ARTICLE IN PRESS

Prostaglandins, Leukotrienes and Essential Fatty Acids 81 (2009) 133–136

Contents lists available at ScienceDirect

Prostaglandins, Leukotrienes and

Essential Fatty Acids
journal homepage: www.elsevier.com/locate/plefa

Docosahexaenoic acid: Measurements in food and dietary exposure$

Jay Whelan , Lisa Jahns 1, Katherine Kavanagh
Department of Nutrition, The University of Tennessee, 229 Jessie Harris Building, 1215 West Cumberland Avenue, Knoxville, TN 37996-1920, USA

abstract

The Dietary Reference Intakes (DRIs) were established to be an indicator of adequacy of dietary
nutrients as well as providing levels for adequacy in reducing risk of chronic diseases such as
neurodegenerative diseases, cardiovascular diseases, cancers, diabetes mellitus, etc. One particular
nutrient that is increasingly discussed as a potential candidate for the generation of a DRI is the omega-
3 (n-3) fatty acid docosahexaenoic acid (DHA) due to its potential benefits in reducing risk for
cardiovascular disease, role in resolution of inflammation, its importance in cognitive function in infants
and inhibiting the progression of neurodegenerative diseases in the elderly. Each reference value refers
to and is predicated on estimates of daily nutrient intake and the goal of this paper is to review these
intakes. The confidence of these values is critical in establishing dose–response relationships. This paper
reports intake values for DHA and examines how these data were generated and the relative confidence
in these values. The adult US population is estimated to consume 80–100 mg/d of DHA based on a
nationally representative sample of 48400 individuals as part of the National Health and Nutrition
Examination Survey (NHANES). This value and those presented for women and men at various ages
appear reasonable and should be used as the basis for establishing an Adequate Intake (AI) for DHA.
& 2009 Elsevier Ltd. All rights reserved.

1. Introduction example, the Estimated Average Requirement (EAR) and Adequate

Over the last half century, the realization that omega-3 (n-3)
polyunsaturated fatty acids (PUFA) from marine sources possess
unique biological properties that are distinct from PUFA of the
omega-6 (n-6) family compels us to revisit these compounds from
the context of their ability to promote health and prevent disease.
The dietary reference intakes (DRI) were developed to serve as a
guide for good nutrition focusing on the general, healthy, North
American population [1]. The DRI committee acknowledged that
the physiological essentiality of a nutrient, even if it was not
clearly identified as an essential dietary nutrient (in the classical
sense) could be considered as essential to one’s diet and could be
eligible for evaluation in the future within this context [1]. For
example, a DRI was set for dietary fiber based on its cardiopro-
tective benefits. At the time (September 2007), the nutrient
of discussion by the DRI committee was docosahexaenoic acid
(DHA, 22:6 n-3).
A key objective of the DRI is to establish accurate intake values
that eventually can be compared with outcome measures. For
$
Sources of support: (JW) Supported by Tennessee Agricultural Experiment
Station, Institute of Agriculture, The University of Tennessee, Knoxville, TN, USA.
 Corresponding author. Tel.: +1865 974 6260; fax: +1865 974 3491.
E-mail address: jwhelan@utk.edu (J. Whelan).
1
Current address: Grand Forks Human Nutrition Research Center, USDA-ARS,
2420 2nd Avenue North, Grand Forks, ND 58202-9034, USA.
Intake (AI) are defined as
‘‘Estimated Average Requirement (EAR): The average nutrient
intake level that is estimated to meet the requirements of half
the healthy individuals in a particular life stage and gender
group.
Adequate Intake (AI): The recommended average daily intake
level based on observed or experimentally determined approx-
imations or estimates of nutrient intake by a group (or groups)
of apparently healthy people that are assumed to be adequate;
used when an RDA cannot be determined’’.
An EAR value needs to be established in order to determine a
Recommended Dietary Allowance (RDA). When a RDA cannot be
established, an AI can be set based on less data, incorporating
more scientific judgment than that used to establish an EAR.
DHA is unique with regards to other n-3 PUFA in that it appears
to be the biologically essential n-3 PUFA. All tissue membranes
contain DHA and all precursors are preferentially converted to
DHA up to a certain point. Within the membrane lipid micro-
domains (i.e., lipid rafts), DHA may uniquely enhance extracellular
signals from cell surface receptors to downstream intracellular
signaling molecules [2]. Modifying tissue DHA levels would
theoretically impact these signaling pathways, favorably modify-
ing risk of certain chronic diseases (i.e., cancer, inflammation).
While metabolic precursors, such as a-linolenic acid and
eicosapentaenoic acid, can be theoretically converted to DHA

0952-3278/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.plefa.2009.05.008
 ARTICLE IN PRESS
134 J. Whelan et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 81 (2009) 133–136
and subsequently incorporated into tissue phospholipids, within
the context of a typical US diet the only way to enrich tissue levels
of DHA is with dietary DHA [3]. It is within this context that
only DHA can modify tissue DHA levels in a general, healthy
population, optimizing health and minimizing disease risk.
This paper reviews the literature outlining current DHA
intakes—the underlying basis for establishing DRI reference levels
of intake for DHA. It reviews the major dietary sources, the typical
levels of DHA found in the US diet, and the challenges we face in
collecting and maintaining accurate data within an ever-changing
commercial environment [4]. It does not address the dose–benefit
relationship with various chronic diseases, a topic covered by
other papers in this series.
2. DHA and diet
2.1. Marine sources
The richest sources of LC n-3 (omega-3) PUFA in the US diet are
from fish and fish products. Oily fish, such as tuna, salmon,
herring, etc., are the richest sources with estimated levels
between 862 and 1840 mg/100 g [5]. Of 37 commonly consumed
types of fish products, DHA is the primary LC n-3 PUFA
(on average 65% of total LC n-3 PUFA) [5]. Typically the level of
DHA in fish varies with the overall content of LC n-3 PUFA. Orange
roughy, tilapia, mahi-mahi, cod, flatfish (flounder) and catfish
have DHA levels in the range of 113–257 mg/100 g. Concentrations
of DHA in halibut, pollock and sea bass vary between 374 and
556 mg/100 g. Sardines, tuna, mackerel, herring and salmon
typically have the highest levels of DHA, 509–1429 mg/100 g. Of
note are the equivalent levels of LC n-3 PUFA, and in particular
DHA, between feral and farmed salmon [5,6].
2.2. Terrestrial sources
In addition to fish, terrestrial (land-based) meats (i.e., beef,
pork, chicken, etc.) can potentially contribute significant amounts
of LC n-3 PUFA, including DHA, although in relatively low
quantities [7]. DHA is not typically found in commonly consumed
plant products. All animal meats contain LC n-3 PUFA as part of
their membrane phospholipids. These levels are typically
o40 mg/100 g serving. DHA appears to be an essential component
of tissue phospholipids as it is found in all animal tissues tested;
however, the levels of DHA in tissue phospholipids of land-based
meats are quite low compared with total content of total LC n-3
PUFA (o10%). For example, a 100 g sample of rib eye contains
approximately 2.4 mg of DHA [7]. Enrichment of tissue DHA
appears to be dependent upon dietary sources of DHA and not on
its precursors [3]. This probably accounts for the enriched tissue
levels of DHA in fish due to the fact that much of the fish
Table 1
Non-traditional foods enriched/fortified with LC n-3 PUFAs (EPA+DHA)a.
Food Serving EPA+DHA (mg)
Breads and Pasta 100 8–80
Milk 250 ml 10–190
Eggs 1 (50 g) 86–150
Processed Meats 100 g 88–190
Salad Dressings 14–31 g 60–700
Margarine and Spreads 10–100 g 60–150
Pizza 1 slice 32
Nutrition bars 50 g 3–115
Juices 170 ml 100
a
Not all foods are available in the US (see Ref. [4]).
consumed in the US diet are carnivorous, feeding off marine
sources rich in DHA (i.e., smaller fish).
2.3. Alternative dietary sources
In addition to foods that typically contain DHA, there is a
growing industry designed to enrich or fortify foods with LC n-3
PUFA, in particular DHA. A representative list of these foods and
their reported levels of LC n-3 PUFA and DHA are listed in Table 1
[4]. Fortification and enrichment of foods not traditionally
containing LC n-3 PUFA include pasta, bread, milk and dairy
products, fruit juice, snack foods, processed meat, salad dressing,
margarine and eggs, and this does not include nutrient
supplements.
The primary ways these products are being enriched/fortified
with LC n-3 PUFA [4]:
 Bio-delivery: feeding an animal the LC n-3 PUFA (or the
precursor) and enriching their tissues with LC n-3 PUFA
(i.e., EPA, DPA in meats, and DHA in eggs).
 Post-harvest modification of the foods: adding the n-3 PUFA
rich/enriched oils directly to foods.
 Post-harvest modification of oils: Micro-encapsulation of the oil
to maintain stability and mask flavors, prior to incorporation.
Furthermore, a variety of plant sources of LC n-3 PUFA have
been and are being developed. For example, DHA is generated
from algae, and genetically modified plant oils are being enriched
with LC n-3 PUFA, (i.e., rapeseed and soybean) [8,9]; however, as
opposed to the algal sources, these latter sources do not contain
DHA. These products, while gaining popularity in the US, never-
theless are much more popular outside of the US [4], and thus, it is
impossible to estimate what contributions these products have on
total DHA content because no reliable information has been
published, as yet. As such, there is no evidence that the overall
contribution of DHA from these sources in the US diet is of overall
importance at this time.
3. DHA intakes
In the US, the estimated intake values for LC n-3 PUFA are, for
the most part, generated from the nationally representative What
We Eat in America/National Health and Nutrition Examination
Survey, which estimates intake using databases based fundamen-
tally upon NHANES III data, using the USDA National Nutrient
Database for Standard Reference [10]. It has been estimated that
women and men in the US consume 110 and 170 mg/d of LC n-3
PUFA, respectively (Table 2) [11]. In the latest NHANES’ ‘‘What We
Eat in America Report’’ [12], the average intake of DHA for
the general population is 70 mg/d for the general population
(X2 years old), and 80 and 100 mg/d for females and males,
respectively, for those X20 years old) (Table 3) [13].
Understandably, children consume a fraction of what adults
consume due to differences in the amount and type of food
DHA (mg)
Table 2
36
Current estimates for the amounts of EPA+DHA (mg/d) in the diets of various
60
Westernized countries for men and women.
150
135
Country EPA+DPA+DHA (women/men) DHA (women/men) Ref.
60–700
60–150
USA 110/170 (+DPA) [11]
32
Canada 135 (pregnant women) 82 [23]
115
Australia 195/298 83/117 [15]
100
Germany 215/315 140/200 [17]
France 400/497 226/273 [18]
 ARTICLE IN PRESS
J. Whelan et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 81 (2009) 133–136 135

Table 3 comparison has yet to be published to help explain ecological

DHA intake (mg/d) by gender and age in the US. diversity in DHA intakes.
Total population (males+females) Females Males

X2 years 70 4. Assessment of DHA intakes in the US

2–5 years 30 20
6–11 40 30 Recently, a number of reports by the Agency for Healthcare
12–19 30 50
20–29 50 90
Research and Quality (US Department of Health and Human
30–39 80 100 Services) have been published evaluating the effects of n-3 PUFA
40–49 90 100 on various chronic diseases [19,20]. These reviews involved
50–59 70 80 randomized control trials, case-control, cohort and nested case-
60–69 90 120
control studies. These studies utilized 24-h recalls and food
X70 80 80
X20 80 100 frequency questionnaires to assess food intake and on nutrient
databases based upon the USDA National Nutrient Database for
See Ref. [13]. Standard Reference.
There is a concern we may be underreporting the amount of
Table 4 DHA intake in the US because a number of terrestrial meats in SR-
DHA intake (mg/d) by race and age in the US. 21 do not contain values. For example, SR-21 reports that lean cuts
of rib eye steak (NDB no: 13952) and pork loin (NDB no: 10043) do
Years All Non- Non- Mexican
individuals Hispanic Hispanic American not contain DHA when in fact, upon direct analysis, DHA is clearly
White Black present at relatively low levels (2.4 mg/100 g sample) [7]. The
reasons for this missing information have been reviewed else-
2–5 30 20 40 30 where [7] and involves types of analytical techniques used
6–11 40 30 50 30
12–19 40 40 50 40
previously, the use of mathematical formulas versus direct
X20 90 80 100 90 analysis, and the fact that fatty acid composition data are
X2 70 70 90 70 commonly reported in gram quantities because lipids are
macronutrients. It is not uncommon for a fatty acid to be reported
See Ref. [14].
as zero if the sample levels are less than 100 mg/100 g sample
even if these fatty acids (i.e., LC n-3 PUFA) can be quantitated [7].
eaten. Individuals in the 20–70-year-old age group consume the The impact of SR-21 on accurately assessing DHA intakes in the
most DHA. Non-Hispanic adult Blacks (X20 years) consume more US population is unknown at this time despite the fact that a
DHA than non-Hispanic Whites and Mexican Americans, with an number of meat products do not have DHA values; however,
average intake of 100 mg/d versus 80 and 90 mg/d, respectively current estimates may be reasonably close for the following
(Table 4) [14]. These levels for all genders and ethic/racial groups reasons. The annual/capita terrestrial meat consumption in the US
have been revised upwards by 10–33% since the 2007 report is 6 times that of fish, but the level of DHA in fish is 4100 fold
covering the years 2003–2004 [12]. [21]. Therefore, the major source of DHA in the US diet is most
Regarding other countries, the levels of LC n-3 PUFA in the likely derived from marine sources. The missing values for DHA
Australian population are almost double those in the US with an for land-based meats are found within a select group of food
estimated average intake of 195 mg/d for women and 298 mg/day products (some beef, pork and veal). The absence of data is not
for men (Table 2) [15,16]. Some of these differences can be universal. More importantly, the missing values are relatively low.
explained by differences in nutrient composition databases and For example, as previously mentioned, DHA levels in rib eye
the inclusion of the n-3 fatty acid docosapentaenoic acid (DPA) in accounted for only 2.4 mg/100 g sample or 8% of the total LC n-3
the Australian database [15]. While DPA is a metabolic precursor PUFA in the sample. Therefore, if there is underreporting of LC n-3
of DHA, there is no evidence that consumption of DPA contributes PUFA due to missing values in SR-21, it is more likely to have an
to tissue DHA levels in humans. Forty-three percent of the LC n-3 impact on DPA and EPA than it does on DHA. In addition, the
PUFA in the Australian diet is derived from terrestrial meat intake, estimates for intakes of DHA have been increased by 10–33% in
of which DPA is a predominant component. DHA intake is 2008 (years 2005–2006) compared with the previously published
estimated to be 90 mg/d on average in the Australian population, report in 2007 (years 2003–2004) [12].
with women and men consuming 83 and 117 mg/d, respectively.
In comparison, the intakes of DHA for Germans (women and men)
are reported to be 140 and 200 mg/d, respectively [17], and for 5. Summary
women and men in France, 226 and 273 mg/d, respectively [18].
Given the reported levels of LC n-3 PUFA by the French (400 and In summary, in the absence of sufficient scientific evidence to
497 mg/d for women and men, respectively), this could, in part, establish an EAR and thus a RDA in an apparently healthy
account for the ‘‘French Paradox’’, which refers to the French population, an AI is set based on less data, incorporating more
having lower rates of cardiovascular morbidity despite diets scientific judgment than that used to establish an EAR [22]. There
relatively rich in saturated fats. appears to be a relationship between dietary DHA, tissue DHA
There is no known explanation for the large differences in the levels, cellular function and potential health outcomes. The levels
reported intakes of DHA among Western countries. N-3 PUFA of DHA in the US diet are influenced by the amount of food we eat
intake is calculated by linking food intake to a food composition (i.e., children versus adults, females versus males), and for the
database and many studies rely on the USDA National Nutrient most part, the amount of fish in the diet. The only n-3 PUFA that
Database for Standard Reference (the most recent release being can enrich tissues with DHA within the context of a healthy diet is
SR-21), or upon the countries own analysis of indigenous foods, DHA. There appears to be incomplete data for LC n-3 PUFA in
complicating comparisons across countries. Certainly, differences SR-21, suggesting we maybe underreporting the amount of these
in data collection, databases and cultural dietary differences and bioactive nutrients by as much as 20% based upon the revised
varying food sources are potential factors, but a systematic figures with land-based meats in the Australian database [15].
 ARTICLE IN PRESS
136 J. Whelan et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 81 (2009) 133–136
However, of all the LC n-3 PUFA, DHA would be the least affected
as a result of underreporting in SR-21. The most recent revisions
on DHA intakes more accurately reflect nutrient intake. In
conclusion, it is estimated that the average intake of DHA in the
US population is 70; 80 and 100 mg/d for women and men,
respectively, and at this time these figures appear to be very
reasonable and should be considered as the initial basis for
establishing a value for Adequate Intake (AI).
Conflict of interest: This manuscript was presented as part of
a Workshop on DHA as a Required Nutrient by JW, sponsored by
Martek Biosciences Corporation, June 20–21, 2008, in Baltimore,
MD. JW received travel expenses and an honorarium for his
participation. LJ and KK have no conflicts of interest.
Acknowledgments
The authors’ responsibilities were as follows: JW conceived the
paper and was primary author. LJ and KK provided research and
input on intake data and contributed to the editing and writing of
the manuscript. All authors reviewed the final version of the
manuscript.
References
[1] Institute of Medicine (IOM), The Development of DRIs 1994–2004: Lessons
Learned and New Challenges—Workshop Summary, The National Academies
Press, Washington, DC, 2008.
[2] R.S. Chapkin, D.N. McMurray, L.A. Davidson, B.S. Patil, Y.-Y. Fan, J.R. Lupton,
Bioactive long chain fatty acids: emerging mechanisms of action, Br. J. Nutr.
100 (2008) 1152–1157.
[3] L.M. Arterburn, E.B. Hall, H. Oken, Distribution, interconversion, and dose
response of n-3 fatty acids in humans, Am. J. Clin. Nutr. 83 (suppl) (2006)
1467S–1476S.
[4] J. Whelan, C. Rust, Innovative dietary sources of n-3 fatty acids, Annu. Rev.
Nutr. 26 (2006) 75–103.
[5] W.S. Harris, P.M. Kris-Etherton, K.A. Harris, Intakes of long chain omega-3
fatty acid associated with reduced risk for death from coronary heart disease
in healthy adults, Curr. Atheroscler Rep. 10 (2008) 503–509.
[6] C. Blanchet, M. Lucas, P. Julien, R. Morin, S. Gingras, E. Dewailly, Fatty acid
composition of wild and farmed atlantic salmon (Salmo salar) and rainbow
trout (Oncorhynchus mykiss), Lipids 40 (2005) 531–539.
[7] L. Taber, C.H. Chiu, J. Whelan, Assessment of the arachidonic acid content
in foods commonly consumed in the American diet, Lipids 33 (1998)
1151–1157.
[8] V.M. Ursin, Modification of plant lipids for human health: develop-
ment of functional land-based omega-3 fatty acids, J. Nutr. 133 (2003)
4271–4274.
[9] A. Meyer, H. Kirsch, F. Domergue, et al., Novel fatty acid elongases and their
use for the reconstitution of docosahexaenoic acid biosynthesis, J. Lipid Res.
45 (2004) 1899–1909.
[10] Nutrient Data Laboratory ARS, USDA National Nutrient Database for Standard
Reference, Unites States Department of Agriculture. Internet: /http://
www.nal.usda.gov/fnic/foodcomp/search/S (accessed 11 February 2009).
[11] T.L. Psota, S.K. Gebauer, P. Kris-Etherton, Dietary omega-3 fatty acid intake
and cardiovascular risk, Am. J. Cardiol. 98 (2006) 3i–18i.
[12] US Department of Agriculture, Agricultural Research Service. 2008. What We
Eat in America, available at /http://www.ars.usda.gov/ba/bhnrc/fsrgS.
[13] US Department of Agriculture, Agricultural Research Service, Nutrient Intakes
from Food: Mean Amounts Consumed per Individual, One Day, 2005–2006.
What We Eat in America, NHANES, 2005–2006, available at /www.ars.usda.
gov/ba/bhnrc/fsrgS, 2008.
[14] US Department of Agriculture, Agricultural Research Service, Nutrient Intakes
from Food: Mean Amounts Consumed per Individual, by Race/Ethnicity and
Age, One Day, 2005–2006. What We Eat in America, NHANES, 2005–2006,
individuals 2 years and over (excluding breast-fed children), Day 1 dietary
sampling weights, available at /www.ars.usda.gov/ba/bhnrc/fsrgS, 2008.
[15] P. Howe, B. Meyer, S. Record, K. Baghurst, Dietary intake of long-chain omega-
3 polyunsaturated fatty acids: contribution of meat sources, Nutrition 22
(2006) 47–53.
[16] P.R. Howe, B.J. Meyer, S. Record, K. Baghurst, Contribution of red meat to very
long chain omega-3 fatty acid (VLCOmega3) intake, Asia Pac. J. Clin. Nutr. 12
(2003) S27.
[17] J. Linseisen, M.B. Schulze, M. Saadatian-Elahi, A. Kroke, A.B. Miller, H. Boeing,
Quantity and quality of dietary fat, carbohydrate, and fiber intake in the
German EPIC cohorts, Ann. Nutr. Metab. 47 (2003) 37–46.
[18] P. Astorg, N. Arnault, S. Czernichow, N. Noisette, P. Galan, S. Hercberg, Dietary
intakes and food sources of n-6 and n-3 PUFA in French adult men and
women, Lipids 39 (2004) 527–535.
[19] Agency for Healthcare Research and Quality, Health effects of omega-3 fatty
acids and cardiovascular disease, 2008, available at /http://www.ahrq.gov/
clinic/tp/o3cardtp.htmS.
[20] Agency for Healthcare Research and Quality, Effects of Omega-3 Fatty Acids
on Cancer, 2008, available at /http://www.ahrqgov/clinic/tp/o3cantp.htmS.
[21] R. York, M.H. Gossard, Cross-national meat and fish consumption: exploring
the effects of modernization and ecological context, Ecol. Econ. 48 (2004)
293–302.
[22] Institute of Medicine IOM, Dietary Reference Intakes. The Essential Guide to
Nutrient Requirements, The National Academies Press, Washington, DC, 2006.
[23] J. Denomme, K.D. Stark, B.J. Holub, Directly quantitated dietary (n-3) fatty
acid intakes of pregnant Canadian women are lower than current dietary
recommendations, J. Nutr. 135 (2005) 206–211.