in umbilical plasma, respectively).
1.2
1.1 Maternal plasma (n=90)
1.0 Umbilical plasma (n=77)
0.9 *
0.8
0.7
*
0.6
0.5
0.4
0.3 *
0.2
*
0.1 *
0.0
TFA trans trans trans 9c13t 9c12t 9t12c
16:1 18:1 18:2 + 9t12t 18:2 18:2
*: p<0.001 18:2
Figure 1. Trans fatty acid composition (% total fatty acids) of maternal and umbilical plasma total
lipids (adapated from Boué et al., 2001).
The prevalence of trans 18:2 isomers and more precisely of 9t12c-18:2 isomer
in umbilical plasma may be explained either by a preferential transfer of this
isomer across the placenta, as a consequence of sequestration process, or by an
accumulation in cord plasma linked to poor metabolism of this isomer by the
foetus compared with the other trans 18:2 isomers (Combe et al., 1997).
2. Incorporation into cord tissues

As for other fatty acids, TFA in maternal plasma lipids are positively correlated
with the levels in infant plasma and umbilical cord lipid at birth (Innis, 2005;
Elias & Innis, 2001). These data clearly suggest that the TFA transfer is related
to the maternal TFA content, which itself is dependent on the dietary intake of
TFA.
With the exception of trans 16:1 acids, all trans isomers present in umbilical
plasma were detected in both parietal and vessel phospholipids of umbilical
cord. The TFA content (expressed as total isomers) of cord wall (0.31% of total
fatty acids) did not differ significantly from that found in cord vein (0.28%) or
artery (0.35%) (Boue et al., 2001).
D. Effects of TFA on infant growth and development

Koletzko et al. (1992) reported an inverse relationship between birth weight
and elaidic acid content in cholesteryl esters in premature newborn babies.
Similarly, an inverse correlation has been reported between TFA and birth
weight in term born babies (Decsi & Koletzko, 1995), which may be correlated
with metabolism of essential fatty acids (see below). Thus, it has been sug-
gested that exposure to high levels of TFA during pregnancy may impair the
growth of the foetus.
However, the recent Aquitaine study failed to demonstrate any correlation
(Boue et al., 2000) between either newborn weight or head circumference
and the TFA level of maternal adipose tissue, as an indicator of the usual
TFA intake level. This intake was assessed at 3.1 g/day as mean value,
ranging from 0.9 to 6.5 g/day (or 0.5 to 2.5% of total energy intake). More-
over, no relationship was observed between the TFA present in maternal
plasma and either newborn weight or head circumference, even after multi-
factorial corrections.
E. TFA and essential fatty acid metabolism in newborns

TFA were reported to be inversely correlated with the content of n–6 PUFA in
foetus from abortion (Van Houwelingen and Hornstra, 1994). Five human
studies have reported inverse correlations between TFA and long-chain PUFA
(Koletzko et al., 1992; Elias & Innis, 2001; Decsi & Koletzko, 1995; Decsi
et al., 2001, 2002). In another study (Combe et al., 1997), no relation was
found between trans 18:1 and long-chain PUFA in artery of umbilical cord;
however, the total percentage of two trans 18:2 isomers (9c13t + 9t12t) was
inversely correlated (r = –0.703, p = 0.003) with arachidonic acid content, but
not with that of linoleic acid, the precursor of arachidonic acid. This is
consistent with data obtained in piglets from mothers fed TFA. The tissue
content of linoleic acid is increased whereas the content in long-chain PUFA
(i.e. arachidonic and docosahexaenoic acids) decreased (Pettersen & Opstvedt,
1992). This may be due to an alteration in bioconversion of C18 precursors, as
suggested in several studies (Van Houwelingen & Hornstra, 1994; Decsi &
Koletzko, 1995). However, the direct demonstration of such an alteration in
rodents was not convincing because animals were deficient in PUFA (Kurata &
Privett, 1980) or TFA intake was outside that of nutritional situations (De
Schrijver & Privett, 1982).
In a recently published cohort study which included 769 human milk
samples collected in Germany, Szabo and co-workers (2007) reported that
the availability of TFA may be inversely related to the bioavailability of
long-chain PUFA in human milk. This study confirms that decreasing the
TFA intake of mothers could be beneficial for infant development, through a
better absorption of long-chain PUFA from milk. A recent study carried out
in rats showed that the DHA status was altered by a very high TFA intake in
pregnant animals. However, the brain DHA content was not affected (Larque
et al., 2000).
F. Conjugated linoleic acid (CLA) isomers in infant nutrition

Research on this topic has been carried out only recently, in parallel with the
growing interest in CLA. In the first report, it was documented that CLA intake
significantly decreases fat content in human milk (Masters et al., 2002). This
deleterious effect has been confirmed in lactating rats (Ringseis et al., 2004),
where it was associated with an alteration of growth in the newborn rats. As
reported for other animal species, trans-10,cis-12 18:2 acid isomer depresses
the fat content of milk. As a consequence, the energy intake of the newborns is
strongly decreased and this further impairs development. Recently, Hayashi
et al. investigated the impact of intake of a CLA mixture on lipogenic enzymes
in lactating rats. They demonstrated clearly an inhibition of fatty acid synthase,
glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase
activities (Hayashi et al., 2007).
Elias and Innis (2001) reported that CLA is able to cross the placenta and that
CLA concentration in foetal plasma is twice that of maternal plasma. On the
other hand, CLA content in plasma triacylglycerols is inversely correlated to
duration of gestation and weight and length at birth.
Some information is available on CLA content in human milk. In a recent
study from Spain, Luna et al. (2007) reported that human milk contains 0.12–
0.15% total fatty acids as CLA, with rumenic acid being the major component.
This is close to the 0.14–0.28% reported earlier for the USA (Jensen et al.,
1998) and 0.22–0.54% for Russia (McGuire et al., 1997) but values are lower
for Africa (< 0.1%) (Glew et al., 2006). Compared to human milk, infant
formulas contain less CLA (McGuire et al., 1997), but no recent data are
available.
G. Conclusion
High intakes of TFA have been reported to have deleterious effects on infant
development. This could be related to interference with the metabolism of
essential fatty acids. However, under nutritional conditions, the impact seems
to be limited, except in the case of CLA supplementation. As a result of the
decreased use of hydrogenated fats in food products and the consequent fall in
the dietary intake of TFA, interference of dietary trans fatty acids with infant
development is not a major issue nowadays.
References
Anderson, RL, Fullmer, CSJ and Hollenbach, EJ (1975) Effects of the trans isomers of
linoleic acid on the metabolism of linoleic acid in rats. J. Nutr., 105, 393–400.
Berdeaux, O, Blond, JP, Bretillon, L, Chardigny, JM, Mairot, T, Vatèle, JM, Poullain, D and
Sébédio, JL (1998) In vitro desaturation or elongation of monotrans isomers of linoleic
acid by the rat liver microsomes. Mol. Cell. Biochem., 85, 17–25.
Blond, JP, Chardigny, JM, Sébédio, JL and Grandgirard, A (1995) Effects of dietary 18:3 n–
3 trans isomers on the Δ6 desaturation of α-linolenic acid. J. Food Lipids, 2, 99–106.
Boue, C, Combe, N, Billeaud, C and Entressangles, B (2001) Nutritional implications of trans
fatty acids during perinatal period in French pregnant women. Ol. Corps Gras, Lipides,
8, 68–72.
Boue, C, Combe, N, Billeaud, C, Mignerot, C, Entressangles, B, Thery, G, Geoffrion, H,
Brun, JL, Dallay, D and Leng, JJ (2000) Trans fatty acids in adipose tissue of French
Brenner, RR and Peluffo, RO (1969) Regulation of unsaturated fatty acid biosyntheses. I.
Effect of unsaturated fatty acid of 18 carbons on the microsomal desaturation of linoleic
acid into γ-linolenic acid. Biochim. Biophys. Acta, 176, 471–479.
Carlson, SE, Clandinin, MT, Cook, HW, Emken, EA and Filer, LJ (1997) Trans fatty acids:
infant and fetal development Am. J. Clin. Nutr., 66, S717–S736.
Carlson, SE and Neuringer, M (1999) Polyunsaturated fatty acid status and neurodevelopment:
a summary and critical analysis of the literature Lipids, 34, 171–178.
Chardigny, JM, Wolff, RL, Mager, E, Bayard, CC, Sébédio, JL, Martine, L and Ratnayake,
WMN (1996) Fatty acid composition of French infant formulas with emphasis on the
content and detailed profile of trans fatty acids J. Am. Oil Chem. Soc., 73, 1595–1601.
Chardigny, JM, Wolff, RL, Mager, E, Sébédio, JL, Martine, L and Juanéda, P (1995) Trans
mono- and poly-unsaturated fatty acids in human milk Eur. J. Clin. Nutr., 49, 523–531.
Chen, ZY, Pelletier, G, Hollywood, R and Ratnayake, WM (1995) Trans fatty acid isomers
in Canadian human milk Lipids, 30, 15–21.
Clandinin, MT (1999) Brain development and assessing the supply of polyunsaturated fatty
acid Lipids, 34, 131–137.
Combe, N, Billeaud, C, Mazette, B, Entressangles, B and Sandler, B (1995) TFAs in human
milk reflect animal or vegetable TFA consumption in France. Pediatr. Res., 37, 304A.
Combe, N, Judde, A, Billeaud, C, Boue, C, Turon, F, Entressangles, B, Dallay, D and Leng,
JJ (1997) In: Essential Fatty Acids and Eicosanoids (Riemersma, RA, Armstrong, RA,
Kelly, RW and Wilson, R, eds) AOCS Press, Champaign, illinois, USA, pp.239–242.
Cook, HW and Emken, EA (1990) Geometric and positional fatty acid isomers interact
differently with desaturation and elongation of linoleic acid in cultured glioma cells.
Cellular Biology, 68, 653–660.
Craig-Schmidt, MC (1992) In: Fatty Acids in Foods and their Health Implications (Chow,
CK, ed.) Marcel Dekker, New York, USA, pp.365–398.
De Schrijver, R and Privett, OS (1982) Interrelationship between dietary trans fatty acids and
the 6- and 9-desaturases in the rat. Lipids, 17, 27–34.
Decsi, T, Boehm, G, Tjoonk, HM, Molnar, S, Dijck-Brouwer, DA, Hadders-Algra, M,
Martini, IA, Muskiet, FA and Boersma, ER (2002) Trans isomeric octadecenoic acids are
related inversely to arachidonic acid and DHA and positively related to mead acid in
umbilical vessel wall lipids. Lipids, 37, 959–965.
Decsi, T, Burus, I, Molnar, S, Minda, H and Veitl, V (2001) Inverse association between trans
isomeric and long-chain polyunsaturated fatty acids in cord blood lipids of full-term
infants. Am. J. Clin. Nutr., 74, 364–368.
Decsi, T and Koletzko, B (1994) Fatty acid composition of plasma lipid classes in healthy
subjects from birth to young adulthood. Eur. J. Pediatr., 153, 520–525.
Decsi, T and Koletzko, B (1995) Do trans fatty acids impair linoleic acid metabolism in
children? Ann. Nutr. Metab., 39, 36–41.
Elias, SL and Innis, SM (2001) Infant plasma trans, n–6, and n–3 fatty acids and conjugated
linoleic acids are related to maternal plasma fatty acids, length of gestation, and birth
weight and length. Am. J. Clin. Nutr., 73, 807–814.
Friesen, R and Innis, SM (2006) Trans fatty acids in human milk in Canada declined with the
introduction of trans fat food labeling. J. Nutr., 136, 2558–2561.
Glew, RH, Herbein, JH, Moya, MH, Valdez, JM, Obadofin, M, Wark, WA and Vanderjagt,
DJ (2006) Trans fatty acids and conjugated linoleic acids in the milk of urban women and
nomadic Fulani of northern Nigeria. Clin. Chim. Acta, 367, 48–54.
Hayashi, AA, de Medeiros, SR, Carvalho, MH and Lanna, DP (2007) Conjugated linoleic
acid (CLA) effects on pups growth, milk composition and lipogenic enzymes in lactating
rats. J. Dairy Res., 74, 160–6.
Hwang, DH and Kinsella, JE (1979) The effects of trans trans methyl linoleate on the
concentration of prostaglandins and their precursors in rat. Prostaglandins, 17, 543–59.
Innis, SM (2005) Essential fatty acid transfer and fetal development. Placenta, 26 Suppl A,
S70–S75.
Innis, SM, Sprecher, H, Hachey, D, Edmond, J and Anderson, RE (1999) Neonatal
polyunsaturated fatty acid metabolism. Lipids, 34, 139–149.
Jensen, RG, Lammi-Keefe, CJ, Hill, DW, Kind, AJ and Henderson, R (1998) The
anticarcinogenic conjugated fatty acid, 9c, 11t-18:2, in human milk: confirmation of its
presence. J. Hum. Lact., 14, 23–27.
Johnston, PV, Kummerow, FA and Walton, CH (1958) Origin of the trans fatty acids in
human tissue. Proc. Soc. Exp. Biol. Med., 99, 735–736.
Koletzko, B (1992) Trans fatty acids may impair biosynthesis of long-chain polyunsaturates
and growth in man. Acta Paediatr., 81, 302–306.
Koletzko, B, Mrotzek, M and Bremer, HJ (1988) Fatty acid composition of mature human
milk in Germany. Am. J. Clin. Nutr., 47, 954–959.
Koletzko, B and Muller, J (1990) Cis- and trans-isomeric fatty acids in plasma lipids of
newborn infants and their mothers. Biol. Neonate, 57, 172–178.
Koletzko B, Thiel I and Abiodun PO (1991) Fatty acid composition of mature human milk
in Nigeria. Z. Ernahrungswiss, 30, 289–297.
Koletzko, B, Thiel, I and Springer, S (1992) Lipids in human milk: a model for infant
formulae? Eur. J. Clin. Nutr., 46, S45–S55.
Kurata, N and Privett, O (1980) Effects of dietary trans acids on biosynthesis of arachidonic
acid in rat liver microsomes. Lipids, 15, 1029–1036.
Larque, E, Perez-Llamas, F, Puerta, V, Giron, MD, Suarez, MD, Zamora, S and Gil, A (2000)
Dietary trans fatty acids affect docosahexaenoic acid concentrations in plasma and liver
but not brain of pregnant and fetal rats. Pediatr. Res., 47, 278–283.
Luna, P, Juarez, M and De la Fuente, MA (2007) Fatty acid and conjugated linoleic acid
isomer profiles in human milk fat. Eur. J. Lipid Sci. Technol, 109, 1160–1166.
Mahfouz, MM, Smith, TL and Kummerow, FA (1984) Effect of dietary fats on desaturase
activities and the biosynthesis of fatty acids in rat-liver microsomes. Lipids, 19, 214–
222.
Masters, N, McGuire, MA, Beerman, KA, Dasgupta, N and McGuire, MK (2002) Maternal
supplementation with CLA decreases milk fat in humans. Lipids, 37, 133–138.
Mc Guire, MK, Park, Y, Behre, RA, Harrison, LY, Shultz, TD and Mcguire, MA (1997)
Conjugated linoleic acid concentrations of human milk and infant formula. Nutrition
Research, 17, 1277–1283.
Mojska, H, Socha, P, Socha, J, Soplinska, E, Jaroszewska-Balicka, W and Szponar, L (2003)
Trans fatty acids in human milk in Poland and their association with breastfeeding
mothers’ diets. Acta Paediatr., 92, 1381–1387.
Moore, CE and Dhopeshwarkar, GA (1980) Placental transport of trans fatty acids in the rat.
Lipids, 15, 1023–1028.
Mosley, EE, Wright, AL, McGuire, MK and McGuire, MA (2005) Trans fatty acids in milk
produced by women in the United States. Am. J. Clin. Nutr., 82, 1292–1297.
Permanyer, JJ, Pinto, BA, Hernandez, N and Boatella, J (1990) Trans isomer content in infant
milk formulas. Lait, 70, 307–311.
Pettersen, J and Opstvedt, J (1989) Trans fatty acids. 3. Fatty acid composition of the brain
and other organs in the newborn piglet. Lipids, 24, 616–624.
Pettersen, J and Opstvedt, J (1992) Trans fatty acids. 5. Fatty acid composition of lipids of
the brain and other organs in suckling piglets. Lipids, 27, 761–769.
including conjugated cis delta 9, trans delta 11 linoleic acid (CLA) as well as total fat
composition of German human milk lipids. Nahrung, 43, 233–244.
Privett, OS, Stearns, EM and Wickell, EC (1967) Metabolism of the geometrical isomers of
linoleic acid in the rat. J. Nutr., 92, 303–310.
Ringseis, R, Saal, D, Muller, A, Steinhart, H and Eder, K (2004) Dietary conjugated linoleic
acids lower the triacylglycerol concentration in the milk of lactating rats and impair the
growth and increase the mortality of their suckling pups. J. Nutr., 134, 3327–34.
Rosenthal, MD and Doloresco, MA (1984) The effects of trans fatty acids on fatty acyl delta 5
desaturation by human skin fibroblasts. Lipids, 19, 869–874.
Sellmayer, A and Koletzko, B (1999) Long-chain polyunsaturated fatty acids and eicosanoids
in infants — physiological and pathophysiological aspects and open questions. Lipids,
34, 199–205.
Shimp, JL, Bruckner, G and Kinsella, JE (1982) The effects of dietary trilinelaidin on fatty
acid and acyl desaturases in rat liver. J. Nutr., 112, 722–735.
Szabo, E, Boehm, G, Beermann, C, Weyermann, M, Brenner, H, Rothenbacher, D and Decsi,
T (2007) Trans octadecenoic acid and trans octadecadienoic acid are inversely related to
long-chain polyunsaturates in human milk: results of a large birth cohort study. Am. J.
Clin. Nutr., 85, 1320–1326.
Tinoco, SM, Sichieri, R, Setta, CL, Moura, AS and Tavares do Carmo, MD (2008) Trans fatty
acids from milk of Brazilian mothers of premature infants. J. Paediatr. Child Health, 44,
50–56.
Van Houwelingen, AC and Hornstra, G (1994) Trans fatty acids in early human development.
World Rev. Nutr. Diet, 75, 175–178.
Vaysse, C, Couedelo, L, Guesnet, P, Alessandri, JM, Billeaud, C, Putet, G and Combe, N
(2008). Abstract of a paper presented at the 6th EuroFedLipid Congress, 7–10 September
2008, Athens, Greece.
Wolff, RL (1993) Heat-induced geometrical isomerization of linolenic acid: effect of
temperature and heating time on the appearance of individual isomers. J. Am. Oil Chem.
Soc., 70, 425–430.
Wolff, RL, Precht, D and Molkentin, J (1998) In: Trans Fatty Acids in Human Nutrition, First
Edition (Sebedio, JL and Christie, WW, eds) The Oily Press, Bridgwater, UK, pp.1–33.
Zevenbergen, JL and Haddeman, E (1989) Lack of effects of trans fatty acids on eicosanoid
biosynthesis with adequate intakes of linoleic acid. Lipids, 24, 555–63.
Zevenbergen, JL, Houtsmuller, UM and Gottenbos, JJ (1988) Linoleic acid requirement of
rats fed trans fatty acids. Lipids, 23, 178–86.
CHAPTER 13
Evolution of worldwide consumption of trans
fatty acids
MARGARET C. CRAIG-SCHMIDT AND YINGHUI RONG
Department of Nutrition and Food Science, Auburn University, Auburn, USA
Trans fatty acids (TFA) from industrial processing of vegetable oils entered the
food supply in the USA in 1911 with the commercialization of ‘Crisco’, the first
partially hydrogenated shortening marketed for frying and baking. This prod-
uct was developed in part from the work of Paul Sabatier, a French chemist who
won the 1912 Nobel Prize in Chemistry for his method of hydrogenating
organic compounds in the presence of metal catalysts. Wilhelm Normann, a
German chemist, extended Sabatier’s work and in 1902 patented the process by
which liquid oils could be hydrogenated. In 1909, the Proctor and Gamble
company acquired the US rights to Normann’s patent and within two years
began marketing Crisco. This product became widely used in households in the
USA as a result of the distribution of free cookbooks in which every recipe
called for Crisco.
Partial hydrogenation in which liquid vegetable oils are converted to solid or
semi-solid fats, such as margarine and shortening, had several advantages for
the food industry. The consistency of the fat could be controlled by the degree
of hydrogenation, and use of these semi-solid fats resulted in desirable qualities
for baked goods. The partially hydrogenated fats were less susceptible to
oxidation and thus afforded a longer shelf life to commercial products.
Prior to the early 1900s, the only sources of trans isomers in the food supply
were meat and dairy products. Biohydrogenation in the rumen of animals
results in the formation of vaccenic acid (trans-11 18:1), the predominant trans
fatty acid in milk and meat, as well as smaller amounts of other TFA. Ruminant
products contain 1 to 8% of the fat as TFA.
Industrial processing of soybean and other vegetable oils results in products
such as margarine, shortening and frying fats that contain TFA at levels up to
40 to 50% of total fatty acids (Craig-Schmidt & Teodorescu, 2008). The
predominant trans isomers in partially hydrogenated vegetable fat are trans-9
18:1 and trans-10 18:1. In addition to trans isomers of 18:1 in the diet, smaller
amounts of trans isomers of 18:2 and even smaller amounts of trans-16:1 have
329
been reported (Lemaitre et al., 1998). Marine oils have also been hydrogen-
ated, resulting in a variety of trans isomers. Deep-fried fast foods, as well as
bakery goods, snack foods and other commercial products made with industri-
ally hydrogenated fat, are major sources of dietary TFA (Craig-Schmidt and
Teodorescu, 2008). Thus, populations consuming relatively large amounts of
commercially hydrogenated fat have a greater dietary TFA intake than do
populations whose dietary fat is primarily derived from ruminant sources.
Populations who use olive oil and other non-hydrogenated vegetable oils as the
primary dietary fat have relatively minor amounts of TFA in the diet.
Health professionals originally promoted margarine and other partially hydro-
genated fats as part of a diet to lower the risk for cardiovascular disease and other
chronic diseases because these products were derived from vegetable rather than
animal sources. However, since the early 1990s, evidence has accumulated that
TFA have a detrimental effect on risk for cardiovascular disease and other
chronic diseases. The landmark work of Mensink and Katan (1990) and Zock and
Katan (1992) led to a revision in thinking about TFA. These investigators
demonstrated that dietary TFA had an adverse effect on lipoprotein profile,
decreasing high-density lipoproteins or HDL and increasing low-density lipo-
proteins or LDL. This finding, as well as a myriad of other results on the harmful
effects of TFA (as reviewed by Hunter, 2006; Mozaffarian, et al. 2006; Stender
& Dyerberg, 2003; and many others), has led to efforts to remove industrially
produced TFA from the food supply either by regulation of industry, by changing
food labelling laws, or by local regulation of restaurants and fast food chains. A
variety of ‘trans-free’ foods are now available to the consumer and alternative
fats for frying and baked goods are beginning to enter the market place.
The purpose of this review is to summarize data on the worldwide consump-
tion of TFA from an historical perspective, with an emphasis on trends in
consumption since the 1990s. Evidence is presented that trans fatty acid
consumption in North America and Europe has decreased over the last 10
to15 years. Earlier reviews or government reports which summarize estimates
of TFA in the diet include a report by the Life Sciences Research Office for the
US Food and Drug Administration compiled by Senti (1985), the International
Life Sciences Institute Expert Panel on Trans Fatty Acids and Coronary Heart
Disease (1995), the American Society for Clinical Nutrition/American Insti-
tute of Nutrition Task Force on Trans Fatty Acids (1996), the International Life
Sciences Institute Expert Panel on Trans Fatty Acids and Early Development
(1997) and Aro (1998). A comprehensive review of worldwide consumption of
TFA through the late 1990s was published by Craig-Schmidt (1998), and the
present chapter seeks to update this earlier review. In addition, a brief overview
of worldwide consumption was published as part of a report on the ‘First
International Symposium on Trans Fatty Acids and Health’ (Craig-Schmidt,
2006). A review dealing with trans fatty acid consumption in only North
America has been published recently (Craig-Schmidt, 2007).
EVOLUTION OF WORLDWIDE CONSUMPTION OF TRANS FATTY ACIDS 331
A. Methods used to estimate trans fatty acid consumption
Methods used to estimate TFA in the diet include: (1) estimates based on ‘food
disappearance’ or market share data; (2) analysis of dietary consumption data
of a representative population; (3) laboratory analysis of duplicate portion or
composite diets; and (4) biomarkers. Each of these methods has inherent
advantages and disadvantages; however, collectively these methods result in
reasonable estimates of trans fatty acid consumption which are useful in
comparing consumption among various populations or identifying the relation-
ship between trans fatty acid intake and heath effects.
Estimates based on food disappearance or market share data tend to overes-
timate trans fatty acid consumption. This method has the advantage that it
accounts for fat in the total population. However, estimates from market share
data can vary widely depending on the ‘typical’ value for the trans fatty acid
content of each commodity or on the ‘wastage’ factor used in the estimates.
Data from US government commodity sales information were used in an early
estimation of trans fatty acid intake in the US population (Senti, 1985). The per
capita estimates at this time were 8 g/day from vegetable fats and 2.21 g/day
from animal and dairy fats, giving a total of 10.21 g/day for total trans fatty acid
intake. After a wastage factor was applied, trans fatty acid intake in the 1980s
was believed to be 8.3 g/day/person in the USA. Estimates using this method
are not commonly reported in the literature at the present time; however,
Johansson et al. (2006) used this method to estimate per capita trans fatty acid
consumption in Norway as 1.6 g/day based on a survey among Norwegian food
manufacturers and importers in 2003.
Laboratory analysis of duplicate portion or composite diets can also be used
to estimate trans fatty acid intake of a population. In this method, diets are
collected from a representative population or formulated based on a diet
composition considered to be ‘typical’ of a population and then analysed in a
laboratory. The advantage of this method is that dietary TFA can be determined
with a high degree of accuracy, especially if sophisticated analytical method-
ology is used; a disadvantage is that laboratory analyses are very expensive
and, thus, only a small number of diets can be used. Additionally, this method
is useful only to the extent that the subset of the population from which the diets
are collected is representative of the entire population or only to the extent that
the composite diets are reflective of the diet of the population as a whole. This
method was used successfully in the Seven Countries Study (Kromhout et al.,
1995) in which the trans fatty acid content of food composites was determined
for The Netherlands, Finland, the USA, former Yugoslavia, Italy, Greece and
Japan. The differences in consumption of TFA among these countries were
clearly illustrated. Kromhout and colleagues (1995) showed that in 1987 intake
of TFA was greater (4 to 6 g/day/person) in northern European countries and
the USA than in eastern European or Mediterranean countries (less than 1.5 g/
day/person) and much greater than in Japan (less than 0.5 g/day/person). Thus,
this method appears to be particularly useful in cross-country comparisons in
which the diets of the countries vary widely.
The most commonly used method to estimate TFA consumption in the
decade from 1995 to 2005 was based on analysis of dietary consumption data
of a representative population. Diet recalls, diet records, and food frequency
questionnaires can all be used to collect data on the diets of individuals. The
amounts of various food items consumed can then be combined with data from
food composition tables to give the average daily consumption of a nutrient or
other component in the diet. This method has the advantage that consumption
by the individual rather than the population is assessed. The main problem with
this method in light of a rapidly changing food supply is that values for trans
fatty acid content in the food composition tables are outdated quickly. These
values are subject to inaccuracies not only because of changes in food process-
ing, but also because of the wide variability in trans fatty acid content of similar
types of foods and differences in the analytical methods used to determine the
trans fatty acid content of each food item. The inherent problems with
assessing the diet of an individual, as well as the limitations of food composi-
tion databases for the trans fatty acid values, tend to result in underestimations
of the trans fatty acid content of the diet compared to values obtained by other
methods.
Innis et al. (1999) have clearly illustrated the problems with dietary method-
ology in assessing the intake of TFA. By analysing over 200 Canadian food
items, these investigators demonstrated the variability in trans fatty acid
content among food items or brands within a product category. For example, in
17 brands of crackers, trans fatty acid content ranged from 23 to 51% of total
fatty acids or differences of 1 to 13 g TFA per 100 g of crackers. The
significance of this variability to the estimation of trans fatty acid consumption
from dietary intake data was illustrated by analyses of a single diet record.
Using minimum and maximum values for TFA within a given food category,
the investigators reported that values for trans fatty acid intake could range
from a low of 1.4 to a high of 25.4 g a day. Thus, the wide variability in trans
fatty acid content of different food items within a category may result in large
errors in the estimation of trans fatty acid intake of individuals and groups.
Because of these unavoidable inaccuracies for trans fatty acid values in food
composition tables and the inherent problems in assessing the diet of individu-
als, the use of methods requiring a food composition database to estimate
dietary intake of TFA has decreased. Instead, the use of biomarkers is believed
to reflect better the actual individual consumption of TFA.
In recent years, the use of biomarkers of trans fatty acid consumption has
become increasingly popular in studies designed to relate the consumption of
TFA to health risk for various diseases. Several biological markers have been
used to estimate trans fatty acid intake. Adipose tissue is believed to reflect
long-term consumption, i.e., that of the previous year (Beynen et al., 1980;
Hirsch et al., 1960), whereas the fatty acid composition of erythrocytes reflects
consumption over the previous months and that of plasma or serum over the
previous weeks. Human milk, on the other hand, reflects the trans fatty acid
composition of the maternal diet of the previous day (Craig-Schmidt et al.,
1984). The tissue values can be used directly in making comparisons between
groups; however, if one wants a value for dietary intake, regression equations
relating tissue values to dietary values must be used. Enig et al. (1990) derived
an equation for estimating trans fatty acid intake from the concentration of
trans-18:1 in adipose tissue. In a similar manner, Craig-Schmidt et al.(1984)
and Wolff (1995) derived equations applicable to converting human milk
values to dietary values for TFA.
Advantages for using biomarkers as surrogate values for dietary intake
include the fact that they are directly reflective of the diet of the individual from
which the sample is obtained, estimates are not dependent on food composition
tables which are constantly changing, and biomarkers are good for compari-
sons among countries or for relating trans fatty acid consumption to risk for
various diseases. One disadvantage is that the amounts of TFA in human milk
or adipose tissues may be influenced by factors such as caloric intake. Com-
parisons are best if the analytical methods are the same, i.e., if the samples are
analysed in the same laboratory. Differences in analytical methodology may
make data difficult to compare among different laboratories. Additionally, the
equations used to convert tissue values to dietary values may have limited
applicability.
Aro et al. (1995) used the fatty acid composition of adipose tissue as a
biomarker of consumption in the EURAMIC study for a cross-country com-
parison of nine countries in studying the relationship of TFA to risk of
myocardial infarction. In 1991–92, trans-18:1 in adipose tissue in men from
The Netherlands, Norway, UK and Israel was over 2%, compared to about
1.5% for Switzerland, Finland, Russia and Germany. Men from Spain, on the
other hand, had much less trans-18:1 in their adipose tissue (less than 0.5%).
Thus, the trans-18:1 content of adipose tissue seems to reflect consumption of
TFA in various countries.
Other investigators have used the trans fatty acid content of adipose tissue
as a marker of trans fatty acid consumption in relationship to the risk for
sudden cardiac death (Roberts et al., 1995) and myocardial infarction
(Clifton et al., 2004; Baylin et al., 2002, 2003). Adipose tissue aspirates can
be used also to indicate the intake of exogenous fatty acids, distinguishing
between the consumption of vegetable fats and those of animal origin (Gar-
land et al. 1998). For example, Combe et al. (1998) used the trans isomer
distribution in adipose tissue of pregnant French women to study the relative
amounts of ruminant fats versus industrially produced hydrogenated fats in
the French diet compared with that of other Western countries. Thus, adipose
tissue can serve as a valuable biomarker of both the amount and type of TFA
in the habitual diet.
Human milk as a biomarker has been particularly useful in cross-country
comparisons of trans fatty acid consumption (Craig-Schmidt, 2006; Wolff,
1995) as well as in documenting a decrease in trans fatty acid intake response
to an effort to decrease trans fat in the food supply (Friesen & Innis, 2006).
Also, distribution of positional trans isomers of the fatty acids in breast milk,
like adipose tissue, appears to reflect the predominant source of TFA in the
maternal diet. The presence of relatively large amounts of vaccenic acid (trans-
11 18:1) and the relative amount of trans-16 18:1 indicate that milk and other
dairy products are the major sources of TFA in the maternal diet (Chardigny
et al. 1995; Wolff, 1995). On the other hand, an almost equal distribution of
trans-11 18:1 and trans-10 18:1 indicates that the TFA in the maternal diet are
derived primarily from commercially hydrogenated fats (Chen et al., 1995a,b).
Values for the trans fatty acid content of human milk in various countries
reported prior to 1998 have been summarized in several reviews (Wolff et al.,
1998; Jensen, 1999; Craig-Schmidt, 2001). More recent values for TFA in
human milk as percent of total fatty acids (wt/wt) include 7.1 ± 0.32% in 103
Canadian women (Innis & King, 1999), 3.81 ± 0.97% with a range of 2.38 to
6.03% in 40 German women (Precht & Molkentin, 1999), 2.80 ± 1.75% in 19
Kuwaiti mothers (Hayat et al., 1999), 2.81 ± 0.61% in 35 women from Prague,
Czech Republic (Dlouhy et al., 2002), 1.00 to 4.31% depending on individual
variation, season of the year, and stage of lactation in 19 Polish women (Mojska
et al., 2003), 11.3 ± 3.4 in 52 women from Western Iran (Bahrami & Rahimi,
2005), 7.0 ± 2.3% with a range of 2.5–13.8% in 81 women in southwestern
USA (Mosley, et al. 2005), 0.22 and 0.34% in 41 nomadic Fulani and 41 urban
women, respectively, in Nigeria (Glew et al., 2006), and 6.2 ± 0.48, 5.3 ± 0.49
and 4.6 ± 0.32 in samples collected in three consecutive periods from late 2004
to early 2006 in 87 Canadian women (Friesen & Innis, 2006). Differences in
these values probably reflect cultural differences in diet and changes in the food
supply over time, as well as differences in the methodology used in collection
and analysis of the samples.
The concentration of trans fatty acid isomers in various fractions of the blood
has been used increasingly as a biomarker to relate the incidence of disease to
trans fatty acid consumption. In some of these studies, correlations have been
found with some of the minor trans isomers and not with trans-18:1 or overall
TFA. For example, Lemaitre et al. (2006) reported that greater trans-18:2, but
not trans-18:1 in plasma phospholipid was associated with greater risk of fatal
ischaemic heart disease and sudden cardiac death in older adults. Also, van de
Vijver et al. (1996) have used TFA in plasma phospholipids as a biomarker for
dietary trans fatty acid consumption in studying risk for cardiovascular dis-
ease. Other investigators, such as Lemaitre, et al. (2002), Mozaffarian et al.
(2004a,b), and Sun et al. (2007a,b), have studied TFA in the red blood cell
membrane in relationship to risk of heart disease and inflammation. King et al.

(2005) used serum, rather than plasma, TFA as a surrogate of diet to investigate
the relationship between trans fatty acid intake and the risk of prostate cancer.
In general, the fatty acid composition of these fractions of blood is believed to
reflect the diet. However, Skeaff & Gowans (2006) reported that plasma TFA
could be used to distinguish between users of margarine made with partially
hydrogenated vegetable oils and users of butter in New Zealand homes, but that
the intake of fat from fast foods, bakery products, meat or meat products did not
appear to be associated with plasma phospholipid trans isomeric composition.
B. Estimates of worldwide consumption of trans fatty acids

All of the methods described above have been used to estimate the intake of
TFA throughout the world. Taken collectively, these estimates can give us a
reasonable idea of the relative consumption of TFA in various countries. In a
previous contribution (Craig-Schmidt, 1998) estimates of trans fatty acid
intake in different regions of the world were summarized. An update to this
previous work is given in the present chapter (Tables 1–8). Values in the tables
are listed for each country chronologically by the date of publication. Data on
biomarkers used as a surrogate of trans fatty acid intake are not included unless
the authors extrapolated to consumption values.
1. North America
Estimates of trans fatty acid intake in the USA and Canada published from the
late 1990s to the present are summarized in Table 1. In the mid 1990s several
groups in the USA published ‘consensus’ estimates of TFA in the US diet.
These values for trans fatty acid intake were: ILSI Expert Panel on Cardiovas-
cular Disease (1995), 2–4% of energy and 4–12% of total fatty acids; ASCN/
AIN Task Force (1996), 2–4% of energy and 8.1–12.8 g/day; and ILSI Expert
Panel on Development (1997), 8% of total fatty acids and 6.4 g/day. Values
published since that time are generally somewhat less.
In results published since 1997, the semi-quantitative food frequency ques-
tionnaire, particularly that of Willett and colleagues, is the most commonly
used method to collect dietary data for determination of trans fatty acid intake
in the USA (Table 1). Typically, mean or median values using food frequency
methodology range from 2 to 3 g/person/day (Garland et al., 1998; Lemaitre
et al., 1998; Lopez-Garcia et al., 2004; Mozaffarian et al., 2004a; Lopez-
Garcia et al., 2005; Tsai et al., 2005) and 1 to 2% of calories (Hu et al., 1997;
Hu et al., 2000; Lopez-Garcia et al., 2005; Lu et al. 2005; Sun et al., 2007a),
whereas values obtained using diet recalls or diet records for the adult pop-
ulation are generally higher, 4 to 8 g/person/day (Allison et al., 1999; Harnack,
et al., 2003) or 2 to 3% of energy (Harnack et al., 2003; Zhou et al., 2003;
Table 1. Estimates of trans fatty acid consumption in North America published late-1990s to present.
336
Country Dietary data Population Trans fatty acid (TFA) intake Method used Reference
g/person/day energy % % total FA
USA Habitual diet Women (n = 80,082) Willett semi-quanti- Hu et al., 1997

(1980 plus in Nurses’ Health tative food frequency
follow-up in Study questionnaire
1984, 1986
and 1990) total TFA 2.2
• Lowest quintile 1.3
• Highest quintile 2.9
USA Habitual diet Women (n = 140) Willett semi-quanti- Garland et al., 1998
(1986, 1987) in Nurses’ Health tative food frequency
Study questionnaire
1986
total TFA 2.8 ± 1.3 4.7 ± 1.4
18:1t 2.2 ± 1.1 3.7 ± 1.3
18:2t 0.4 ± 0.2 0.7 ± 0.2
1987
total TFA 2.8 ± 1.4 4.8 ± 1.5
18:1t 2.3 ± 1.2 3.8 ± 1.4
18:2t 0.4 ± 0.2 0.7 ± 0.2
USA Habitual diet Subjects (n = 51) Self-administered food Lemaitre et al., 1998
(1996) from greater Seattle, frequency question-
Washington area, naire modified from
aged 51–78 y NCI/Block combined
with revised US
Men (n = 24) Department of
total TFA 2.57 ± 0.69 5.46 ± 1.68 Agriculture Food
16:1t 0.02 ± 0.002 Composition Database
18:1t 2.24 ± 0.60
18:2t 0.29 ± 0.08
Women (n = 27)
total TFA 1.99 ± 0.88 4.98 ± 2.39
16:1t 0.01 ± 0.002
18:1t 1.78 ± 0.78
18:2t 0.24 ± 0.09
USA Individual diet Subjects (n = 11,258) 24-hour recall and Allison et al., 1999
(1989–1991) from Continuing 2-day diet record
Survey of Food methodology combined
Intakes by Individuals with US Department
(CSFII) of Agriculture
Average intake for 5.3 ± 0.08 2.6 ± 0.02 7.4 ± 0.06 Database
US population
Boys & girls, 3–5 y 4.1 ± 0.12 2.6 ± 0.05 7.5 ± 0.13
(n = 615)
Boys & girls, 6–11 y 5.3 ± 0.13 2.7 ± 0.05 7.7 ± 0.13
(n = 1,172)
Boys, 12–19 y 7.1 ± 0.37 2.8 ± 0.06 7.6 ± 0.14
(n = 618)
Girls, 12–19 y 5.1 ± 0.16 2.7 ± 0.06 7.6 ± 0.13
(n = 672)
Men, 20–49 y 6.6 ± 0.16 2.6 ± 0.03 7.1 ± 0.08
(n = 2,076)
Women, 20–49 y 4.6 ± 0.10 2.6 ± 0.03 7.4 ± 0.09
(n = 2,787)
Men, 50–69 y 5.8 ± 0.19 2.6 ± 0.06 7.3 ± 0.14
(n = 879)
Women, 50–69 y 4.2 ± 0.13 2.6 ± 0.05 7.5 ± 0.13
EVOLUTION OF WORLDWIDE CONSUMPTION OF TRANS FATTY ACIDS
(n = 1,226)
Men & women, 70+ 4.5 ± 0.12 2.6 ± 0.05 7.9 ± 0.12
(n = 1,213)
337
338
Table 1. continued
USA Habitual diet Women (n = 85,941) Willett semi-quanti- Hu et al., 2000

(1980, 1984, from Nurses’ Health tative food-frequency
1986, 1990) Study, aged 34–59 y questionnaire
1980 2.20
1984 1.90
1986 1.68
1990 1.52
USA Individual diets Adults from Minn- 24-hour diet recall Harnack et al., 2003
(1980–1997) esota Heart Survey combined with the
in Minneapolis-St University of Minne-
Paul, Minnesota, sota Nutrition Coordin-
metropolitan area, ating Center Food and
aged 25–74 y Nutrient Database
Men (n = 3,766)
1980–1982 8.4 3.0
1985–1987 8.0 2.8
1990–1992 7.0 2.5
1995–1997 6.4 2.2

Women (n = 4,183)
1980–1982 5.4 2.8
1985–1987 4.9 2.6
1990–1992 4.5 2.4
1995–1997 4.7 2.2
USA Habitual diet US population in 4 standardized 24-h Zhou et al., 2003
(1997–1999) the INTERMAP dietary recalls combined
study, aged 40–59 y with country-specific
Men (n = 1103) 2.0 ± 0.8 food composition data-
Women (n = 1092) 1.9 ± 0.8 base updated by the
Nutrition Coordinating
Center, University of
Minnesota
USA Usual diet Generally healthy 2 semi-quantitative Mozaffarian et al.,
(1986 and 1990 women (n = 823) food frequency ques- 2004a
for NHS, 1995 in the Nurses’ Study tionnaires and US
and 1999 for Study (NHS and Department of Agri-
NHSII). NHS II) culture Food Com-
total TFA 2.7 4.7 position Database and
(1.5 to 9.2) the Harvard University
Food Composition
Database
USA Individual diet Women (n = 727) Validated Willett food Lopez-Garcia et al.,
(1986, 1990) from the Nurses’ frequency question- 2004
Health Study I naire
cohort, aged 43–69 y
total TFA 3±1
3rd quintile (n–3)
USA Habitual diet Women (n = 730) Willett semi-quanti- Lopez-Garcia et al.,
(Mean of 1986 from Nurses’ tative food frequency 2005
& 1990) Health Study, questionnaire
aged 43–69 y
total TFA 2.4 ± 0.1 1.4 ± 0.1 4.3 ± 0.6
16:1t 0.2 ± 0.04

18:1t 1.9 ± 0.1
18:2tt 0.1 ± 0.02
339
18:2tc,ct 0.2 ± 0.1

Table 1. continued
340
USA Usual diet Boston women 5 quantitative food Lu et al., 2005

(1980 until (n = 440) in Nurses’ frequency question-
1993–1995) Health Study, naires and US Depart-
aged 53–73 y ment of Agriculture
total TFA (median) Food Composition
• Lowest quartile 1.29 (0.5–1.4) Database
• Highest quartile 2.3 (2.1–3.2)
USA Habitual diet US men Validated semi- Tsai et al., 2005
(1986) (n = 45,912) quantitative food
in the Health frequency questionnaire
Professionals and US Department of
Follow-up Study Agriculture Food
total TFA (mean) Composition Database.
• Lowest quintile 1.4
• Middle quintile 2.7
• Highest quintile 4.5
USA Habitual diet US women Food frequency Sun et al., 2007
(1990) (n = 306) in Nurses’ questionnaire and US
Health Study, Department of Agri-

aged 43–69 y culture Food
total TFA 1.48 ± 0.54 4.67 ± 1.42 Composition Database
16:1t 0.08 ± 0.03 0.26 ± 0.06
Total 18:1t 1.18 ± 0.47 3.71 ± 1.27
9t,12t-18:2 0.08 ± 0.03 0.25 ± 0.08
9c,12t-18:2 0.14 ± 0.06 0.45 ± 0.18
Total 18:2t 0.22 ± 0.09 0.69 ± 0.23
USA Individual diets Adults in Minne- 24-h diet recall Lee et al., 2007
sota Heart Survey combined with the
(n = 6,070); subjects University of
residing in Minne- Minnesota Nutrition
apolis/St. Paul, Coordinating Center
Minnesota Food and Nutrient
metropolitan area Database
Men
1980–1982 3.0
1985–1987 2.8
1990–1992 2.5
1995–1997 2.2
2000–2002 2.3
Women
1980–1982 2.8
1985–1987 2.6
1990–1992 2.4
1995–1997 2.2
2000–2002 2.2
USA Habitual diet Men (n = 1012) Validated semi-quanti- Liu et al., 2007
(2001–2004) Case-control study tative food frequency
of prostate cancer questionnaire; analysed
Caucasians at Fred Hutchinson
• Cases (n = 406) 5.66 ± 3.41 Cancer Research Center
• Controls (n = 408) 4.64 ± 3.03
African Americans
• Cases (n = 83) 6.42 ± 4.90
• Controls (n = 82) 5.97 ± 4.97
341
342
Table 1. continued
Canada Typical Sample Canadian ‘Market basket’ diet Innis et al., 1999
Vancouver diet diet used to show analysed with database
variability in of 200 food items pur-
estimated trans chased in Vancouver,
fatty acid intake British Columbia,
depending on brand Canada, analysed in
or database chosen Innis’ laboratory or
total TFA 1.4 – 25.4 FOOD PROCESSOR
database (ESHA
Research, Salem,
Oregon)
Canada Individual diets Canadian lactating 3-day diet records; Innis and King, 1999
women (n = 21) analysed using FOOD
total TFA 6.87 ± 1.07 2.46 ± 0.38 PROCESSOR database
(ESHA Research, Salem,
Oregon) modified to
include 300 brand name
food products analysed

by the laboratory of Innis
Canada Individual diets Healthy, pregnant Cross-sectional pros- Elias and Innis, 2002
women (n = 60) pective study using a
recruited in Vancouver, food intake question-
British Columbia, naire designed to
Canada; (mean ± SE) estimate trans fatty acid
Second trimester 3.8 ± 0.3 1.3 ± 0.1 intakes from a trans
Third trimester 3.4 ± 0.3 1.3 ± 0.1 fatty acid food database
developed by food
analysis
Canada Habitual diet Canadian children Validated 178-item Innis et al., 2004
(n = 84), food frequency ques-
aged 18–60 mo tionnaire designed to
All children 4.81 ± 3.12 assess patterns of fat
18–24 mo old 3.47 ± 1.86 and fatty acid intakes
25–36 mo old 5.27 ± 3.86 in detailed, in-person
37–60 mo old 4.98 ± 3.05 interview with the
primary caregiver;
FOOD PROCESSOR
food composition data-
base (ESHA Research,
Salem, Oregon) &
Canadian nutrient file
modified to include
500 food products
analysed by the
laboratory of Innis
Canada Habitual diet Canadian lactating Estimate based on Friesen and Innis,
(1998 and women trans fatty acid content 2006
2004–2006) 1998 (n = 103) 4.0 (0.51–12.32) of human milk and
2004–2005 (n = 24) 3.4 (1.36–8.72) equation of Craig-
mid-2005 (n = 24) 2.7 (1.07–9.29) Schmidt et al., 1984
2005–2006 (n = 39) 2.2 (0.56–7.65)
Data expressed as total trans fatty acids, and as mean ± SD or as mean (range) unless otherwise specified.
343
Lee et al., 2007). An exception is Liu et al., 2007, who reported total trans fatty
acid consumption values of 4 to 6 g/day even though a food frequency
questionnaire was used to obtain dietary data. In general, these estimates of
trans fatty acid consumption using both types of dietary data are less than the
‘consensus’ values published in the mid 1990s.
As one might expect, there are documented differences in trans fatty acid
consumption among various groups. Men, because of greater caloric intake,
have a greater trans fatty acid consumption than do women (Lemaitre et al.,
1998; Allison et al., 1999; Harnack et al, 2003). Expressed as energy percent,
values for men and women are somewhat similar (Zhou et al., 2003; Lee et al.,
2007). In a comparison of various age and gender groups, Allison et al. (1999)
found that teenage boys had the greatest consumption of TFA (7.1 g/day) of
any age group studied. Children had a surprisingly high level of TFA in the diet
(4.1 g/day for 3–5 year old boys and girls and 5.3 g/day in 6–11 year old boys
and girls). Also, trans fatty acid consumption appears to be greater in African
Americans than in Caucasians (Liu et al., 2007).A direct comparison of TFA
intake in the USA with that of three other countries was done by Zhou et al.
(2003). In this study, consumption in the USA and the UK (2.0 and 1.6% of
energy, respectively) was considerably higher than that in China or Japan (0.2
and 0.3% of energy, respectively).
There is evidence that dietary intake of TFA in the USA has decreased since
about 1980. For example, Hu et al. (2000) reported decreases in total TFA
consumption of 2.20, 1.90, 1.68 and 1.52% of calories for 1980, 1984, 1886,
and 1990, respectively. Similarly, Harnack et al. (2003) documented a de-
crease in consumption of TFA in US men from 8.4 g/day in 1980 to 1982 to
6.4 g/day in 1995 to 1997 and in US women from 5.4 g/day to 4.7 g/day in the
same periods of time. This trend was confirmed by Lee et al. (2007) who
reported a drop in TFA consumption from 3.0 to 2.3% of energy in men and 2.8
to 2.2% of energy in women in the period from 1980 to 1982 until 2000 to 2002.
The decrease in TFA consumption in the USA is due to the response of
industry to food label regulations and increased consumer awareness of the
deleterious effects of TFA in foods. In 2003, the US Food and Drug Adminis-
tration (FDA) issued a ruling requiring the declaration of trans fat on food
labels (DHHS/FDA 2003; Schrimpf-Moss & Wilkening, 2007). Manufacturers
were given until 1 January 2006, to implement the regulation. ‘Trans-free’
packaged foods are now prevalent in the US marketplace, although baked
goods and fried foods in restaurants may still contain relatively high amounts
of TFA. Stender and colleagues (Stender et al., 2006) demonstrated that the
TFA content of fast food in the USA was among the greatest in the 20 countries
from which samples were collected. French fries and chicken nuggets, pur-
chased from McDonald’s or Kentucky Fried Chicken establishments between
November, 2004, and September, 2005, were analysed by capillary gas chro-
matography in a single laboratory using a standard method (Leth et al, 2003).
For fast food from McDonald’s restaurants, the USA was the greatest in TFA
(10 g or 23% of total fat in French fries and 11% in chicken nuggets) with
Denmark being the least (less than 1 g or 1% of total fat in both products). It is
likely that recent reformulation of fats (Tarrago-Trani et al., 2006; Eckel et al.,
2007) has resulted in decreased use of fat and oils rich in TFA by the fast food
industry in the USA, but this change has not been well documented in the
scientific literature.
Trends in TFA consumption in Canada are roughly parallel to that of the
USA (Craig-Schmidt, 2007). Brisson (1981) was one of the first to estimate
TFA intake in the Canadian population, calculating that a typical Canadian
person would consume an average of 9.1 g/day, with a maximal value of
17.5 g/day. These estimates, based on the average TFA content of selected food
items and Canadian government consumption data from 1977, were in the same
range as early estimates of TFA consumption by the population of the USA.
Estimates in the early 1990s for TFA consumption in Canada were also fairly
high. Using the equation of Craig-Schmidt et al. (1984) in combination with
human breast milk analysis of 198 samples collected in 1992, Ratnayake and
colleagues (Chen et al., 1995a) calculated that TFA intake by lactating women
ranged from 3.0 g/day to 20.3 g/day, with the majority of women consuming
about 10 g/day. These same investigators (Ratnayake and Chen, 1995) also
estimated TFA intake to be 10.4% of the total dietary fat or 5.2 g/day for elderly
women to 12.5 g/day for young men in Nova Scotia.
More recently, Innis and King (1999) reported the average TFA intake by
Canadian lactating women (n = 21) to be 6.87 ± 1.07 g/person/day or 2.46 %
of calories (Table 1), a value somewhat less than that reported earlier by Chen
et al. (1995a). Moreover, in a cross-sectional prospective study of healthy,
pregnant women in Vancouver, Canada, Elias and Innis (2002) reported values
of 3.8 ± 0.3 g/day for women in the second trimester of pregnancy and
3.4 ± 0.3 g/day for women in the third trimester of pregnancy. These values
were based on a food intake questionnaire designed specifically to estimate
TFA intakes from a database of food items analysed in the investigator’s own
laboratory. Using a similar approach, Innis et al. (2004) reported values for the
TFA intake of Canadian children aged 18–60 months to be 4.81 ± 3.12 g/day
for this vulnerable population. This intake for Canadian children is similar to
the value of 4.1 g/day reported by Allison et al. (1999) for US children, 3–5
years old.
The recent values for TFA consumption reported by Innis’ group (Elias &
Innis, 2002; Innis et al., 2004) are less than those reported in the 1990s. Using
Canadian human milk as a biomarker and the equation of Craig-Schmidt et al.
(1984), Innis and colleagues (Friesen & Innis, 2006) confirmed that the
consumption of TFA had declined with the introduction of trans fat food
labelling regulations in Canada (Table 1). Like the USA, most Canadian food
products must carry the amount of trans fat in grams on the food label. In
346
Table 2. Estimates of trans fatty acid consumption in UK and Republic of Ireland published late 1990s to present.
UK Household UK men and women 7-day household con- Hulshof et al., 1999
consumption (n = 7921) in sumption survey
(1996) TRANSFAIR study, combined with Market
aged 0–75+ y Basket analyses of 100
total TFA 2.8 1.3 foods collected
14:1t9 0.11 1995–1996 plus other
16:1t9 0.18 country-specific data
18:1t 2.00
18:2t 0.28
18:3t +20:1 0.17
20:2t11,14 0.02
22:1t 0.06
UK Habitual diet UK population in the 4 standardized 24-h Zhou et al., 2003
(1997–1999) INTERMAP study, dietary recalls
aged 40–59 y combined with

Men (n = 266) 1.6 ± 4.1 country-specific food
Women (n = 235) 1.3 ± 0.6 composition database
updated by the Nutrition
Coordinating Center,
University of
Minnesota, USA
Ireland Habitual diet Irish healthy adults, 88-item fat intake Cantwell et al.,
aged 23–63 y questionnaire (FIQ) 2005a,b
(n = 105) combined with food
total TFA 5.42 ± 3.28 1.89 ± 0.92 database including
Males (n = 62) 225 foods
Natural TFA 1.8 ± 1.49
Industrial TFA 4.53 ± 3.79
Mean TFA 6.26 ± 3.64 1.96 ± 0.94
Females (n = 43)
Natural TFA 1.34 ± 1.29
Industrial TFA 2.90 ± 2.29
Mean TFA 4.21 ± 2.22 1.78± 0.89
347
Canada, there must also be the percent Daily Value for trans fat and saturated
fat combined (Canadian Department of Health, 2003). Increased consumer
awareness of TFA in processed foods and reformulation of products by the food
industry will likely result in further decline of TFA in the Canadian food
supply, as well as in that of the USA.
2. UK and Ireland
Estimates of TFA intake in the UK and Republic of Ireland published from the
late 1990s to the present are summarized in Table 2. As part of the TRANSFAIR
study, Hulshof et al. (1999) estimated that TFA consumption for men and
women in the UK was 2.8 g/person/day. This value was less than earlier values
reported for the UK (Craig-Schmidt, 1998), but comparable to some European
countries, such as France, included in the TRANSFAIR study. In the
INTERMAP study, Zhou et al. 2003 reported that consumption of TFA in the
UK was 1.3–1.6en%, values which were in the same range as the TRANSFAIR
study. However, Irish adults appeared to consume greater amounts of TFA,
5.42 g/day or 1.89en%, than their counterparts in the UK (Cantwell et al.,
2005a,b).
Cantwell et al. (2005a,b) have studied the proportion of ruminant or ‘natu-
ral’ TFA compared to industrially produced TFA in the diets of healthy Irish
adults (Table 2). These investigators estimate that about 30% of the TFA
consumption in Ireland is from ruminant fats. Irish men consumed 1.8 g/day
from meat or dairy fats out of a total of 6.26 g/day, whereas Irish women
consumed 1.34 g/day from meat or dairy fats out of a total of 4.21 g/day as
dietary TFA.
The Food Standards Agency of the UK makes declaration of trans fat on food
labels mandatory if a claim is made regarding trans fat (Food Standards
Agency, 2003; Schrimpf-Moss & Wilkening, 2007). In the UK, as in North
America, it is probable that food labelling regulations eventually will make
industrially synthesized TFA in foods a thing of the past.
3. Continental Europe
Estimates of TFA intake in continental Europe published from the late 1990s to
the present are summarized in Table 3. The TRANSFAIR study (Hulshof et al.,
1999; van de Vijver, 2000) is the major study comparing TFA consumption in
European countries. Hulshof et al. (1999) reported that intakes for continental
Europe were low in the Mediterranean/southern European countries with
Greece, Italy, Portugal and Spain having values of 1.2 to 2.1 g/day for men
compared to northern European countries, with values ranging from 2.4 g/day
in Germany to 4.8 g/day in The Netherlands. Intermediate values were reported
for men in France (2.7 g/day) and Belgium (4.4 g/day). Other investigators
Table 3. Estimates of trans fatty acid consumption in continental Europe published late-1990s to present.
Belgium Individual diet Belgium population 3-day diet records Hulshof et al., 1999
(1991–1992) (n = 492) in combined with Market
TRANSFAIR study, Basket analyses of
aged 18–65 y 100 foods collected
Men (n = 323) 1995–1996 plus other
total TFA 4.4 ± 2.0 1.4 ± 0.5 country-specific data
14:1t9 0.19 ± 0.10
16:1t9 0.30 ± 0.16
18:1t 3.2 ± 1.6
18:2t 0.47 ± 0.19
18:3t +20:1 0.16 ± 0.08
20:2t11,14 0.04 ± 0.06
22:1t 0.02 ± 0.02
Women (n = 169)
total TFA 3.6 ± 1.8 1.5 ± 0.5
14:1t9 0.16 ± 0.09
16:1t9 0.27 ± 0.16
18:1t 2.6 ± 1.3
18:2t 0.37 ± 0.17
18:3t +20:1 0.14 ± 0.08
20:2t11,14 0.03 ± 0.04
22:1t 0.02 ± 0.02
France Individual diet French population in 7-day diet records Hulshof et al., 1999
(1993–1994) TRANSFAIR study, combined with Market

aged 19–64 y Basket analyses of
Men (n = 300) 100 foods collected
349
total TFA 2.7 ± 1.2 1.1 ± 0.4 1995–1996 plus other

Table 3. continued
350
14:1t9 0.25 ± 0.12 country-specific data

16:1t9 0.33 ± 0.16
18:1t 1.5 ± 0.8
18:2t 0.40 ± 0.21
18:3t +20:1 0.04 ± 0.04
20:2t11,14 0.02 ± 0.02
22:1t 0.16 ± 0.12
Women (n = 463)
total TFA 2.1 ± 0.8 1.2 ± 0.3
14:1t9 0.19 ± 0.08
16:1t9 0.26 ± 0.13
18:1t 1.13 ± 0.50
18:2t 0.30 ± 0.13
18:3t + 20:1 0.03 ± 0.02
20:2t11,14 0.02 ± 0.02
22:1t 0.14 ± 0.14
France Dietary history French healthy Dietary history van de Vijver et al.,
population, as part method combined 2000
of TRANSFAIR with Market Basket

study, aged 50–65y analyses of 100 foods
Men (n = 32) collected 1995–1996
Total TFA 1.02 ± 0.36 plus other country-
14:1t9 0.08 ± 0.03 specific data
16:1t9 0.16 ± 0.08
18:1t 0.55 ± 0.21
18:2t 0.13 ± 0.04
22:1t 0.05 ± 0.03
Women (n = 11)
Total TFA 1.08 ± 0.41
14:1t9 0.09 ± 0.05
16:1t9 0.18 ± 0.10
18:1t 0.58 ± 0.23
18:2t 0.13 ± 0.05
22:1t 0.05 ± 0.03
France Typical French French women Estimated based on Boué et al., 2000
diet (1997–1998) (n = 71) the 18:1 t content in
free of cancer and adipose tissue of
diabetes, women and equation
aged 37 ± 10 y of Enig et al., 1990
total TFA 1.93 (0.12–4.10) and daily per capita
total fat intake
France Individual diet French non-pregnant 7-day diet record Combe et al., 2000
(1996–1999) (n = 97) and
pregnant (n = 90)
women, aged 18–50y
total TFA 2.7 1.3
Germany Individual diet German population Dietary history Hulshof et al., 1999
(1991) in TRANSFAIR combined with Market
study, aged 19–64 y Basket analyses of
16:1t9 0.36 ± 0.16
18:1t 1.35 ± 0.52
18:2t 0.34 ± 0.12
18:3t +20:1 0.03 ± 0.03

20:2t11,14 0.02 ± 0.01
22:1t 0.00 ± 0.00
351
Table 3. continued
352
Women (n = 881)
total TFA 1.9 ± 0.7 0.9 ± 0.2
14:1t9 0.26 ± 0.12
16:1t9 0.29 ± 0.13
18:1t 1.07 ± 0.40
18:2t 0.28 ± 0.10
18:3t +20:1 0.03 ± 0.02
20:2t11,14 0.02 ± 0.01
22:1t 0.00 ± 0.00
Greece Individual diet Greek population 24-h recall combined Hulshof et al., 1999
(1995) in TRANSFAIR with Market Basket
study, aged 23–64y analyses of 100 foods
Men (n = 141) collected 1995–1996
total TFA 1.2 ± 1.2 0.5 ± 0.4 plus other country-
16:1t9 0.12 ± 0.14
18:1t 0.82 ± 0.89
18:2t 0.18 ± 0.18
18:3t +20:1 0.06 ± 0.13

20:2t11,14 0.01 ± 0.02
22:1t 0.00 ± 0.00
Women (n = 107)
total TFA 1.7 ± 1.3 0.8 ± 0.6
14:1t9 0.05 ± 0.05
16:1t9 0.13 ± 0.10
18:1t 1.17 ± 1.03
18:2t 0.23 ± 0.17
18:3t +20:1 0.08 ± 0.14
20:2t11,14 0.01 ± 0.02
22:1t 0.00 ± 0.00
Greece Dietary history Greek healthy Dietary history method van de Vijver et al.,
population as part combined with Market 2000
of TRANSFAIR Basket analyses of 100
study, aged 50–65 y foods collected
Men (n = 45) 1995–1996 plus other
Total TFA 0.52 ± 0.26 country-specific data
14:1t9 0.02 ± 0.01
16:1t9 0.05 ± 0.03
18:1t 0.35 ± 0.21
18:2t 0.07 ± 0.04
22:1t 0.00 ± 0.00
Women (n = 37)
Total TFA 0.61 ± 0.27
14:1t9 0.02 ± 0.01
16:1t9 0.06 ± 0.03
18:1t 0.42 ± 0.21
18:2t 0.08 ± 0.04
22:1t 0.00 ± 0.00
Italy Household Italian men 7-day household con- Hulshof et al., 1999
consumption (n = 10,000) in sumption combined
(1980–1984) TRANSFAIR with Market Basket
study, aged 1–80 y analyses of 100 foods
total TFA 1.6 0.5 collected 1995–1996
14:1t9 0.14 plus other country-
16:1t9 0.17 specific data
18:1t 1.00
18:2t 0.23
18:3t +20:1 0.02
20:2t11,14 0.02
353
22:1t 0.00
Table 3. continued
354
The Individual diet Dutch population 2-day diet record Hulshof et al., 1999
Nether- (1992) in TRANSFAIR combined with Market
lands study, aged 19–64 y Basket analyses of 100
Men (n = 1822) foods collected 1995–
total TFA 4.8 ± 2.3 1.5 ± 0.6 1996 plus other
16:1t9 0.14 ± 0.08
18:1t 3.9 ± 2.0
18:2t 0.55 ± 0.22
18:3t +C20:1 0.09 ± 0.06
20:2t11,14 0.0 ± 0.0
22:1t 0.02 ± 0.04
Women (n = 2203)
total TFA 3.8 ± 2.0 1.6 ± 0.7
14:1t9 0.04 ± 0.04
16:1t9 0.11 ± 0.06
18:1t 3.1 ± 1.8
18:2t 0.43 ± 0.19
18:3t +C20:1 0.07 ± 0.05
20:2t11,14 0.0 ± 0.0

22:1t 0.01 ± 0.03
The Dietary history Dutch healthy Dietary history com- van de Vijver et al.,
Nether- population in bined with Market 2000
lands TRANSFAIR Basket analyses of
study, aged 50–65 y 100 foods collected
Men (n = 48) 1995–1996 plus other
14:1t9 0.02 ± 0.02
16:1t9 0.07 ± 0.06
18:1t 0.71 ± 0.28
18:2t 0.15 ± 0.04
22:1t 0.01 ± 0.05
Women (n = 43)
Total TFA 0.98 ± 0.31
14:1t9 0.02 ± 0.02
16:1t9 0.06 ± 0.02
18:1t 0.73 ± 0.26
18:2t 0.14 ± 0.04
22:1t 0.00 ± 0.01
The Dietary history Men (n = 667) of Dietary history Oomen et al., 2001
Nether- (1985–1995) Zutphen Elderly combined with time-
lands Study aged 64–84 y specific Dutch food
1985 10.9 ± 6.3 4.3 ± 2.2 composition tables
1990 6.9 ± 4.0 2.9 ± 1.5
1995 4.4 ± 1.7 1.9 ± 0.6
The Habitual diet Women in Validated 150-item Voorrips et al., 2002
Nether- (1986–1992) Netherlands food frequency ques-
lands Cohort Study tionnaire combined
Breast Cancer Cases with DUTCH food
(n = 941) composition table and
Total TFA 2.5 ± 0.9 TRANSFAIR data
18:1t11 0.8 ± 0.4
18:1t6 or t9 1.2 ± 0.8
18:2c9,t11 or t9,c11 0.2 ± 0.1
Subcohort members
(n = 1598)
Total TFA 2.5 ± 0.9

18:1t11 0.7 ± 0.4
18:1t6,9 1.2 ± 0.8
355
18:2c9,t11 or t9,c11 0.2 ± 0.1

356
Table 3. continued
Spain Household Spanish men and 7d household con- Hulshof et al., 1999
consumption women (n = 21,155) sumption combined
(1991) in TRANSFAIR with Market Basket
study, aged 0–70+ y analyses of 100 foods
total TFA 2.1 0.7 collected 1995–1996
14:1t9 0.10 plus other country-
16:1t9 0.17 specific data
18:1t 1.40
18:2t 0.25
18:3t +C20:1 0.06
20:2t11,14 0.48
22:1t 0.03
Spain Dietary history Spanish healthy Dietary history com- van de Vijver et al.,
population as part bined with Market 2000
of TRANSFAIR Basket analyses of
Men (n = 36) 1995–1996 plus other

14:1t9 0.03 ± 0.02
16:1t9 0.05 ± 0.02
18:1t 0.35 ± 0.15
18:2t 0.07 ± 0.03
22:1t 0.00 ± 0.01
Women (n = 47)
Total TFA 0.76 ± 0.28
14:1t9 0.04 ± 0.02
16:1t9 0.08 ± 0.05
18:1t 0.52 ± 0.21
18:2t 0.09 ± 0.03
22:1t 0.01 ± 0.02
Portugal Individual diet Portuguese men 24 h recall combined Hulshof et al., 1999
(1988–1989) (n = 78) in with Market Basket
TRANSFAIR analyses of 100 foods
study, aged 38 y collected 1995–1996
total TFA 1.6 ± 0.8 0.6 ± 0.3 plus other country-
16:1t9 0.17 ± 0.10
18:1t 1.10 ± 0.60
18:2t 0.17 ± 0.11
18:3t +20:1 0.05 ± 0.04
20:2t11,14 0.05 ± 0.11
22:1t 0.01 ± 0.02
Portugal Dietary history Portuguese healthy Dietary history com- van de Vijver et al.,
population in bined with Market 2000
TRANSFAIR Basket analyses of
Men (n = 41) 1995–1996 plus other
14:1t9 0.04 ± 0.02
16:1t9 0.05 ± 0.02
18:1t 0.37 ± 0.14
18:2t 0.05 ± 0.02

22:1t 0.00 ± 0.01
357
358
Table 3. continued
Women (n = 39)
Total TFA 0.57 ± 0.26
14:1t9 0.04 ± 0.02
16:1t9 0.06 ± 0.03
18:1t 0.39 ± 0.18
18:2t 0.05 ± 0.02
22:1t 0.00 ± 0.01
Portugal Habitual diet Portuguese men Semi-quantitative food Lopes, et al., 2007
aged ≥40 y frequency ques-
(*data = % total fat tionnaire combined
intake) with Portuguese food
Myocardial infarction composition table
cases (n = 297) 1.27 ± 0.83*
Population Control
(n = 310) 1.16 ± 0.78*
estimated TFA consumption to average 1.9 g/day in French women based on

adipose tissue trans-18:1 (Boue et al., 2000) and 2.7 g/day in French pregnant
women and non-pregnant controls based on 7-day diet records (Combe et al.,
2000). In a study of breast cancer in women in The Netherlands, Voorrips et al.
(2002) reported a TFA intake of 2.5 g/day, whereas Lopes et al. (2007) in
studying myocardial infarction in Portuguese men found an intake of 1.16 to
1.27 % of total dietary fat.
Oomen et al. (2001) have documented the decrease in TFA consumption in
The Netherlands. In following men in the Zutphen Elderly study over time,
these investigators reported a progressive decline in TFA intake from 10.9 g/
day in 1985 to 6.9 g/day in 1990 to 4.4 g/day in 1995. Government regulations
in the Netherlands are minimal and much of the decrease in trans fat that has
occurred in this country has been due to the efforts of industry (Katan, 2006).
Voluntary nutrition labelling of trans fat is practised by the countries in the
European Union (Schrimpf-Moss & Wilkening, 2007).
4. Nordic Countries
Estimates of TFA intake in the Nordic countries published from the late 1990s
to the present are summarized in Table 4. The TRANSFAIR study (Hulshof
et al., 1999; van de Vijver, 2000) is the major study comparing TFA consump-
tion in these countries. Hulshof et al. (1999) reported that intakes for men in the
Nordic countries ranged from 2.3 g/day in Finland to 6.7 g/day in Iceland.
Intermediate values were reported for men in Denmark (2.9 g/day) and Norway
(4.8 g/day).
Jakobsen, et al. (2006, 2008) have studied the amount of ruminant fat
consumed by the Danish population. They report that adult men consume 2.0 g/
day and adult women consume 1.5 to 1.6 g/day of TFA from meat and dairy
products. With recent legislation targeting elimination of industrially produced
TFA from the Danish diet (Leth et al. 2006), these values could represent total
TFA consumption in Denmark.
In addition to Denmark, TFA consumption has decreased in recent years in
several of the other Nordic countries. Pedersen et al. (1998) have documented
this decrease in the Norwegian population. Trans fatty acid consumption has
decreased in Norway from 15 g/day/person in 1958 to 4 g/day/person in 1996.
A 2003 market survey conducted by Johansson et al. (2006) indicates that TFA
consumption from industrially produced sources is 1.6 g/day/person or 0.6en%.
5. Australia
Estimates of TFA intake in Australia published from the late 1990s to the
present are summarized in Table 5. Clifton et al. (2004) reported that in an
Australian adult population with a mean age of 56 years, the TFA intake was
Table 4. Estimates of trans fatty acid consumption in Nordic countries published late-1990s to present.
360
Denmark Individual diet Danish population 7-day diet records Hulshof et al., 1999
16:1t9 0.14 ± 0.07
18:1t 2.0 ± 1.0
18:2t 0.30 ± 0.13
18:3t + 20:1 0.09 ± 0.05
20:2t11,14 0.10 ± 0.06
22:1t 0.02 ± 0.01
Women (n = 702)
total TFA 2.3 ± 1.2 1.0 ± 0.5
14:1t9 0.16 ± 0.07
16:1t9 0.11 ± 0.06
18:1t 1.6 ± 0.8
18:2t 0.22 ± 0.10
18:3t + 20:1 0.06 ± 0.03
20:2t11,14 0.06 ± 0.04

22:1t 0.01 ± 0.01
Denmark Habitual diet Danish population 7-day diet records Jakobsen, et al., 2006
(1995) (n = 3,098), combined with TFA
aged 1–80 y content in dairy
Ruminant TFA 1.7 0.7 products and ruminant
All ages (0.9–2.7) (0.5–1.0) meat products
General population
1–6 y (n = 551) 1.4 0.8
7–14 y (n = 710) 1.6 0.6
15–29 y (n = 489) 1.8 0.7
30–80 y (n = 1348) 1.8 0.7
Male
1–6 y (n = 265) 1.4 0.8
7–14 y (n = 347) 1.8 0.7
15–29 y (n = 221) 2.0 0.7
30–80y (n = 683) 2.0 0.7
Female
1–6 y (n = 286) 1.3 0.8
7–14 y (n = 363) 1.5 0.6
15–29 y (n = 268) 1.5 0.7
30–80 y (n = 665) 1.6 0.7
Denmark Habitual diet Danish population 7-day weighed food Jakobsen, et al., 2008
(n = 3686), record and dietary
aged 30–71 y history interview com-
Ruminant TFA bined with TFA content
3rd quintile (range) in dairy products and
Women 1.5 (0.7–2.7) ruminant meat products/
Men 1.8 (1.8–3.4) Danish food com-
position table
Finland Individual diet Finnish population 3-day diet records Hulshof et al., 1999
16:1t9 0.14 ± 0.08
18:1t 1.6 ± 0.9
18:2t 0.26 ± 0.12
18:3t +20:1 0.08 ± 0.04

20:2t11,14 0.06 ± 0.10
22:1t 0.00 ± 0.00
361
Table 4. continued
362
Women (n = 991)
total TFA 1.9 ± 0.9 0.9 ± 0.3
14:1t9 0.11 ± 0.05
16:1t9 0.11 ± 0.05
18:1t 1.3 ± 0.7
18:2t 0.21 ± 0.09
18:3t +20:1 0.06 ± 0.03
20:2t11,14 0.05 ± 0.08
22:1t 0.00 ± 0.00
Finland Dietary history Finnish healthy Dietary history com- van de Vijver et al.,
population in bined with Market 2000
TRANSFAIR study, Basket analyses of
aged 50–65 y 100 foods collected
Men (n = 38) 1995–1996 plus other
14:1t9 0.04 ± 0.02
16:1t9 0.04 ± 0.02
18:1t 0.67 ± 0.25
18:2t 0.08 ± 0.03

22:1t 0.00 ± 0.00
Women (n = 39)
Total TFA 0.94 ± 0.34
14:1t9 0.05 ± 0.02
16:1t9 0.05 ± 0.02
18:1t 0.72 ± 0.31
18:2t 0.09 ± 0.03
22:1t 0.00 ± 0.00
Iceland Dietary history Icelandic population Dietary history com- Hulshof et al., 1999
(1990) in TRANSFAIR bined with Market
16:1t9 0.77 ± 0.41
18:1t 4.23 ± 2.27
18:2t 0.65 ± 0.35
18:3t +C20:1 0.49 ± 0.35
20:2t11,14 0.03 ± 0.03
22:1t 0.36 ± 0.38
Women (n = 510)
total TFA 4.1 ± 2.2 1.9 ± 0.6
14:1t9 0.21 ± 0.11
16:1t9 0.47 ± 0.23
18:1t 2.60 ± 1.59
18:2t 0.39 ± 0.20
18:3t +C20:1 0.30 ± 0.20
20:2t11,14 0.02 ± 0.01
22:1t 0.22 ± 0.21
Iceland Dietary history Icelandic healthy Dietary history com- van de Vijver et al.,
Men (n = 45) 1995–1996 plus other
14:1t9 0.07 ± 0.03
16:1t9 0.19 ± 0.07
18:1t 0.94 ± 0.52

18:2t 0.14 ± 0.06
22:1t 0.03 ± 0.04
363
Table 4. continued
364
Women (n = 45)
Total TFA 1.65 ± 0.82
14:1t9 0.08 ± 0.03
16:1t9 0.21 ± 0.09
18:1t 1.05 ± 0.65
18:2t 0.16 ± 0.05
22:1t 0.04 ± 0.05
Norway Norwegian population Pedersen et al., 1998
1958 15 5.5
1996 4
Norway Habitual diet Norwegian population Quantitative food Hulshof et al., 1999
(1993–1994) in TRANSFAIR frequency questionnaire
study, aged 19–64 y combined with Market
Men (n = 1257) Basket analyses of
total TFA 4.8 ± 2.9 1.5 ± 0.6 100 foods collected
14:1t9 0.17 ± 0.10 1995–1996 plus other
18:1t 3.49 ± 2.40
18:2t 0.24 ± 0.12
18:3t +C20:1 0.31 ± 0.20

20:2t11,14 0.02 ± 0.01
22:1t 0.14 ± 0.12
Women (n = 1330)
total TFA 3.2 ± 1.9 1.4 ± 0.5
14:1t9 0.12 ± 0.07
16:1t9 0.33 ± 0.19
18:1t 2.29 ± 1.48
18:2t 0.17 ± 0.08
18:3t +C20:1 0.22 ± 0.15
20:2t11,14 0.01 ± 0.01
22:1t 0.11 ± 0.09
Norway Market survey Norwegian population 1.6 0.6 Survey among Johansson et al., 2006
(2003) Norwegian food
manufacturers and
importers and Norweg-
ian food composition
data base
Sweden Individual diet Swedish population 7-day diet record com- Hulshof et al., 1999
(1989) in TRANSFAIR bined with Market
16:1t9 0.20 ± 0.11
18:1t 2.3 ± 1.1
18:2t 0.3 ± 0.1
18:3t + 20:1 <0.1
20:2t11,14 <0.05
22:1t <0.05
Women (n = 697)
total TFA 2.3 ± 1.0 1.1 ± 0.4
14:1t9 0.10 ± 0.05
16:1t9 0.14 ± 0.07
18:1t 1.8 ± 0.9
18:2t 0.20 ± 0.09
18:3t + 20:1 0.06 ± 0.03
20:2t11,14 <0.05
22:1t <0.05
365
Table 4. continued
366
Sweden Dietary history Swedish healthy Dietary history com- van de Vijver et al.,
Men (n = 42) 1995–1996 plus other
14:1t9 0.05 ± 0.02
16:1t9 0.06 ± 0.03
18:1t 0.72 ± 0.26
18:2t 0.09 ± 0.03
22:1t 0.00 ± 0.00
Women (n = 38)
Total TFA 1.01 ± 0.44
14:1t9 0.05 ± 0.02
16:1t9 0.06 ± 0.02
18:1t 0.77 ± 0.40
18:2t 0.09 ± 0.03
22:1t 0.00 ± 0.00
Table 5. Estimates of trans fatty acid consumption in Australia published late-1990s to present (FFQ, food-frequency questionnaire; WFR,
weighed food record).
Australia Individual diet Australian population, 300-item food- Clifton et al., 2004
(1995–1997) mean age 56 y frequency questionnaire
median intake and Australian food
3rd quintile 3.03 composition database
` (range) (1.55–5.46)
Myocardial infarction
cases (n = 209)
Total TFA 3.52 ± 1.82
18: 1t 1.21 ± 0.4
Control (n = 174)
Total TFA 3.01 ± 1.29
18:1t 1.17 ± 0.4
Australia Habitual/ Queensland men 129-item FFQ/WFR McNaughton
Individual diet (n = 18) and women and Australian fatty et al., 2007
(n = 25), aged 28–75 y acids in foods com-
FFQ 0.11 ± 0.31 position database
WFR 0.01 ± 0.01 (Mann, et al., 2003)
367
3.03 g/day in 1995 to 1997. Toward the end of the study, ‘no-trans’ margarines
were introduced in Australia, and these investigators had evidence that TFA
consumption decreased in their study population after 1996. McNaughton
et al. (2007) reported very low TFA intakes in men and women of Queensland,
Australia, based on a food frequency questionnaire or a weighed food record
combined with an Australian food database by Mann, et al. (2003). These
investigators report 0.11 and 0.01 g/day for TFA intake based on a food
frequency questionnaire or on weighed food records, respectively. Recent
regulations for food labelling of trans fat have been instituted in Australia and
New Zealand (Schrimpf-Moss & Wilkening, 2007).
6. Asian/Pacific Region
Estimates of TFA intake in the Asian/Pacific region published from the late
1990s to the present are summarized in Table 6. In the study of Zhou et al.
(2003), intakes of TFA in China and Japan were low, 0.2en% and 0.3en%,
respectively. The populations in both countries were middle-aged (40 to 59
years old), and it is possible that some younger people would have a greater
TFA intake as they adopt a more Western diet.
7. Central and South America

Estimates of TFA intake in Central and South America published from the late
1990s to the present are summarized in Table 7. The consumption of TFA in
Costa Rica has been studied by Baylin et al. (2002), Monge-Rojas et al. (2005)
and Colon-Ramos et al. (2006). In the latter two studies, a food composition
database modified for Costa Rican foods was used. Estimates of TFA consump-
tion in these studies range from 4 to 5 g/day, and the urban Costa Rican
population had a greater TFA intake than the rural population (Monge-Rojas
et al., 2005).
Trans fatty acid consumption in Brazil has been studied by Bertolino, et al.
(2006). Values of 10.2 and 11.5 g/day were reported for men and women,
respectively, for 1993. Consumption of TFA appeared to fall somewhat by
2000, but the estimates still appear to be high compared to estimates of
consumption in the USA and Europe at the same time. The authors attributed
the high values, in part, to the food composition database which was based
primarily on foods from other countries. However, ‘high trans’ fats are being
used in Central and South America at least by fast-food chains. Stender et al.
(2006) reported that Peru was second from the top in TFA in ‘McDonald’s’
French fries and chicken collected between November 2004 and September
2005. For French fries and chicken obtained from ‘Kentucky Fried Chicken’,
samples from Peru were third from the top and samples from the Bahamas were
fifth from the top.
Table 6. Estimates of trans fatty acid consumption in the Asian/Pacific region published late-1990s to present
China Individual diet Chinese population 4 standardized 24-h Zhou et al., 2003
in the INTERMAP dietary recalls combined
Minnesota
Japan Individual diet Japanese population 4 standardized 24-h Zhou et al., 2003
in the INTERMAP dietary recalls combined
Minnesota
369
Table 7. Estimates of trans fatty acid consumption in Central and South America published late-1990s to present.
370
Costa Habitual diet Costa Rican people 135-item food- Baylin et al., 2002
Rica (n = 503) frequency question-
Total TFA 3.98 ± 1.85 4.76 ± 2.24 naire combined with
Trans 18:1 2.47 ± 1.23 2.94 ± 1.43 modified USDAfood
composition tables
Costa Individual diet Costa Rican 3-day food record and Monge-Rojas et al.,
Rica adolescents fatty acid content of 2005
(n = 275), Costa Rican foods
aged 12–19 y
Male (n = 144)
Rural (n = 75)
Total TFA 4.04 ± 0.78
16:1t 0.07 ± 0.01
18:1t 3.14 ± 0.44
18:2t 0.57 ± 0.69
Urban (n = 69)
Total TFA 4.96 ± 0.82
16:1t 0.10 ± 0.03
18:1t 3.66 ± 0.78

18:2t 0.83 ± 0.59
Female (n = 131)
Rural (n = 69)
Total TFA 4.35 ± 0.67
16:1t 0.08 ± 0.02
18:1t 3.36 ± 0.15
18:2t 0.66 ± 0.16
Urban (n = 62)
Total TFA 4.75 ± 0.68
16:1t 0.10 ± 0.01
18:1t 3.64 ± 0.28
18:2t 0.76 ± 0.12
Costa Habitual diet Costa Rican Food frequency ques- Colón-Ramos et al.,
Rica (1994–1999) Population tionnaire and nutrient 2006
(n = 954) database specific for
Total TFA 4.1 Costa Rican foods
median intake
Brazil Habitual diet Japanese Brazilians Two cross-sectional Bertolino, et al., 2006
(1993, 2000) (n = 328) in San surveys on health and
Paulo, Brazil, nutritional status; food
aged 40–79 y frequency questionnaire
Women (n = 165) (US National Cancer
1993 10.2 ± 6.5 5.1 ± 2.6 Institute) combined
2000 7.0 ± 4.8 3.4 ± 1.7 with USDA database
Men (n = 150) and some Brazilian
1993 11.5 ± 7.4 4.7 ± 2.4 trans fatty acid food
2000 8.3 ± 5.1 3.3 ± 1.9 composition
371
Recent regulations may result in changes in consumption patterns in Central

and South America. Since 2006, the Mercosur countries (including Brazil,
Argentina, Paraguay and Uruguay) have included a mandatory trans fat
declaration on their food labels (ANVISA, 2003).
8. Iran
A recent estimate of TFA intake in Iran is summarized in Table 8. The intake
of industrially produced TFA has been estimated by Mozaffarian et al. (2007)
to be 4.2% of energy. Even though this estimate does not include ruminant fats,
it is quite high compared to other countries, such as the USA, in which values
are about 2% of energy (Table 1). That TFA consumption in Iran is greater than
in most countries is confirmed by an analysis of human milk samples from 52
Western Iranian women. The trans-18:1 content of human milk was 11.3% of
total milk fatty acids, a value much higher than currently observed in North
American or European women (Craig-Schmidt, 2001).
C. Consumption of trans fatty acids: future trends

Government regulation of TFA, either by requiring a trans fat declaration on
food labels or by regulating the amount of industrially produced TFA in
processed foods (Schrimpf- Moss & Wilkening, 2007), will eventually result in
the demise of commercial sources of TFA. These regulations have dealt
primarily with packaged foods; however, some local regulation of restaurants
and fast food establishments is beginning to take place. Cities, such as New
York in the USA, have enacted legislation to restrict trans fat in restaurants
within the city. Fast-food chains worldwide have been reluctant to modify their
source of frying fats. Stender and colleagues (Stender et al., 2006) have
compared the TFA content of fast food in 20 countries. French fries and
chicken nuggets were purchased from McDonald’s or Kentucky Fried Chicken
establishments between November 2004 and September 2005, and analysed by
capillary gas chromatography in a single laboratory. For fast food from
McDonald’s restaurants, the USA was the greatest in TFA (10 g or 23% of total
fat in French fries and 11% in chicken nuggets) with Denmark being the least
(less than 1 g or 1% of total fat in both products). If one averages the data for
French fries in terms of percentage of total fat as TFA for McDonald’s and
Kentucky Fried Chicken, one can obtain a rough profile of trans fat consump-
tion in fast food restaurants throughout much of the world in 2004 to 2005.
Poland, Peru, the Bahamas and South Africa were the highest (averaging 26 to
34% of the fat as TFA), followed by the USA (19%) and the Czech Republic
(17%). Italy was next with 14% at one establishment. Restaurants in Finland,
Norway, Sweden, France, Austria and the UK produced French fries with an
average of 10 to 12% of the total fat as TFA, whereas the level in French fries
Table 8. Estimates of trans fatty acid consumption in the Middle East published late-1990s to present
Iran Habitual diet 7158 urban and Estimates of con- Mozaffarian et al.,
rural households sumption of industrial 2007
containing 35,924 TFA by detailed in-
individuals home assessments of
Industrial TFA 4.2 dietary intake and lab-
oratory analysis of
commonly consumed fats
373
in Hungary, Russia, Germany, The Netherlands, Portugal and Spain were more
moderate (averaging 6 to 8%). Denmark was the least with only 1–2% of the
total fat as TFA in the French fries. It is anticipated that as appropriate frying
fats become available and as fast-food chains respond to consumer pressure,
the trans fat content of fast-foods will decrease as have the trans contents of
packaged foods.
Within a few years, the TFA in the food supply should be limited to the
‘natural’ ruminant fats in meat and dairy products. This lower limit for Western
countries should be in the neighbourhood of 1 to 2 g/day and even lower in
countries where dairy products are not routinely consumed. The data of
Hulshof et al. (1999) can be used to calculate trans fat consumption from
ruminant sources (milk, cheese, meat and butter) as 0.9 in Finland to 2.1 in
Iceland. Other estimates, such as 1.7 g/day for ruminant fat consumption in the
Danish population (Jakobsen, et al., 2006) are within this range. Whether or
not TFA isomers in ruminant fats have the same effects as trans isomers in
partially hydrogenated vegetable oils is currently under investigation
(Weggemans et al., 2004; Chardigny et al., 2006; Willett and Mozaffarian,
2008; Motard-Belanger, et al., 2008; Chardigny, et al., 2008) and the issue is
discussed elsewhere in this book.
The trans fatty acid saga is near its end. These food ingredients have had a
long journey from being the new, popular shortening in bakery products in the
early twentieth century to being one of the ‘Bad Fat Boys’ of the American
Heart Association in the twenty-first century.
Acknowledgement
Appreciation is expressed to Elias J. Bungenstab for translation of Portuguese
material into English.
References
Allison, DB, Egan, K, Barraj, LM, Caughman, C, Infante, M and Heimbach, JT (1999)
Estimated intakes of trans fatty and other fatty acids in the US population. J. Am. Diet.
Assoc., 99, 166–174.
American Society for Clinical Nutrition /American Institute for Nutrition (ASCN/AIN) Task
Force on Trans Fatty Acids (1996), Position paper on trans fatty acids. Am. J. Clin. Nutr.,
63, 663–670.
ANVISA (Agência Nacional de Vigilância Sanitária), National Agency for Sanitary Moni-
toring Brazil. Mercosul Regulamento Téchnico Sobre Rotulagem Nutricional de Alimentos
Embalados. RDS No.360, 23 December 2003. (In Portuguese.)
Aro, A (1998) Epidemiology of trans fatty acids and coronary heart disease in Europe. Nutr.
Metab. Cardiovasc. Dis., 8, 402–407.
Aro, A, Kardinaal, AF, Salminen, I, Kark, JD, Riemersma, RA, Delgado-Rodriguez, M,
Gomez-Aracena, J, Huttunen, JK, Kohlmeier, L, Martin, BC, Martin-Moreno JM,
Mazaev, VP, Ringstad, J, Thamm, M, Van’t Veer, P, Kok, FJ (1995) Adipose tissue
isomeric trans fatty acids and risk of myocardial infarction in nine countries: the
EURAMIC study. Lancet, 345, 273–278.
Bahrami, G and Rahimi, Z (2005) Fatty acid composition of human milk in Western Iran. Eur.
J. Clin. Nutr., 59, 494–497.
Baylin, A, Kabagambe, EK, Ascherio, A, Spiegelman, D and Campos, H (2003) High 18:2
trans-fatty acids in adipose tissue are associated with increased risk of nonfatal acute
myocardial infarction in Costa Rican adults. J. Nutr., 133, 1186–1191.
Bertolino, CN, Castro, TG, Sartorelli, DS, Ferreira, SR and Cardoso, MA (2006) Dietary
trans fatty acid intake and serum lipid profile in Japanese-Brazilians in Bauru, São Paulo,
Brazil. Cadernos de Saúde Pública, 22, 357–364. (In Portuguese.)
Beynen, AC, Hermus, RJ and Hautvast, JG (1980) A mathematical relationship between the
fatty acid composition of the diet and that of the adipose tissue in man. Am. J. Clin. Nutr.,
33, 81–85.
Boué, C, Combe, N, Billeaud, C, Mignerot, C, Entressangles, B, Thery, G, Geoffrion, H,
Brun, JL, Dallay, D and Leng, JJ (2000) Trans fatty acids in adipose tissue of French
Brisson, J (1981) Lipids in Human Nutrition: An Appraisal of Some Dietary Concepts. Jack
K. Burgess Inc, Eaglewood, New Jersey, USA, pp.41–71.
Canadian Department of Health. (2003) Regulations amending and food and drug regulations
(Nutrition labeling, nutrient content claims and health claims). Canada Gazette, Part II,
137 (1), p.154, 1 January 2003.
Cantwell, MM, Flynn, MA, Cronin, D, O’Neill, JP and Gibney, MJ (2005a) Contribution of
foods to trans unsaturated fatty acid intake in a group of Irish adults. J. Hum. Nutr. Diet.,
18, 377–385.
Cantwell, MM, Gibney, MJ, Cronin, D, Younger, KM, O’Neill, JP, Hogan, L and Flynn, MA
(2005b) Development and validation of a food-frequency questionnaire for the determi-
nation of detailed fatty acid intakes. Public Health Nutrition, 8, 97–107.
Chardigny, JM, Wolff, RL, Mager, E, Sébédio, J-L, Martine, L and Juanéda, P (1995)
Trans mono- and polyunsaturated fatty acids in human milk. Eur. J. Clin. Nutr., 49,
523–531.
Chardigny, JM, Malpuech-Brugere, C, Dionisi, F, Bauman, DE, German, B, Mensink, RP,
Combe, N, Chaumont, P, Barbano, DM, Enjalbert, F, Bezelgues, J-B, Cristiani, I, Moulin,
J, Boirie, Y, Golay, P-A, Giuffrida, F, Sebedio, J-L, Destaillats, F (2006) Rationale and
design of the TRANSFACT project phase I: a study to assess the effect of the two different
dietary sources of trans fatty acids on cardiovascular risk factors in humans. Contempo-
rary Clinical Trials, 27, 364–373.
Barbano, DM, Mensink, RP, Bezelgues, J-B, Chaumont, P, Combe, N, Cristiani, I, Joffre,
F, German, JB, Dionisi, F, Boirie, Y, and Sebedio, J-L. (2008) Do trans fatty acids from
vascular disease risk factors in healthy subjects? Results of the trans Fatty Acids
Chen, ZY, Pelletier, G, Hollywood, R and Ratanayke, WM (1995a) Trans fatty acid isomers
in Canadian human milk. Lipids, 30, 15–21.
Chen, ZY, Ratnayake, WM, Fortier, L, Ross, R and Cunnane, SC (1995b) Similar distribution
of trans-fatty acid isomers in partially hydrogenated vegetable oils and adipose tissue of
Canadians. Can. J. Physiol. Pharmacol., 73, 718–723.
Clifton, PM, Keogh, JB and Noakes, M (2004) Trans fatty acids in adipose tissue and the food
supply are associated with myocardial infarction. J. Nutr., 134, 874–879.
Colón-Ramos, U, Baylin, A and Campos, H (2006) The relation between trans fatty acid
levels and increased risk of myocardial infarction does not hold at lower levels of trans
fatty acids in the Costa Rican food supply. J. Nutr., 136, 2887–2892.
Combe, N, Boué, C and Entressangles, B (2000) Consommation en acides gras trans et risqué
cardio-vasculaire: Étude Aquitaine. Ol. Corps Gras, Lipides, 7, 17–24. (In French.)
Combe, N, Judde, A, Boué, C, Billeaud, C, Entressangles, B, Dallay, D, Leng, JJ and Baste,
JC (1998) Composition en acides gras trans du tissu adipeux d’une population francaise
et origines alimentaires de ces acides gras trans. Ol. Corps Gras, Lipides, 5, 142–148. (In
French.)
Craig-Schmidt, MC (1998) Worldwide consumption of trans fatty acids. In: Trans-Fatty
Acids in Human Nutrition (JL Sébédio and WW Christie, eds), The Oily Press,
Craig-Schmidt, MC (2001) Isomeric fatty acids: evaluating status and implications for
maternal and child health. Lipids, 36, 997–1006.
Craig-Schmidt, MC (2006) World-wide consumption of trans fatty acids. Atherosclerosis
Supplements, 7, 1–4.
Craig-Schmidt, MC (2007) Consumption of trans fatty acids in North America. In: Trans Fats
in Foods (GR List, D Kritchevsky, and N Ratnayake, eds), AOCS Press, Urbana. Illinois,
USA, pp.127–153.
Craig-Schmidt, MC, Weete, JD, Faircloth, SA, Wickwire, MA and Livant, EJ (1984) The
effect of hydrogenated fat in the diet of nursing mothers on lipid composition and
prostaglandin content of human milk. Am. J. Clin. Nutr., 39, 778–786.
Craig-Schmidt, MC and Teodorescu, CA (2008) Trans-fatty acids in foods. In: Fatty Acids
in Foods and Their Health Implications, Third edition (CK Chow, ed.), CRC Press, New
York, USA, pp.377–437.
DHHS/FDA, Department of Health and Human Services, Food and Drug Administration.
(2003) 21 CFR Part 101 (Docket No. 94P-0036). Food labeling: Trans fatty acids in
nutrition labeling, nutrient content claims and health claims. Washington, DC, USA, July
2003, p.254.
Dlouhý, P, Tvrzická, E, Stanková, B, Buchtiková, M, Pokorný, R, Wiererová, O, Bílková, D,
Rambousková, J and Andel, M (2002) Trans fatty acids in subcutaneous fat of pregnant
women and in human milk in the Czech Republic. Ann. NY Acad. Sci., 967, 544–547.
Eckel, RH, Borra, S, Lichtenstein, AH, Yin-Piazza, SY and Trans Fat Conference Planning
Group. (2007) Understanding the complexity of trans fatty acid reduction in the
American diet: American Heart Association Trans Fat Conference 2006: report of the
Trans Fat Conference Planning Group. Circulation, 115, 2231–2246.
Elias, SL and Innis, SM (2002) Bakery foods are the major dietary source of trans-fatty acids
among pregnant women with diets providing 30 percent energy from fat. J. Am. Diet.
Assoc., 102, 46–51.
Enig, MG, Atal, S, Keeney, M and Sampugna, J (1990) Isomeric trans fatty acids in the US
Diet. J. Am. Coll. Nutr., 9, 471–486.
Friesen, R and Innis, SM (2006) Trans fatty acids in human milk in Canada declined with the
introduction of trans fat food labeling. J. Nutr., 136, 2559–2561.
FSA, UK Food Standards Agency (2003) The Food Labeling Regulations 1996: Guidance
Notes. 2003. Internet address: www.food.gov.uk/foodindustry/guidancenotes/
labelregsguidance/foodlabelregsguid
Garland, M, Sacks, FM, Colditz, GA, Rimm, EB, Sampson, LA, Willett, WC and Hunter, DJ
(1998) The relation between dietary intake and adipose composition of selected fatty
acids in US women. Am. J. Clin. Nutr., 67, 25–30.
Glew, RH, Herbein, JH, Moya, MH, Valdez, JM, Obadofin, M, Wark, WA and Vanderjagt,
DJ (2006) Trans fatty acids and conjugated linoleic acids in the milk of urban women and
nomadic Fulani of northern Nigeria. Clin. Chim. Acta, 367, 48–54.
Harnack, L, Lee, S, Schakel, SF, Duval, S, Luepker, RV and Arnett, DK (2003) Trends in the
trans-fatty acid composition of the diet in a metropolitan area: the Minnesota Heart
Survey. J. Am. Diet. Assoc., 103, 1160–1166.
Hayat, L, al-Sughayer, MA and Afzal, M (1999) Fatty acid composition of human milk in
Kuwaiti mothers. Comp. Biochem. Physiol. Part B: Biochemistry & Molecular Biology,
124, 261–267.
Hirsch, J, Farquhar, JW, Ahrens Jr, EH, Peterson, ML and Stoffel, W (1960) Studies of
adipose tissue in man: a microtechnic for sampling and analysis. Am. J. Clin. Nutr., 8,
499–511.
Hu, FB, Stampler, MD, Manson, JE, Rimm, E, Colditz, GA, Rosner, BA, Hennekens, CH and
Willett, WC (1997) Dietary fat intake and the risk of coronary heart disease in women.
New Eng. J. Med., 337, 1491–1499.
Hu, FB, Stampfer, MJ, Manson, JE, Grodstein, F, Colditz, GA, Speizer, FE and Willett, WC
(2000) Trends in the incidence of coronary heart disease and changes in diet and lifestyle
in women. New Eng. J. Med., 343, 530–537.
Hulshof, KF, van Erp-Baart, MA, Anttolainen, M, Becker, W, Church, SM, Couet, C,
Herman-Kunz, E, Kesteloot, H, Leth, T, Martins, I, Moreiras, O, Moschandreas, J,
Pizzoferrato, L, Rimestad, AH, Thorgeirsdottir, H, van Amelsvoort, JM, Aro, A, Kafatos,
AG, Lanzmann-Petithory, D and van Poppel, G (1999) Intake of fatty acids in Western
Europe with emphasis on trans fatty acids: the TRANSFAIR Study. Eur. J. Clin. Nutr.,
53, 143–157.
Hunter, JE (2006) Dietary trans fatty acids: review of recent human studies and food industry
responses. Lipids, 41, 967–992.
Innis, SM, Green, TJ and Halsey, TK (1999) Variability in the trans fatty acid content of foods
within a food category: implications for estimation of dietary trans fatty acid intake.
J. Am. Coll. Nutr., 18, 255–260.
Innis, SM and King, DJ (1999) Trans fatty acids in human milk are inversely associated with
concentrations of essential all-cis n–6 and n–3 fatty acids and determine trans, but not n–
6 and n–3, fatty acids in plasma lipids of breast-fed infants. Am. J. Clin. Nutr., 70,
383–390.
Innis, SM, Vaghri, Z and King, DJ (2004) n–6 Docosapentaenoic acid is not a predictor of
low docosahexaenoic acid status in Canadian preschool children. Am. J. Clin. Nutr., 80,
768–773.
International Life Sciences Institute (ILSI) Expert Panel on Trans Fatty Acids and Coronary
Heart Disease (1995) Trans fatty acids and coronary heart disease risk. Am. J. Clin. Nutr.,
62, 655S–708S.
International Life Sciences Institute (ILSI) Expert Panel on Trans Fatty Acids and Early
Development (1997) Trans fatty acids: infant and fetal development. Am. J. Clin. Nutr.,
66, 715S–736S.
Jakobsen, MU, Bysted, A, Andersen, NL, Heitmann, BL, Hartkopp, HB, Leth, T, Overvad,
K and Dyerberg, J (2006) Intake of ruminant trans fatty acids in the Danish population
aged 1–80 years. Eur. J. Clin. Nutr., 60, 312–318.
fatty acids and risk of coronary heart disease. Int. J. Epidemiology, 37, 173–182.
Johansson, L, Borgejordet, A and Pedersen, JI (2006) Trans fatty acids in the Norwegian diet.
Tidsskrift for den Norske Legeforening,126, 760–763. (In Norwegian.)
Katan, MB (2006) Regulation of trans fats: the gap, the Polder, and McDonald’s French fries.
Atherosclerosis Supplements, 7, 63–66.
King, IB, Kristal, AR, Schaffer, S, Thornquist, M and Goodman, GE (2005) Serum trans-
fatty acids are associated with risk of prostate cancer in beta-carotene and retinol efficacy
trial. Cancer Epidemiology, Biomarkers & Prevention, 14, 988–992.
Kromhout, D, Menotti, A, Bloemberg, B, Aravanis, C, Blackburn, H, Buzina, R, Dontas, AS,
Fidanza, F, Giampaoli, S, Jansen, A, Karvonen, M, Katan, M, Nissinen, A, Nedeljkovic,
S, Pekkanen, J, Pekkarinen, M, Punsar, S, Rasanen, L, Simic, B and Toshima, H (1995)
Dietary saturated and trans fatty acids and cholesterol and 25-year mortality from
coronary heart disease: The Seven Countries Study. Preventive Medicine, 24, 308–315.
Lee, S, Harnack, L, Jacobs, DR Jr, Steffen, LM, Luepker, RV and Arnett, DK (2007) Trends
in diet quality for coronary heart disease prevention between 1980–1982 and 2000–2002:
The Minnesota Heart Survey. J. Am. Diet. Assoc., 107, 213–222.
Lemaitre, RN, King, IB, Mozaffarian, D, Sotoodehnia, N, Rea, TD, Kuller, LH, Tracy, RP
and Siscovick, DS (2006) Plasma phospholipid trans fatty acids, fatal ischemic heart
disease, and sudden cardiac death in older adults: the cardiovascular health study.
Circulation, 114, 209–215.
Lemaitre, RN, King, IB, Patterson, RE, Psaty, BM, Kestin, M and Heckbert, SR (1998)
Assessment of trans-fatty acid intake with a food frequency questionnaire and validation
with adipose tissue levels of trans-fatty acids. Am. J. Epidemiology, 148, 1085–1093.
Lemaitre, RN, King, IB, Raghunathan, TE, Pearce, RM, Weinmann, S, Knopp, RH, Copass,
MK, Cobb, LA and Siscovick, DS (2002) Cell membrane trans-fatty acids and the risk
of primary cardiac arrest. Circulation, 105, 697–701.
Leth, T, Bysted, A, Hansen, K and Ovesen, L (2003) Trans fatty acid content in Danish
margarines and shortenings. J. Am. Oil Chem. Soc., 80, 475–478.
Leth, T, Jensen, HG, Mikkeksen, AA, and Bysted, A (2006) The effect of the regulation on
trans fatty acid content in Danish food. Atherosclerosis Supplements 7, 53–56.
Liu, X, Schumacher, FR, Plummer, SJ, Jorgenson, E, Casey, G and Witte, JS (2007) Trans-
fatty acid intake and increased risk of advanced prostate cancer: modification by
RNASEL R462Q variant. Carcinogenesis, 28, 1232–1236.
Lopes, C, Aro, A, Azevedo, A, Ramos, E and Barros, H (2007) Intake and adipose tissue
composition of fatty acids and risk of myocardial infarction in a male Portuguese
community sample. J. Am. Diet. Assoc., 107, 276–286.
Lopez-Garcia, E, Schulze, MB, Manson, JE, Meigs, JB, Albert, CM, Rifai, N, Willett, WC
and Hu, FB (2004) Consumption of (n–3) fatty acids is related to plasma biomarkers of
inflammation and endothelial activation in women. J. Nutr., 134, 1806–1811.
Lopez-Garcia, E, Schulze, MB, Meigs, JB, Manson, JE, Rifai, N, Stampfer, MJ, Willett, WC
and Hu, FB (2005) Consumption of trans fatty acids is related to plasma biomarkers of
inflammation and endothelial dysfunction. J. Nutr., 135, 562–566.
Lu, M, Taylor, A, Chylack, LT Jr, Rogers G, Hankinson, SE, Willett, WC, Jacques, PF (2005)
Dietary fat intake and early age-related lens opacities. Am. J. Clin. Nutr., 81, 773–779.
Mann, NJ, Sinclair, AJ, Percival, P, Lewis, JL, Meyer, BJ and Howe, PRC (2003) Develop-
ment of a database of fatty acids in Australian foods. Nutr. Diet., 60, 42–45.
McNaughton, SA, Hughes, MC and Marks, GC (2007) Validation of a FFQ to estimate the
intake of PUFA using plasma phospholipid fatty acids and weighed foods records. Brit.
J. Nutr., 97, 561–568.
Mensink, RP and Katan, MB (1990) Effect of dietary trans fatty acids on high-density and
low-density lipoprotein cholesterol levels in healthy subjects. New Eng. J. Med., 323,
439–445.
Mojska, H, Socha, P, Socha, J, Soplinska, E, Jaroszewska-Balicka, W and Szponar, L (2003)
Trans fatty acids in human milk in Poland and their association with breastfeeding
mothers’ diets. Acta Paediatrica, 92, 1381–1387.
Monge-Rojas, R, Campos, H and Fernández Rojas, X (2005) Saturated and cis- and trans-
unsaturated fatty acids intake in rural and urban Costa Rican adolescents. J. Am. Coll.
Nutr., 24, 286–293.
Mosley, EE, Wright, AL, McGuire, MK and McGuire, MA (2005) Trans fatty acids in milk
produced by women in the United States. Am. J. Clin. Nutr., 82, 1292–1297.
P, and Lamarche, B (2008) Study of the effect of trans fatty acids from ruminants on blood
Mozaffarian, D, Abdollahi, M, Campos, H, Houshiarrad, A and Willett, WC (2007)

Consumption of trans fats and estimated effects on coronary heart disease in Iran. Eur.
J. Clin. Nutr., 61, 1004–1010.
Mozaffarian, D, Katan, MB, Ascherio A, Stampfer, MJ and Willett, WC (2006) Trans fatty
acids and cardiovascular disease. New Eng. J. Med., 354, 1601–1613.
Mozaffarian, D, Pischon, T, Hankinson, SE, Rifai, N, Joshipura, K, Willett, WC and Rimm,
EB (2004a) Dietary intake of trans fatty acids and systemic inflammation in women. Am.
J. Clin. Nutr., 79, 606–612.
Mozaffarian, D, Rimm, EB, King, IB, Lawler, RL, McDonald, GB and Levy, WC (2004b)
Trans fatty acids and systemic inflammation in heart failure. Am. J. Clin. Nutr., 80, 1521–
1525.
Oomen, CM, Ocké, MC, Feskens, EJ, van Erp-Baart, MA, Kok, FJ and Kromhout, D (2001)
Association between trans fatty acid intake and 10-year risk of coronary heart disease in
the Zutphen Elderly Study: a prospective population-based study. Lancet, 357, 746–751.
Pedersen, JI, Johansson, L, and Thelle, DS (1998) Trans fatty acids and health. Tidsskr Nor
Laegeforen, 118, 3474–3480. (in Norwegian.)
including conjugated cis delta 9, trans delta 11 linoleic acid (CLA) as well as total fat
composition of German human milk lipids. Die Nahrung, 43, 233–244.
Ratnayake, WMN and Chen, ZY (1995) Trans fatty acids in Canadian breast milk and diet.
In: Development and Processing of Vegetable Oils for Human Nutrition (R Przybylski
and BE McDonald, eds), AOCS Press, Champaign, Illinois, USA, pp.20–35.
Roberts, TL, Wood, DA, Riemersma, RA, Gallagher, PJ and Lampe, FC (1995) Trans
isomers of oleic and linoleic acids in adipose tissue and sudden cardiac death. Lancet, 345,
278–282.
Schrimpf-Moss, J and Wilkening, V (2007) Labeling of trans fatty acids. In: Trans Fats in
Foods (GR List, D Kritchevsky, and N Ratnayake, eds), AOCS Press, Urbana, Illinois,
pp.177–191.
Senti, FR, ed. (1985) Health Aspects of Dietary Trans Fatty Acids. Life Science Research
Office, Federation of American Societies for Experimental Biology; Bethesda, Mary-
land, USA
Skeaff, CM and Gowans, S (2006) Home use of margarine is an important determinant of
plasma trans fatty acid status: a biomarker study. Brit. J. Nutr., 96, 377–383.
Stender, S and Dyerberg, J (2003) The Influence of Trans Fatty Acids on Health. A Report
From the Danish Nutrition Council, Fourth edition. Publication no.34.
Stender, S, Dyerberg, J and Astrup, A (2006) High levels of industrially produced trans fat
in popular fast foods. New Eng. J. Med., 354, 1650–1652.
Sun, Q, Ma, J, Campos, H, Hankinson, SE and Hu, FB (2007a) Comparison between plasma
and erythrocyte fatty acid content as biomarkers of fatty acid intake in US women. Am.
J. Clin. Nutr., 86, 74–81.
Sun, Q, Ma, J, Campos, H, Hankinson, SE, Manson, JE, Stampfer, MJ, Rexrode, KM, Willett,
WC and Hu, FB (2007b) A prospective study of trans fatty acids in erythrocytes and risk
of coronary heart disease. Circulation, 115, 1858–1865.
Tarrago-Trani, MT, Phillips, KM, Lemar, LE and Holden, JM (2006) New and existing oils
and fats used in products with reduced trans-fatty acid content. J. Am. Diet. Assoc., 106,
867–880.
Tsai, CJ, Leitzmann, MF, Willett, WC and Giovannucci, EL (2005) Long-term intake of
trans-fatty acids and risk of gallstone disease in men. Arch. Int. Med., 165, 1011–1015.
van de Vijver, LPL, van Poppel, G, van Houwelingen, A, Kruyssen, DA and Hornstra, G
(1996) Trans unsaturated fatty acids in plasma phospholipids and coronary heart disease:
a case-control study. Atherosclerosis, 126, 155–161.
van de Vijver, LP, Kardinaal, AF, Couet, C, Aro, A, Kafatos, A, Steingrimsdottir, L, Amorim
Cruz, JA, Moreiras, O, Becker, W, van Amelsvoort, JM, Vidal-Jessel, S, Salminen, I,

Moschandreas, J, Sigfússon, N, Martins, I, Carbajal, A, Ytterfors, A and Poppel, G (2000)
Association between trans fatty acid intake and cardiovascular risk factors in Europe: the
TRANSFAIR study. Eur. J. Clin. Nutr., 54, 126–135.
Voorrips, LE, Brants, HA, Kardinaal, AF, Hiddin, GJ, van den Brandt, PA and Goldbohm,
RA (2002) Intake of conjugated linoleic acid, fat, and other fatty acids in relation to
postmenopausal breast cancer: the Netherlands Cohort Study on Diet and Cancer. Am. J.
Clin. Nutr., 76, 873–882.
Weggemans, RM, Rudrum, M and Trautwein, EA (2004) Intake of ruminant versus industrial
trans fatty acids and risk of coronary heart disease — what is the evidence? Eur. J. Lipid.
Sci. Technol., 106, 390–397.
Willett, W, and Mozaffarian, D (2008) Ruminant or industrial sources of trans fatty acids:
public health issue or food label skirmish? Am. J. Clin. Nutr., 87, 515–516.
Wolff, RL (1995) Content and distribution of trans-18:1 acids in ruminant milk and meat fats.
Their importance in European diets and their effect on human milk. J. Am. Oil Chem.
Soc., 72, 259–272.
Wolff, RL, Precht, D and Molkentin, J (1998) Trans-18:1 acid content and profile in human
milk lipids. Critical survey of data in connection with analytical methods. J. Am. Oil
Chem. Soc., 75, 661–671.
Zhou, BF, Stamler, J, Dennis, B, Moag-Stahlberg, A, Okuda, N, Robertson, C, Zhao, L, Chan,
Q, Elliott, P and INTERMAP Research Group (2003) Nutrient intakes of middle-aged
men and women in China, Japan, United Kingdom, and United States in the late 1990s:
the INTERMAP study. J. Hum. Hypertens., 17, 623–630.
Zock, PL and Katan, MB (1992) Hydrogenation alternatives: effects of trans fatty acids and
stearic acid versus linoleic acid on serum lipids and lipoproteins in humans. J. Lipid Res.,
33, 399–410.
CHAPTER 14
Legislation relating to trans fatty acids
KOENRAAD DUHEM
Centre National Interprofessionnel de l’Économie Laitière (CNIEL), Paris,

France
A. Definition of trans fatty acids: a prerequisite to regulation

The chemistry of trans fatty acids (TFA) has been described in an earlier
chapter. The same isomers occur in industrially produced TFA and ruminant
TFA but the pattern of individual isomers differs widely depending on the
source. In ruminants, mainly vaccenic (trans-11 18:1) acid is formed and
accounts for 60% of the TFA in animal fat, whereas it represents less than 20%
of industrially produced TFA. During partial hydrogenation of vegetable oils,
various trans octadecenoic acid isomers such as elaidic (trans-9 18:1) or trans-
10 18:1 acids are formed, but these isomers are present at lower concentrations
in ruminant fat. A conjugated isomer of linoleic acid (CLA) having a trans
double bond and with interesting biological properties is naturally present in
small amount in ruminant fats. CLA isomers have cis/trans or trans/cis double
bonds that are conjugated, i.e. not interrupted by a methylene group. The main
CLA isomers are cis-9,trans-11 18:2 (rumenic acid, because it is found in the
rumen and is the main ruminant CLA isomer) and trans-10,cis-12 18:2 acid.
The choice of regulatory definition has differed from one country to another
and is reviewed in this chapter.
1. Australia and New Zealand

The approach to regulatory definition in Australia and New Zealand closely
follows the chemical definition of trans configuration in fatty acids, and
includes all types of TFA, including ruminant TFA. The regulatory definition
is as follows:
“Trans fatty acids means the total number of unsaturated fatty acids where one or more
of the double bonds are in the trans configuration and declared as trans fat.”
381
2. France
In France too, a similar solely chemical definition was adopted in February
2005 by the Agence Française de Sécurité Sanitaire des Aliments (AFSSA,
2005).
3. Definition excluding conjugated isomers of linoleic acids: rationale

In most of the other countries the choice was made to adopt a regulatory
definition that would not be solely based on chemical definition of TFA but
would also distinguish between different dietary sources of TFA. When the
regulations regarding TFA were developed in the USA, there were a number of
requests that certain ruminant TFA should be excluded from the regulatory
definition. Further, there were some suggestions that the definition should be
based on functional or metabolic aspects of the fatty acids, and not their actual
chemical structure.
As a result several countries have restricted the chemical definition of TFA.
The Food and Drug Administration in the USA (2003), the Canadian govern-
ment and the Danish Nutrition Council include in the definition of TFA all
monounsaturated TFA as well as all polyunsaturated TFA with non-conjugated
(isolated) double bonds or non–methylene interrupted double bonds. This
definition does not encompass the main dietary conjugated linoleic acid
isomers that have trans/cis or cis/trans conjugated ethylenic double bonds. The
regulatory definitions at use in those countries are close but not identical.
4. Denmark
In Denmark, TFA are defined as:
“The sum of all fatty acid isomers with 14, 16, 18, 20 or 22 carbon atoms and one or
more trans double bonds, i.e. C14:1, C16:1, C18:1, C18:2, C18:3, C20:1, C20:2,
C22:1, C22:2 fatty acid trans isomers, but only polyunsaturated fatty acids with
methylene interrupted double bonds.”
5. Canada
In Canada, TFA means:
“Unsaturated fatty acids that contain one or more isolated or non-conjugated double
bonds in a trans configuration”
6. USA
The definition of TFA in the USA is:
LEGISLATION RELATING TO TRANS FATTY ACIDS 383
“Unsaturated fatty acids that contain one or more isolated (i.e, non-conjugated) double
bonds in a trans configuration.”
7. Codex Alimentarius definition

During its 21st session in November 2003, the Codex Committee on Nutrition
and Foods for Special Dietary Uses (CCNFSDU) defined TFA as:
“all geometrical isomers of mono- and poly-unsaturated fatty acids having non
conjugated carbon-carbon double bounds interrupted by at least one methylene group
in trans configuration”.
On its 26th session in November 2004 this definition evolved slightly and in
addition incorporated a mistake because the methylene group was designated
–CH2–CH2– where it should have been –CH2–.
Many regulatory definitions of TFA, while not specifically excluding rumi-
nant TFA, exclude fatty acids with conjugated bonds, even though these acids
have double bonds in trans configuration. These definitions stem from the view
that regulatory definitions try to a certain extent to identify adequately the fatty
acids targeted by the regulation.
8. European Food Safety Authority definition

In its 2004 opinion, the European Food Safety Authority (EFSA, 2004) gives a
broad definition of TFA:
“Trans fatty acids (TFA) are unsaturated fatty acids that have at least one double bond
in the trans configuration. Some polyunsaturated TFA have a conjugated structure (e.g.
conjugated linoleic acid [CLA] in milk fat), i.e. have double bonds which are not
separated by a methylene group, but most have isolated (non-conjugated) double
bonds”.
At the European Commission level the TFA issue is currently being dis-
cussed in the context of the ‘Nutrient profiles’ and for labelling purposes, but
no regulatory definition has been proposed yet (as at November 2007). How-
ever, release has been announced. The fact that regulatory definitions exclude
CLA acknowledges differences between TFA, but only to a limited extent.
B. Regulatory decisions
1. CODEX/WHO/FAO
The Codex Alimentarius Commission (Codex) was created in 1963 to develop
food standards, guidelines and related texts such as codes of practice under the
Joint Food and Agriculture Organization/World Health Organisation Food
Standards Programme. Under the Guidelines for Nutrition Labelling of Codex,

TFA must be declared where the amount and/or type of fatty acids or the
amount of cholesterol is declared on a label.
In its draft action plan for implementation of the Global strategy on diet,
physical activity and health, the WHO (World Health Organisation, 2007) has
requested Codex to consider setting limits on the content of manufactured TFA
in foods. The means of implementation have been referred to the ‘Codex
Committee on Nutrition and Foods for Special Dietary Uses’ (CCNFSU) and
the ‘Codex Committee on Food Labelling’ (CCFL). Since these are the initial
stages of consideration, it is likely to be a matter of years before the outcomes
are finalized as Codex standards. In this draft plan, the WHO has recommended
that governments around the world phase out partially hydrogenated oils if
trans fat labelling alone does not spur significant reductions:
“If the provisions for labelling of, and claims for, trans-fatty acids do not affect a
marked reduction in the global availability of foods containing trans-fatty acids
produced by processing of oils and by partial hydrogenation, consideration should be
given to the setting of limits on the content of industrially produced trans-fatty acids
in foods.”
The Global strategy paper of WHO established no distinction between

industrially produced and naturally produced TFA. A report of the Joint WHO/
FAO/ expert consultation, Diet, Nutrition and Prevention of Chronic Disease
(WHO,2003) assessed the adverse health impact of TFA and recommended a
population nutrient intake goal of less than 1%. This level has been a reference
target for most of the regulatory moves at national level.
As far as nutrition claims on TFA are concerned the current reference in the
Guidelines on Nutrition and Health Claims results from the FAO/WHO Expert
Consultation on Fats and Oils in Human Nutrition held in 1993. Further
scientific advice has been required from FAO/WHO in this area by the
CCNFSDU. A scientific advisory meeting on fats and oils had been convened
in the first half of 2008, but it has been replaced by Joint FAO/WHO expert
consultation on fats and oils in human nutrition scheduled for early 10–14
November 2008. One of the four agenda items is:
“…to assess the risks and the health effects of excessive intakes of fat and fatty acids,
in particular total fat, saturated fatty acids and trans fatty acids”.
A previous authors’ meeting took place on 29–30 October 2007 in Geneva,

Switzerland, on this issue, with a view to working out a ‘WHO Scientific Update
on the Trans Fatty Acids’. The six items under discussion were as follows.
• General historical background of the work related to TFA and the Global
Strategy.
• Health effects of TFA: experimental and observational evidence.

• Quantitative effects on cardiovascular risk factors and coronary heart
disease risk of replacing partially hydrogenated vegetable oils with other
fats and oils.
• Feasibility of recommending certain replacement or alternative fats.
• Assessing approaches to removing TFA in the food supply in industrialized
countries and in developing countries.
• Summary and conclusions of the Scientific Update.
Each of those items will lead to a scientific review paper that will be
published in a supplement to the European Journal of Clinical Nutrition.
The joint FAO/WHO expert consultation is likely to be decisive for deter-
mining the direction of future Codex and national regulations with regard to
labelling of TFA and/or removing of TFA from food.
In the report of the 35th session of CCFL held in Ottawa in May 2007 it was
decided that taking into account the need for additional scientific advice, no
new work would be undertaken on nutrition claims for TFA in CCFL.
2. Denmark and Switzerland

In March 2003, the Danish authorities decided to take a step further by adopting
a regulation to be enforced in June 2003, in order to reduce the levels of
industrially produced TFA in fats and processed foods. The Order sets the
scope of the regulation; it applies to oils and fats including emulsions which
alone or as part of foodstuffs are intended to be consumed by humans. It does
not apply to naturally occurring content of TFA in animal fats. The content of
TFA in the designated oils and fats must not exceed 2%.
This Order of the Danish authorities restricts, by its wording, the regulatory
definition of TFA made previously by the same authorities that included all the
monounsaturated TFA. Denmark has been the first country to implement a
policy of “Consumer protection through a legislative ban on industrially
produced trans fatty acids in foods in Denmark” (Stender et al., 2006). While
Denmark can mandate the composition of products sold, it cannot change the
rules for nutrition labelling on its own initiative independently of the European
Commission; this includes introducing compulsory declaration of TFA. Con-
sequently, there is no mandatory declaration of TFA in Denmark.
On 1 April 2008, Switzerland enforced a law similar to the one in force in
Denmark, with a limit of 2 g of TFA in 100 g of vegetable oil or fat. Producers
have a period of one year to make the changes. The law does not regulate TFA
from animal origin.
3. UK
In the UK, the Food Labelling Regulations require hydrogenated fat to be
identified as such in the ingredient list on the label when it has been used as an
ingredient in food. However, if hydrogenated fat is part of a compound
ingredient that makes up less than 25% of the finished product it is not required
to be mentioned in the ingredient list. In the UK, it appears there are no plans
at this stage to undertake regulatory action over and above current promotion
of a balanced diet within which edible oils of all types should be consumed
sparingly.
4. European Union
In reaction to the Danish decision and taking notice that it fostered observations
from other EU member states with diverging views upon the issue, the
European Commission decided to request the EFSA to issue a scientific
opinion on the presence of TFA in foods and on the effect on human health of
the consumption of TFA. In this context the Authority was asked:
• to take into account all trans fatty acids in foods, including the ingredients
of food products, both those that are naturally present, such as in certain
animal fats, and those occurring as a result of manufacturing processes,
such as hydrogenation of oils;
• to advise whether the evidence indicates any specific effects on health of
trans fatty acids, whether the effects, if any, differ according to the food
source, and how the effects, if any, compare to effects on health of other
types of fatty acids;
• to advise, if there are effects on health, whether the effects are associated
with a specific level of intake of trans fatty acids in the context of the overall
diet.
• to advise if there are any methods of analysis that can distinguish between
trans fatty acids which are naturally present in fats and those formed during
the processing of fats, oils or foods.
At the time of the opinion formulation the experts stated:
“In most of the human intervention studies monounsaturated TFA from hydrogenated
vegetable oils were evaluated. No human intervention studies have been carried out to
evaluate the effects of TFA from ruminant fat, and indeed such studies are not
practicable. Thus it is not possible to determine whether there are differences between
TFA from ruminant fat and TFA from hydrogenated vegetable oils in their effects on
metabolic risk parameters such as LDL-C or HDL-C”.
This prudent statement prompted the European Commission to wait for more
scientific evidence before taking any decisions on labelling or banning TFA.
In early 2007, Members of the European Parliament worked out a ‘Written
declaration on trans fatty acids’ calling on the European Commission and
Council to increase awareness of the dangers of TFA by introducing a manda-
tory labelling system and to introduce the necessary measures to ensure a

movement towards the elimination of TFA from food products. They express
concerns that the lack of existing legislation across the EU poses unnecessary
health risks to citizens, welcome the initiative taken by Denmark regarding
legislation aimed at significantly reducing the amount of TFA in food products,
and regret the move taken by the European Commission to initiate legal
proceedings against Denmark who banned industrially produced TFA.
The TFA issue was to be dealt with under an EU Directive on nutrition
labelling that had to be amended in 2008. Eventually, the TFA labelling issue
in EU will be ruled by ‘Regulation of the European Parliament and of the
council on the provision of food information to consumers’ (Commission of
European Communities, 2008) which is a regulation on both general and
nutrition labelling, still at a proposal stage. After revision by the EU Parliament
in early 2009, the law should be enforced by the end of 2009.
Under this law proposal, TFA levels are required on labels only if a TFA
claim is made. Declarations of the amount of TFA in a food are subject to the
rules on nutrition labelling, which are harmonized at EU level. Nutrition
labelling is voluntary unless a nutrition claim is made. Separate identification
of the amount of TFA, as a component of the total fat content of the food, is only
required if a TFA nutrition claim is made.
5. USA
On 11 July 2003, the US Food and Drug Administration (FDA) issued a
regulation requiring that TFA be declared on the nutrition label of conventional
foods and dietary supplements on a separate line immediately under the line for
the declaration of saturated fatty acids on the Nutrition Facts panel of foods and
some dietary supplements (Food and Drug Administration, 2003). The new
labeling rule allowed for immediate voluntary compliance with mandatory
compliance by 1 January 2006. The regulation allows trans fat levels of less
than 0.5 g per serving to be labeled as 0 g per serving. The FDA did not approve
nutrient content claims such as “trans fat free” or “low trans fat”, as they could
not determine a “recommended daily value”.
It is worthwhile relating that this amendment to FDA regulation on nutrition
labelling came in response to a citizen petition from the Center for Science in
the Public Interest (CSPI) which was received by the FDA dated 14 February
1994. Until 1992, CSPI had defended TFA in preference to saturated fatty
acids.
The US regulatory authorities decided not to exclude TFA derived from
rumen hydrogenation from labelling requirements; consequently, dairy prod-
ucts must be appropriately labelled. However, TFA with conjugated bonds are
not included because they do not meet the US regulatory definition of TFA.
The state of California has recently announced a new legislation which will
ban from 1 January 2010 the use of trans fats in oil, shortening and margarine used
in spreads or for frying in restaurants, and by 2011 in all retail baked goods.
Packaged foods will be exempt. Several major US towns, such as New York and
Seattle, have already adopted similar bans on trans fat use in restaurants.
6. Canada
Trans fats in Canadian foods have been reviewed by the Trans Fat Task Force
in its report ‘TRANSforming the food supply’ submitted to Canada’s Minister
of Health in June 2006 (Health Canada, 2006).
The Canadian Food and Drug Regulations (FDR) specifically prescribe what
information must be displayed on a food label. In December 2002, the Food and
Drug Regulations were amended to make nutrition labelling mandatory on
most food labels, to update requirements for nutrient content claims, and to
permit the use of diet-related health claims on foods (Canada Gazette, 2003) .
As a result, nutrition labelling became mandatory for most pre-packaged foods
in December 2005, with smaller businesses having until December 2007 to
comply with the new regulations. The TFA content of a food is considered core
nutrition information and must be declared in a Nutrition Facts table.
With regard to TFA, the Danish example triggered the Canadian House of
Commons to introduce and vote a motion to ban TFA. In December 2005,
Health Canada required that food labels list the amount of trans fat in the
nutrition facts section for most foods. Products with less than 0.2 g of trans fat
per serving may be labeled as free of trans fats. Other claims are allowed such
as “Reduced in trans fatty acids” provided the food is processed, formulated,
reformulated or otherwise modified, without increasing the content of satu-
rated fatty acids, so that it contains at least 25% less trans fatty acid. The claim
“Lower in trans fatty acids” is also permitted provided the food contains at least
25% less TFA and the content of saturated fatty acids is not higher per reference
amount of the food, than the reference amount of the reference food of the same
food group. These labelling allowances are not widely known and controversy
over truthful labelling has arisen; they are not believed to protect consumers
from high exposure.
Presently, TFA quantities on labels include naturally-occurring trans fats.
CLA isomers are not included in the label declaration because they fall outside
the definition.
In mid-2007, the federal government announced its intention to regulate
trans fats to new standards. These would be based on the June 2006 outcomes
of a ‘Trans Fats Task Force” co-chaired by Health Canada and the Heart and
Stroke Foundation of Canada. The Task Force had a mandate to develop
recommendations and strategies to effectively eliminate or reduce processed
trans fats in Canadian foods to the lowest level possible. It recommended the
following.
• Foods purchased by retailers or food service establishments from a manu-

facturer for direct sale to consumers be regulated on a finished product or
output basis and foods prepared on site by retailers or food service establish-
ments be regulated on an ingredient or input basis.
• For all vegetable oils and soft, spreadable (tub-type) margarines sold to
consumers or for use as an ingredient in the preparation of foods on site by
retailers or food service establishments, the total trans fat content be limited
by regulation to 2% of total fat content.
• For all other foods purchased by a retail or food service establishment for
sale to consumers or for use as an ingredient in the preparation of foods on
site, the total trans fat content be limited by regulation to 5% of total fat
content. This limit does not apply to food products for which the fat
originates exclusively from ruminant meat or dairy products.
• This manner of expression (2%, 5%) helps cover products that are not
subjected to the nutrition labelling regulations, for example non pre-
packaged foods such as those that are sold in food service establishments.
• The Government of Canada and all concerned food industry associations
urge companies affected to use the most healthful oils for their food
applications.
• Facilitate the reformulation of food products with healthier alternatives to
trans fat.
• Help the food industry communicate the healthier nature of its products to
consumers.
• Help small and medium-sized enterprises prepare for compliance.
• Enhance the capacity of the Canadian agri-food industry to take a leader-
ship role in this area.
The Task force further recommended that these regulations be finalized by

June 2008, with appropriate transition periods to allow industry to comply with
the new requirements. This implementation would mean that most of the
industrially produced trans fats would be removed from the Canadian diet, and
about half of the remaining trans fat intake would be of naturally occurring
trans fats.
In fact, in June 2007, the Minister of Health, Canada announced that if
significant progress has not been made over the next two years to achieve the
2% and 5% limits recommended by the Trans Fat Task Force, the Department
would develop regulations to ensure that the levels are met.
To meet those limits, the Trans Fat Monitoring Program was established. It
analyses the trans fat content of foods that were, as indicated by earlier surveys,
significant sources of trans fats, i.e. foods with high levels of trans fats, or
foods with lower levels of trans fats that were consumed in large quantities by
a large number of consumers. The analysis of food samples is on-going and the
results from the program will be posted approximately every six months on the
Health Canada website. The second set of Trans Fat Monitoring data shows an
encouraging downward trend.
7. Australia
The Federal Government of Australia is currently (2008) willing to act against

industrial TFA and asks fast food companies to voluntarily reduce the TFA in
their products. If the industry does not act voluntarily, the Government will
consider the option of mandatory labeling.
So far, trans fatty acid contents must be declared on a food label if a nutrition
claim is made about cholesterol or saturated, trans, polyunsaturated or mono-
unsaturated fatty acids, or omega-3, omega-6 or omega-9 fatty acids
Food Standards Australia New Zealand (FSANZ, 2007) is in charge of the
issue. The Australia New Zealand Collaboration on Trans Fats was established
in early 2007. It includes representatives from the National Heart Foundation
of Australia, the National Heart Foundation of New Zealand, the Dietitians
Association of Australia, the Australian Food and Grocery Council, the New
Zealand Food and Grocery Council, the New Zealand Food Safety Authority
(NZFSA) and FSANZ.
The primary aim of this group is to cooperate in reducing the amount of TFA
in the New Zealand and Australian food supply, without an associated increase
in the amount of saturated fat. The group will promote wide implementation of
current industry and public health initiatives for reducing the levels of TFA and
increasing consumer awareness and understanding.
FSANZ has also conducted a formal scientific review of TFA in the food
supply and reported back to the Australia and New Zealand Food Regulation
Ministerial Council in May 2007. The Report found that the contributions of
TFA to energy intakes of Australians and New Zealanders was below the goal
of 1% proposed by the World Health Organisation, and comparable to or lower
than intake estimates from some countries overseas.
Ministers endorsed the findings of the Review that immediate regulatory
intervention was not required and that non-regulatory measures to further
reduce the levels of TFA in the Australian and New Zealand food supply would
be the most appropriate action.
The Report also recommended that a review would commence in early 2009
of the outcome of non-regulatory measures to reduce TFA in the food supply.
Ministers noted that regular updates on progress in trans fatty acid reduction
would be provided from the work being carried out by the Australia New
Zealand Collaboration on Trans Fats and agreed to consider regulatory action
should sufficient progress not be made.
New South Wales introduced a new bill – the Food Amendment (trans fatty
acids eradication) Bill 2008 – in order to make the Federal government enact
similar legislation for the protection of all Australians by the eradication of

trans fats in our national food supply.
8. Asia
In 2007, the Korea Food and Drug Administration adopted plans to designate
“zero-trans fat” “labeling on food products that contain less than 0.5 g of trans
fatty acids per portion and “non-trans fat” for those with less than 0.2 g.
Korea and Japan, whose citizens eat less TFA than Western countries, have
been less active in legislation efforts. But as the harmful effects of TFA have
been widely publicized around the world, Korea has revised food labeling
regulations to specify cholesterol, trans fats and saturated fat content and is the
first Asian country to do so.
The law takes effect in December 2007. By this date the Korean FDA had
also developed criteria for claims of “low trans fatty acids” and “free of trans
fatty acids”.
In 2007, the Taiwanese Department of Health implemented mandatory
labeling of TFA on pre-packaged foods. Trans fat must be expressed in g per
100 g or per serving. Nutrients content for trans fat may be labeled “0” if
nutrient content is less than 0.3 g for 100 g of solid (semi-solid) food or 100 ml
of liquid food.
C. Advantages and disadvantages of the different regulatory models

In essence, national and international health authorities adopted two main
policies to tackle the issues pertaining to TFA: banning or labelling. These two
options are still under consideration by WHO and FAO in their ‘Draft action
plan for implementation of the Global strategy on diet, physical activity and
health’ since it states that:
“If the provisions for labelling of, and claims for, trans-fatty acids do not affect a
marked reduction in the global availability of foods containing trans-fatty acids
produced by processing of oils and by partial hydrogenation, consideration should be
given to the setting of limits on the content of industrially produced trans-fatty acids
in foods”
The levels of enforcement of the measures accompanying those policies are,

however, fundamentally different, as discussed below.
1. Advantages and disadvantages of banning TFA: lessons from the Danish

experience
The first option is to phase-out the industrially produced TFA, since they are
presumed to be the main health hazard, at the level of the source. This has been
the choice of the Danish Food Administration.
• In Denmark (Stender et al., 2006) this policy proved to be efficient in

markedly reducing the average TFA intake in the populations and even
virtually eliminated industrially produced TFA from foods.
• There were no noticeable effects on availability, price or quality of foods.
• TFA were eliminated from margarines without increasing the amount of
saturated fats and often with an increase in monounsaturated fats.
This option can be challenged if no clear nutritional differences are shown
between industrially and naturally occurring TFA.
2. Advantages and disadvantages of labelling TFA

The second option, labelling, is a more liberal one. Labelling allows the market
and the consumers to make their choice or to take a risk or not. In the long term,
the result will be the same as with the first option; the hazard will eventually be
removed but the process will take longer. This policy will have longer lasting
and more irreversible effects; once the consumers have learnt that “TFA are
bad”, it will be difficult for them to learn that some TFA might be beneficial.
The main concern raised with this option is that the decision is left to the
consumer to whom nutritional labelling appears already very confusing and
complex. As a result, the labelling must be as simple as possible and is at risk
of over-simplification. With an over-simplified message, the consumer can be
induced to make wrong decisions and to refrain from ingesting what might
appear later as very valuable nutrients. Conversely to the consequences of the
first option, even if beneficial effects of natural TFA are clearly demonstrated,
distrust might dissuade consumers from choosing them.
The main advantage is that the consumer makes his own choice but the
disadvantages are:
• nutrition labelling is already very confusing and more confusion might be
added;
• the labelling option takes more time to achieve results than does a ‘ban’ type
of option;
• the labelling option will be more irreversible in its effects because all TFA
will be reputed to be bad, even if some isomers might later prove beneficial;
and
• this option might induce a loss of food diversity which could prove worse
than the primary risk.
D. Remarks and future perspective

Most recent policies adopted to manage the trans fatty acids issue, based on
success of the Danish experience, seek to implement the reduction or the
banning of industrially produced TFA (Canada, Switzerland, California) and to
target the issue at its main source. The trend is also toward considering
naturally occurring TFA as not having the same deleterious effects as the
industrial isomers, at the concentration at which they are found naturally.
Scientific investigation of the biological role of TFA started only very
recently. TFA exist as natural compounds, albeit in small amounts compared to
the cis isomers; they seem to be subject to selection by nature, judging from the
very specific pattern they display compared to the random pattern produced by
catalytic hydrogenation of vegetable oil. Some families of those fatty acids are
present in animal products or vegetables in amounts that compare to the levels
of some biologically active fatty acid families such as omega-3 fatty acids. The
experimental study of their effects – in vitro, on animal models, or in humans
– have so far been hindered by the limited availability of the pure isomers. Most
studies have so far used mixtures of isomers, which resulted in limited or
confusing conclusions.
New, forthcoming studies using pure isomers, at physiological levels, are
starting to demonstrate some coherent effects of those trans compounds. Some
positions of the trans double bounds in the carbon chain of fatty acids seem to
be favoured by biological selection (trans-11 double bound) as are also specific
non-methylene-interrupted double bonds.
It is hoped that these studies will shed light on a broader range of physiologi-
cally active and beneficial fatty acids and help the regulatory authorities take
decisions on a more scientifically founded basis.
References
AFSSA (2005) Risques et bénéfices pour la santé des AG trans apportés par les aliments —
recommandations. Agence Française de Sécurité Sanitaire des Aliments. Online:
www.afssa.fr
Canada Gazette (2003) Part II, Vol.137, No.1 (1 January 2003).
Commission of European Communities (2008) Com2008(40) Final. Proposal for a Regula-
tion of the European Parliament and of the council on the provision of food information
to consumers. Online: ec.europa.eu/food/food/labellingnutrition/foodlabelling/publica-
tions/proposal_regulation_ep_council.pdf
European Food Safety Authority (2004) Opinion of the Scientific Panel on Dietetic Products,
Nutrition and Allergies on a request from the Commission related to the presence of trans
fatty acids in foods and the effect on human health of the consumption of trans fatty acids
(adopted on 8 July 2004). EFSA Journal, 81 , 1–49. Online: www.efsa.eu.int/science/
nda/nda_opinions/588/opinion_nda09_ej81_tfa_en1.pdf
Food and Drug Administration, USA (2003) CFR Part 101. Food Labeling: Trans Fatty Acids
in Nutrition Labeling: Consumer Research to Consider Nutrient Content and Health
Claims and Possible Footnote or Disclosure Statements: Final Rule and Proposed Rule.
National Archives and Records Administration. Federal Register, 68 (133), pages
41433–41506, 11 July 2003.
Food Standards Australia New Zealand, FSANZ (2007) Trans fatty acids in the New Zealand
and Australian food supply. Review Report, FSANZ. Online: www.foodstandards.gov.au/
_srcfiles/Transfat%20report_CLEARED.pdf
Health Canada (2006) TRANSforming the food supply. Report by Canada’s Trans Fat Task
Force (ISBN 0-662-43689-X). Online: www.hc-sc.gc.ca/fn-an/alt_formats/hpfb-dgpsa/
pdf/nutrition/tf-gt_rep-rap_e.pdf (Accessed 11 November 2007)
Stender S, Dyerberg J and Astrup A (2006) High levels of industrially produced trans fat in
popular fast foods. New Engl. J. Med., 354, 1650–1652.
Stender, S, Dyerberg, J and Astrup, A (2006). Consumer protection through a legislative ban
on industrially produced trans fatty acids in foods in Denmark. Scandinavian Journal of
Food & Nutrition, 50, 155–160.
World Health Organisation (2007) Draft action plan for implementation of Global Strategy
on diet, Physical activity and Health. Alinorm 07/30/26 and CL/2006/44 -CAC
CHAPTER 15
Consumer concerns and risk perception related
to trans fatty acids
CLOTILDE AUBERTIN
Only ten years ago few people would have anticipated or even imagined the
enormous impact on our society of worldwide communication over the Internet,
and the profound effect that it has since had. Nowadays, consumers use this
technology daily to help them to gauge potential risks. In the case of trans fatty
acids (TFA), it would be relevant to measure the impact of Internet traffic
pertaining to its consumers. In this chapter, we are mainly concerned with the
following questions. What is the definition of ‘consumer’ in the increasing
information flow? What are consumers’ concerns in terms of foods and health?
How do they clarify such concerns? As information can now be obtained so
rapidly and easily, could the Internet be considered as a relevant way to identify
and measure risk perception by consumers? And how are TFA perceived in this
context?
A. Introduction and context around the consumer

With increasing ways of accessing information, the Internet allows us to
understand what drives consumers and, therefore, how they make their food
choices. The impact of these choices upon health, nutrition and lifestyle is
fundamental. The study of consumers’ concerns and risk perception related
solely to TFA are reviewed below. Furthermore, the different entities, which
generate information and condition the consumers’ mind, are also defined.
1. Entities and interactions conditioning consumers’ concerns and their

risk perception
Different entities have an influence on consumer behaviour and the feedback
from consumers can affect reciprocally the motivation of the same or other
entities. In Figure 1, each entity represents a key action producer, transmitter,
or receiver of information.
Everyone, of course, is a consumer, and the first and central entity considered
395
Figure 1. Interaction between entities generating information perceived and conveyed by consumer.
in this chapter is that of consumer. According to the International Life Sciences

Institute (ILSI, 2007) and its Europe Consumer Science Risk Force, ‘con-
sumer’ is defined as an individual being “reasonably well informed, reasonably
observant and circumspect, taking into account social, cultural and linguistic
factors”.
The second entity is authorities, which manage risks and inform consumers
locally or sometimes globally. Several government websites are available,
which provide an increasing number of interactive and original documents and
offer comprehensive and reliable information on the decisions relating to
critical policies. Of the many sources of food-related information, government
agencies and authorities generally receive the fewest enquiries. This may be
because people did not realize that the actual source of their information is the
government concerned. Alternatively, it could be because the channels used for
government communication, such as leaflets in clinics, are less effective than
the mass media.
The level of trust between various sources of information by consumers is
relatively consistent. In fact the level of trust in health professionals is over
80% in all European member states, and for government sources 70% or more
(Menigault et al., 1998). In a survey by Pieniak et al. (2007) advertising
materials were the least trusted; there was little difference in trust between
males and females, level of education and employment status, while older
people tended to trust less than other age groups.
The third entity is the scientific community publishing experimental data,
reviews and recommendations.
The fourth is press and media, conveying messages produced by government
authorities and scientific communities. As mediators, they translate scientific
results and make the information accessible for consumers. Today, the inter-
action between scientific community and media is closer than ever. Members
CONSUMER CONCERNS AND RISK PERCEPTION RELATED TO TFA 397
of the scientific community increasingly publicise their scientific achieve-

ments by using press releases in order to reach a wider audience.
In this system of four main entities, there is a consensus with several
interactions. In Figure 1, information represents real stimuli from the consum-
ers’ environment that contributes to their knowledge or beliefs. We are
particularly interested in the content of the messages communicated. On one
hand, information could be used to reduce the uncertainty that could cause
distress; while on the other hand, information could increase uncertainty thus
allowing hope or optimism, and inviting reappraisal of uncertainty. Informa-
tion can also increase stress-producing certainty or uncertainty while maintaining
the current state of consumer knowledge or beliefs (Brashers et al., 2002).
2. The Internet as an open window on ‘powerful consumer writing’ of

health concerns and risk perception
Discussions via the Internet add a new dimension to consumer life in the
democratic debate. The Internet could be an opportunity or a threat to democ-
racy. But we might ask if it could be one of the safest ways towards a
‘Balkanization’ of public opinion or the seedbed of new deliberative practices.
Public opinion on the Internet could be compared to a unique large state divided
into several autonomous and small different groups of various opinions.
Consumers are submerged in an overwhelming flow of information, and now
they know where to find responses to their concerns and become active or even
pro-active.
Media sources offer additional opportunities for mass and interpersonal
communication aimed at seekers and providers of health and nutrition informa-
tion. Television programmes, magazines, and newspaper articles on health and
nutrition topics are now very common (Johnson, 1997). The World Wide Web
supporting the Internet is characterized by accumulation, storage and dissemi-
nation of information through web sites, sponsored by individuals or health
information from private companies (Borzekowski & Rickert, 2001).
Listservers, newsgroups, and chatrooms provide information and help with
interpreting and appraising it.
In the mid-1990s, the Web (‘Web 1.0’) began as a repository of information
and static content. With the beginning of the 21st century, the Web became
much more interactive, allowing users to play, stop, rewind and fast-forward
through audio and video content. Web 2.0 makes Web-based applications run
as smoothly as local applications (the term became notable after the first
O’Reilly Media Web 2.0 conference in 2004). In this second wave of the World
Wide Web, blogs provide some interesting angles on questions, particularly
such as from the perspective of power in relation to language in the context of
an ‘information society. Blogs provide an arena where self-expression and
creativity are encouraged. Links to other bloggers establish the same peer-
group relationships as found in the real world (Huffaker, 2004). Some web sites
become influential as they are the most read, the most quoted, and the most
discussed. Blogs created by nutritionists have more weight in the ‘blogosphere’;
as an example, ‘Youtube’ allows dieticians and nutritionists to make their own
video programmes on health benefits or bad effects of nutrients. Later in this
chapter we will analyse how blogs tackle TFA.
The Internet plays an important role in risk communication and the forma-
tion of consumers’ views on an issue. It is an information source that helps the
public to form an opinion on risk (Yankelovich, 1991). Bennett (1999) sug-
gests that while media coverage may in fact amplify the public’s interest in an
issue, it does not create it. Bennett goes on to say that “a good story is one in
which public and media interests reinforce each other”.
The Internet tends to highlight existing concerns, uncertainties and con-
flicts. It rarely questions the legitimacy of any source and presents all sources
on a somewhat equal footing. In this sense, the Internet’s role might be
considered to be ‘non-judgemental’. However, some universities such as the
University of Kentucky, USA, and more particularly its College of Agricul-
ture, allow internet readers the opportunity to use a variety of learning modes
and information retrieval methods. The University’s Agripedia project and
website offer not only instructional material, but also increase the abilities of
users to consider information with more objectivity (see Agripedia refer-
ence).
Most online communities grow slowly at first, due in part to the fact that
the strength of motivation for contributing is usually proportional to the size
of the community. As the size of the potential audience increases, so does
the attraction of writing and contributing. This, coupled with the fact that
organizational culture does not change overnight, means that creators can
expect slow progress at first with a new virtual community. However, as
more people begin to participate, it creates a virtuous cycle in which more
participation brings more participants. Many online community members
describe their participation as “addictive” (Bishop, 2007). The growth in
community adoption is often forecast (estimating the number of users in the
community) by use of the Bass diffusion model. This is a mathematical
formula originally conceived by Frank Bass (1969) to describe the process
by which new products get adopted as an interaction between users and
potential users.
B. Evidence of consumer’s attitudes and concerns about food and

health worldwide
To understand whether TFA are a consumer concern or perceived as a risk, it
is essential to identify the sources of consumer concerns.
1. Individual attributes and collective values: the construction of health

concerns
a. Health concerns and the consumer

In nutrition and health, consumer understanding of a new claim or new
information is likely to be low, but is expected to increase over time with
repeated exposure to the information. Apart from some new health claims, it is
not realistic to expect that full understanding can be developed before the claim
is permitted in the market. This could be an additional complication for the
acceptance of new health claims. A potential solution is to increase efforts in
education, before the claims reach the market.
The overall consumer perception of food is generally positive; for instance
in Europe, food safety is not at the top of the consumers’ mind. Consumers
identify a wide range of concerns and tend to focus on factors which are beyond
their control.
Within the overwhelming mass of information, the educated or curious
consumers pay attention to nutrition and health. Consumers find answers to
their demands more easily than when searching in other areas. They are also
sovereign and choose what they like. They become ‘actors’.
Concerning trends in fat consumption, the US Department of Agriculture
(USDA) provides data based on the ‘disappearance’ of the fats and oils supply,
rather than actual consumption. Disappearance, as defined by the USDA’s
Economics Research Service, means [beginning food stocks + production +
imports] minus [exports + shipments to the US territories + ending stocks]. So
it is a reasonable proxy for consumption, given that data for consumption are
not collected overall. Generally, disappearance data over-estimate consump-
tion. However, by keeping track of disappearance trends over time, researchers
can determine relative changes in fats and oils consumed, and predict the
consumption of TFA and their consequences on health.
King et al. (2000) described the trends in consumption and consumer trends
for fats and oils. They show that roughly 85% of total fat and oil consumption
was comprised of vegetable oils, with animal fats making up the remaining
15%. Consumption per person of animal fats and oils had fallen 28% (from 14.1
to 10.2 pounds) in the previous 25 years. During that same period, vegetable oil
consumption increased 40% (from 38.5 to 53.9 pounds). Despite declines in
animal fats and oils consumption, total fats and oils consumption rose 22%
during the same period. Thus, most fats consumed were added at the discretion
of food processors, consumers, or preparers rather than naturally occurring in
foods. Salad oil and cooking oils were the leading source of added fats,
followed by shortening, margarine and dairy fats. These added fats are con-
sumed in addition to the fats occurring naturally in meats, fish, nuts, eggs and
dairy products.
Food consumers are now more aware of the health consequences of the fats
they choose to include in their diets. Research conducted by USDA suggests

that polyunsaturated and monounsaturated oils, such as safflower, canola
(rapeseed) and olive oil, can be healthy choices for consumers who wish to
reduce the risk of heart disease. Furthermore, consumer awareness of the link
between dietary fat and risk of heart disease increased 25% between 1988 and
1995. Consumer awareness of the relationship between dietary fat and cancer
risk remained unchanged during this same period. Annual consumption per
person of added fats and oils declined 7% between 1993 and 1997, reflecting
consumer concern about fat intake and health.
b. Consumer associations formulating health consumer concerns

Globally, food information organizations have been following consumer atti-
tudes and trends in the perception and use of information on the label, and
consumers’ attitudes to food, nutrition and health. These organizations alert
consumers about the adverse effects that TFA have on health. Thus, they pre-
empt consumer safety beyond authorities or alongside scientific communities.
Thanks to associations or foundations, identification of the top ten consumer
concerns in nutrition and health is accessible and provides a relevant overview
of food safety and risk perception in order to warn authorities. Heart disease
and obesity are the most important health concerns but it is self-image related
to obesity that makes it one of the top five health concerns. The story of TFA
enters into this topic in the basic consumers’ mind.
Among several actions, National, European and international research projects
with special interests or needs for updating or aspects of food quality are
enhanced. An example is the TRANSFAIR project on ‘Trans fatty acids intake
and risk factors for cardiovascular disease’ which ran from 1995 to 1998.
In 1993, as mentioned in Chapter 14, the Center for Science in the Public
Interest (CSPI, Washington, DC, USA) asked the US Food & Drug Adminis-
tration (FDA) to require disclosure of the amount of trans fat in food products.
Moreover, this CSPI asked FDA to stop misleading health claims after more
refined studies done in the early 1990s.
These studies showed that consumption of TFA increased the risk of
coronary heart disease. Finally, in February 1994, CSPI petitioned the FDA to
require label disclosure of the amount of TFA and to prevent misleading label
claims. Some years later, in 2003, the FDA issued a final trans fat labelling rule.
A few months later, the Institute of Medicine (Washington, DC, USA) con-
cluded that it was possible to exclude from the diet trans fat from partially
hydrogenated vegetable oil. A year later, the Nutrition Subcommittee of the
FDA’s Food Advisory Committee concluded that trans fat is “more adverse”
than saturated fat with respect to coronary heart disease.
In conclusion, thanks to a petition managed by an organization that has more
than 900,000 subscribers in the USA and Canada, and which provides the
largest-circulation health newsletter in North America, those countries are now
able to deal with the TFA problem. It took ten years to clarify the situation and
reduce the risk perception of consumers.
During the CSPI story, several consumer groups gave warnings about the
side-effects of consuming too much TFA and explained physiological mecha-
nisms and the regulatory situation using creditable experts on their websites.
Furthermore, heart health foundations grant research projects on this topic in
order to understand the impacts of TFA on human health. Since the 1990s,
several governing bodies, scientific associations and consumer interest groups
have expressed concerns about TFA (e.g. the International Association of
Consumer Food Organizations, IACFA; the European Food Information Coun-
cil, EUFIC; the Trans Atlantic Consumer Dialogue; the Australian Consumers
Association). Turning to science, policy formation should be underpinned by
reliable scientific evidence. However, this remains a potential problem as the
scientific community is still often viewed as being remote from the general
public. In the pursuit of scientific progress, the common interests between
science and society become vital for the benefit of public health.
Consumers International (London, UK) is an independent global campaign-
ing voice for consumers. With over 220 member organizations in 115 countries,
its goal is to build a powerful international consumer movement to help protect
and empower consumers everywhere. Through the lobbying of countries by
Consumers International and its member organizations, it has achieved a
resolution that has committed the World Health Organization (WHO) to:
“promote responsible marketing including the development of a set of recom-
mendations on marketing of foods and non-alcoholic beverages to children”.
Consumers International takes over the message from WHO for explaining that
the exposure to the commercial promotion of foods high in saturated fat, TFA,
sugars or salt has a direct effect on children wanting and eating these unhealthy
foods.
2. Consumer concerns related to trans fatty acids

The analysis of most frequent words written in blogs shows (Figure 2) three
main topics related to TFA. All topics related to health risk and TFA, the
information related to governmental positions (and more particularly Europe
and North America), and the consumers’ awareness focused on food industries
and restaurants.
The analysis of the main topics and messages conveyed by blogs provides
some tracks for defining a typology of blogs concerning TFA. On the basis of
the blogs’ articles the following conclusions may be drawn.
• Consumers are aware of certain food safety issues, especially the conse-
quences of TFA on heart health.
• Although various laws regulate types and amounts that are allowed in
Figure 2. Analysis and interpretation of the most frequently occurring words cited on Internet blogs
since 2007 related to trans fatty acids.
foods, consumers say TFA are one of the greatest food-associated hazards
at present.
• Awareness of consequences is dependent on education and age; the more
educated and older the consumers are, the more aware they are of food
safety issues.
• According to some bloggers, exposure to TFA may be diminished by
consuming organic food, purchasing food at a reliable source or consuming
food of known origin.
We can conclude that the understanding of food safety issues among
consumers is low, which calls for information campaigns to raise consumer
education.
In addition to these blogs, there are forums, chatrooms or discussion lists,
which offer the opportunity to exchange information and remain an important
Internet activity. Surely, consumers do not access these virtual spheres for the
sole pleasure of denouncing abuse (related to consumer rights, for instance).
Online communities have been identified by the Internet founders as “common
interest communities” (Licklider & Taylor, 1990). These communities, very
numerous and often enduring on the Internet, may exchange experiences or
expertise. In the case of TFA, heterogeneous typology of online communities
is observed. In this context, typology is the study or systematic classification of
different types of communities that have characteristics or traits in common and

that could be described.
The ‘epistemic communities’ are particularly interesting as they bring
together diversely skilled participants. They can be defined as transnational
networks of knowledge-based experts who define for decision-makers what
the problems they face are, and what they should do about them. Although
exchanges of information are the main purpose, ‘experts’ often mediate
between the beginners and the developers. These communities can only
survive if they are regulated (Auray, 2004, 2005). A permanent discussion
list is initiated in blogs classified by types. Several nutritionists, mainly in
the USA, are using blogs to try to educate consumers with regard to TFA and
health. They represent an essential channel for receiving and debating new
proposals: few personal blogs reflect the activist’s concerns. Corporate blogs
are also developed to obtain more responses from consumers, and elaborate
voting systems are provided. They provide a directly moderated demo-
cratic source based on technological aptitude and the knowledge of past
history.
Many other examples spring to mind in the field of health and reviews of
cultural products. In all those cases, communities of interest represent a realm
where the Internet can operate as a genuine hub for exchange and productive
public debate (Akrich & Méadel, 2002). It is also worth noting that these
communities are less homogeneous than has often been supposed.
C. The role of risk and benefit perceptions in the construction of

consumer attitude
1. Lessons from the past: food safety risk perception

Slovic developed a psychometric approach (Fischhoff et al., 1978) for evaluat-
ing the number of concerns, such as whether the risk is perceived as involuntary
or as controllable, whether it will affect a large number of people, or seems
natural or unnatural. These hypotheses are likely to be important determinants
of public responses, and partly explain the differences between lay and expert
beliefs about risks (Flynn et al., 1993).
Clearly risk communication is likely to be more effective if it addresses the
real concerns of the public regarding a particular hazard, not just those
concerns which are believed to be important by experts.
A second issue raised by the risk perception literature is ‘optimistic bias’ or
‘unrealistic optimism’ (Weinstein, 1989). However, if the sample of people is
representative of the appropriate population then the mean response over the
sample should be near the centre of the scale. Optimistic bias has been reported
for a wide variety of hazards including a number of food-related hazards
(Frewer et al., 1994; Sparks et al., 1995).
a. Food risks and worldwide reaction

Learning from recent European history, notably bovine spongiform encepha-
lopathy (BSE) outbreaks in the UK and Germany, politicians and industry now
have to strike a balance between giving ‘the facts’ and sounding alarmist or
complacent. Society’s relationship to matters of risk is influenced by five broad
factors: governance, science, society, culture, and media. The relationship
between governments and public institutions and broader society increasingly
affect the important issue of public trust.
Risk perception is generally applied to the attitudes to risks held by a specific
population. It should be distinguished from risk assessment or risk manage-
ment. Consumers are thought to be overly sensitive to risks (Fischhoff, 1989)
and to be heterogeneous with regards to small and major risks (Kasperson
et al., 1988). If we examine different past events within the food domain, the
question is: “What do people think about different types of risks and what are
the important types of characteristics of different hazards?”
Risk communication and risk perception cannot be dissociated. Even if TFA
do not represent exactly the same parameters as the introduction of genetically-
modified (GM) organisms, it presents an interesting case for studying risk
perception.
In relation to the GM debate, several articles have explored the cultural
differences which appear to exist between European and American citizens.
For instance, American citizens put more trust in their government structures
and are more confident that those institutions represent and protect their
interests. On the other hand, they were perhaps less able than their European
counterparts to influence government decisions.
The issue for public authorities is how to transmit clear and accurate risk
messages against the backdrop of some sections of the media that are appar-
ently intent on maximising hysteria, and building up a story and maintaining it.
Consumers’ attitudes about risk are an important endpoint of risk manage-
ment. Considerable resources are directed by risk management agencies to
analysing public opinion towards risk. Some observations have shown that
these attitudes are far from static; rather, they seem to take on a life of their own
and evolve over time. Yankelovich suggests that this evolution is quite an
orderly process and that consumer attitudes start with macroscopic opinions
and move toward consumer judgement in a complicated process that involves
sorting through and coming to terms with conflicting emotions, values and
interests around a given issue. It may reasonably be expected to fall somewhere
between the ultimate goal of ‘wisdom’ and the more common notion of a ‘well-
informed citizenry’ (Yankelovich, 1991).
Individuals may hold seemingly irrational views about certain risks. This
phenomenon is extremely difficult to overcome and can cause risk messages to
fail. Risk communication initiatives must be designed to ensure that the
messages target individual groups within the population. To do this one must
first find ways of segregating individual differences and needs and include the
public’s real concerns in the information provided (Frewer et al., 1998).
b. Typology of food safety risk and risk communication

For a better understanding of how the consumer perceives risk, some important
environmental factors were identified, taking advantage of past events. Differ-
ent approaches to risk could lead to a kind of ‘Deming wheel’, an iterative
process typically used in quality assurance and management. It is also known
as the PDCA cycle (Plan, Do, Check, Act).
In the case of risk assessment, we could differentiate four steps as follows.
• Design or revision, for instance, of the food process components in order to
improve results: the level of consumer risk perception could be estimated.
• implementation of the plan and measurement of its performance: that could
be the risk management.
• assessment of system performance and reporting of results to decision
makers: we could then talk about risk assessment.
• decision on changes needed to improve the process: it is still risk manage-
ment and the plan of communication again modifies the risk perception in
order to reduce the perception of the risk.
In this dynamic process the circle of different factors able to generate a risk
(Figure 3) represents a part of a larger dynamic economy and society. The way
in which we produce and consume food reflects our values, concerns and
activities.
2. Individual knowledge and consumer risks perception: overview and case

of trans fatty acids
Consumer perceptions and behaviour related to food risk or foodborne illness
changed little between 1988 and 1993 (Martin, R, 1990, Fein et al., 1995, Food
Institute, 1996).
If consumers misperceive the nature and origin of foodborne illness, they
underestimate the frequency of serious consequences and are less motivated to
change. Schafer found that motivation for proper food handling requires
viewing the mishandling of food as a direct threat to one’s health (Schafer,
1993). The failure to associate mishandling of food in the home with foodborne
illness interferes with efforts to educate the public on foodborne disease (Fein,
1995).
The idea of ‘availability heuristic’ is probably one of the most important
concepts emerging from the study of risk perception. It describes the observa-
tion that the more frequently a certain message about risk is repeated, the better
the availability of the message, and the greater the number of people who will
be likely to over-estimate its risk. Some studies show that trusted sources such
406
Figure 3. Environment of sources generating a food risk.

as public authorities, medical doctors, scientists, and consumer groups are

perceived to be both knowledgeable and concerned about public wellness
(Frewer et al., 1996) (EFSA, 2006).
For TFA, some actions have been made to prevent the risk associated with its
consumption. Discussions on the risk of consuming TFA have led to a well-
publicized ban on the use of trans fats in restaurants in New York City, USA.
Other US cities such as Boston and Philadelphia are introducing similar
measures. The trend in banning the use of TFA follows action in Denmark,
which effectively abolished the use of partially hydrogenated vegetable oils
four years ago. A growing number of firms, such as the dairy ingredients group
Land O’Lakes, has recently made ingredients available to food manufacturers
that are eager to remove trans fats from their formulations.
The MacDonald’s group has also started to use a blend of canola, corn and
soybean oils to cook French fries and other deep-fried products. Meanwhile
Nestlé (see Nestlé web site reference) published a policy stating that TFA will
not make up more than 3% of a normal consumption of its products, or 1% of
the total daily energy intake as recommended by the World Health Organization.
Many difficulties are identified when a common model is established for
illustrating responses of consumer to a risk factor over time (Figure 4). The
curiosity of consumers encourages them to read and to seek more information.
This information alerts and exacerbates perception of risk. As a result, decreas-
ing consumption is observed.
In a nutshell, Figure 4 shows that it is essential to understand what consum-
ers think about potential adverse health effects when consuming an excessive
amount of TFA. In particular, it is important to identify where attempts are
made to communicate cautiously. The consumers’ reaction could be consid-
ered as irrational and not helping the communication process. However,
depending where consumers live, their perception of TFA could be totally
misinterpreted. We can assume that the consumers will remain passive, leaving
the authorities, scientific community, companies and specialists to manage the
situation. Finally, the more consumers pay attention to food safety, the more
they consume with better choices. This then creates competition among food
manufacturers to provide new products catering to the new trend.
3. Internet: agora or cacophony to clarify concerns about trans fatty acids

and to perceive risk
In the early 1990s, the Internet was often hailed as a new virtual agora (Flichy,
2007), an open free market place online where people buy and trade free from
any government coercion. According to Rheingold (1994), the Internet is a
device with the potential to revitalise democracy. The ‘TFA’ topic requires a
relatively high level of education to be correctly understood by consumers. The
more information is accessible, the more consumers will have the opportunity
Consumption
1. 2. 3. 4.
No information Information Decreasing media Persisting anxiety
communicated communicated attention or Food safety
concerns
Basic curve of consumer reaction in front of the risk

Indicative curve of consumer reaction in front of the
risk linked to trans fatty acids
Time
Figure 4. Comparison of consumer reaction to a food scare and the perception of adverse effects of
trans fatty acids consumption (adapted from Beardworth and Keil, 1996 and Mazzochi, 2004).
to understand the context and its consequences. Online debates do not meet the
standards of debates in the public sphere. In other words, a debate is sought
between equals where rational arguments prevail over a common position.
Internet debates would only match the first feature (Poster, 1997). However,
discussions about TFA by Web users do indeed exchange on an equal footing
but this exchange may not be reasoned with creditable evidence.
For a wider topic than TFA, the debate usually does not move towards the
elaboration of a common position but rather splinters into many contradictory
viewpoints. This splintering is further reinforced by the fact that users’ identi-
ties are vague and shifting. Not only do contributors use aliases and create for
themselves a virtual identity but they may also change their identity or have
several different identities. The coexistence of identities, which was studied by
Turkle (1997), appears to be one of the major causes of difficulty in reaching a
common viewpoint by online communities. In real life, when dealing with each
other face to face, each speaker can perceive the complex arguments of the
other and still reach an agreement. By contrast, virtual communities foster a
multiplicity of fixed viewpoints rather than flexibility (Kolko & Reid, 1998).
Fortunately, TFA offers a simpler situation because the topic is precise and
it concerns specific communities even if the problem affects everybody.
Opinions on the Internet converge on a similar sense and there emerges a
common wish and action tailored to the problem itself.
4. Influence and role of each entity concerning trans fatty acids

In order to measure the impact and the rate of information diffusion, the patent
Figure 5. A timeline of trans fats from the introduction of hydrogenation to the launch of zero-trans
fats.
by Wilhelm Normann (Figure 5) is the critical point. To summarize, in 1903

Normann was trying to find a way of making a substitute for tallow that was
very expensive at the time. He discovered that when he heated cottonseed oil to
260°C in the presence of a catalyst such as nickel, it became hard when cooled.
He had produced cheap candle wax by ‘hydrogenating vegetable oil’. Proctor
& Gamble (P&G) realized the potential and bought the patent from Normann.
By 1911, P&G was producing Crisco in the USA, a hard vegetable fat that was
great for baking and had a long shelf-life. In 1988, the first hypotheses were
formulated concerning the effect of eating TFA on risk of coronary diseases. In
1995, the first concrete results were published opening the way to new
investigations and governmental actions. Today, several scientific projects are
conducted on this topic and some marketing and economics research studies
are carried out for consumers.
Recently, the presentation of a product’s TFA content on the label has been
largely misunderstood by consumers through the lack of interpretive footnotes
or information on recommended daily value.
A study carried out by Howlett et al. at the University of Arkansas, USA
(2008), found that even consumers at risk for heart disease are confused by
current nutritional information on TFA. Even for a well-informed consumer,
general knowledge about TFA is rather poor, not to mention the critical
information that 4 g of TFA is a large amount. Howlett concluded that:
“Nutritionally motivated consumers lacking appropriate prior knowledge make
inappropriate product judgements”. This implies that consumers may not be
able to interpret whether a product contains high levels of TFA or not.
According to Howlett, the results indicate overall that without appropriate
consumer education programmes, the addition of TFA levels to the Nutrition
Facts panel may have limited or even unintended consequences.
This topic has changed rapidly. Recently, researchers have developed Euro-
pean and international scientific projects focusing on consumers’ understanding,
economical impact, market opportunities and new ways of communication.
From the market point of view, business analysts transform TFA into an
opportunity to open new markets. They explain that the effects of TFA on
human health represent the biggest issue facing the edible oils and fats industry
today. Edible oils and fats will also be used on a wide scale in diet therapy,
where they can halt the progress of diseases or reverse the damage caused by
illnesses.
Scientific research is growing faster partly because the scientific community
has benefited form easier information access in recent years.
Media coverage is sometimes difficult to predict and explain. Two effects
are well known: the term ‘snowball effect’ has been used to describe a story that
has become established and where increased frequency of coverage results in
media competition for interest; the ‘ripple effect’ is a form of risk amplification
and it often occurs when stories with an original focus on a single specific risk
are expanded to include related issues. The example of TFA illustrates how risk
perception could be ‘amplified’, or extended, by media coverage to create new
concerns about other fats and oils. Often these effects, which originate from
routine media coverage, can lead to what is termed by some as a ‘media event’.
It has been suggested that media coverage is driven largely by rarity, novelty
and commercial viability but, unfortunately, very little by risk evaluation.
D. Conclusions and perspectives

This chapter attempted to gather published findings that have an influence on
the value of using TFA in food products. Both in terms of willingness by
consumers to accept and to pay a premium cost for TFA-free food, we
examined if the Internet is a trustworthy source for measuring consumers’
perception to TFA and whether there is continuity of information from scien-
tific communities and the press to consumers. Thus, we could underpin the
evidence that the Internet makes information more easily accessed and ex-
changed by each entity in the consumer environment described in the first part
of this chapter, and affects the duration of the impact related to risk perception.
The multi-dimensional nature of risk in a pluralistic society increases the
challenges to communicators and risk communicators specifically.
For TFA, a simple observation is that it is linked to global informational
context that influences and conditions attitudes and concerns of all categories
of consumers. Food safety is a non-negotiable issue. The supply chain is
increasingly global and so solutions, therefore, should focus on collaborative
solutions and oriented approaches beyond national borders. We also need to
promote the most trusted sources of information and find a balanced and
intelligent way to communicate risk because such information influences
critically the consumers’ perception influencing their attitude to TFA.
According to the statistics of world Internet usage (Internet World Stats web
site), capturing data on consumer perceptions through this medium becomes
increasingly possible. Indeed, 22% of the world population uses the Internet
and growth in usage since 2000 is 306%. The Internet is used by 74% of the
population in North America, 60% in Oceania/Australia and 48% in Europe.
The majority of Internet users interested in nutrition are active on the Web.
Groups, such as scientists, authorities, foundations and associations, as well as
individuals, share their expertise, opinions and web video and thus amplify the
mass of information on health effects of TFA. The TFA issue, in spite of the
relative level of knowledge expected from consumers, was taken into account
without significant drawbacks. In conclusion, the Internet might increasingly
constitute one way of evaluating risk as perceived by consumers. The impact on
them is potentially measurable.
Analysis of consumer perception regarding a specific food issue through the
Internet could offer one approach. For example, the issue of intense sweeteners
might be examined. In fact, more than ever, people are consuming large
amounts of sugar in their daily diet, which adds extra calories, which in turn
could cause weight gain. Artificial sweeteners are the subject of several stories
or comments, presented in the press and on the Internet, claiming that they
cause a variety of health problems, including cancer. Numerous studies con-
firm that artificial sweeteners are safe for the general population. However
‘Aspartame’ does carry a cautionary note, particularly for people who have the
rare hereditary disease phenylketonuria.
New scientific investigations are, or could be, undertaken with natural
intense sweeteners extracted from plants. Moreover science is progressing and
food companies are working to integrate these into products. Thus, it might be
interesting to measure consumer perceptions on the consumption of artificial
sweeteners or introduction of natural intense sweeteners into food products.
Finally, we can now see how risk perception data regarding these issues might
be captured from the Internet as already done for TFA.
References
Agripedia (1996) College of Agriculture, University of Kentucky, USA. Online:
www.ca.uky.edu/agripedia
Akrich, M and Méadel, C (2002) Prendre ses médicaments, prendre la parole: les usages des
médicaments par les patients dans les listes de discussion électroniques [To take a pill,
to take a stand: arguing about medication on electronic patient discussion lists], Sciences
sociales et santé, 1, 89–116
Auray, N (2004) La régulation de la connaissance: arbitrage sur la taille et gestion aux
frontières dans la communauté Debian [The regulation of knowledge: size and scope
control for the Debian community], Revue d’économie politique, March 2004, pp.160–
182.
Auray, N (2005) Le sens du juste dans un noyau d’experts: debian et le puritanisme civique
[Debian and civic puritanism]. In: Internet, une utopie limitée. Nouvelles régulations,
nouvelles solidarités [Internet a Limited Utopia. New Regulations, New Solidarities]
(Proulx, S, Massit-Folléa, F and Conein, B, eds) Les presses de l’Université Laval,
Québec, Canada, pp.71–94.
Bass, F (1969). A new product growth model for consumer durables. Management Science,
15 (5), 215–227.
Bennett, P (1999) Understanding responses to risk: some basic findings. In: Risk Communi-
cation and Public Health (Bennett, P and Calman, K, eds), Oxford University Press, New
York, USA, pp.3–19.
Bishop, J (2007) Increasing participation in online communities: a framework for human–
computer interaction. Computers in Human Behavior, 23, 1881–1893.
Brashers, DE, Goldsmith, D J, and Hsieh, E (2002) Information seeking and avoiding in
health contexts. Human Communication Research, 28, 258–271.
Consumers International, London, UK; Online: www.consumersinternational.org
EFSA (2006) ‘The Metabolic Syndrome Summit’, European Food Safety Authority, London,
UK.
Fein, SB, Jordan, C and Levy, AS (1995) Foodborne illness: perceptions, experience, and
preventive behaviors in the United States. Journal of Food Protection, 58, 1405–1411.
Fischhoff, B, Watson, SR, and Hope, C (1984) Defining risk. Policy Sciences, 17, 123–139.
Flichy, P (2007) Internet Imaginaire, Chapter 7: On Cyberdemocracy. MIT Press, Cam-
bridge, Massachusetts, USA.
Flynn, L, Slovic, P and Mertz, CK (1993) Decidedly different: expert and public views of
risks from a radioactive waste repository. Risk Analysis, 13, 643 648.
Food Institute (1996) Population trends 1996–2050. Food Institute, Fair Lawn, New Jersey,
USA, p.31.
Frewer, LJ, Howard, C, Hedderley, D and Shepherd, R (1998) Methodological approaches
to assessing risk perceptions associated with food-related hazards. Risk Analysis, 18, 95–
102.
Frewer, LJ, Raats, MM and Shepherd, R (1994) The interrelationship between perceived
knowledge, control and risk associated with a range of food related hazards targeted at
the self, other people and society. Journal of Food Safety, 14, 19–40.
Howlett E, Burton S and Kozup J (2008) How modification of the nutrition facts panel
influences consumers at risk for heart disease. Journal of Public Policy & Marketing, 27
(1), 83–97.
Huffaker, D (2004) Spinning yarns around a digital fire: storytelling and dialogue among
youth on the Internet. First Monday, 9 (1) (January 2004). Online: www.firstmonday.org/
issues/issue9_1/huffaker
ILSI (2007) ‘Consumer understanding of health claims’, ILSI Report Series. International
Life Sciences Institute, Brussels, Belgium.
Institute of Food Technologists’ Expert Panel on Food Safety and Nutrition (1995) Scientific
status summary, foodborne illness: role of home food-handling practices. Food Technol-
ogy,49, 119–131.
Internet World Stats. Online: www.internetworldstats.com/stats.htm
Kasperson, RE, Renn, O, Slovic, P, Brown, HS, Emel, J, Goble, R, Kasperson, JK and Ratick,
S (1998) The social amplication of risk: a conceptual framework. Risk Analysis, 8, 177–
187.
King, BS, Tietyen, JL and Vickner, SS (2000) Food and Agriculture: Consumer Trends and
Opportunities Fats, Oils, and Sweets, Cooperative Extension Service, University of
Kentucky, College of Agriculture, USA. Online: www.ca.uky.edu/agc/pubs/ip/ip58g/
ip58g.htm
Kolko, B and Reid, E (1998) Dissolution and fragmentation: problems in on-line communi-
ties. In Cybersociety 2.0 (Steven Jones, ed.), Sage, Thousand Oaks, California, USA,
p.212–229.
Latouche K, Rainelli, P and Vermersch D (1998) Food safety issues and the BSE scare: some
lessons from the French case. Food Policy, 23, 347–356.
Licklider, J, and Taylor, R (1990) The computer as a communication device. In: Science and
Technology, April 1968, Reprinted in In Memoriam: JCR Licklider 1915–1990, Digital

Systems Research Center, Palo Alto, California, p.38.
Menigault, E, Berson, M, Vieyres, P, Lepoivre, B, Pourcelot, D, Pourcelot, L, Lappalainen,
R, Kearney, JL, Gibney, M (1998) A pan EU survey of consumer attitudes to food,
nutrition and health: an overview. Food Quality and Preference, 9, 467–478.
Martin, R (1990) Foodborne disease threatens industry. National Restaurant News, pp.24 –
27, 30.
Nestlé, FAQs, Current Issues. Online: www.nestle.com/AllAbout/FAQs/CurrentIssues/
FAQs.htm
Pieniak, Z, Verbeke, W, Scholderer, J, Brunsø, K and Ottar Olsen S (2007) European
consumers’ use of and trust in information sources about fish. Food Quality and
Preference, 18, 1050–1063.
Poster, M (1997) Cyberdemocracy : the internet and the public sphere. In: Virtual Politics,
Identity and Community in Cyberspace (David Holmes, ed.), Sage, London, UK, pp.212–
228.
Rheingold, H (1994) The Virtual Community. Homesteading on the Electronic Frontier,
Harper Perennial, New York, USA.
Schafer, RB, Schafer, E, Bultena, GL and Hoiberg, EO (1993) Food safety: an application of
the health belief model. Journal of Nutritional Education, 25, 17–24.
Slovic, P (1987) Perception of risk, Science, 236, 280–285.
Sobey, BA, Simpson, ACD and Ives, DP (1994) Managing food related risks: integrating
public and scientific judgements. Food Control, 5, 9–19.
Thompson, M (1992) Risk communication: dealing with the spectrum of environmental and
health risks in Europe. Environmental Risk Assessment Unit, Norwich, University of
East Anglia, Research Report No.11. Risk and Rationality: Approaches to the Manage-
ment of Ignorance About Environmental Health Risks. Proceedings of the WHO
Consultation in Ulm, Germany, 28–30 November 1992.
Turkle, S (1997) Life on the Screen, Touchstone, New York, USA.
Tversky, A, and Kahneman, D (1973) Availability: a heuristic for judging frequency and
probability. Cognitive Pychology, 4, 207–232.
USDA, Economic Research Service (1998) America’s eating habits: changes and conse-
quences (Frazao, E, ed.) Agricultural Information Bulletin No.750, USDA, Washington,
DC, USA
Weinstein, ND (1998) Optimistic biases about personal risks. In: Risk and Modern Society
(Lofstedt, R and Frewer, L, eds) Earthscan Publications, London, UK, pp.239–241.
Yankelovich. D (1991) Coming to Public Judgement: Making Democracy Work in a Complex
World. Syracuse University Press, New York, NY, USA.
Index
Abomasum 20, 201 isomeric profile 4

absorption 164–165, 324 isomerase 13–19, 178, 205–206, 232, 235
activation energy 70 reductase 14, 17, 198
α-linolenic acid 186–188, 196, 198–201, rumen 1–24, 28–30, 129, 196–207, 221, 232–
205–206, 260–279, 298, 309–310, 315, 319 234
algae (see marine algae) body composition 237–239, 247
America, Central and South, trans fatty acids butter 130, 150, 153, 196, 212–220, 310, 335,
consumption 368 374
analysis
argentation chromatography 73, 106, 113, Cancer 210–216, 220–222, 241–247, 288, 335,
121–122, 124–126, 128–129, 134–136, 340, 351, 370, 400, 411
138–140 canola (rapeseed) oil 1, 3, 20–21, 25–27, 50, 56,
dimethyloxazoline (DMOX) 127–128 58–59, 65, 67, 69, 108–110, 113, 118–119,
fatty acid methyl esters (FAME) 106–140, 122, 148, 152–153, 156–157, 168, 186, 400,
176 407
Fourier transform near-infrared (FT-NIR) cardiovascular diseases (see also ischaemic heart
spectroscopy 44, 106 diseases and coronary heart diseases) 244–
gas chromatography (GC) 106–141, 175–176, 245, 255–258, 260, 265, 278, 284, 294, 296,
321 307–308, 313–314, 330, 334–335, 400
gas chromatography-mass spectrometry (GC- catalytic hydrogenation
MS) 106, 123, 126, 128 copper-based catalysts 49–50, 58
methylation 106–107, 129, 131, 140 heterogeneous catalysts 57, 58
silver ion-high performance liquid chromatog- homogeneous catalysts 58
raphy (Ag+- HPLC) 106, 113, 121–122, nickel catalysts 46–59
124–126, 129, 134–135, 138–140 palladium catalysts 57–59
silver ion-thin layer chromatography (Ag- partial hydrogenation 49, 59, 108, 115, 148,
TLC) 73, 106, 113, 121–124, 129, 132, 155, 329, 381, 384, 391
135, 136–138, 140–141 platinum catalysts 57–59
ultraviolet spectroscopy 125–126 ruthenium catalysts 58
arachidonic acid 167–168, 173–174–184, 186, Central and South America, trans fatty acids
188, 324 consumption 368
Asian/Pacific Region, trans fatty acids chemical interesterification 149–154
consumption 368 chemical synthesis
atherosclerosis 216–222, 244, 310–314 Horner-Emmons reaction 80
argentation chromatography 73, 106, 113, olefination 77–78 ,80–82, 85–87, 90, 100
121–122, 124–126, 128–129, 134–136, 138– phase transfer catalysis 81, 83, 86–87
140 Wadsworth-Emmons reaction 80
Australia, trans fatty acids consumption 359 Wittig reaction 77–100
Wittig-Horner reaction 80–81, 83, 85
Bacteria 2, 6 ,9–12 ,14 ,17–18, 22, 25, 29, 198, ylides 78–81, 83
204–206, 232–235 cholesterol
beef 232 high-density lipoprotein (HDL) choles-
benefit perceptions 403 terol 44, 186–187, 217–220, 256–257, 265,
β-oxidation 171–172, 177–179, 186, 188, 239, 283, 299, 309–310, 313, 386
245, 298 low-density lipoprotein (LDL) choles-
biohydrogenation terol 165, 186, 217–220, 257, 265, 280,
intermediates 2, 4 , 11, 20 , 22–25, 198, 206 283, 308–310, 313, 330, 386
415
chromatography (see analysis) Emersol process 95–96

coagulation 309–310 encapsulation 20
Codex Alimentarius 383 endothelial cell 183, 310, 313
concentrates 1, 16–17, 27, 206 energy metabolism 237
conjugated linoleic acid (CLA) enzymatic interesterification 149–152
chemical synthesis 235 epidemiological studies 195, 215, 220, 255–302,
production by microorganisms 233 307
rumenic acid 30, 129, 185, 195–222, 235, 381 Europe, Continental, trans fatty acids consump-
rumenic acid enriched milk fat 209, 219 tion 348
trans-10,cis-12 18:2 acid isomer 197, 200, extraction of lipids 106–107, 123–124, 129–131,
203–204, 206–207, 210–212, 218, 220–221, 140
231–246, 325, 381 extrusion 15, 21–22, 25
consumer attitude 403
consumer concerns 395, 398, 400–401 Fatty acid methyl esters (FAME) 106–140, 176
consumption fibrinolysis 309–311
Asian/Pacific Region 368 fish oil 6–8, 10, 15, 19–20, 24–25, 29, 51, 70–
Australia 359 74, 201, 206–207, 210, 310
Central and South America 368 forages 1, 206–207
Continental Europe 348 Fourier transform near-infrared (FT-NIR)
Iran 372 spectroscopy 44, 106
Nordic Countries 359 fractionation 50, 59, 94, 96–100, 121, 123–124,
North America 335 134–135, 147, 236–237
UK and Ireland 348 free fatty acids 65, 70, 107, 128, 131, 151, 164,
Continental Europe, trans fatty acids consump- 186, 214, 235–236
tion 348 fungi 11, 150, 234
copper-based catalysts 49–50, 58
copper supplementation 25,27 Galactolipids 1–2, 8–9
coronary heart diseases (see also ischaemic heart gas chromatography (GC) 106–141, 175–176,
diseases and cardiovascular diseases) 244– 321
245, 255–258, 260, 265, 278, 284, 294, 296, gas chromatography-mass spectrometry (GC-
307–308, 313–314, 330, 334–335, 400 MS) 106, 123, 126, 128
cord 322–324 genetically-modified oils 156–159, 234
crystallization geometrical isomerization 3–4, 43–44, 54, 59,
Emersol process 95–96 65–67, 69–70, 72, 90–94
Lanza process 98–99 glucose tolerance 245, 315
cyclic fatty acid monomers (CFAM) 70, 72 glycolipids (see galactolipids)
growth and development 185, 323
Dairy cows 3, 16, 21, 24, 29, 201–202, 205, 209
dairy products (see milk and butter)
deodorization 43, 65–74, 110, 164 Haemostatic function 309–310
desaturation 4–5, 15, 23, 30, 167, 169, 172–175, heterogeneous catalysts 57, 58
177, 179–180, 188, 199, 204–205, 232, 319 high-density lipoprotein (HDL) cholesterol 44,
diabetes 314–316 186–187, 217–220, 256–257, 265, 283, 299,
diet (see consumption) 309–310, 313, 386
digestion 1, 11, 13–14, 130, 164, 175 high-oleic oils 156–158
dimethyloxazoline (DMOX) 127–128 homogeneous catalysts 58
docosahexaenoic acid (DHA) 70–74, 164, 169– Horner-Emmons reaction 80
171, 206, 324 human milk 185–186, 320–321, 324–325, 333–
duodenal flow 2–3, 5, 7, 9, 11, 20, 23, 25, 27, 334, 336, 345, 372
28, 30–31 hydrogenated oil 52, 58–59, 159, 181
duodenum 5, 6, 8–9, 11, 21, 23, 24, 164 hydrogenation (see catalytic hydrogenation)
hydroxy acids 5, 13, 129
Eicosanoids 180, 182, 184, 216, 247, 319
eicosapentaenoic acid (EPA) 70–74, 164, 169 Immune function 231, 243–244, 247
elaidic acid 3, 46, 58, 109, 148, 171, 173, 178– incorporation 164–166, 169–171, 174–175, 182,
180, 185, 255, 301, 305, 322–323, 381 199, 323
elongation 167–168, 173, 175, 177–179, 319, infant formulas 321, 325
325 infant nutrition 320, 325
INDEX 417
inflammatory markers 310, 312 Non-genetically modified oils 156–157

insulin sensitivity 231, 245–246, 314–316 Nordic Countries, trans fatty acids consump-
interesterification tion 359
chemical interesterification 149–154 North America, trans fatty acids consump-
enzymatic interesterification 149–152 tion 335
Iran, trans fatty acids consumption 372 nickel catalysts 46–59
Ireland and UK, trans fatty acids consump-
tion 348 Obesity 210, 237–239, 246–247, 314, 400
ischaemic heart diseases (see also coronary heart olefination 77–78 ,80–82, 85–87, 90, 100
diseases and cardiovascular diseases) 244–
245, 255–258, 260, 265, 278, 284, 294, 296, Palladium catalysts 57–59
307–308, 313–314, 330, 334–335, 400 palm oil 1, 20, 50–51, 54–55, 65, 98–100, 148–
isomerase 13–19, 178, 205–206, 232, 235 149, 151–155, 157
isomerization partial hydrogenation 49, 59, 108, 115, 148,
geometrical 3–4, 43–44, 54, 59, 65–67, 69– 155, 329, 381, 384, 391
70, 72, 90–94 phase transfer catalysis 81, 83, 86–87
positional 3, 43, 47, 50, 56, 90, 94, 108 phospholipids 1–2, 8–9, 14, 107, 128, 131, 165–
167, 169–171, 173–175, 181–183, 185–186,
Labelling 44, 105, 108–109, 158, 172, 222, 330, 323, 334
345, 348, 359, 368, 383–392, 400 physical properties 43–44, 50, 52, 54, 59, 153–
Lanza process 98–99 154
lipolysis 1–2, 7–9, 11, 16–18, 21–22, 313 pigs 173, 244
lipoproteins (see high-density and low-density) placental transfer 322
linseed oil 5–6, 19, 27–29, 92, 156, 169, 174– platelet aggregation 182–183, 309–310
175, 200 platinum catalysts 57–59
long-chain polyunsaturated fatty acids positional isomerization 3, 43, 47, 50, 56, 90,
arachidonic acid 167–168, 173–174–184, 94, 108
186, 188, 324 protozoa 11
docosahexaenoic acid (DHA) 70–74, 164,
169–171, 206, 324
docosapentaenoic acid (DPA) 73, 377 Rapeseed (canola) oil 1, 3, 20–21, 25–27, 50,
eicosapentaenoic acid (EPA) 70–74, 164, 169 56, 58–59, 65, 67, 69, 108–110, 113, 118–119,
low-density lipoprotein (LDL) cholesterol 165, 122, 148, 152–153, 156–157, 168, 186, 400,
186, 217–220, 257, 265, 280, 283, 308–310, 407
313, 330, 386 reductase 14, 17, 198
low-linolenic acid soybean oil 157 regulatory models 391
replacement of hydrogenated oils in foods 147–
162
Mammary gland 30, 185, 199, 205–206, 210,
risk communication 398, 403–405
212–213
risk perception 395, 397, 400–401, 403–405,
margarines 108, 150–155, 157, 159, 255–256,
410–411
298, 301, 368, 389, 392
rumen 1–24, 28–30, 129, 196–207, 221, 232–
marine algae 7, 19, 25, 206
234
mass spectrometry (see analysis)
ruminant fats 106, 128–129, 163, 165, 319, 333,
metabolism of trans fatty acids
348, 372, 374, 381
absorption 164–165, 324
ruthenium catalysts 58
β-oxidation 171–172, 177–179, 186, 188,
239, 245, 298
desaturation 4–5, 15, 23, 30, 167, 169, 172– Safflower oil 9, 166
175, 177, 179–180, 188, 199, 204–205, 232, shortenings 43, 108, 150–153, 255–256, 298,
319 301, 319
elongation 167–168, 173, 175, 177–179, 319, silages 1, 16
325 silver ion-high performance liquid chromatogra-
incorporation 164–166, 169–171, 174–175, phy (Ag+- HPLC) 106, 113, 121–122, 124–
182, 199, 323 126, 129, 134–135, 138–140
methylation (see analysis) silver ion-thin layer chromatography (Ag-
milk (see human milk and butter and rumenic TLC) 73, 106, 113, 121–124, 129, 132, 135,
acid) 136–138, 140–141
soybean oil 4, 18, 20–21, 27, 29–30, 50, 54–56, linseed oil 5–6, 19, 27–29, 92, 156, 169, 174–
98, 109–110, 152, 157–158, 171, 173, 178, 175, 200
181, 236, 238, 310 non-genetically modified 156–157
stearidonic acid 180 palm oil 1, 20, 50–51, 54–55, 65, 98–100,
stearoyl-CoA desaturase 202 148–149, 151–155, 157
sterculic acid 201 rapeseed (canola) oil 1, 3, 20–21, 25–27, 50,
sunflower oil 5–6, 19, 24, 28–30, 46, 156–157 56, 58–59, 65, 67, 69, 108–110, 113, 118–
119, 122, 148, 152–153, 156–157, 168, 186,
Transgenic (see genetically-modified) 400, 407
transmethylation 107 safflower oil 9, 166
type 2 diabetes 314–316 soybean oil 4, 18, 20–21, 27, 29–30, 50, 54–
56, 98, 109–110, 152, 157–158, 171, 173,
UK and Ireland, trans fatty acids consump- 178, 181, 236, 238, 310
tion 348 sunflower oil 5–6, 19, 24, 28–30, 46, 156–157
venturi loop reactor 53
Vaccenic acid 2, 6, 30, 86, 88–89, 109, 167, Wadsworth-Emmons reaction 80
185, 196–222, 232, 234, 255, 301, 315, 329, Wittig reaction 77–100
334, 381 Wittig-Horner reaction 80–81, 83, 85
vegetable oil
genetically-modified 156–159, 234 ylides 78–81, 83

(The Oily Press Lipid Library) Frederic Destaillats, Jean-Louis Sebedio, Fabiola Dionisi, Jean-Michel Chardigny - Trans Fatty Acids in Human Nutrition-Woodhead Publishing (2009)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(The Oily Press Lipid Library) Frederic Destaillats, Jean-Louis Sebedio, Fabiola Dionisi, Jean-Michel Chardigny - Trans Fatty Acids in Human Nutrition-Woodhead Publishing (2009)

Uploaded by

Copyright:

Available Formats

Trans fatty acids in

Trans fatty acids in

Oxford Cambridge Philadelphia New Delhi

Woodhead Publishing, 1518 Walnut Street, Suite 1100, Philadelphia,

First published by The Oily Press, 2009

© PJ Barnes & Associates, 2009; © Woodhead Publishing Limited, 2012

Trademark notice: Product or corporate names may be trademarks or registered trade-

British Library Cataloguing in Publication Data

ISBN 978-0-9552512-3-8 (print)

This book is Volume 23 in The Oily Press Lipid Library

Typeset by Ann Buchan (Typesetters), Middlesex, UK

Frédéric Destaillats (Lausanne, Switzerland)

1 Biosynthesis of trans fatty acids in ruminants 1

2 Formation of trans fatty acids during catalytic hydrogenation 43

3 Formation of trans fatty acids during deodorization of edible oils 65

4 Chemical synthesis of monounsaturated trans fatty acids 77

5 Analysis of trans fatty acids of partially hydrogenated 105

4. Analysis of cis and trans isomers of partially hydrogenated

6 Replacement of partially hydrogenated oils in food products: 147

7 Metabolism of trans fatty acid isomers 163

8 Biosynthesis and biological activity of rumenic acid: 195

9 Biosynthesis, synthesis and biological activity of 231

10 Observational epidemiological studies on intake of trans 255

2. Quality of the studies

11 Dietary Trans Fatty Acids and Cardiovascular Disease Risk 307

13 Evolution of worldwide consumption of trans fatty acids 329

14 Legislation relating to trans fatty acids 381

15 Consumer concerns and risk perception related to 395

Clotilde Aubertin, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

Dale E. Bauman, Department of Animal Science, Cornell University, Ithaca,

Jean-Baptiste Bezelgues, Nestlé Product Technology Center, 809 Collins

Laure Candy, Laboratoire de Chimie AgroIndustrielle; ENSIACET, 4 Allées

Jean-Michel Chardigny , INRA, UMR1019, Clermont-Ferrand, 63000 France;

William W. Christie, Scottish Crop Research Institute and Mylnefield Lipid

Nicole Combe, ITERG, Département de Nutrition, 33000 Bordeaux, France

Margaret C. Craig-Schmidt, Department of Nutrition and Food Science,

Cristina Cruz-Hernandez, Nutrient Bioavailability, Nestlé Research Centre,

Frédéric Destaillats, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

Albert J. Dijkstra, Carbougnères, 47210 St Eutrope-de-Born, France

Koenraad Duhem, Centre National Interprofessionnel de l’Économie Laitière

Francis Enjalbert, Université de Toulouse, INPT, ENVT, UMR 1289 Tan-

J. Bruce German, Department of Food Science and Technology, University of

Francesca Giuffrida, Nestlé Research Center, PO Box 44, CH-1000 Lausanne

Marianne Uhre Jakobsen, Department of Clinical Epidemiology, Aarhus

Jana Kraft, Department of Animal Science, University of Vermont, 219

Adam L. Lock, Department of Animal Science, University of Vermont, 219

Corinne Malpuech-Brugère, Clermont Université, UFR Médecine, UMR

Ronald P. Mensink, Maastricht University, Nutrition and Toxicology Re-

Béatrice Morio, Clermont Université, UFR Médecine, UMR 1019 Nutrition

Zephirin Mouloungui, Laboratoire de Chimie AgroIndustrielle; ENSIACET,

Guillermo Napolitano, Nestlé Product Technology Center, 809 Collins Av-

Kim Overvad, Department of Clinical Epidemiology, Aarhus University

W.M. Nimal Ratnayake, Nutrition Research Division, Food Directorate,

Beth H. Rice, Department of Animal Science, University of Vermont, 219

Yinghui Rong, Department of Nutrition and Food Science, Auburn Univer-

Jean-Louis Sébédio, UMR 1019, Unité de Nutrition Humaine, Plateforme

Delphine Tissot-Favre, Nestlé Purina PetCare, Checkerboard Square - 1RS,

Annabelle Troegeler-Meynadier, Université de Toulouse, INPT, ENVT,

University of California, Davis, California, USA, and Nestlé Research Center,

Few innovations in the brief history of industrialized food production have

FRANCIS ENJALBERT1,2 AND ANNABELLE TROEGELER-

B. Physiology of rumen biohydrogenation

1. Lipolysis of dietary lipids: a prerequisite for biohydrogenation