EFSA Journal - 2008 - Polycyclic Aromatic Hydrocarbons in Food Scientific Opinion of The Panel On Contaminants in The

The EFSA Journal (2008) 724, 1-114
Polycyclic Aromatic Hydrocarbons in Food1
Scientific Opinion of the Panel on Contaminants in the Food Chain
(Question N° EFSA-Q-2007-136)
Adopted on 9 June 2008
SCIENTIFIC PANEL MEMBERS

Jan Alexander, Diane Benford, Andrew Cockburn, Jean-Pierre Cravedi, Eugenia Dogliotti,
Alessandro Di Domenico, María Luisa Fernández-Cruz, Johanna Fink-Gremmels, Peter Fürst,
Corrado Galli, Philippe Grandjean, Jadwiga Gzyl, Gerhard Heinemeyer, Niklas Johansson,
Antonio Mutti, Josef Schlatter, Rolaf van Leeuwen, Carlos Van Peteghem, Philippe Verger.
SUMMARY
Polycyclic aromatic hydrocarbons (PAHs) constitute a large class of organic compounds that are
composed of two or more fused aromatic rings. They are primarily formed by incomplete
combustion or pyrolysis of organic matter and during various industrial processes. PAHs
generally occur in complex mixtures which may consist of hundreds of compounds. Humans are
exposed to PAHs by various pathways. While for non-smokers the major route of exposure is
consumption of food, for smokers the contribution from smoking may be significant. Food can be
contaminated from environmental sources, industrial food processing and from certain home
cooking practices.
In the past decade PAHs were evaluated by the International Programme on Chemical Safety
(IPCS), the Scientific Committee on Food (SCF) and by the Joint FAO/WHO Expert Committee
on Food Additives (JECFA). SCF concluded that 15 PAHs, namely benz[a]anthracene,
benzo[b]fluoranthene, benzo[j]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene,
benzo[a]pyrene, chrysene, cyclopenta[cd]pyrene, dibenz[a,h]anthracene, dibenzo[a,e]pyrene,
1 For citation purposes: Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the
European Commission on Polycyclic Aromatic Hydrocarbons in Food. The EFSA Journal (2008) 724, 1-114.
© European Food Safety Authority, 2008
Polycyclic Aromatic Hydrocarbons in Food
dibenzo[a,h]pyrene, dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3-cd]pyrene and 5-

methylchrysene show clear evidence of mutagenicity/genotoxicity in somatic cells in
experimental animals in vivo and with the exception of benzo[ghi]perylene have also shown clear
carcinogenic effects in various types of bioassays in experimental animals. Thus, SCF reasoned
that these compounds may be regarded as potentially genotoxic and carcinogenic to humans and
therefore represent a priority group in the assessment of the risk of long-term adverse health
effects following dietary intake of PAHs. SCF suggested to use benzo[a]pyrene as a marker of
occurrence and effect of the carcinogenic PAHs in food, based on examinations of PAH profiles
in food and on evaluation of a carcinogenicity study of two coal tar mixtures in mice.
Using the assessments of IPCS and SCF as starting points and taking into account newer studies,
the JECFA re-evaluated PAHs in 2005. Overall, the JECFA concluded that 13 PAHs are clearly
genotoxic and carcinogenic. Except benzo[ghi]perylene and cyclopenta[cd]pyrene the
compounds were the same as those stated by SCF. The JECFA also concluded that
benzo[a]pyrene could be used as a marker of exposure to, and effect of, the 13 genotoxic and
carcinogenic PAHs. In addition, the JECFA recommended to include benzo[c]fluorene as a
further compound into future analyses as data on its occurrence in food are still scarce but rat
studies indicate that benzo[c]fluorene may contribute to the formation of lung tumours after oral
exposure to coal tar.
Following a recommendation on the further investigation into levels of PAHs in certain foods
(2005/108/EC)2, eighteen Member States submitted almost 10,000 results for PAH levels in
different food commodities. An evaluation of these data performed by EFSA in June 2007 and
updated in June 2008 demonstrated that benzo[a]pyrene could be detected in about 50% of the
samples. However, in about 30% of all the samples other carcinogenic and genotoxic PAHs were
detected despite testing negative for benzo[a]pyrene. Of the individual PAHs, chrysene was most
commonly found in the samples negative for benzo[a]pyrene with the highest level of 242 µg/kg.
In view of these findings, the Commission requested a full review of the 2002 SCF opinion on
PAHs.
The EFSA Panel on Contaminants in the Food Chain (CONTAM Panel) reviewed the available
data on occurrence and toxicity of PAHs. As no new toxicological data could be identified that
would lead to the inclusion of further compounds into the list of priority PAHs, the CONTAM
Panel decided to cover the 15 PAHs identified by SCF in 2002 and additionally benzo[c]fluorene
as suggested by the JECFA in 2005 in the present opinion. Special attention was paid to those
eight carcinogenic and genotoxic PAHs that were measured in the coal tar mixtures used in the
carcinogenicity studies, which provided the basis of the SCF and JECFA risk assessments.
2 OJ L 34, 8.2.2005, p.43

The CONTAM Panel explored whether a toxic equivalency factor (TEF) approach in the risk
characterisation of the PAH mixtures in food could be applied and concluded that the TEF
approach is not scientifically valid because of the lack of data from oral carcinogenicity studies
on individual PAHs, their different modes of action and the evidence of poor predictivity of the
carcinogenic potency of PAH mixtures based on the currently proposed TEF values. Therefore
the CONTAM Panel concluded that the risk characterisation should be based upon the PAHs for
which oral carcinogenicity data were available, i.e. for benzo[a]pyrene and the other PAHs that
were measured in the two coal tar mixtures used in the carcinogenicity studies of Culp et al.
(1998): benz[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene,
chrysene, dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene. The CONTAM Panel concluded
that these eight PAHs (PAH8), either individually or in a combination, are currently the only
possible indicators of the carcinogenic potency of PAHs in food.
In total, results from 9714 PAH analyses in 33 food categories/subcategories were evaluated. As
in about 30% of the samples analysed for all 15 priority PAHs as recommended by SCF other
carcinogenic and genotoxic PAHs were detected despite testing negative for benzo[a]pyrene,
individual compounds were grouped and summed in order to check whether their sums would
better reflect the occurrence of carcinogenic and genotoxic PAHs in different food categories.
The selection of the individual PAHs was based on the frequency of their results above the limit
of detection (LOD).
Besides the sum of the above mentioned eight PAHs (PAH8), the sum of benzo[a]pyrene,
chrysene, benz[a]anthracene and benzo[b]fluoranthene (PAH4) as well as the sum of
benzo[a]pyrene and chrysene (PAH2) were calculated. The correlation between PAH2 and PAH4
or PAH8 was 0.92 and between PAH4 and PAH8 was 0.99. Of samples negative for PAH2, 26%
and 18% identified concentrations above the LOD for at least one other PAH for samples tested
for all PAH15 or all PAH8, respectively. The frequency varied between 2% and 9% for the
individual PAHs or PAH combinations. Of samples negative for PAH4, 14% and 6% identified
concentrations above the LOD for at least one other PAH for samples tested for all 15 PAHs or
all PAH8, respectively. The frequency varied between 1% and 6% for the individual PAHs or
PAH combinations. Overall, the Panel concluded that PAH4 and PAH8 were better indicators of
the occurrence of PAHs than PAH2.
For different food categories and subcategories, the data on PAH8, PAH4 and PAH2 were then
used for the exposure calculation as well as the estimation of margins of exposure (MOEs) based
on the bench mark dose lower confidence limit for a 10% increase in the number of tumour
bearing animals compared to control animals (BMDL10).
The median dietary exposure across European countries was calculated both for mean and high
dietary consumers and varied between 235 ng/day (3.9 ng/kg b.w. per day) and 389 ng/day (6.5
ng/kg b.w. per day) respectively for benzo[a]pyrene alone, 641 ng/day (10.7 ng/kg b.w. per day)
and 1077 ng/day (18.0 ng/kg b.w. per day) respectively for PAH2, 1168 ng/day (19.5 ng/kg b.w.
per day) and 2068 ng/day (34.5 ng/kg b.w. per day) respectively for PAH4 and 1729 ng/day (28.8
ng/kg b.w. per day) and 3078 ng/day (51.3 ng/kg b.w. per day) respectively for PAH8. The two
highest contributors to the dietary exposure were cereals and cereal products, and sea food and
sea food products.
The CONTAM Panel used a MOE approach based on dietary exposure for average and high level
consumers to benzo[a]pyrene, PAH2, PAH4 and PAH8, respectively and their corresponding
BMDL10 values derived from the two coal tar mixtures that were used in the carcinogenicity
studies of Culp et al. (1998). The resulting MOEs for average consumers were 17,900 for
benzo[a]pyrene, 15,900 for PAH2, 17,500 for PAH4 and 17,000 for PAH8. For high level
consumers, the respective MOEs were 10,800, 9,500, 9,900 and 9,600. These MOEs indicate a
low concern for consumer health at the average estimated dietary exposures. This applies to the
full range of estimates of average exposures across EU Member States (3.1-4.3 ng/kg b.w. per
day, MOEs: 16,300-22,600 for benzo[a]pyrene alone and 23.6-35.6 ng/kg b.w. per day, MOEs:
13,800-20,800 for PAH8). However, for high level consumers the MOEs are close to or less than
10,000, which as proposed by the EFSA Scientific Committee indicates a potential concern for
consumer health and a possible need for risk management action. Comparison of the MOEs
calculated for benzo[a]pyrene, PAH2, PAH4 and PAH8, indicates that PAH2, PAH4 and PAH8
can be used as alternatives to benzo[a]pyrene alone as markers of the carcinogenicity of the
genotoxic and carcinogenic PAHs, and would be equally effective.
The CONTAM Panel concluded that benzo[a]pyrene is not a suitable indicator for the occurrence
of PAHs in food. Based on the currently available data relating to occurrence and toxicity, the
CONTAM Panel concluded that PAH4 and PAH8 are the most suitable indicators of PAHs in
food, with PAH8 not providing much added value compared to PAH4.
Keywords: Polycyclic aromatic hydrocarbons (PAHs), food, occurrence, indicators, exposure,

risk assessment, benchmark dose lower confidence limit (BMDL), margin of exposure (MOE),
toxic equivalency factor (TEF)

TABLE OF CONTENTS
SCIENTIFIC PANEL MEMBERS ......................................................................................................... 1
SUMMARY........................................................................................................................................ 1
BACKGROUND AS PROVIDED BY THE EUROPEAN COMMISSION .................................................... 7
TERMS OF REFERENCE AS PROVIDED BY THE EUROPEAN COMMISSION ...................................... 9
ACKNOWLEDGEMENT ..................................................................................................................... 9
ASSESSMENT ................................................................................................................................. 10
1. Introduction ............................................................................................................................... 10
2. Legislation................................................................................................................................. 14
3. Sampling and methods of analysis ............................................................................................ 17
3.1 Sampling.............................................................................................................................. 17
3.2 Methods of analysis............................................................................................................. 18
4. Sources and environmental fate ................................................................................................ 21
4.1 Formation and production ................................................................................................... 21
4.2 Environmental fate .............................................................................................................. 23
4.3 Sources of food contamination............................................................................................ 26
5. Occurrence and patterns of PAHs in food................................................................................. 27
5.1 Factors influencing the levels of PAHs in food .................................................................. 42
5.1.1 Commercial processing techniques.............................................................................. 42
5.1.2 Home cooking and other small scale cooking practices .............................................. 47
5.2 Suitable indicators for occurrence of total PAHs................................................................ 49
6. Food consumption ..................................................................................................................... 52
7. Human exposure assessment..................................................................................................... 56
7.1 Dietary intakes..................................................................................................................... 56
7.2 Contributions of different food groups to PAH exposure ................................................... 58
7.2.1 Mean dietary exposure to PAHs................................................................................... 59
7.2.2 Dietary exposure to PAHs for high consumers............................................................ 60
7.3 Sensitivity analysis of dietary exposure .............................................................................. 61
7.4 Estimates of non-dietary exposure to PAHs ....................................................................... 63
7.4.1 Ambient air................................................................................................................... 63
7.4.2 Occupational exposure ................................................................................................. 63
7.4.3 Smoking ....................................................................................................................... 64
8. Hazard identification and characterisation................................................................................ 65
8.1 Summary of key data........................................................................................................... 65
8.1.1 Toxicokinetics .............................................................................................................. 65
8.1.1.1 Absorption............................................................................................................. 65
8.1.1.2 Distribution............................................................................................................ 66
8.1.1.3 Metabolism............................................................................................................ 67
8.1.1.4 Excretion ............................................................................................................... 71
8.1.1.5 Biomarkers of exposure ........................................................................................ 72
8.1.1.6 Biomarkers of effect.............................................................................................. 75
8.1.2 Toxicological studies.................................................................................................... 75
8.1.2.1 Acute and short-term toxicity................................................................................ 75
8.1.2.2 Reproductive toxicity ............................................................................................ 76
8.1.2.3 Immunotoxicity ..................................................................................................... 76

8.1.2.4 Carcinogenicity ..................................................................................................... 76

8.1.2.5 Genotoxicity .......................................................................................................... 80
8.1.2.6 Observations in humans ........................................................................................ 80
8.1.3 Dose-response modelling ............................................................................................. 81
8.1.3.1 Benchmark dose modelling................................................................................... 82
8.1.3.2 Benchmark dose calculations for benzo[a]pyrene ................................................ 85
8.1.3.3 Benchmark dose calculations for PAH2 ............................................................... 86
8.1.4 Use of a TEF concept in the risk assessment of mixtures of carcinogenic PAHs........ 88
9. Risk characterisation ................................................................................................................. 90
10. Suitable indicators for occurrence and toxicity of genotoxic and carcinogenic PAHs ........... 91
11. Uncertainty analysis ................................................................................................................ 92
11.1 Assessment objectives....................................................................................................... 92
11.2 Exposure scenario ............................................................................................................. 92
11.3 Exposure model................................................................................................................. 93
11.4 Model input (parameters) .................................................................................................. 93
11.5 Other uncertainties ............................................................................................................ 94
CONCLUSIONS ............................................................................................................................... 94
RECOMMENDATIONS (INCL. KNOWLEDGE/DATA GAPS) .............................................................. 97
REFERENCES ................................................................................................................................. 97
LIST OF ABBREVIATIONS AND ACRONYMS ................................................................................. 111

BACKGROUND AS PROVIDED BY THE EUROPEAN COMMISSION

Polycyclic aromatic hydrocarbons (PAHs) form a class of diverse organic compounds, each of
them containing two or more aromatic rings. Hundreds of different such compounds may be
formed and released during a variety of combustion and pyrolysis processes and thus the natural
and anthropogenic sources of PAHs in the environment are numerous. PAH compounds are
emitted from the processing of coal, crude oil, petroleum, and natural gas, from production of
aluminium, iron and steel, from heating in power plants and homes (oil, gas, charcoal-fired
stoves, wood stoves), burning of refuse, wood fires, and from motor vehicle exhausts.
Humans can be exposed to PAHs through different routes. For the general population, the major
routes of exposure are from food and inhaled air, while in smokers, the contributions from
smoking and food may be of a similar magnitude. Food can be contaminated by environmental
PAHs that are present in air, soil or water, by industrial food processing methods (e.g. heating,
drying and smoking processes) and by home food preparation (e.g. grilling and roasting
processes).
The Scientific Committee on Food (SCF) reviewed the presence and toxicity of PAHs in food
and issued an opinion on 4 December 20023. SCF concluded that a number of PAHs are
genotoxic carcinogens and recommended that exposure to PAHs should be as low as reasonably
achievable. Fifteen substances were identified as a priority due to their potential genotoxicity
and/or carcinogenicity in humans. SCF also concluded that benzo[a]pyrene may be used as a
marker of occurrence and effect of the carcinogenic PAHs in food, but stressed that data
collection on the whole PAH profile should continue in order to be able to evaluate the
contamination of food commodities and any future change in the PAH profile.
The Joint FAO/WHO Expert Committee on Food Additives (JECFA) performed a risk
assessment on PAHs in 2005 in which it estimated Margins of Exposure (MOE) for PAHs. Based
on these Margins of Exposure the JECFA concluded that PAHs were of low concern for human
health. The JECFA also identified an additional substance, benzo[c]fluorene, as a priority
substance.
In the framework of Council Directive 93/5/EEC, a specific SCOOP task on data collection for
PAH has been performed in 20044. On the basis of SCF opinion and the SCOOP data, the
Commission introduced maximum levels for benzo[a]pyrene for some commodities that are
3 Opinion of the Scientific Committee on Food on the risks to human health of polycyclic aromatic hydrocarbons in
food expressed on 4 December 2002, available at: http://ec.europa.eu/food/fs/sc/scf/out153_en.pdf
4 Report on SCOOP task 3.2.12 "Collection of occurrence data on polycyclic aromatic hydrocarbons in food,
October 2004, available at: http://ec.europa.eu/food/food/chemicalsafety/contaminants/scoop_3-2-
12_final_report_pah_en.pdf. The SCOOP task was carried out in the framework of scientific cooperation with
Member States under Council Directive 93/5/EEC, OJ L 52, 4.3.1993, p.18-21

significant for human exposure and/or in which high PAH levels were found (e.g. oils and fats,
smoked meats and smoked meat products, smoked fish and smoked fish products, muscle meat of
fish other than smoked, crustaceans, cephalopods, bivalve molluscs and infant foods). The
current Regulation setting maximum levels for PAHs in foodstuffs is Commission Regulation
(EC) No. 1881/20065.
For some foodstuffs high levels were found in the SCOOP report, but data were inconclusive as
to the levels that were reasonably achievable (e.g. dried fruits, food supplements and cocoa
butter). The Commission therefore asked Member States to monitor PAHs (and in particular the
15 priority substances identified by SCF) in Commission Recommendation 2005/108/EC6. EFSA
collected the data submitted in the framework of this monitoring recommendation and reported
the findings in the recent Report "Findings of the EFSA Data collection on Polycyclic Aromatic
Hydrocarbons in Food" of 29 June 2007.
The EFSA report stated that the conclusion made by SCF that benzo[a]pyrene is a good indicator
for PAH occurrence could not be demonstrated by the recent monitoring data from the Member
States. Moreover, in October 2005 EFSA adopted the Margin of Exposure (MOE) approach7 for
substances which are both genotoxic and carcinogenic.
In order to reflect these changes, it is appropriate to request a full review of the 2002 SCF
opinion. Within this general review, a special focus should be given to the question whether or
not benzo[a]pyrene can be maintained as a marker for both occurrence and toxicity of the most
relevant PAHs. The relevance of the 15 priority PAHs as identified by SCF and the additional
PAH identified by the JECFA should be confirmed in the light of any new scientific data that
may have become available since the SCF opinion of 2002.
5 OJ L 364, 20.12..2006, p. 5
6 Commission Recommendation of 4 February 2005 on the further investigation into the levels of polycyclic
aromatic hydrocarbons in certain foods, OJ L 34, 8.2.2005, p. 43-45
7 The EFSA Journal (2005) 282, 1-31

TERMS OF REFERENCE AS PROVIDED BY THE EUROPEAN COMMISSION

In accordance with Article 29 (1) (a) of Regulation (EC) No 178/2002 the European Commission
asks the European Food Safety Authority to review the scientific opinion on PAH given by the
Scientific Committee on Foods (SCF) of December 2002.
The updated opinion should take into account
• the new occurrence data that were collected by EFSA and reported in the "Report on Findings
of the EFSA Data collection on Polycyclic Aromatic Hydrocarbons in Food" of 29 June 2007
and should contain an updated exposure assessment on basis of these data,
• any other relevant scientific information that may have become available since the SCF
opinion in 2002, including any new toxicological studies,
• the margin of exposure approach (MOE) as adopted by the EFSA Scientific Committee in the
Opinion related to substances which are both genotoxic and carcinogenic. In the updated risk
assessment the MOE approach should be used, if appropriate.
Within the general review, the following more specific questions should also be addressed:
• Can benzo[a]pyrene still be considered as a suitable marker for both occurrence and
carcinogenic effects for the most relevant PAHs ?
• In the event that benzo[a]pyrene can not be maintained as a marker, could other suitable
markers or concepts (e.g. TEF concept) be recommended for the occurrence as well as
toxicity of the most relevant PAHs to be applied in European monitoring and official controls
in order to protect human health ?
• Which are the food commodities in the different Member States that contribute most to the
exposure to the most relevant PAHs and which specific groups of the population are most
exposed?
ACKNOWLEDGEMENT
EFSA wishes to thank the working group members Diane Benford, Jean-Pierre Cravedi, Peter
Fürst, Niklas Johansson, John Christian Larsen, Dieter Schrenk, Rolaf van Leeuwen and Philippe
Verger.

ASSESSMENT
1. Introduction
Polycyclic aromatic hydrocarbons (PAHs) constitute a large class of organic compounds that are
composed of two or more fused aromatic rings. They solely consist of carbon and hydrogen and
do not contain heteroatoms. PAHs are primarily formed by incomplete combustion or pyrolysis
of organic matter and during various industrial processes. Consequently, the natural and
anthropogenic sources in the environment are numerous. PAHs generally occur in complex
mixtures which may consist of hundreds of compounds. The composition of these mixtures varies
with the generating process.
Humans are exposed to PAHs by various pathways. While for non-smokers the major route of
exposure is consumption of food, for smokers the contribution from smoking may be significant.
Food can be contaminated from environmental sources, industrial food processing and from
certain home food preparation.
In the past decade PAHs were evaluated by the International Programme on Chemical Safety
(IPCS) (WHO/IPCS, 1998), the Scientific Committee on Food (SCF) (EC, 2002) and by the Joint
FAO/WHO Expert Committee on Food Additives (JECFA) (FAO/WHO, 2005).
SCF concluded that 15 out of the 33 PAHs which were considered in their assessment, namely
benz[a]anthracene, benzo[b]fluoranthene, benzo[j]fluoranthene, benzo[k]fluoranthene,
benzo[ghi]perylene, benzo[a]pyrene, chrysene, cyclopenta[cd]pyrene, dibenz[a,h]anthracene,
dibenzo[a,e]pyrene, dibenzo[a,h]pyrene, dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3-
cd]pyrene and 5-methylchrysene show clear evidence of mutagenicity/genotoxicity in somatic
cells in experimental animals in vivo and with the exception of benzo[ghi]perylene, have also
shown clear carcinogenic effects in various types of bioassays in experimental animals.
Furthermore, SCF concluded that these compounds may be regarded as potentially genotoxic and
carcinogenic to humans and thus represent a priority group in the assessment of the risk of long-
term adverse health effects following dietary intake of PAHs. SCF estimated a maximum daily
intake of benzo[a]pyrene from food of approximately 6-8 ng/kg b.w. per day for a person
weighing 70 kg. It suggested to use benzo[a]pyrene as a marker of occurrence and effect of the
carcinogenic PAHs in food, based on examinations of PAH profiles in food and on evaluation of
a carcinogenicity study of coal tar mixtures in mice (Culp et al., 1998). Based on this latter study,
a conservative assumption would imply that the carcinogenic potency of total PAHs in foods
would be 10 times of that contributed by benzo[a]pyrene alone. SCF finally stressed that though
it considered benzo[a]pyrene as a marker of carcinogenic PAHs in food, chemical analyses
should continue to collect data on other PAHs in order to be able to evaluate the contamination of
food commodities and any future change in the PAH profile.

Using the risk assessments of IPCS and SCF as starting points and taking into account newer
studies, the JECFA re-evaluated PAHs in 2005. Overall, the JECFA concluded that 13 PAHs are
clearly genotoxic and carcinogenic. Except benzo[ghi]perylene and cyclopenta[cd]pyrene the
compounds are the same as those stated by SCF. The JECFA also decided to apply a surrogate
approach to the evaluation, in which benzo[a]pyrene was used as a marker of exposure to, and
effect of, the 13 genotoxic and carcinogenic PAHs. Comparisons of representative mean and high
level intakes with the BMDL108 equivalent to 100 µg of benzo[a]pyrene/kg body weight per day
indicated margins of exposure (MOE) of 25,000 and 10,000, respectively. Based on these MOEs,
the JECFA concluded that the estimated dietary intakes of PAHs were of low concern for human
health. Like the SCF the JECFA also recommended that efforts should be made to collect data on
concentrations in the major food groups for those PAHs which were clearly identified as being
genotoxic and carcinogenic. In addition, the JECFA recommended to include benzo[c]fluorene as
a further compound into the analysis as data on its occurrence in food are still scarce but levels of
benzo[c]fluorene-derived adducts were much higher than those of benzo[a]pyrene-derived
adducts in the lung of rats fed with coal tar, which might indicate that benzo[c]fluorene may
contribute to the formation of lung tumours after oral exposure to coal tar.
Following the suggestion of SCF the Commission issued a recommendation on the further
investigation into levels of polycyclic aromatic hydrocarbons in certain foods (2005/108/EC)9.
These investigations should especially include those PAHs considered to be carcinogenic by
SCF. It was also recommended that Member States should investigate the production and
processing methods used for food oils, fats as well as smoking and drying foods. Meanwhile
eighteen Member States have already submitted approximately 10,000 results for PAH levels in
food. An evaluation of these data by EFSA, published in June 2007 and updated in June 2008
showed that samples belonging to those food categories that are covered by Commission
Regulation (EC) No 1881/2006 exceeded the respective maximum levels for benzo[a]pyrene in
6.4% of cases. Certain PAHs, such as chrysene were found in some food samples that were tested
negative for benzo[a]pyrene (chrysene levels up to 242 µg/kg). In these cases benzo[a]pyrene can
not be an indicator of PAH contamination. In view of these findings as well as the MOE
approach adopted by EFSA in 2005 (EFSA, 2005a) for substances which are both genotoxic and
carcinogenic, the Commission requested a full review of the 2002 SCF opinion on PAHs (see
terms of reference).
The EFSA Panel on Contaminants in the Food Chain (CONTAM Panel) reviewed the available
data on occurrence and especially toxicology of PAHs. As no new toxicological data could be
identified that would lead to the inclusion of further compounds into the list of priority PAHs, the
CONTAM Panel decided to cover the 15 PAHs identified by SCF in 2002 and additionally
8 BMDL10: 95% lower confidence limit of the benchmark dose for a 10% increase in tumor bearing animals
9 OJ L 34, 8.2.2005, p.43

benzo[c]fluorene as suggested by the JECFA in 2005 in the present opinion. The 16 compounds
are given in Table 1 along with their molecular weight, CAS number and structure. Special
attention was paid to the eight carcinogenic and genotoxic PAHs that were measured in the coal
tar mixtures used in the carcinogenicity studies of Culp et al. (1998), which provided the basis of
the SCF and JECFA risk assessments. These eight PAHs are underlined in Table 1.
Table 1: Polycyclic aromatic hydrocarbons considered in the present opinion.

Compound Abbr. MW CAS Structure
Benz[a]anthracene BaA 228.3 56-55-3
Benzo[b]fluoranthene BbFA 252.3 205-99-2
Benzo[j]fluoranthene BjFA 252.3 205-82-3
Benzo[k]fluoranthene BkFA 252.3 207-08-9
Benzo[ghi]perylene BghiP 276.3 191-24-2
Benzo[a]pyrene BaP 252.3 50-32-8

Chrysene CHR 228.3 218-01-9
Cyclopenta[cd]pyrene CPP 226.3 27208-37-3
Dibenz[a,h]anthracene DBahA 278.3 53-70-3
Dibenzo[a,e]pyrene DBaeP 302.3 192-65-4
Dibenzo[a,h]pyrene DBahP 302.3 189-64-0
Dibenzo[a,i]pyrene DBaiP 302.3 189-55-9

Dibenzo[a,l]pyrene DBalP 302.3 191-30-0
Indeno[1,2,3-cd]pyrene IP 276.3 193-39-5
5-methylchrysene MCH 242.3 3697-24-3
CH3
Benzo[c]fluorene BcFL 216.3 205-12-9
Abbr.: Abbreviation, MW: Molecular weight, CAS: Chemical Abstract Service Number
2. Legislation
In view of disparities caused by different maximum levels (ML) for PAHs in food in several
Member States, the Commission set harmonized ML for benzo[a]pyrene for the first time in 2005
by Commission Regulation (EC) No 208/200510 amending Regulation (EC) No 466/200111 as
regards PAHs. Benzo[a]pyrene was chosen because SCF concluded in its opinion of 4 December
2002 that this compound can be used as a marker for the occurrence and effects of carcinogenic
PAHs in food. Moreover, the data on occurrence and relative proportions of other carcinogenic
PAHs than benzo[a]pyrene in food were considered insufficient by the Commission for setting
further ML. The current MLs are laid down in the Annex, Section 6 of Commission Regulation
(EC) No 1881/2006 setting maximum levels for certain contaminants in foodstuffs12. Maximum
levels are especially set for foodstuffs containing fats and oils and foods where smoking and
10 OJ L 34, 8.2.2005, p. 3
11 OJ L 77, 16.3.2001, p. 1
12 OJ L 364, 20.12.2006, p. 5

drying processes or environmental pollution might cause high levels of contamination. The
lowest MLs are set for food for infants and young children. All MLs are expressed as µg/kg wet
weight (Table 2) unless more specific footnote applies.
Table 2: Maximum levels (MLs) for benzo[a]pyrene as laid down in Regulation (EC) No.
1881/2006. See the original Regulation for further definitions and explanations of individual food
commodities.
ML (µg/kg wet
Foodstuff
weight)
Oils and fats (excluding cocoa butter) intended for direct human
2.0
consumption or use as an ingredient in foods
Smoked meats and smoked meat products 5.0
Muscle meat of smoked fish and smoked fishery products, excluding

bivalve molluscs. The maximum level applies to smoked crustaceans,
5.0
excluding the brown meat of crab and excluding head and thorax meat of
lobster and similar large crustaceans (Nephropidae and Palinuridae).
Muscle meat of fish, other than smoked fish 2.0
Crustaceans, cephalopods, other than smoked. The maximum level applies

to crustaceans, excluding the brown meat of crab and excluding head and
5.0
thorax meat of lobster and similar large crustaceans (Nephropidae and
Palinuridae)
Bivalve molluscs 10.0
Processed cereal-based foods and baby foods for infants and young
1.0
children*
Infant formulae and follow-on formulae, including infant milk and follow-on
1.0
milk*
Dietary foods for special medical purposes intended specifically for infants* 1.0
*
The maximum level applies to the product as sold.
In the past 20 years the traditional smoking process of food has been replaced more and more by
the application of smoke flavourings. Because smoke flavourings are produced from smoke
which is subjected to fractionation and purification processes, their use is generally considered to
be of less health concern than the traditional smoking process. In this context, already in 1988 the

Council of the European Communities set a ML of 0.03 µg/kg for benzo[a]pyrene in foodstuffs
as a result of the use of flavourings by Council Directive 88/388/EEC on the approximation of
the laws of the Member States relating to flavourings for use in foodstuffs and to source materials
for their production13. This directive defines a “smoke flavouring” as a “smoke extract used in
traditional foodstuffs smoking processes”. Directive 88/388/EEC also provides for the adoption
of appropriate provisions concerning source materials used for the production of smoke
flavourings and reaction conditions under which they are prepared. These provisions are laid
down in Regulation (EC) No 2065/2003 of the European Parliament and of the Council on smoke
flavourings used or intended for use in or on foods14. This Regulation lays down a Community
procedure for the:
(a) evaluation and authorisation of primary smoke condensates and primary tar fractions for
use as such in or on foods or in the production of derived smoke flavourings for use in or on
foods,
(b) establishment of a list of primary smoke condensates and primary tar fractions
authorised to the exclusion of all others in the Community and their conditions of use in or
on foods.
To this end Regulation (EC) No 2065/2003 sets MLs of 10 µg/kg for benzo[a]pyrene and 20
µg/kg for benz[a]anthracene in water-based primary smoke condensates as well as fractions
obtained after physical processing from water-insoluble high density tar phases.
Requirements for several PAHs in drinking water are set by Council Directive 98/83/EC on the
quality of water intended for human consumption15. This Directive stipulates that Member States
shall set limit values of 0.01 µg/L for benzo[a]pyrene and 0.1 µg/L for the sum of
benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene.
Commission Directive 95/45/EC laying down specific purity criteria concerning colours for use
in foodstuffs16 and Commission Directive 96/77/EC laying down specific purity criteria on food
additives other than colours and sweeteners17 both provide maximum levels for PAHs as
impurities in some food additives without referring to specific PAH.
13 OJ L 184, 15.7.1988, p.61

14 OJ L 309, 26.11.2003, p. 1
15 OJ L 330, 5.12.1998, p. 32
16 OJ L 226, 22.9.1995, p. 1
17 OJ L 339, 30.12.1996, p. 1

3. Sampling and methods of analysis
3.1 Sampling
As sampling, sample preparation and analytical procedures play an important role for a reliable
determination of PAHs in foodstuffs, harmonized general criteria were at first set in the European
Union (EU) in 2005 by Commission Directive 2005/10/EC laying down the sampling methods
and the methods of analysis for the official control of the levels of benzo[a]pyrene in foodstuffs18
simultaneously with the first setting of MLs for this compound. The present provisions are set in
Commission Regulation (EC) No. 333/2007 laying down the methods of sampling and analysis
for the official control of the levels of lead, cadmium, mercury, inorganic tin, 3-
monochloropropane-1,2-diol (3-MCPD) and benzo[a]pyrene in foodstuffs19. Besides definitions
and general provisions this Regulation stipulates detailed requirements for methods of sampling
for the different food commodities. No requirements exist for the number of samples that have to
be analysed.
The requirements for sampling methods differ especially depending on type and weight of lot, in
order to obtain samples that are representative for the respective lot. Each lot which is to be
examined shall be sampled separately. Large lots shall be subdivided into defined sub lots which
have to be sampled separately. In any case, as far as possible, incremental samples shall be taken
at various places throughout the lot or sub lot. By combining the incremental samples an
aggregate sample is formed which after homogenization represents the base material for
enforcement, defence and reference purposes.
In the course of sampling and preparation of the samples, precautions shall be taken to avoid any
changes in composition of the sample. As a specific and detailed requirement it is laid down that
the analyst shall ensure that samples do not become contaminated during sample preparation.
Containers shall be rinsed with high purity acetone or hexane before use to minimise the risk of
contamination. Wherever possible, apparatus and equipment coming into contact with the sample
shall be made of inert materials such as aluminium, glass or polished stainless steel. Plastics such
as polypropylene or polytetrafluoroethylene (PTFE) shall be avoided because the PAHs can
adsorb onto these materials.
Losses of PAHs may also occur if the sample is exposed to light and high temperatures during the
sample collection and storing, or PAHs react with other matrix substances during a long-term
storage. Exposure to tobacco smoke may increase the PAH levels in the sample (FAO/WHO,
2006).
18 OJ L 34, 8.2.2005, p. 15
19 OJ L 88, 29.3.2007, p. 29

3.2 Methods of analysis

Commission Regulation (EC) No 333/2007 also contains strict requirements with which the
methods of analysis have to comply in order to ensure that control laboratories use procedures
with comparable levels of performance. The Regulation follows the “criteria approach”. This
means that no prescribed fixed official methods have to be followed but laboratories can use each
method of analysis, provided it can be demonstrated in a traceable manner that they strictly fulfil
the analytical requirements laid down in the respective legislation. As a general requirement,
methods for benzo[a]pyrene analysis used for food control purposes must comply with the
provisions of points 1 and 2 of Annex III (characterisation of methods of analysis) to Regulation
(EC) No 882/2004 of the European Parliament and of the Council of on official controls
performed to ensure the verification of compliance with feed and food law, animal health and
animal welfare rules20. While Regulation (EC) No 882/2004 contains the general provisions, the
specific requirements for the official control of benzo[a]pyrene in foodstuffs are laid down in
Commission Regulation (EC) No 333/2007. Annex, Table 7 of the latter one sets performance
criteria for applicability, limits of detection (LOD) and quantification (LOQ), precision, recovery
and specificity. The methods used for the determination should be applicable to those food
specified in Regulation (EC) No 1881/2006. Requirements for LOD and LOQ are given as less
than 0.3 µg/kg and less than 0.9 µg/kg, respectively. The recovery should be 50-120%.
Concerning precision, it is required that the HORRATr21and HORRATR22 values are less than 2.
The requirement for specificity is given as “free from matrix or spectral interferences,
verification of positive detection”.
Finally, the Commission Regulation (EC) No 333/2007 sets requirements for the reporting of
results and the assessment of compliance of the lot or sub lots. For this, the analytical result
corrected for recovery shall be used for checking compliance. The analytical result must be
reported as x ± U whereby x is the analytical result and U is the expanded measurement
uncertainty, using a coverage factor of 2 which gives a level of confidence of approximately
95%. The lot or sublot is accepted if the analytical result of the laboratory sample does not
exceed the respective maximum level as laid down in Regulation (EC) No 1881/2006 taking into
account the expanded measurement uncertainty and correction of the result for recovery.
20 OJ L 191, 28.5.2004, p. 1
21 HORRATr:: The observed relative standard deviation calculated from results generated under repeatability
conditions (RSDr) divided by the RSDr value estimated from the Horwitz equation (using the assumption that the
repeatability r = 0.66R (reproducibility). The Horwitz equation is a generalised precision equation which has been
found to be independent of analyte and matrix but solely dependent on concentration for most routine methods of
analysis.
22 HORRATR: The observed relative standard deviation calculated from results generated under reproducibility
conditions (RSDR) divided by the RSDR value calculated from the Horwitz equation.

The extraction method used for extracting PAHs from food samples greatly depends on the
nature of the food matrix. Saponification followed by liquid-liquid extraction and extraction with
organic solvent are the most often used methods for solid food samples (e.g. meat, fish and their
products), while liquid-liquid extraction is often used for liquid samples (e.g. vegetable oils). For
solid food matrices automated extraction techniques, such as pressurised liquid extraction (PLE)
and supercritical fluid extraction (SFE), are also applied, but less frequently. Column
chromatography, solid phase extraction (SPE) and gel permeation chromatography (GPC) are the
main sample purification techniques used for isolating PAHs from interfering matrix substances.
Recently especially the automatic GPC technique has been applied for cleaning up PAH sample
extracts (Fontcuberta et al., 2006; Llobet et al., 2006; Fromberg et al., 2007; Reinik et al., 2007).
Nowadays the two main analytical techniques for determining PAHs in foods are high
performance liquid chromatography (HPLC) coupled to a fluorescence detector (FLD) and gas
chromatography-mass spectrometry (GC-MS). Both of these methods are sufficiently sensitive
for determining PAH concentrations usually found in foods. Earlier HPLC with an ultraviolet
(UV) or a photo-diode array (PDA) detector and GC with a flame ionisation detector (FID) were
also methods often applied but today they are not state of the art anymore due to their poorer
selectivity and sensitivity. Programmed temperature vaporization (PTV) injection used in the
GC-methods allows injection of high sample volumes and hence low detection limits are
achieved for PAHs. On-line methods simplify the analytical procedure of the PAH analysis
because several separate manual sample preparation steps can be omitted. In the on-line HPLC-
method the sample extract is injected into a donor-acceptor complex chromatographic column
which is coupled on-line to an analytical column in the HPLC-FLD system (van der Wielen et
al., 2006; ISO, 2007).
The GC-MS methods have become popular methods for analysing PAHs in foods. This is due to
the selectivity of the MS-detector, the use of mass spectrum data for reliable confirmation of
PAHs, and the possibility to use isotope labelled PAHs as internal standards. Single quadrupole
MS is the main detection technique applied but the use of tandem mass spectrometry is increasing
because it gives more specific mass fragments (daughter ions) and hence improves specificity and
sensitivity of the MS-methods (Varlet et al., 2007; Veyrand et al., 2007).
LODs and recoveries for PAHs are compound, food matrix and method dependent. The LODs for
PAHs typically range from 0.1 to 1 µg/kg. However, for some PAHs lower and higher LODs,
even up to 20 µg/kg, are determined (EFSA, 2007). Due to very specific tandem mass
spectrometry, LODs lower than 0.1 µg/kg can be obtained by GC-MS/MS methods (Varlet et al.,
2007). The recoveries for PAHs in different food samples are mostly over 70% with both HPLC
and GC methods but for some PAHs recoveries are lower (FAO/WHO, 2006).
A number of PAH compounds, such as benzofluoranthene isomers, have been reported to be

difficult to separate chromatographically (FAO/WHO, 2006; Rose et al., 2007). In the first inter-
laboratory test organised by the Community Reference Laboratory for PAHs (CRL PAH) on the
determination of the 15 SCF priority PAHs and benzo[c]fluorene in solvent, the participating
National Reference Laboratories (NRLs) reported difficulties to analyse cyclopenta[cd]pyrene
and some of the dibenzopyrenes (dibenzo[a,h]pyrene and dibenzo[a,i]pyrene) (EC-DG-JRC-
IRMM/CRL, 2007). The second inter-laboratory comparison test organised by the CRL PAH for
the consortium of NRLs was in that respect much more promising. Nearly all NRLs were able to
determine all 15 SCF priority PAHs and benzo[c]fluorene both in solvent solution and in spiked
olive oil. The overall result of the measurements of the 15 SCF priority PAHs and
benzo[c]fluorene in the edible oil was satisfactory. All together 345 analysis results for individual
PAHs were reported to the CRL PAH. About 76% of them were rated with z-scores < ±1. The
majority of the 29 unsatisfactory results were reported by 4 laboratories only. However,
scrutinising their results revealed that mainly instrument calibration was responsible for the
deviations, which underpins the importance of proper standard preparation (EC-DG-JRC-
IRMM/CRL, 2008). Interferences coming from non-target PAHs such as triphenylene that may
interfere with chrysene, should also be considered. For these reasons care must be taken when the
results for PAHs are calculated.
Isotope labelled PAHs (deuterium-labelled and 13C-labelled PAHs) are commonly used as
internal standards to ensure the reliability of the final results. The 13C-labelled PAHs are
preferred in most of the GC-MS methods due to the instability of the deuterated standards
(Fromberg et al., 2007; Veyrand et al., 2007). Although a number of 13C-labelled PAH standards
are available on the market at the moment there is still a need for 13C-labelled analogues for
various PAHs such as benzo[j]fluoranthene, dibenzo[a,h]pyrene, dibenzo[a,l]pyrene
cyclopenta[cd]pyrene, 5-methylchrysene and benzo[c]fluorene. The lack of the 13C-labelled
standards may partly explain the high variation of the results reported for some of the PAH
compounds in foods (Rose et al., 2007). Some certified reference materials (CRMs) for PAHs are
available on the market and they are used for method validation studies to assess the performance
of the method. In addition, the performance of the analytical methods is tested in the proficiency
tests organised for PAHs. However, only a limited number of CRMs containing only some of the
PAH compounds are on the market today and the few proficiency tests, which are organised
presently, often cover only a narrow range of PAHs in some food products23. Only recently
proficiency tests for the full range of 15 SCF priority PAHs and benzo[c]fluorene (olive oil) have
been provided23. The lack of CRMs and proficiency tests are clearly limitations when the method
performance for the analytical methods for analysis of PAHs in foods is assessed.
23 http://irmm.jrc.ec.europa.eu/rmcatalogue/searchResultrmcatalogue.do,
http://www.fapas.com/prog.cfm?currsch=fapas and
http://ts.nist.gov/measurementservices/referencematerials/index.cfm

4. Sources and environmental fate

Data on the sources, fate and degradation of PAHs are extensively covered in a review made by
IPCS (WHO/IPCS, 1998). PAHs have low solubility in water but are readily soluble in organic
solvents or organic acids. Thus, in aqueous environments, PAHs are generally found adsorbed on
particulates and on humic matter, or dissolved in any oily contaminant that may be present in
water, sediment and soil. The solubility of PAHs in water is inversely proportional to the number
of rings the PAH molecule contains.
PAHs are solids at room temperature. Since PAHs tend to have low vapour pressure, they are
usually adsorbed on particulate matter in the atmosphere. The vapour pressure of PAH is
inversely proportional to the number of rings contained and thus almost all five-ring PAH
compounds are particulate bound, while three-ring PAHs are also present as vapour in the
atmosphere.
4.1 Formation and production

There is no known use for the carcinogenic and genotoxic PAHs considered in this opinion
except as research chemicals.
A plethora of different PAHs may be formed and released during a variety of combustion and
pyrolysis processes. Thus the natural and anthropogenic sources of PAHs in the environment are
numerous. So far about 500 PAHs have been detected in ambient air. The emission of PAHs
during industrial production and processing in developed countries are not thought to be
important in comparison with the release of PAHs from incomplete combustion processes, since
closed systems and recycling processes are usually used (WHO/IPCS, 1998). The primary natural
sources of airborne PAHs are forest fires and volcanoes. The most important stationary
anthropogenic sources include residential burning of wood, oil, gas and charcoal as well as
industrial power generation, incineration, production of aluminium, iron and steel, petroleum
catalytic cracking and production of asphalt, coal tar and coke (EC, 2002). Stationary sources
account for approximately 80% of total annual PAH emissions. The most important mobile
sources are vehicular exhausts from gasoline and diesel-powered engines (ATSDR, 1995). In
combustion processes the formation of PAHs is reduced when combustion is more thoroughly
performed but this will increase the formation of nitrogen oxides.
Limited data are available on emission factors of PAHs, and the available data are often reported
in different ways which means that comparison of data for verification purposes is difficult. The
reasons for this are usually:
• many of the reported emissions of PAHs only give a figure for ‘total PAHs’, without
indicating which PAH compounds are included in the total PAHs,

• where emissions of individual PAHs are given, there is a lack of consistency between
reports on which PAHs are included in the measurements taken,
• most of the reported emissions of individual PAH only give data for one or two
compounds (usually including benzo[a]pyrene).
Different main sources exhibit different PAH profiles regarding their PAH emissions. This is
exemplified in Tables 3 and 4 where the estimations for ratios of benzo[b]fluoranthene,
benzo[k]fluoranthene and indeno[1,2,3-cd]pyrene in relation to benzo[a]pyrene are given.
Table 3: PAH profiles for main stationary sources, estimated in a ratio to benzo[a]pyrene.
PAH Coal Wood Natural fires/ Anode baking
combustion combustion agricultural
(industrial and (industrial and biomass
domestic) domestic) burning
Benzo[b]fluoranthene 0.05 1.2 0.6 2.2 BbFA + BkFA
Benzo[k]fluoranthene 0.01 0.4 0.3
Benzo[a]pyrene 1.0 1.0 1.0 1.0
Indeno[1,2,3-cd]pyrene 0.8 0.1 0.4 0.5
Profiles in the above table taken from http://reports.eea.europa.eu/EMEPCORINAIR5/en/BPAH.pdf.
Table 4: PAH profiles for main mobile sources, estimated in a ratio to benzo[a]pyrene.
PAH Passenger Passenger Passenger Passenger Heavy
cars, cars, closed cars, diesel, cars, diesel, duty
conventional loop catalyst direct indirect vehicles
injection injection
Benzo[b]fluoranthene 1.2 0.9 0.9 0.9 5.6
Benzo[k]fluoranthene 0.9 1.2 1.0 0.8 8.2
Benzo[a]pyrene 1.0 1.0 1.0 1.0 1.0
Indeno[1,2,3-cd]pyrene 1.0 1.4 1.1 0.9 1.4
Profiles in the above table taken from http://reports.eea.europa.eu/EMEPCORINAIR5/en/BPAH.pdf.
As seen in Tables 3 and 4 the PAH profiles vary considerably between the stationary sources
investigated. Whereas benzo[a]pyrene is of major importance in coal combustion activities, it is
less dominating in the anode baking process. On the other hand, the variation in PAH profiles for
mobile sources is less pronounced with the exception of high relative emissions of
benzo[b]fluoranthene and benzo[k]fluoranthene from heavy duty vehicles.

Estimations of annual PAH emissions due to residential heating are available for a few countries
only. In western Germany, the benzo[a]pyrene emissions were about 10 tons in 1981, 7 tons in
1985, and 2.5 tons in 1988, mainly resulting from coal heating. The reduction in the release of
PAHs into the atmosphere due to domestic heating resulting from increasing use of oil and gas
during the last 40-50 years was estimated to be 90-99%. In the Netherlands, the estimated release
in 1985 was less than 1 ton for each benzo[k]fluoranthene and indeno[1,2,3-cd]pyrene, less than
10 tons per year each for benz[a]anthracene, chrysene, benzo[a]pyrene and other PAHs, mainly
resulting from wood heating. The total PAH emission, mainly from coal and wood heating, was
in 1985 about 63 tons in Norway and 130 tons in Sweden (Björseth and Ramdahl, 1985). In
Canada in 1990, the total PAH release due to residential heating, mainly wood burning, was
about 500 tons (WHO/IPCS, 1998). Kochbach et al. (2006) has demonstrated that particles from
wood burning could contain 30 times more PAHs compared to particles from road traffic.
Compared by amount of PAHs per unit of energy, this difference is even more pronounced
(Boström et al., 1998).
The European Pollutant Emission Register (EPER)24, has collected information on emissions of
PAHs in Europe in 2001 (15 Member States) and 2004 (25 Member States). Under EPER the
following PAHs are included: benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene,
benzo[a]pyrene, indeno[1,2,3-cd]pyrene and fluoranthene. For 2004, EPER reported aluminium
industry and other metal industries to be major emitters of PAHs and the total amount emitted to
air in Europe to be approximately 545 tons. The amount emitted to water is reported to be much
less, or totally about 25 tons per year released by e.g. petrochemical and textile industry, but also
by metal industry. Among the facilities that reported on the PAH emissions to EPER in both 2001
and 2004 the total emissions of PAHs increased by 13.2%.
4.2 Environmental fate

PAHs released to the atmosphere are subject to short- and long-range transport and are removed
from the atmosphere by wet and dry deposition onto soil, water, and vegetation. In surface water,
PAHs can be volatilised, photolysed, biodegraded and bind to suspended particles or sediments,
or accumulate in aquatic organisms. The bioconcentration factors vary between 10 and 10,000
with species and type of PAH. PAHs in soil can be volatilised or undergo biotic or abiotic
degradation, mainly photolysis and oxidation. PAHs in soil can also enter groundwater and be
transported within an aquifer. Based on experimental results, the estimated half-lives of PAHs in
soil are e.g. for chrysene 371-387 days, benzo[a]pyrene 229-309 days and dibenz[a,h]anthracene
361-420 days (Park et al., 1990). In the atmosphere, PAHs are present in the gaseous phase or
24 http://www.eper.ec.europa.eu

adsorbed to particles. In general, PAHs having two or three rings are present in air predominantly
in the vapour phase. PAHs that have four rings exist both in the vapour and particulate phase, and
PAHs having five or more rings are found predominantly bound to particles. Atmospheric
residence time and transport distance depend on the size of particles to which PAHs are adsorbed
and on climatic conditions. About 90-95% of particulate PAHs are associated with particle
diameters less than 3.3 μm. Particles with diameter range of 0.1-3.0 μm, with which airborne
PAHs are principally associated, remain airborne for a few days or longer and can thus be
transported over long distances.
In the air PAHs can undergo a number of chemical reactions. The most important are reactions
between PAHs adsorbed on the particle surfaces and oxidant gases like NO2, O3, and SO3, and
photo oxidation of PAHs irradiated under solar radiation (ATSDR, 1995). Photolysis is the most
important factor in the decay of PAHs adsorbed to particles in the atmosphere. Many PAHs such
as benzo[a]pyrene are, however, rapidly destroyed by UV light. In the absence of local sources a
pronounced seasonal trend has been demonstrated by the Co-operative Programme for
Monitoring and Evaluation of the Long-range Transmissions of Air Pollutants in Europe (EMEP)
(2001).
0,1
BaP (ng/m )
3
0,01
0,001
2 4 6 8 10 12
M o nth
K osetice, C zech R epublic
S pitsbergen
Figure 1: Sum of benzo[a]pyrene concentrations in air and aerosol at the EMEP air sampling
stations in Kosetice, Bohemian Highlands and in Zeppelinfjell, Spitsbergen in 1999. Based on
data from EMEP (2001). Note logarithmic scale.

Butler and Crossley (1981) estimated the following half-lives for degradation of a number of
PAHs: benzo[a]pyrene 7 days, benzo[ghi]perylene 8 days, benz[a]anthracene 11 days and
chrysene 26 days when PAH was adsorbed to soot particles and exposed to sunlight in air
containing approximately 20 mg/m3 nitrogen oxides.
There is a relatively large body of data characterizing PAH levels in air at a variety of sites.
Caution must be used in interpreting and comparing results of different studies, because of the
different sampling methods used. To fully characterize atmospheric PAH levels, both particle-
and vapour-phase samples must be collected.
In the 1960s, the benzo[a]pyrene levels were sometimes higher than 100 ng/m3 in many
European cities but during the past 35 years concentrations in urban ambient air have decreased
considerably. Greenberg et al. (1985) evaluated atmospheric concentrations of particulate phase
PAHs at four New Jersey sites (three urban and one rural) over two summer and winter seasons
during 198l-82. Urban PAH concentrations were approximately 3-5 times higher than those at the
rural site. In addition, winter PAH concentrations were approximately 5-10 times higher than
summer PAH concentrations. In urban areas, as illustrated in Figure 1, the concentrations of
PAHs such as benzo[a]pyrene in ambient air show a cyclic pattern with often more than 10 times
higher concentrations during the winter. In extremely remote areas this cyclic pattern is less
pronounced or even absent.
Background concentration and deposition of PAHs have been studied at Rörvik on the Swedish
west coast and Pallas in the northern Finland. At these stations, PAH concentrations in air and
deposition rates have been performed with identical methods. The PAHs measured include eleven
compounds: phenanthrene, anthracene, fluoranthene, pyrene, benz[a]anthracene, chrysene,
benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, benzo[ghi]perylene, and
indeno[cd]pyrene. The background deposition of the sum of these eleven PAHs at Rörvik during
1994 to 1998 showed no temporal trend. The mean air concentration for the sum of eleven PAHs
at Rörvik, gas and particle phases combined, was around 4 ng/m3 and for benzo[a]pyrene 0.09
ng/m3. At Pallas in the northern Finland, the levels were found to be lower by a factor of 3-5.
In the study of Pham et al. (1993), raw water samples from 5 areas in the St. Lawrence River and
its tributaries were analyzed for 12 PAHs. The highest mean total PAH concentrations were
observed in samples collected in the spring (27.3 ng/L) and autumn (21.03 ng/L), which was
attributed to snow melt and increased runoff during these respective seasons. The lowest mean
total PAH concentration was observed in summer (14.63 ng/L). High molecular weight PAHs
were detected more frequently in the spring and autumn samples. Moreover, a general decrease in
concentration with increasing molecular weight was observed.
The profound variation in half life and ability to undergo transport in atmospheric and other

environmental media cause large differences between the composition of PAHs released from
different sources and the composition that can be found in different kinds of foods and feeds as a
result of environmental contamination.
The Marine Chemistry Working Group of the International Council for the Exploration of the Sea
(ICES) has recently examined available data on PAHs in marine biota including mussels from
“remote” areas in Scotland and Spain. A large additional dataset from France for mussels and
oysters was also used although this was not screened to select remote areas. Sediment core data
were available for one Scottish remote site. From these datasets, the median of the 10th percentile
for the different datasets was proposed as a “low concentration”. However, the working group
suggested that natural background concentrations would be lower than these proposed “low
concentrations”, see Table 5.
Table 5: Proposed “low concentrations” of selected PAHs in shellfish such as mussels and
oysters. From the draft of the Protection of the Marine Environment of the North-East Atlantic
(OSPAR) (2008).
PAH Proposed ”low concentrations” of individual PAHs
(μg/kg wet weight)
Benzo[a]anthracene 0.2
Chrysene/Triphenylene 0.8
Benzo[b]fluoranthene 0.6
Benzo[j]fluoranthene 0.6
Benzo[k]fluoranthene 0.2
Benzo[a]pyrene 0.1
Indeno[1,2,3-cd]pyrene 0.2
Benzo[ghi]perylene 0.3
4.3 Sources of food contamination

The waxy surface of vegetables and fruits is able to concentrate low molecular mass PAHs
through surface adsorption and particle-bound high molecular mass PAHs can contaminate the
surface due to atmospheric fallout. PAHs can also contaminate foods during industrial smoking,
heating and drying processes that allow combustion products to come into direct contact with
food. Contamination of cereals and vegetable oils (including seed oils and olive residue oils) with
PAHs usually occurs during technological processes like direct fire drying, where combustion
products may come into contact with the grain, oil seeds or the oil (Speer et al., 1990; EC, 2001).
PAHs are also formed as a result of certain home food preparation methods, such as grilling,
roasting and smoking. High PAH concentrations have been reported in charcoal

grilled/barbecued foods (such as fatty meat and meat products grilled under prolonged and severe
conditions), in foods smoked by traditional techniques (fish in particular), and in mussels and
other seafood from polluted waters (Guillén et al., 1997; Phillips, 1999). Smoked and grilled food
may contribute significantly to the intake of PAHs if such foods are a large part of the usual diet.
The presence of PAHs in coffee has also been reported and it has been suspected to be due to
either a contamination of green coffee beans during the drying step or an endogenous formation
in the coffee beans during the roasting process (Houessou et al., 2005; Houessou et al., 2008).
5. Occurrence and patterns of PAHs in food

With Commission Recommendation 2005/108/EC25 the Commission asked the Member States to
monitor PAHs, in particular the 15 priority substances identified by SCF, in the foods listed in
Regulation (EC) No 208/2005 and in other foods that can potentially contain high levels of
PAHs, such as dried fruits and food supplements. EFSA collected and evaluated these occurrence
data and reported the first findings in the report "Findings of the EFSA Data collection on
Polycyclic Aromatic Hydrocarbons in Food" of 29 June 2007. The following chapter gives an
update of these findings with additional data analysed until the end of 2007.
In response to the Commission request of 2005 for the monitoring of PAHs in food, 18 countries
submitted analytical results covering analyses of a variety of different food products mainly in
2005-2007, but in some cases including results from as early as 2000. The submission from
Hungary contained aggregated data only and even data in ranges so they could not be included in
the database. Some 38 results with a LOD above 1 µg/kg for any PAH (except
cyclopenta[cd]pyrene, see below) were also not included since they would have unduly
influenced calculations. Valid results included in the analyses comprised testing of 9,714
products as illustrated in Figure 2.
Germany submitted the most results consisting of 64% of the overall material. However, 2,584 of
these results provided only benzo[a]pyrene values, which have been used when looking at the
overall contamination level but cannot be used when comparing relationships between the
different PAHs.
25 OJ L 34, 8.2.2005, p. 43-45

7,000
6,249
6,000
Number of samples
5,000
4,000
3,000
2,000
912
1,000
369 362 510
203 342
45 162 65 189 101 43
35 43 75 9
0
Austria Cyprus Denmark Finland Germany Ireland Latvia Spain UK
Belgium Czech Estonia France Greece Italy Slovakia Sweden
Country
Figure 2: Number of results reported from the respective Member State.
The products were categorised into food categories relevant for the exposure calculations. The
product mix varied considerably between submissions from the different Member States. In
Figure 3 the foods sampled are presented in 13 broad food categories. Most samples belonged to
the fish and seafood category, followed by meat and meat products, and fats and oils. Sampling to
a large extent covered food sub-categories with MLs in legislation and there is thus not
necessarily a representative distribution of foods even within food categories. Some bias is also
possible related to the sampling method. The samples originated from both targeted and random
sampling.

2,989
3,000
2,500
Number of samples
2,205
2,145
2,000
1,500
1,000 847
500 365
261 310
229
117 111 84
4 47
0
10. Meat and meat

06. Fruits
13. Milk and dairy based
14. Miscellaneous/Food
07. Non-alcoholic
03. Fats and oils
02. Sugar and
pulses
tubers
01. Cereals & cereal
products
products
products
beverages
11. Fish and seafood

04. Vegetables, nuts,
for special use

confectionery
05. Starchy roots and
08. Coffee, tea, cocoa
09. Alcoholic beverages
Food category
Figure 3: Number of results reported in broad food categories.
The lower (LB) and upper (UB) bounds were used when calculating descriptive statistics, that is
for values below the LOD, a value of zero or the numerical value of the LOD was entered,
respectively. Some laboratories, but not all, reported values as less than LOD for samples with no
detectable levels and as less than LOQ for samples with non-quantifiable traces. Other
laboratories consistently reported either the LOD or the LOQ for any such case. LOD is most
often defined as three times the standard deviation of a blank or a low concentration sample while
LOQ is ten times the standard deviation or 3.3 times the LOD. To standardise the material the
LOQ was divided by the factor of 3.3 to be entered as the LOD for all samples with non-detected
or non-quantifiable concentrations and then treated the same way as above for calculating lower
and upper bounds. In Figure 4, reported or calculated LODs varied between 0.0002 and 1 μg/kg
for the respective PAH, except for cyclopenta[cd]pyrene with a maximum of 6 μg/kg (38 samples
were initially excluded in full because the LOD for either benz[a]anthracene, chrysene or
indeno[1,2,3-cd]pyrene was reported above 1 μg/kg). Also 322 individual results for
cyclopenta[cd]pyrene with a reported LOD above 6 µg/kg were marked as missing only for this
compound to not unduly influence the exposure assessment. There were four methods reported in
use – GC-FID, GC-MS, HPLC-FLD alone or in combination with HPLC-UV (see 3.2 Methods
of analysis). The lower LODs were reported when using HPLC except for cyclopenta[cd]pyrene
where the use of GC was more sensitive and the preferred method. The median LOD across all
PAHs was 0.091μg/kg with water at the lowest end of the spectrum. As an example, the median
for the benzo[a]pyrene detection level in water was 0.0003 μg/kg with the second lowest more
than ten times higher at 0.007 μg/kg. There is evidence that the quality of the methodology
employed is improving over time with increased sensitivity of the instrumentation.
10.0000
Concentration µg/kg (note log scale)
1.0000
0.1000
0.0100
0.0010
0.0001
BaP BbFA BghiP DBahA BjFA DBaeP DBaiP MCH
BaA BkFA CHR IP CPP DBahP DBalP BcFL
PAH compound
Figure 4: Limit of detections (LODs) for the different PAHs as reported by Member States. The
box indicates lower and upper quartiles with a line at the median. The whiskers indicate
minimum and maximum values with individual values marked for outliers (o) and extremes (*).
Analytical results reported for individual samples covered between one and 16 different PAHs
(Table 6). Only 1,375 of the food products were analysed for the full range of the 15 SCF priority
PAHs (abbreviated as PAH15) and 823 if also benzo[c]fluorene, as listed by the JECFA, is
included. Thus concentrations in Table 6 should not be compared between the different PAHs
since they represent different ranges of food categories. Of the 9,714 samples analysed for the
presence of one or more of the priority PAHs, 3,086 samples (31.8%) had no result for any PAH
analysed above the LOD. For only one of the PAHs, chrysene, did the proportion of results above
the LOD exceed 50% and as a consequence the median reflects in most cases only the actual
LOD. A maximum result of 1,064 µg/kg was recorded for benz[a]anthracene in a tinned sprats in
oil product and a second highest result of 690 µg/kg was recorded for benzo[b]fluoranthene in a
food supplement. The result for the sprats in oil product could be an aberration since no other

PAH in this sample had a high value, however, the food supplement also recorded the maximum
of 590 μg/kg for chrysene, and the highest value also for several other PAHs. No other products
came within half of the high values recorded for these two products.
Table 6: Descriptive statistics for the concentration (μg/kg) for up to 16 PAHs in 9,714 food
products.
Concentration in µg/kg
PAH N >LOD P05 Median Mean P95 Maximum
LB UB LB UB LB UB LB UB LB UB
BaP 9714 47.3% 0 0.007 0 0.12 1.02 1.08 3.60 3.60 270 270
BaA 5706 49.4% 0 0.010 0 0.15 1.97 2.06 6.00 6.00 1064 1064
BbFA 6728 47.4% 0 0.003 0 0.15 1.48 1.55 6.00 6.00 690 690
BkFA 6630 41.5% 0 0.001 0 0.10 0.57 0.62 2.44 2.44 150 150
BghiP 6406 35.8% 0 0.001 0 0.10 0.77 0.84 2.90 2.90 220 220
CHR 5421 60.8% 0 0.020 0.13 0.30 3.20 3.27 10.00 10.00 590 590
DBahA 6119 14.0% 0 0.010 0 0.09 0.12 0.20 0.53 0.54 37 37
IP 6908 24.8% 0 0.010 0 0.10 0.56 0.69 2.49 2.49 100 100
BjFA 2184 38.8% 0 0.010 0 0.05 0.58 0.66 2.43 2.43 57 57
CPP 1641 40.0% 0 0.010 0 0.09 1.00 1.36 1.96 6.00 112 112
DBaeP 2538 12.4% 0 0.050 0 0.10 0.16 0.28 0.9 1.00 27 27
DBahP 2341 2.2% 0 0.010 0 0.10 0.02 0.15 0 0.49 3 3
DBaiP 2471 4.9% 0 0.033 0 0.10 0.04 0.17 0 0.91 4 4
DBalP 2702 8.1% 0 0.010 0 0.10 0.06 0.16 0.30 0.68 14 14
MCH 2227 6.9% 0 0.010 0 0.04 0.07 0.15 0.10 0.45 18 18
BcFL 1148 46.5% 0 0.010 0 0.10 1.60 1.65 6.53 6.53 231 231
N: Number of samples, LOD: Limit of detection
A comparison was also made for the 1,375 food products analysed for all the 15 SCF priority
PAHs (Table 7). The highest mean value was recorded for chrysene with two further PAHs,
benz[a]anthracene and benzo[b]fluoranthene recording mean values higher than the mean value
for benzo[a]pyrene.

Table 7: Descriptive statistics for the concentration in μg/kg for 1,375 food products analysed for
all of the 15 SCF priority PAHs.
Concentration in μg/kg
PAH N >LOD P05 Median Mean P95 Maximum
LB UB LB UB LB UB LB UB LB UB
BaP 1375 54.9% 0 0.01 0.04 0.07 0.79 0.81 3.23 3.23 67 67
BaA 1375 70.3% 0 0.01 0.08 0.09 1.33 1.35 4.59 4.59 147 147
BbFA 1375 63.1% 0 0.01 0.05 0.08 1.40 1.42 6.43 6.43 116 116
BkFA 1375 52.7% 0 0.01 0.01 0.05 0.60 0.62 2.90 2.90 52 52
BghiP 1375 56.4% 0 0.01 0.02 0.05 0.63 0.65 2.82 2.82 38 38
CHR 1375 77.8% 0 0.01 0.17 0.19 2.79 2.81 7.90 7.90 353 353
DBahA 1375 18.3% 0 0.01 0 0.03 0.15 0.18 0.72 0.72 10 10
IP 1375 36.7% 0 0.01 0 0.05 0.57 0.59 2.40 2.40 45 45
BjFA 1375 51.9% 0 0.01 0.01 0.05 0.70 0.72 2.74 2.74 57 57
CPP 1375 39.3% 0 0.01 0 0.05 0.68 1.01 1.10 2.95 112 112
DBaeP 1375 12.2% 0 0.03 0 0.10 0.11 0.20 0.78 0.78 6 6
DBahP 1375 3.6% 0 0.01 0 0.10 0.03 0.12 0 0.20 3 3
DBaiP 1375 7.6% 0 0.03 0 0.10 0.06 0.15 0.24 0.35 3 3
DBalP 1375 10.3% 0 0.01 0 0.10 0.08 0.16 0.57 0.57 14 14
MCH 1375 4.0% 0 0.01 0 0.01 0.04 0.07 0 0.15 17 17
Total 0 0.19 0.38 1.12 9.96 10.86 36.42 38.73 1040 1040
Using the lower or the upper bound had a little impact on the mean and the 95th percentile values
with an average increase for the upper bound compared to the lower bound of 9% and 6%,
respectively. The difference was larger for the median.
The proportions of the different PAHs in relation to the sum of the total concentration of 15 SCF
priority PAHs were calculated for mean values from Table 7 for the total of the 1,375 food
products (Figure 5). Chrysene comprised more than a quarter of the total mean concentration
followed by benzo[b]fluoranthene, benz[a]anthracene, cyclopenta[cd]pyrene and benzo[a]pyrene
at 13%, 12%, 9% and 8%, respectively for the upper bound results with very similar proportions
for lower bound results.

30.0%
Relative proportion to total PAH
25.0%
20.0%
15.0%
10.0%
5.0%
Upper bound
Lower bound
0.0%
BaA
BbFA
BkFA
BjFA
DBahA
MCH
IP
BaP
BghiP
CPP
DBaeP
DBahP
DBaiP
DBalP
CHR
Figure 5: Relative proportion for the mean concentrations of the different PAHs for 1,375 food
products analysed for all of the 15 SCF priority PAHs.
As in about 30% of the samples analysed for all 15 SCF priority PAHs other carcinogenic and
genotoxic PAHs were detected despite testing negative for benzo[a]pyrene, individual
compounds were grouped and summed in order to check whether their sums would better reflect
the occurrence of carcinogenic and genotoxic PAHs in different food categories. The selection of
the individual PAHs was based on the frequency of their results above the LOD. Results covering
at least the eight carcinogenic and genotoxic PAHs measured in the coal tar mixtures studied by
Culp et al. (1998) (benzo[a]pyrene, benz[a]anthracene, benzo[b]fluoranthene,
benzo[k]fluoranthene, benzo[ghi]perylene, chrysene, dibenz[a,h]anthracene and indeno[1,2,3-
cd]pyrene, referred in this opinion as PAH8) were reported in 4,065 samples. The mean
concentrations of these eight PAHs were calculated for each of the 33 food
categories/subcategories for 4,065 food products (Table 8). Of the total number of samples, 24%
had no detectable levels of any of the eight PAHs.
Besides the sum of the above mentioned eight PAHs, the sum of benzo[a]pyrene, chrysene,
benz[a]anthracene and benzo[b]fluoranthene (referred in this opinion as PAH4) as well as the
sum of benzo[a]pyrene and chrysene (referred in this opinion as PAH2) were calculated to get
respective PAH loads for each category and subcategory which can then be used for the exposure
calculations. The mean concentrations, calculated as lower bound and upper bound values, for the

three sums and the corresponding data for benzo[a]pyrene in the different food categories are
shown in Table 9.

Table 8: Lower (LB) and upper (UB) bounds for the mean of the concentration in µg/kg of those eight individual PAHs that were identified in coal tar mixtures, and the proportion of samples above the limit of detection
(>LOD) for the respective PAHs in 4,065 food products.
BaP BaA BbFA BkFA BghiP CHR DBahA IP
N LB UB >LOD LB UB >LOD LB UB >LOD LB UB >LOD LB UB >LOD LB UB >LOD LB UB >LOD LB UB >LOD
Cereals 25 0.20 0.40 40% 0.41 0.58 56% 0.26 0.33 52% 0.07 0.14 44% 0.63 0.69 56% 0.30 0.36 56% 0.10 0.19 4% 0.01 0.40 8%
Processed 9 0.03 0.26 44% 0.15 0.32 67% 0.13 0.18 67% 0.02 0.09 56% 0.03 0.10 44% 0.20 0.25 67% 0.00 0.08 0% 0.02 0.26 22%
Unprocessed 16 0.30 0.47 38% 0.56 0.73 50% 0.34 0.41 44% 0.10 0.17 38% 0.97 1.02 63% 0.35 0.42 50% 0.15 0.25 6% 0.00 0.49 0%
Chocolate 148 0.30 0.32 83% 0.35 0.38 60% 0.39 0.44 76% 0.15 0.17 74% 0.43 0.45 68% 0.57 0.62 83% 0.00 0.05 1% 0.18 0.33 30%
Fats and oil 1140 0.97 1.00 56% 1.23 1.28 52% 1.26 1.37 57% 0.52 0.57 52% 0.75 0.83 41% 3.59 3.69 66% 0.12 0.20 19% 0.68 0.86 27%
Cocoa butter 128 1.99 1.99 96% 2.92 2.93 89% 3.37 3.37 97% 1.60 1.61 91% 1.94 1.96 84% 3.76 3.78 87% 0.24 0.28 49% 2.55 2.74 64%
Other 899 0.65 0.69 52% 0.74 0.80 45% 0.75 0.89 52% 0.30 0.36 46% 0.43 0.53 36% 2.14 2.26 65% 0.07 0.16 13% 0.38 0.59 22%
Pomace oil 113 2.38 2.39 44% 3.24 3.24 62% 2.91 2.92 50% 1.05 1.06 50% 1.91 1.96 30% 14.97 14.99 52% 0.36 0.38 37% 0.93 0.96 26%
Vegetables and nuts 80 0.16 0.26 26% 0.14 0.23 29% 0.21 0.27 36% 0.08 0.16 29% 0.11 0.27 21% 0.31 0.38 30% 0.00 0.15 1% 0.11 0.23 20%
Dried 40 0.03 0.15 13% 0.02 0.11 10% 0.07 0.15 20% 0.02 0.14 10% 0.00 0.24 5% 0.07 0.15 13% 0.00 0.23 0% 0.01 0.14 8%
Other 40 0.29 0.38 40% 0.25 0.35 48% 0.34 0.38 53% 0.14 0.18 48% 0.22 0.29 38% 0.55 0.61 48% 0.00 0.08 3% 0.20 0.32 33%
Fruit 134 0.13 0.17 36% 0.91 0.95 59% 0.27 0.32 57% 0.07 0.11 44% 0.07 0.12 39% 1.19 1.23 66% 0.00 0.08 1% 0.03 0.11 17%
Dried 133 0.13 0.17 36% 0.92 0.95 59% 0.28 0.32 57% 0.07 0.11 44% 0.07 0.12 39% 1.20 1.23 67% 0.00 0.08 2% 0.03 0.11 17%
Processed 1 0.00 0.05 0% 0.00 0.05 0% 0.00 0.05 0% 0.00 0.05 0% 0.00 0.10 0% 0.00 0.05 0% 0.00 0.05 0% 0.00 0.05 0%
Coffee, tea and cocoa 63 5.02 5.12 57% 5.21 5.28 60% 5.86 5.93 59% 2.58 2.64 57% 4.39 4.45 59% 9.55 9.61 60% 0.19 0.29 24% 3.97 4.07 56%
Cocoa powder 3 0.02 0.35 33% 1.19 1.36 67% 0.04 0.15 33% 0.01 0.11 33% 1.69 1.74 67% 0.08 0.18 33% 0.00 0.11 0% 0.00 0.34 0%
Coffee powder 28 2.33 2.46 21% 3.09 3.22 25% 2.57 2.69 25% 1.20 1.32 21% 1.98 2.09 25% 3.76 3.87 29% 0.04 0.19 4% 1.96 2.10 21%
Dried tea 30 8.37 8.38 97% 7.90 7.91 93% 9.91 9.91 97% 4.30 4.30 97% 7.20 7.21 93% 16.48 16.49 93% 0.36 0.42 47% 6.51 6.51 97%
Other 2 0.00 0.50 0% 0.42 0.67 50% 0.00 0.16 0% 0.00 0.15 0% 0.00 0.15 0% 0.80 0.88 50% 0.00 0.15 0% 0.00 0.50 0%
Alcoholic beverages 30 0.00 0.01 3% 0.00 0.01 7% 0.00 0.01 3% 0.00 0.01 3% 0.00 0.03 0% 0.01 0.02 30% 0.00 0.03 0% 0.00 0.04 0%
Meat 777 0.29 0.32 41% 0.39 0.43 40% 0.20 0.23 32% 0.14 0.16 40% 0.15 0.21 21% 0.47 0.49 61% 0.03 0.09 4% 0.10 0.17 12%
Barbequed meat 39 1.92 1.92 100% 0.70 0.72 38% 1.86 1.87 77% 0.56 0.57 74% 0.71 0.73 44% 0.72 0.74 36% 0.51 0.53 41% 0.86 0.87 64%
Grilled meat 53 0.61 0.63 57% 0.56 0.57 68% 0.32 0.33 62% 0.15 0.17 32% 0.49 0.53 36% 0.79 0.79 68% 0.01 0.07 4% 0.35 0.39 38%
Other 123 0.05 0.10 27% 0.06 0.11 37% 0.04 0.10 46% 0.02 0.06 34% 0.03 0.10 17% 0.10 0.15 56% 0.00 0.06 1% 0.02 0.11 5%
Smoked meat 562 0.20 0.23 39% 0.42 0.47 37% 0.10 0.14 23% 0.13 0.15 39% 0.11 0.16 19% 0.50 0.52 63% 0.01 0.07 2% 0.04 0.12 7%
Fish and seafood 1031 1.26 1.30 51% 3.96 4.01 54% 1.77 1.84 54% 0.69 0.75 45% 0.57 0.64 38% 4.24 4.31 62% 0.17 0.25 19% 0.58 0.71 32%
Cephalopods 5 0.00 0.10 0% 0.00 0.09 0% 0.00 0.09 20% 0.00 0.09 0% 0.00 0.09 20% 0.01 0.10 20% 0.00 0.09 0% 0.00 0.09 0%
Fresh fish 69 0.02 0.11 7% 0.01 0.11 10% 0.02 0.13 14% 0.01 0.11 13% 0.01 0.12 1% 0.03 0.13 23% 0.00 0.11 0% 0.01 0.24 1%
All processed fish 757 1.36 1.41 45% 4.93 4.98 49% 1.41 1.50 48% 0.44 0.50 37% 0.40 0.48 29% 4.76 4.85 59% 0.15 0.24 9% 0.49 0.64 21%
Preserved fish 421 2.21 2.26 49% 8.42 8.47 46% 2.30 2.39 50% 0.69 0.77 40% 0.58 0.68 27% 7.97 8.05 56% 0.24 0.35 11% 0.76 0.95 23%
Smoked fish 336 0.29 0.34 40% 0.56 0.61 52% 0.30 0.39 46% 0.13 0.18 33% 0.16 0.22 30% 0.75 0.84 63% 0.03 0.09 6% 0.14 0.26 19%
All bivalve molluscs 200 1.34 1.35 88% 1.75 1.77 88% 3.75 3.77 90% 1.89 1.90 90% 1.42 1.44 89% 3.81 3.82 88% 0.32 0.37 64% 1.15 1.16 86%
Fresh bivalve molluscs 187 1.34 1.36 87% 1.75 1.77 88% 3.81 3.82 89% 1.92 1.93 89% 1.43 1.45 88% 3.78 3.80 88% 0.33 0.37 64% 1.16 1.17 86%
Smoked bivalve molluscs 13 1.28 1.28 100% 1.77 1.77 92% 2.99 2.99 100% 1.41 1.41 100% 1.24 1.24 100% 4.15 4.16 92% 0.25 0.28 54% 1.00 1.01 92%
Dairy 61 0.08 0.13 43% 0.06 0.12 23% 0.04 0.10 26% 0.04 0.08 39% 0.01 0.09 13% 0.10 0.14 39% 0.00 0.08 0% 0.00 0.11 3%
Other dairy 29 0.03 0.12 10% 0.05 0.14 14% 0.04 0.12 21% 0.01 0.09 10% 0.00 0.11 0% 0.05 0.13 28% 0.00 0.11 0% 0.00 0.16 0%
Smoked cheese 32 0.13 0.14 72% 0.07 0.10 31% 0.05 0.07 31% 0.06 0.07 66% 0.03 0.07 25% 0.14 0.15 50% 0.00 0.05 0% 0.01 0.06 6%
Miscellaneous 576 1.45 1.48 39% 2.77 2.80 55% 2.84 2.86 58% 0.89 0.91 42% 1.16 1.20 50% 5.12 5.13 74% 0.25 0.29 10% 0.91 0.96 28%
Food supplements 283 2.78 2.82 48% 5.31 5.34 64% 5.61 5.63 64% 1.73 1.76 53% 2.29 2.32 59% 10.03 10.05 74% 0.49 0.54 19% 1.80 1.85 39%
Other 13 0.27 0.46 62% 0.22 0.42 38% 0.40 0.46 62% 0.17 0.23 46% 0.21 0.28 54% 0.42 0.48 54% 0.00 0.10 8% 0.16 0.38 31%
Spices 10 3.50 3.51 60% 8.21 8.24 70% 3.54 3.55 80% 1.89 1.91 60% 1.50 1.53 60% 7.94 7.98 30% 0.39 0.42 10% 1.23 1.27 50%
Infant food 270 0.02 0.05 27% 0.03 0.05 46% 0.04 0.05 50% 0.01 0.03 29% 0.02 0.05 41% 0.09 0.10 77% 0.00 0.03 0% 0.01 0.05 16%
Total 4065
N: Number of samples
Of the food product categories/subcategories, benzo[a]pyrene was consistently found in barbequed meat, dried tea, cocoa butter, chocolate and bivalve molluscs. The high concentrations found in spices will not impact much
on the exposure assessment since spices are consumed in low amounts. This is equally true for food supplements. The high concentrations found in cocoa butter are equivalent to the levels found in chocolate if the amount of
cocoa butter in chocolate is considered since chocolate must contain at least 18% cocoa butter.

Table 9: Lower (LB) and upper (UB) bounds for the mean of the concentration in µg/kg of
benzo[a]pyrene (BaP), benzo[a]pyrene+chrysene (PAH2), benzo[a]pyrene+chrysene+
benz[a]anthracene+benzo[b]fluoranthene (PAH4) and PAH4+benzo[k]fluoranthene+
benzo[ghi]perylene+dibenz[a,h]antracene+indeno[1,2,3-cd]pyrene (PAH8) in Table 8 in 4,065
food products.
BaP PAH2 PAH4 PAH8
Food group/subgroup N LB UB LB UB LB UB LB UB
Cereals 25 0.20 0.40 0.50 0.75 1.17 1.66 1.97 3.08
Processed 9 0.03 0.26 0.23 0.50 0.51 1.00 0.57 1.53
Unprocessed 16 0.30 0.47 0.65 0.89 1.54 2.03 2.76 3.96
Chocolate 148 0.30 0.32 0.87 0.93 1.61 1.75 2.37 2.76
Fats and oil 1140 0.97 1.00 4.56 4.70 7.04 7.35 9.10 9.81
Cocoa butter 128 1.99 1.99 5.75 5.77 12.03 12.07 18.36 18.65
Other 899 0.65 0.69 2.78 2.95 4.27 4.65 5.44 6.28
Pomace oil 113 2.38 2.39 17.35 17.37 23.50 23.54 27.74 27.90
Vegetables and nuts 80 0.16 0.26 0.47 0.64 0.82 1.14 1.12 1.95
Dried 40 0.03 0.15 0.11 0.30 0.20 0.55 0.24 1.30
Other 40 0.29 0.38 0.84 0.99 1.43 1.72 2.00 2.60
Fruit 134 0.13 0.17 1.32 1.40 2.51 2.66 2.68 3.08
Dried 133 0.13 0.17 1.33 1.41 2.53 2.68 2.71 3.10
Processed 1 0.00 0.05 0.00 0.10 0.00 0.20 0.00 0.45
Coffee, tea and cocoa 63 5.02 5.12 14.57 14.72 25.64 25.94 36.77 37.39
Cocoa powder 3 0.02 0.35 0.10 0.53 1.33 2.03 3.02 4.33
Coffee powder 28 2.33 2.46 6.09 6.33 11.76 12.24 16.95 17.94
Dried tea 30 8.37 8.38 24.85 24.87 42.66 42.69 61.02 61.14
Other 2 0.00 0.50 0.80 1.38 1.22 2.20 1.22 3.16
Alcoholic beverages 30 0.00 0.01 0.01 0.03 0.01 0.06 0.01 0.18
Meat 777 0.29 0.32 0.76 0.81 1.34 1.48 1.77 2.11
Barbequed meat 39 1.92 1.92 2.64 2.66 5.20 5.25 7.84 7.96
Grilled meat 53 0.61 0.63 1.40 1.42 2.28 2.32 3.29 3.48
Other 123 0.05 0.10 0.15 0.25 0.25 0.46 0.32 0.79
Smoked meat 562 0.20 0.23 0.70 0.75 1.23 1.36 1.52 1.86
Fish and seafood 1031 1.26 1.30 5.49 5.61 11.22 11.46 13.23 13.82
Cephalopods 5 0.00 0.10 0.01 0.20 0.01 0.39 0.01 0.77
Fresh fish 69 0.02 0.11 0.04 0.24 0.08 0.48 0.11 1.05
All processed fish 757 1.36 1.41 6.12 6.25 12.46 12.74 13.93 14.60
Preserved fish 421 2.21 2.26 10.17 10.31 20.89 21.17 23.17 23.92
Smoked fish 336 0.29 0.34 1.04 1.17 1.90 2.17 2.36 2.93
All bivalve molluscs 200 1.34 1.35 5.14 5.18 10.65 10.71 15.42 15.58
Fresh bivalve molluscs 187 1.34 1.36 5.12 5.16 10.68 10.75 15.52 15.68
Smoked bivalve molluscs 13 1.28 1.28 5.43 5.44 10.19 10.20 14.09 14.14
Dairy 61 0.08 0.13 0.18 0.27 0.28 0.49 0.34 0.84
Other dairy 29 0.03 0.12 0.08 0.25 0.16 0.51 0.17 0.98
Smoked cheese 32 0.13 0.14 0.27 0.29 0.39 0.47 0.48 0.72
Miscellaneous 576 1.45 1.48 6.56 6.61 12.18 12.28 15.39 15.64
Food supplements 283 2.78 2.82 12.81 12.87 23.73 23.84 30.03 30.31
Other 13 0.27 0.46 0.69 0.94 1.31 1.83 1.85 2.81
Spices 10 3.50 3.51 11.43 11.49 23.18 23.27 28.19 28.40
Infant food 270 0.02 0.05 0.12 0.15 0.19 0.25 0.23 0.41
Sample mean 4065 0.95 0.99 4.06 4.16 7.37 7.58 9.21 9.73
Group mean 0.90 0.96 3.21 3.32 5.80 6.02 7.70 8.24

To better illustrate high values, the concentrations for PAH8 split into the 33 food
categories/subcategories have been plotted in Figure 6.
70
60
Concentration of eight PAHs µg/kg
50
40
30
20
10
Upper bound
Low er bound
0
Smoked meat
Pomace oil
- Smoked bivalve molluscs
Smoked cheese
Grilled meat
Other meat
Food supplements
Other miscellaneous
- Smoked fish
Other vegetables and
Chocolate
Dried tea
Barbequed meat
Cocoa butter
- Bivalve molluscs
Dried vegetables and nuts
Other dairy
Cephalopods
Processed cereals
Unprocessed cereals
Fresh fish
- Preserved fish
Dried fruit
Processed fruit
Other fats and oils
All bivalve molluscs
Spices
Other coffee
Infant food
Cocoa powder
Alcoholic beverages
All processed fish

Coffee powder
Figure 6: Lower and upper bounds for the mean of the sum of the eight PAHs in 4,065 samples
split across 33 food categories/subcategories (note grouping of fish and bivalve molluscs).
Nine food categories (not counting the fish and seafood groupings) had upper bound values for
the PAH8 exceeding 10 µg/kg. The 30 tealeaf samples varied between 1 and 173 µg/kg with 20%
exceeding 100 µg/kg. However, preparation of the tea will dilute the level found in the beverage
before consumption. Barbecued meat had more than double the contamination level of grilled
meat with smoked meat products in third place. Modern smoking methods are designed to avoid
PAH contamination and in some cases artificial smoke might have been used reducing
contamination levels even further. Information on the use of artificial smoke was not provided.
Fresh fish had very little PAH contamination, while the bivalve mollusc average exceeded 15
µg/kg. However, processed fish had approximately a 50% higher mean than the bivalve mollusc
average. A closer look at this subcategory revealed some very high contamination levels in sprats

in oil. Out of 155 samples tested, 46% exceeded 10 µg/kg, 13% exceeded 100 µg/kg with a
maximum combined sum of 1,086 µg/kg in one sample. The highest contribution to the sum in
this last case came from benz[a]anthracene.
To further study the relationships between the different PAHs, the relative contributions of each
of the individual PAH8 compounds to the total concentration of PAH8 were calculated per food
category/subcategory for upper bound mean values and are illustrated in Table 10 and Figure 7.
The lowest and highest relative contributions in any food category are also indicated in Table 10.
For comparison, the Table also contains the percentage of these eight PAHs compounds that were
measured in the two coal tar mixtures used in the carcinogenicity studies of Culp et al. (1998),
which provided the basis of the SCF and JECFA risk assessments.
Table 10: Summary of relative contributions of upper bound mean concentrations for the
individual PAH8 for 4,065 food samples and the composition of the two coal tar mixtures.
% in the
Overall Lowest Highest
Compound Subcategory Subcategory coal tar
average average average
mixtures
Benzo[a]pyrene 10% 6% Dried fruit 24% Barbequed meat 15-16%
Benz[a]anthracene 20% 8% Alcoholic beverages 35% Processed fish 19%
Benzo[b]fluoranthene 15% 3% Cocoa powder 24% Bivalve molluscs 16-17%
Benzo[k]fluoranthene 6% 2% Cocoa powder 12% Bivalve molluscs 6%
Benzo[ghi]perylene 7% 3% Processed fish 40% Cocoa powder 12-13%
Chrysene 33% 4% Cocoa powder 54% Pomace oil 17-19%
Dibenz[a,h]anthracene 2% 1% Dried tea 18% Dried vegetables 2%
Indeno[1,2,3-cd]pyrene 7% 3% Pomace oil 24% Alcoholic beverages 11%

Cereals processed (9)

Cereals unprocessed (16)
Chocolate (148)
Cocoa butter (128)
Other fats and oils (899)
Pomace oil (113)
Dried vegetables and nuts (40)
Other vegetables and nuts
Dried fruit (133)
Processed fruit (1)
Cocoa pow der (3)
Coffee pow der (28)
Dried tea (30)
Other coffee (2)
Alcoholic beverages (30)
Barbecued meat (39)
Grilled meat (53)
Other meat (123)
Smoked meat (562)
Cephalopods (5)
Fresh fish (69)
All processed fish (757)
- Preserved fish (421)
- Smoked fish (336)
All bivalve molluscs (200)
- Bivalve molluscs (187)
- Smoked bivalve molluscs (13)
Other dairy (29)
Smoked cheese (32)
Food supplements (283)
Other miscellaneous (13)
Spices (10)
Infant foods (270)
Sample mean (4065)
Group mean
Coal tar mix 1
Coal tar mix 2
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
BaP BaA BbFA BkFA BghiP CHR DBahA IP
Figure 7: Relative contribution calculated for mean upper bound concentrations of eight PAHs in
4,065 samples across 33 food categories/subcategories (number of samples in brackets) and in the
two coal tar mixtures (food groupings as in Figure 6).
Chrysene is a dominating component with the highest average of 33% followed by

benz[a]anthracene at 20%. Dibenz[a,h]anthracene at 2% had the lowest average presence
followed by benzo[k]fluoranthene at 6%. Compared to the coal tar mixtures the largest overall
deviation was found for chrysene. A mean for the relative contribution of the eight PAHs was
calculated as an average across all samples (“sample mean”). In addition, an average of the mean
values for the respective food category/subcategory was calculated (“group mean”). The sample
mean is biased by the number of samples tested in each food category/subcategory while the
group mean is not.
Although much less accurate because of the low or missing sample numbers in some categories a
comparison was made for 1,375 samples analysed for all SCF priority PAHs to also check the
relative concentration of the remaining seven PAHs other than the PAH8. The concentrations of
the respective PAH or combination of PAHs is presented in Table 11 and Figure 8 for lower

bound and Table 11 and Figure 9 for upper bound values. Food categories/subcategories with less
than five sample results are not shown but are included in the overall sample mean.
Table 11: Lower (LB) and upper (UB) bound concentrations in µg/kg for the respective PAH
and PAH combinations. Sample and group means are calculated for all samples or subgroups,
respectively, for 1,375 sample results.
BaP PAH2 PAH4 PAH8 PAH15
Subgroup N LB UB LB UB LB UB LB UB LB UB
Processed cereal 5 0.06 0.06 0.19 0.20 0.32 0.33 0.43 0.51 0.49 0.99
Chocolate 17 0.17 0.17 0.67 0.67 1.13 1.13 1.45 1.49 1.75 2.19
Other fats and oils 121 0.39 0.39 2.72 2.73 3.62 3.65 4.58 4.69 6.93 10.05
Dried vegetables 20 0.06 0.14 0.17 0.33 0.23 0.58 0.26 1.03 0.27 1.83
Other vegetables 11 0.02 0.04 0.18 0.28 0.29 0.42 0.33 0.59 0.36 1.01
Dried fruit 67 0.06 0.11 0.62 0.71 0.96 1.14 1.06 1.44 1.17 2.19
Barbequed meat 31 2.27 2.27 3.02 3.04 5.80 5.86 8.91 9.02 12.18 12.48
Grilled meat 21 1.12 1.14 1.96 1.99 3.23 3.27 5.25 5.34 11.27 11.77
Other meat 49 0.09 0.11 0.28 0.30 0.45 0.50 0.60 0.70 0.96 1.43
Smoked meat 167 0.23 0.24 0.98 1.00 1.84 1.88 2.13 2.23 2.97 3.61
Fresh fish 6 0.00 0.18 0.10 0.42 0.12 0.77 0.12 1.50 0.19 2.82
Molluscs 167 1.50 1.50 5.72 5.72 11.91 11.92 17.31 17.36 19.57 20.14
Processed fish 29 2.20 2.27 8.52 8.70 16.11 16.42 18.71 19.51 31.90 33.86
Smoked fish 153 0.33 0.34 1.35 1.37 2.32 2.34 2.78 2.86 3.51 4.03
Smoked molluscs 11 1.43 1.43 6.24 6.24 11.52 11.52 15.70 15.73 18.23 18.65
Other dairy 17 0.01 0.09 0.01 0.18 0.04 0.37 0.04 0.71 0.04 1.42
Smoked cheese 11 0.04 0.04 0.06 0.07 0.19 0.20 0.27 0.31 0.27 2.10
Food supplements 232 2.16 2.19 11.65 11.69 19.55 19.62 24.85 25.03 28.00 28.63
Infant food 213 0.03 0.04 0.13 0.15 0.20 0.23 0.25 0.33 0.33 0.79
Spices 7 0.71 0.71 1.19 1.23 5.80 5.85 7.44 7.55 8.64 10.29
Sample mean 1375 0.79 0.81 3.58 3.62 6.32 6.38 8.26 8.42 9.97 10.86
Group mean 0.57 0.60 2.09 2.14 3.91 4.01 5.09 5.32 6.64 7.63
N: Number of samples

Processed cereal (5)

Chocolate (17)
Dried vegetables (20)
Other vegetables (11)
Dried fruit (67)
Barbecued meat (31)
Grilled meat (21)
Other meat (49)
Smoked meat (167)
Fresh fish (6)
Bivalve molluscs (167)
Preserved fish (29)
Smoked fish (153)
Smoked bivalve molluscs (11)
Other dairy (17)
Smoked cheese (11)
Infant food (213)
Spices (7)
Sample mean (1375)
Group mean
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
BaP PAH2-BaP PAH4-PAH2 PAH8-PAH4 PAH15-PAH8
Figure 8: Mean lower bound relative contribution of one, two, four, eight and 15 PAHs across 26
food categories/subcategories for 1,375 samples (number of samples in brackets). PAH2-BaP
indicates the additive component to reach PAH2 etcetera for each grouping.
Processed cereal (5)

Chocolate (17)
Dried vegetables (20)
Other vegetables (11)
Dried fruit (67)
Barbecued meat (31)
Grilled meat (21)
Other meat (49)
Smoked meat (167)
Fresh fish (6)
Bivalve molluscs (167)
Preserved fish (29)
Smoked fish (153)
Smoked bivalve molluscs (11)
Other dairy (17)
Smoked cheese (11)
Infant food (213)
Spices (7)
Sample mean (1375)
Group mean
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
BaP PAH2-BaP PAH4-PAH2 PAH8-PAH4 PAH15-PAH8
Figure 9: Mean upper bound relative contribution of one, two, four, eight and 15 PAHs across 26
food categories/subcategories for 1,375 samples (number of samples in brackets). PAH2-BaP
indicates the additive component to reach PAH2 etcetera for each grouping. The upper bound
values will be influenced by sample results below the LOD.

The proportion of PAHs other than the PAH8 varied from 0% to 53% for the lower bound results
and from 13% to 85% for the upper bound results. Both a sample mean, that is the mean across
all samples, and a group mean, that is the mean of the group/subgroup results, were calculated.
For the lower bound results the combined sum of the seven PAHs other than PAH8 comprised
between 17% (sample mean) and 23% (group mean). For the upper bound results the equivalent
numbers were 22% and 30%, respectively. The number of samples in each food group (shown in
brackets beside the names in the two figures) will influence the sample mean, while the group
mean is independent of the sample number in each group. Further, the upper bound results will be
influenced by differences in the LOD for different PAHs and the number of sample results below
the LOD. This is particularly true for the seven PAHs other than PAH8 that in several cases will
be unduly influenced by the large majority of samples below the LOD (see Table 7). Thus, as a
best measure based on the lower bound the relative proportion of the combined sum of the seven
PAHs other than PAH8 can be estimated at about 20%.
There was not a sufficient number of benzo[c]fluorene data submitted to allow a detailed
analysis. However, it was noted from Table 6 that benzo[c]fluorene showed the third highest
mean result and the second highest 95th percentile results indicating some very high values with
the fourth highest maximum value recorded.
5.1 Factors influencing the levels of PAHs in food

There is a variety of literature on the factors that influence the levels of PAHs in food due to
commercial or home cooking processing techniques. The following summary gives a short
overview of the key findings.
5.1.1 Commercial processing techniques

Traditional commercial smoking techniques, in which smoke from incomplete wood burning
comes into direct contact with the product, can lead to its considerable contamination with
various PAHs if the process is not adequately controlled. Critical parameters include temperature,
time, humidity, types of control and smoke used (natural or generated), and the design and types
of smokehouses or kilns. Traditional smokehouses or kilns allow natural airflow or convection,
and their use requires a lot of expertise. Modern smokehouses or kilns control the airflow
mechanically or electrically and generally have temperature monitors and controls; and
microprocessor control systems that can regulate smoke exposure, humidity, drying, and
temperature. This permits strict control of the complex smoking operation. Wood smoke can be
generated by burning wood or, more commonly, by heating sawdust or small wood chips. A wide
variety of woods has been used for smoking, including oak, hickory, mahogany, pine,

whitewood, cherry, apple, alder, mesquite, beech, birch, and maple to impart various flavours and
colours (Jahncke and Herman, 2001).
In the past two decades the traditional smoking process in commercial food production has been
increasingly replaced by the use of liquid smoke flavourings. Because smoke flavourings are
produced from smoke that is subjected to fractionation and purification processes their use is
generally considered to result in lower PAH levels and thus to be of less health concern compared
to the traditional smoking. Other benefits derived from smoke flavourings besides the possibility
of better controlling the PAH content of the smoked products are flavour reproducibility, uniform
distribution of flavour throughout the product, intensifying the flavour of traditionally smoked
foods, application to a wide range of foodstuffs, savings in costs because wood and smoking
equipment are not required, less environmental pollution associated with the use of smoke
flavourings; and a variety of application methods such as spraying on the surface, dipping, or
mixing with the food (Guillén et al., 2000). Moreover, the concentrations of certain PAHs are
regulated by the European Commission.
The influence of different smoking techniques on the PAH levels in the processed food
commodities has been investigated by numerous authors. Karl and Leinemann (1996) studied the
influence of the type of smoking kilns on the PAH levels in processed products by determining
the levels of 13 PAHs in smoked fishery products from modern smoking kilns with external
smoke generation and from traditional smoking kilns. The average benzo[a]pyrene concentration
of all 35 samples from commercial smoking kilns with external smoke generation was 0.1 µg/kg
(wet weight) and the sum of the carcinogenic compounds determined in the study, i.e.
benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene,
dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene, did not exceed 4.5 µg/kg (wet weight). The
benzo[a]pyrene levels of the 27 smoked fish samples from traditional kilns ranged from 0.2
µg/kg to 4.1 µg/kg, with a mean value of 1.2 µg/kg. The average concentration of the sum of the
carcinogenic compounds was 9.0 µg/kg. Large variations were found in the content of the non-
carcinogenic PAHs phenanthrene, anthracene, fluoranthene and pyrene in all samples from both
types of smoking kilns.
Comprehensive research programmes in Latvia inter alia showed that there is a difference of up
to 6 times in PAH concentrations in smoked fish which have been processed by applying
woodchips of different trees for smoking. The highest concentrations were found when using
spruce, hazel-tree, plum-tree and aspen, whereas the lowest concentrations were found when
apple-tree, alder and maple were used for smoking. Moreover, it was demonstrated that in canned
smoked fish a benzo[a]pyrene diffusion occurs from the smoked fish to the oil fraction resulting
in an average reduction of 73.1% in the fish fraction. Finally, tests showed that the
benzo[a]pyrene content in semi-processed fish (smoked and beheaded) and in finished smoked
fish in cans at one and the same establishment, in the same lot of products, under the same

technological smoking regime and in one smoking container can vary between 0.1 and 10
µg/kg26.
Chen and Lin (1997) studied the effects of the processing methods, smoking, roasting, steaming,
charcoal grilling, and liquid smoke flavouring (LSF), on the formation of PAHs in duck breast
steak. Results showed that with processing time from 0.5 to 1.5 hours, charcoal grilling of duck
samples with skin contained the highest amount of total PAHs, followed by charcoal grilling of
duck samples without skin, smoking, roasting, steaming, and LSF. For carcinogenic PAHs,
smoking contained the highest amount, followed by charcoal grilling and roasting. No
carcinogenic PAHs were observed for steaming and LSF-treated duck samples. Also, the highest
amounts of both total and carcinogenic PAHs were found after smoking duck samples for 3
hours.
Visciano et al. (2006) studied the levels of PAHs in fresh and cold-smoked Atlantic salmon fillets
and found no differences between raw and smoked samples in the concentrations of several
carcinogenic PAHs, but significant differences for some low weight molecular PAHs, which
confirms that PAH concentrations in smoked fish are the result of both sea pollution and the
smoking process.
Guillén and Sopelana (2004) investigated the possible presence of PAHs in 7 commercial types
of cheese smoked by traditional techniques, in order to provide an indication of the PAH
contamination in the cheeses that are commonly consumed. The samples were manufactured with
milk from cows, sheep, goats, or a mixture of them, and purchased in local supermarkets. The
results revealed the presence of several PAHs in the exterior zone of the samples, some of the
PAHs having methyl groups. In all cases, the concentrations of PAH compounds of low
molecular weight were much higher than those of high molecular weight. Significant differences
in the number and concentration of PAHs in smoked cheese were observed from rind to interior.
A similar gradient between rind and interior was also found by Anastasio et al. (2004) who
analysed “mozzarella di bufala campana”, a typical Italian stretched cooked cheese,
experimentally smoked according to traditional procedures using straw, cardboard and wood
shavings or aromatized with smoke flavouring. In the cheeses smoked with straw and cardboard
the benzo[a]pyrene levels were statistically higher than those of the cheeses smoked with wood
shavings and aromatized with liquid smoke. Independent of the type of smoking all cheese
samples showed the highest benzo[a]pyrene levels in the rind, followed by slice and core.
Another factor that may influence the PAH levels in smoked foods is the type of casing. Garcia
Falcon et al. (1996; 1999) analysed the PAH contents in chorizo samples, a Spanish pork
sausage, which is usually cased in natural casings and smoked, and in various other types of
26 Letter of the Ministry of Agriculture of the Republic of Latvia to the European Commission from 28.02.2007

smoked sausages cased in cellophane casings. The levels of benzo[a]pyrene in the latter samples
were lower than those found in the chorizo samples. The authors explained the difference by the
hydrophilic nature of cellophane, which renders it essentially impermeable to the PAHs. In
contrast, the natural casings, being primarily protein and lipid, are more likely to absorb such
hydrophobic compounds, thus allowing diffusion and hence penetration to take place. This
phenomenon was also observed earlier by Gilbert and Knowles (1975). In a comprehensive
review Simko (2005) described the physico-chemical factors that affect the PAH content in
smoked meat products, such as light, additional cooking, and packaging, which are able to
decrease considerably the PAH content in some meat products. The most important effect on
PAH concentration decrease in liquid smoke flavouring was low-density polyethylene (LDPE)
package due to sorption processes on a surface of the plastic with subsequent diffusion into the
plastic bulk. A less effective material is polyethylene terephthalate (PET), when only a surface
adsorption process comes into account.
Dependent on the technological process the PAH levels in smoked food may be subject to
changes over time. Simko et al. (1991) investigated the benzo[a]pyrene content in fermented
salami during the technological processes of cold smoking, ripening and storage. They found that
the changes in benzo[a]pyrene content were not so great as in products smoked with hot smoke.
However, the decrease in benzo[a]pyrene content caused by photodegradation was compensated
by successive dehydration of the fermented product. After recalculating the benzo[a]pyrene
content on a dry weight basis, a concentration decrease was evident. Significant changes were
registered during smoking and ripening, while storage under different conditions caused no
further effect on these changes.
Other commercial processing techniques, such as coffee roasting or treating of wooden barrels
for aging of alcoholic beverages may also lead to direct or indirect contamination of food
commodities.
Houessou et al. (2007) studied the formation of PAHs during the roasting process of Arabica
green coffee beans from Cuba under controlled conditions. They showed that the PAH
concentration in the roasted coffee beans is dependent on roasting temperature and roasting time.
The content of low molecular PAHs, such as phenanthrene, fluoranthene, and pyrene increased
during the roasting process under elevated temperatures above 220°C. Strong roasting conditions
(260°C) led to significant levels of pyrene, chrysene and benzo[a]anthracene. Benzo[a]pyrene
and dibenzo[a,h]anthracene, were either not detected or found only at very low levels
(approximately 0.20 µg/kg at 240-250°C). Moreover, the authors concluded that the results of the
study seem to indicate a possible transformation of low molecular PAHs to high molecular PAHs
as the roasting degree is increased.
While the presence of PAHs in coffee has been attributed to their formation during the roasting
step, the transfer of PAHs from coffee powder to coffee brew during coffee making has been
reported to be relatively low (Hischenhuber and Stijve, 1987; de Kruijf et al., 1987; García-
Falcón et al., 2005). This is due to the low solubility of PAHs in water. 1% or less of
benzo[a]pyrene initially present in the solid coffee has been reported to transfer to a coffee brew
(Hischenhuber and Stijve, 1987; de Kruijf et al., 1987). However, since infusion strength
influences the transferred benzo[a]pyrene concentrations, higher transfer percentages up to 26%
with the mean value of 5% were found by Maier (1991). More recently Houessou et al. (2007)
reported a transfer percentage to vary from 0.6 to 30.9% for benz[a]anthracene, chrysene and
benzo[b]fluoranthene depending on the PAH compound, with a slightly lower extractability for
dark-roasted coffee as compared to light-roasted coffee. The concentrations for individual PAHs,
benzo[a]pyrene, benzo[b]fluoranthene, benz[a]anthracene, chrysene, dibenz[a,h]anthracene and
benzo[ghi]perylene, in coffee brew have been reported to be <140 ng/L (Hischenhuber and
Stijve, 1987; de Kruijf et al., 1987; Kayali-Sayadi et al., 1999; Houessou et al., 2005; Houessou
et al., 2007).
Chatonnet and Escobessa (2007) investigated the impact of toasting oak barrels on the presence
of PAHs in wine. Toasting Quercus sp. oak wood is one of the key stages in manufacturing
barrels intended for aging wines and spirits. During this operation, the increase in temperature
causes variable modifications in the physical structure and, more importantly, the chemical
composition of the wood. Therefore, wood toasted to different levels under different conditions,
as well as wines aged in barrels made using different methods, was analyzed for the identification
of the main PAH present. The results clearly showed that the heating processes associated with
barrel production actually resulted in the formation of various PAHs. The formation was highly
dependent on temperature, with greater accumulations in the surface layers directly exposed to
the heat of the fire. The PAH profiles were mainly dominated by naphthalene, phenanthrene and
fluoranthene. In contrast, benzo[a]anthracene, chrysene, benzo[b]fluoranthene,
benzo[j]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, dibenzo[a,h]antracene,
benzo[ghi]perylene, indeno[1,2,3-cd]pyrene, dibenzo[a,i]pyrene, dibenzo[a,e]pyrene,
dibenzo[a,l]pyrene, and dibenzo[a,h]pyrene only showed in a few cases concentrations above the
limit of detection given at 1 µg/kg. When investigating the impact of barrel aging on the transfer
of PAHs into the wine, it was observed that, for identical types of wood and similar toasting
levels, the toasting technique used by the cooperage had a significant impact on both the quantity
and type of PAH that migrated into the wine with almost all of the carcinogenic and genotoxic
PAHs below the limit of detection. Similar results were also found by García-Falcón and Simal-
Gándara (2005) who determined the contents of 7 PAHs in alcoholic beverages of variable
alcoholic strength that had been aged in charred barrels for different times. In all samples the
levels of PAHs were very low. While benzo[a]pyrene was not detected at a limit of detection of
0.01 µg/L the highest level of 0.17 µg/L was found for a brandy. Nevertheless, the results
indicated that the way the tree raw material is toasted influences the PAH levels in alcoholic
drinks. Thus traditional charring produces increased amounts of PAHs from the wood relative to
convective toasting.

Galinaro et al. (2007) investigated the profiles of PAHs in the Brazilian sugar cane spirit
“cachacas” produced from non burned and burned sugar cane crops. The main difference
between these two classes of cachacas was found to be in the amount of benzo[a]pyrene, which is
more abundant in cachaca produced from burned sugar cane crops than in cachaca produced from
non burned crops.
5.1.2 Home cooking and other small scale cooking practices

Considerable differences in the PAH levels can be found in grilled products depending on the
heat source as well as the type and geometry of the grill. This is especially important for home
cooking practices. Lijinsky and Ross (1967) studied the effect of variation in methods of cooking
meat on the content of benzo[a]pyrene and other PAHs in the prepared food. They showed that
PAHs arise from the pyrolysis of fat which drips to the heat source. The amount produced
increases with increased fat content, longer exposure of the food to the flames and closeness to
the heat source. In contrast, the total absence of PAHs from meat broiled in electric or gas
broilers (in which the heat source is above the food) or cooked above charcoal in a vented pan,
which catches the melted fat and prevents contact with the flames, shows that deposition of PAHs
on the meat can be prevented by sequestering the meat surface from the flames.
Lintas et al. (1979) found no difference in content of benzo[a]pyrene between grilled meat and
sausage and the respective raw product when the meat was cooked on a specially designed
electric grill, having a water-filled pan underneath the electric resistance. In this particular type of
grill water creates an air vortex which prevents the dripping from falling on the incandescent coil,
thus avoiding the formation of smoke and pyrolysis.
The strong influence of the cooking method and the type of heat source on the PAH levels in the
processed food was also demonstrated by Larsson et al. (1983). The grilling of frankfurters in the
flames of a log fire resulted in extremely high PAH levels, up to 212 µg/kg benzo[a]pyrene.
When the grilling was carried out over the embers, i.e. when flames no longer emerged from the
fire, the average level of benzo[a]pyrene was only 7.7 µg/kg. Relatively high benzo[a]pyrene
levels, (average: 17.6 µg/kg) were found in frankfurters grilled over smouldering spruce or pine
cones. In charcoal-grilled frankfurters the benzo[a]pyrene levels did not exceed 1 µg/kg when
grilled alone. However, when frankfurters were grilled together with pork chops, the PAH levels
in the frankfurters were noticeably enhanced which confirms the theory of Lijinsky and Ross that
PAHs in charcoal grilled products originate from the pyrolysis of the fat that drips down on the
hot coals during grilling. Frying or electric broiling of frankfurters did not lead to any appreciable
increase of the original trace levels.
The geometry of the barbecue on the PAH formation was investigated by Saint-Aubert et al.
(1992) who analysed several samples of meat and fish grilled on two geometrically different gas

barbecues. After cooking on the vertical barbecue, in which dripping of fat onto the heat source is
prevented, levels of fluoranthene, regularly detected, were lower or equal to 1 µg/kg and amounts
of benzo[a]pyrene and other PAHs were very low, under, or near the detection limit of the
analytical method at 0.1 µg/kg. On the other hand, benzo[a]pyrene and other PAH levels were
10-30 times higher with the horizontal lava-rocks barbecue and varied with the kind of food
sample and particularly with the cooking time.
The influence of the fat content of the meat upon the amount of PAHs formed during charcoal
grilling was also demonstrated by Mottier et al. (2000) who found significantly higher levels of
carcinogenic PAHs in high fat containing lamb sausages compared to other meat products with
lower fat content.
In a recent study White et al. (2008) investigated the effects of frying, grilling, barbecuing,
toasting and roasting on the formation of PAHs in foods prepared in the home and from catering
outlets. The study comprised 77 retail cooked food samples and 256 samples from in-house
cooking experiments. The key findings can be summarized as follows:
Only three cooked retail products, all beef burgers, contained benzo[a]pyrene at concentrations of
8.4 μg/kg, 9.8 μg/kg and 20.4 μg/kg.
With respect to home cooking practices, in general there was little evidence of PAH formation
during the grilling, frying, roasting and toasting experiments. Comparison with the raw materials
used in the experiments showed little or no increase in PAH concentrations for all of the sample
types, distances from the heat source, cooking mediums and severity of cooking conditions.
In contrast, barbecuing with charcoal plus wood chips gave the highest benzo[a]pyrene levels in
each type of food. For beef burgers only, barbecuing over charcoal gave the highest levels. In
general PAH levels increased when the food was cooked closer to the heat source. However, for
sausages cooked over briquettes, and for beef burgers, beef and salmon cooked over charcoal, the
concentration of PAHs was lower when the food was closer to the heat source. Cooking time may
increase PAH levels moderately in some foods, although levels in beef burgers appeared to fall
when cooking time was extended by 50-100%.
In summary, the type of food processing may have a significant influence on the PAH levels in
food commodities. The increasing replacement of traditional smoking processes by liquid smoke
flavouring has certainly led to a decrease of PAH levels in commercially smoked food
commodities over the past two decades. On the other hand, a number of investigations, especially
on various barbecuing practices have demonstrated that certain home cooking practices may lead
to PAH levels that are more than one magnitude higher compared to PAH concentrations in foods
prepared with proper practices. Thus, a considerable potential for a reduced formation of PAHs
and consequently for lowering the exposure to these compounds seems feasible for certain home

cooking practices if techniques that avoid the pyrolysis of fat that drops into the flames are
applied.
5.2 Suitable indicators for occurrence of total PAHs

With the current legislation prescribing benzo[a]pyrene to be used as an indicator for general
PAH contamination, the relevance of such use was tested by calculating concentrations of the
other PAHs in the absence of benzo[a]pyrene.
Out of 4,065 samples analysed for all PAH8, 2,090 tested negative for benzo[a]pyrene. Of the
latter 2,090 samples, 54% identified concentrations above the LOD for at least one other PAH out
of the PAH8. Only 1,375 samples out of the 4,065 were analysed for all PAH15 of which 620
tested negative for benzo[a]pyrene. Of the latter 620 samples, 74% identified concentrations
above the LOD for at least one other PAH out of the PAH15. The difference between the two
comparisons is partly due to differences in the representation of products in respective food
category but also to the more complete range of PAHs tested in the latter case.
Table 12 illustrates the situation for individual PAHs and for the group of seven PAHs (PAH15-
PAH8) not included in PAH8. Of the individual PAHs, chrysene at 43% was most commonly
found in samples negative for benzo[a]pyrene while dibenz[a,h]anthracene and indeno[1,2,3-
cd]pyrene were found in only 2% of such samples. The mean concentrations of individual PAHs
in those samples varied from 0.32 to 1.59 µg/kg. The highest maximum for such individual PAHs
was recorded for chrysene with 242 µg/kg and the second highest for benz[a]anthracene with 153
µg/kg. Of samples negative for PAH2 (benzo[a]pyrene+chrysene), 26% and 18% identified
concentrations above the LOD for at least one other PAH for samples tested for all PAH15 or all
PAH8, respectively. The frequency varied between 2% and 9% for the individual PAHs or PAH
combinations. Of samples negative for PAH4 (PAH2+benz[a]anthracene+benzo[b]fluoranthene),
14% and 6% identified concentrations above the LOD for at least one other PAH for samples
tested for all PAH15 or all PAH8, respectively. The frequency varied between 1% and 6% for the
individual PAHs or PAH combinations. Thus the likelihood of results above the LOD when using
the PAH2 measure rather than the benzo[a]pyrene measure alone was considerably reduced for
all PAHs except for dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene. Using PAH4 removed
the samples above the LOD for benz[a]anthracene and benzo[b]fluoranthene but had little impact
on the other PAHs. The same 12 samples in the PAH15 group other than PAH8 still identified
concentrations above the LOD for samples negative to both PAH4 and PAH8. The slight increase
in the percentage above the LOD is an anomaly caused by the fact that the overall number of
samples remaining when applying the two measures is reduced from 191 to 176.

Table 12: Comparison of upper bound concentrations of seven individual PAHs and of the
remaining PAHs combined (PAH15-PAH8) in samples with values above the LOD with
benzo[a]pyrene, PAH2, PAH4 and PAH8 results less than the LOD, respectively.
Concentration µg/kg
Limitation N >LOD
Mean (>LOD) Min (>LOD) Max (>LOD)
CHR All 4065 64% 4.85 0.01 590
BaP<LOD 2090 43% 1.56 0.01 242
BaA All 4065 50% 3.91 0.01 1064
BaP<LOD 2090 26% 1.59 0.01 153
PAH2<LOD 1182 9% 0.62 0.01 10.7
BbFA All 4065 51% 2.68 0.01 690
BaP<LOD 2090 19% 0.33 0.01 21
PAH2<LOD 1182 6% 0.67 0.02 7.5
BkFA All 4065 46% 1.14 0.001 150
BaP<LOD 2090 12% 0.35 0.001 16
PAH2<LOD 1182 4% 0.60 0.001 7.5
PAH4<LOD 1024 2% 0.58 0.001 7.5
BghiP All 4065 38% 1.70 0.01 220
BaP<LOD 2090 11% 0.32 0.01 5
PAH2<LOD 1182 4% 0.86 0.01 5
PAH4<LOD 1024 3% 0.90 0.01 5
DBahA All 4065 13% 0.95 0.01 36
BaP<LOD 2090 2% 0.70 0.1 3
PAH2<LOD 1182 2% 0.72 0.1 2.6
PAH4<LOD 1024 1% 0.69 0.22 2.3
IP All 4065 25% 2.22 0.01 100
BaP<LOD 2090 2% 0.65 0.01 5
PAH2<LOD 1182 2% 0.76 0.03 2.4
PAH4<LOD 1024 1% 1.05 0.4 2.4
PAH15-PAH8 All 1375 63% 3.37 0.18 169
BaP<LOD 620 35% 0.51 0.18 2.4
PAH2<LOD 222 8% 0.82 0.23 2.4
PAH4<LOD 191 6% 0.60 0.23 1.9
PAH8<LOD 176 7% 0.60 0.23 1.9
As a further assessment of the relationships between the different PAHs, a correlation matrix
covering the eight PAHs and their combinations as outlined in Table 9 was developed for the
4,065 products as shown in Table 13 using upper bound values.

Table 13: Correlation coefficient matrix for eight PAHs found in the coal tar mixtures and for
combinations of the PAHs calculated for 4,065 upper bound sample results from different food
categories (see Table 8).
BaP BaA BbFA BkFA BghiP CHR DBahA IP PAH2 PAH4 PAH8
BaP 0.52 0.92 0.90 0.90 0.79 0.59 0.82 0.87 0.87 0.90
BaA 0.52 0.49 0.46 0.44 0.54 0.29 0.40 0.56 0.82 0.78
BbFA 0.92 0.49 0.93 0.95 0.71 0.60 0.78 0.79 0.84 0.88
BkFA 0.90 0.46 0.93 0.94 0.70 0.63 0.87 0.78 0.80 0.86
BghiP 0.90 0.44 0.95 0.94 0.69 0.62 0.85 0.77 0.79 0.85
CHR 0.79 0.54 0.71 0.70 0.69 0.42 0.65 0.99 0.89 0.88
DBahA 0.59 0.29 0.60 0.63 0.62 0.42 0.57 0.48 0.50 0.55
IP 0.82 0.40 0.78 0.87 0.85 0.65 0.57 0.71 0.71 0.77
PAH2 0.87 0.56 0.79 0.78 0.77 0.99 0.48 0.71 0.92 0.92
PAH4 0.87 0.82 0.84 0.80 0.79 0.89 0.50 0.71 0.92 0.99
PAH8 0.90 0.78 0.88 0.86 0.85 0.88 0.55 0.77 0.92 0.99
Benzo[a]pyrene was not a good indicator for the concentration of benz[a]anthracene or

dibenz[a,h]anthracene in particular. By combining four PAHs (PAH4) correlations improved but
the concentration of dibenz[a,h]anthracene is still not well correlated with the other four.
However, judging from the number of samples positive for any of the PAHs when PAH4 is
below the LOD (Table 12), only 3% of samples at the most carried measurable quantities of a
PAH from the group of eight.
Similarly, to test the correlation for all 16 priority PAHs a comparison was made for the 1,375
samples (823 for benzo[c]fluorene) analysed for the full range of PAHs (Table 14). It should be
acknowledged that this range of samples covered a limited product range and thus is much less
accurate than the correlations presented in Table 13. Out of the 33 food categories/subcategories
11 are missing and three of the categories covered included less than 10 samples each.

Table 14: Correlation coefficient matrix calculated for 1,375 upper bound samples analysed for
all 16 priority PAHs (only 823 samples for benzo[c]fluorene) and combinations of the PAHs.
DBahA
DBahP
DBaeP
DBaiP
DBalP
BghiP
BbFA
PAH2
PAH4
PAH8
BkFA
BcFL
BjFA
MCH
CHR
CPP
BaA
BaP
IP
BaP 0.92 0.89 0.89 0.94 0.80 0.88 0.96 0.90 0.41 0.74 0.04 0.39 0.65 0.08 0.24 0.86 0.90 0.92
BaA 0.92 0.92 0.91 0.87 0.91 0.81 0.91 0.90 0.38 0.65 0.02 0.35 0.51 0.05 0.22 0.94 0.97 0.97
BbFA 0.89 0.92 0.95 0.91 0.91 0.89 0.93 0.90 0.23 0.71 0.03 0.37 0.52 0.02 0.14 0.94 0.96 0.97
BkFA 0.89 0.91 0.95 0.87 0.86 0.84 0.91 0.88 0.23 0.68 0.05 0.37 0.58 0.02 0.12 0.89 0.92 0.94
BghiP 0.94 0.87 0.91 0.87 0.82 0.89 0.97 0.87 0.33 0.77 0.04 0.40 0.59 0.04 0.16 0.87 0.89 0.92
CHR 0.80 0.91 0.91 0.86 0.82 0.76 0.84 0.87 0.30 0.64 0.01 0.32 0.36 0.14 0.36 0.99 0.98 0.96
DBahA 0.88 0.81 0.89 0.84 0.89 0.76 0.92 0.81 0.18 0.81 0.15 0.49 0.60 -0.01 0.04 0.81 0.84 0.87
IP 0.96 0.91 0.93 0.91 0.97 0.84 0.92 0.90 0.28 0.80 0.08 0.45 0.63 0.02 0.12 0.89 0.92 0.94
BjFA 0.90 0.90 0.90 0.88 0.87 0.87 0.81 0.90 0.44 0.69 0.12 0.43 0.50 0.24 0.63 0.90 0.92 0.93
CPP 0.41 0.38 0.23 0.23 0.33 0.30 0.18 0.28 0.44 0.20 0.13 0.20 0.04 0.39 0.71 0.33 0.33 0.32
DBaeP 0.74 0.65 0.71 0.68 0.77 0.64 0.81 0.80 0.69 0.20 0.49 0.78 0.51 0.00 0.03 0.68 0.69 0.72
DBahP 0.04 0.02 0.03 0.05 0.04 0.01 0.15 0.08 0.12 0.13 0.49 0.86 0.04 0.01 0.01 0.01 0.02 0.03
DBaiP 0.39 0.35 0.37 0.37 0.40 0.32 0.49 0.45 0.43 0.20 0.78 0.86 0.30 0.02 0.02 0.34 0.35 0.37
DBalP 0.65 0.51 0.52 0.58 0.59 0.36 0.60 0.63 0.50 0.04 0.51 0.04 0.30 -0.01 0.04 0.43 0.47 0.51
MCH 0.08 0.05 0.02 0.02 0.04 0.14 -0.01 0.02 0.24 0.39 0.00 0.01 0.02 -0.01 0.67 0.13 0.10 0.08
BcFL 0.24 0.22 0.14 0.12 0.16 0.36 0.04 0.12 0.63 0.71 0.03 0.01 0.02 0.04 0.67 0.34 0.28 0.25
PAH2 0.86 0.94 0.94 0.89 0.87 0.99 0.81 0.89 0.90 0.33 0.68 0.01 0.34 0.43 0.13 0.34 0.99 0.99
PAH4 0.90 0.97 0.96 0.92 0.89 0.98 0.84 0.92 0.92 0.33 0.69 0.02 0.35 0.47 0.10 0.28 0.99 1.00
PAH8 0.92 0.97 0.97 0.94 0.92 0.96 0.87 0.94 0.93 0.32 0.72 0.03 0.37 0.51 0.08 0.25 0.99 1.00
Correlation coefficients vary from -0.01 to 0.96 with dibenzo[a,h]pyrene in particular deviating
from most of the other PAHs except benzo[c]fluorene. Additional calculations for
benzo[c]fluorene showed that in general it constituted close to 25% of the total when added to the
sum of the eight PAHs found in the coal tar mixtures. However, this varied from 1.7% for spices
to 61.5% for dried fruit samples explaining the low correlation coefficients found between
benzo[c]fluorene and the eight PAHs.
Looking mainly at the number of samples above the LOD for any individual PAH or combination
of PAHs when any of the four different measures (benzo[a]pyrene, PAH2, PAH4 or PAH8) is
below the LOD (Table 12), the PAH4 combination seems to give a sufficient accuracy that is not
much improved by using the PAH8 combination. Likewise there is little change in the correlation
coefficients between the PAH4 and PAH8 combinations.
6. Food consumption
The EFSA colloquium on a European Food Consumption database held in Brussels in April 2005
(EFSA, 2005b) made it clear that a common database on food consumption would improve the
consistency and reliability of exposure assessments carried out by the various EFSA Panels and
other experts in Europe. Following the recommendations of the colloquium, in 2007 EFSA
started the development of the Concise European Food Consumption Database27 and published
information collected from 16 countries in March 2008. Information covering further countries
will be published as the data become available. For this opinion food consumption data covering
the whole population and for consumers only was extracted from the Concise European Food
Consumption Database on the EFSA website27 and presented in Tables 15 to 17. The principles
for using the data as described in the accompanying guidelines on the same website were
followed. Only food categories to be used in the exposure calculations are presented in the three
Tables. The full food consumption datasets for each country are available on the EFSA website27.
27 http://www.efsa.europa.eu/EFSA/ScientificPanels/DATEX/efsa_locale-
1178620753812_ConciseEuropeanConsumptionDatabase.htm

Table 15: Mean food consumption (g/day) in total population as recorded in 16 European countries.
BE BG CZ DK FI FR DE HU IS IE IT NL NO SK SE GB
01 Cereals and cereal products 245 257 274 217 153 317 287 252 276 227 271 220 192 345 291 249
02 Sugar and sugar products including chocolate 31 40 39 43 42 31 45 39 31 41 19 43 47 69 28 27
03 Fats (vegetable and animal) 46 39 48 36 40 28 29 54 33 36 36 48 41 29 24 20
04 Vegetables, nuts, pulses 230 210 131 166 135 210 252 191 125 244 249 193 140 164 118 163
06 Fruits 113 70 122 150 121 132 190 180 71 106 203 107 119 116 119 95
08 Coffee, tea, cocoa (expressed as liquid) 432 120 559 836 580 282 691 176 429 714 124 887 604 461 575 724
09 Alcoholic beverages 214 102 413 292 139 163 231 76 104 335 126 206 123 64 191 313
10A Meat and meat products and substitutes 121 108 175 128 120 134 127 173 93 122 134 141 103 179 150 49
11A Seafood and seafood products 0 0 0 3 5 1 0 4 1 10 2 49 1 10 3
11B Fish and fish products 19 20 17 15 27 29 18 9 32 23 33 10 3 6 24 21
13C Cheese 33 31 28 29 40 42 14 39 14 56 33 37 18 30 15
BE: Belgium, BG: Bulgaria, CZ: Czech Republic, DK: Denmark, FI: Finland, FR: France, DE: Germany, HU: Hungary, IS: Iceland, IE: Ireland, IT: Italy, NL: The
Netherlands, NO: Norway, SK: Slovakia, SE: Sweden, GB: Great Britain
Table 16: Average consumption (g/day) in consumers only as recorded in 16 European countries.
04 Vegetables, nuts, pulses including carrots, tomato 238 219 132 166 137 212 252 191 137 244 249 196 140 253 119 164
06 Fruits 168 185 147 165 143 156 195 204 150 122 210 141 121 235 132 120
13C Cheese 43 57 42 29 43 43 24 54 17 56 40 37 92 31 20
Netherlands, NO: Norway, SK: Slovakia, SE: Sweden, GB: Great Britain

Table 17: 97.5th percentile consumption (g/day) in consumers only as recorded in 16 European countries.
04 Vegetables, nuts, pulses 652 613 370 408 373 551 599 406 443 759 555 470 382 700 296 409
06 Fruits 456 720 463 529 462 521 599 568 430 442 596 429 465 750 376 403
13C Cheese 128 200 126 76 124 115 77 182 56 125 120 107 400 81 60
Netherlands, NO: Norway, SK: Slovakia, SE: Sweden, GB: Great Britain.

7. Human exposure assessment
7.1 Dietary intakes

The dietary exposure assessment is based on the occurrence data on PAH concentrations (Table
9) and consumption of the corresponding food categories reported in the Concise European Food
Consumption Database (Tables 15 to 17). As a first step the food categories used for occurrence
were matched as well as possible with the corresponding food consumption categories (Table
18). Coffee powder was selected to represent the food category “coffee, tea and cocoa (expressed
as liquid)” since the dilution factor for coffee (7 g of coffee in 125 ml of liquid) is less than for
tea (2 g in 125 ml of liquid) and the transfer rate of PAH from coffee to the liquid has been
published (see Chapter 5.1.1).
As expected the aggregated categories used for food consumption were broader than the
corresponding categories used for reporting the analytical contaminant results. The Concise
European Food Consumption Database is intended for preliminary exposure assessment use only.
Sampling targeted suspect foods, using the above consumption data without adjustment will
result in a conservative overestimate of the real exposure situation. Thus an attempt was made to
calculate the impact of refining the food consumption categories to better reflect the food-
sampling pattern. Fruit consumption was split into dried fruit and other fruit. Data from Denmark
indicated an annual consumption of about 2.3 kg/capita of dried fruit or 4% of total fruit
consumption (DIPP, 2007). This is the same as the dried fruit market share indicated for the
USA28 and was used for the adjusted calculations. Most cereal products are consumed after
processing, however, the cereals category comprises occurrence data from both unprocessed and
processed cereals. As an alternative the processed cereals occurrence data were used. Sugar and
sugar products including chocolate was split into chocolate and other products. Average EU
annual chocolate consumption was estimated at 5.2 kg/capita in 2004 by CAOBISCO29 or about a
third of the overall consumption recorded for the category. Finally fish and fish products are
represented by the PAH concentrations recorded for “all processed fish”. However, according to
information from FAO (FAO, 2004) canned and cured fish, the bulk of the occurrence data,
comprise about a third of the total fish consumption worldwide and the information was adjusted
accordingly.
The above adjustment factors, as summarised in Table 18, were used for calculating exposure but
it must be kept in mind that in reality they will vary across populations. The impact of universally
applying the adjustment factors is estimated in the sensitivity analysis in Section 7.3.
28 http://www.statpub.com/stat/open/2000/01av9kj5.html
29 http://www.caobisco.com/page.asp?p=213

Table 18: Allocation of food categories from the classifications used for food consumption to
represent the food categories used for PAH occurrence. Adjustment factors for refinement of the
food consumption categories are shown in brackets.
Category food consumption Category occurrence
01 Cereals and cereal products Cereals
02 Sugar and sugar products including chocolate Chocolate
(chocolate 33%, other products 67%)
03 Fats (vegetable and animal) Fats and oil excluding cocoa
butter and pomace oil
04 Vegetables, nuts, pulses including carrots, tomato Vegetables and nuts
06 Fruits (dried fruit 4%, other fruit 96%) Dried fruit
08 Coffee, tea, cocoa (expressed as liquid) Coffee powder30
09 Alcoholic beverages Alcoholic beverages
10A Meat and meat products and substitutes Meat
11A Seafood and seafood products All molluscs
11B Fish and fish products All processed fish
(processed fish 30%, fresh and frozen fish 70%)
13C Cheese Smoked cheese
Regarding the level of contamination, because of the potential for chronic effects of PAHs, the
use of the median concentrations would represent in theory the most realistic picture for long
term exposure. Nevertheless, because of the high proportion of samples below the LOD or the
LOQ, the use of the median would not allow calculation of dietary exposure. Therefore all the
calculations below are based on the mean concentrations to represent the central tendency of the
distribution, although this is expected to be an overestimation of long-term dietary exposure.
Exposure to four distinct groups of PAHs were calculated:
• Benzo[a]pyrene only,
• PAH2 (benzo[a]pyrene + chrysene),
• PAH4 (PAH2 + benz[a]anthracene + benzo[b]fluoranthene), and
• PAH8 (PAH4 + benzo[k]fluoranthene + benzo[ghi]perylene + dibenz[a,h]anthracene +
indeno[1,2,3-cd]pyrene).
It appears that due to the high sensitivity of the analytical methods and despite the relatively high
proportion of samples below the LOD/LOQ, there is a very limited impact of using the lower or
30 The concentration of PAHs in coffee was divided by 18 to take into account the conversion from solid coffee to
beverage and multiplied by 26% to take into account the transfer rate of PAHs from the solid coffee to the liquid
coffee.

the upper bound31. Therefore, all the exposure scenarios detailed below are based on the upper
bound of the mean concentrations of the four groups of PAHs under consideration. This
assumption represents a small but real overestimation. The food consumption information used in
the model includes mean consumption and the 97.5th percentile of the distribution.
The assessment was undertaken in three steps according to the guidance for the use of the concise
database: determination of the main food vehicles, estimation of the average dietary exposure and
estimation of the dietary exposure of the highly exposed consumers (EFSA, 2008).
7.2 Contributions of different food groups to PAH exposure

Due to the ubiquitous presence of PAHs in food, the food categories that contribute significantly
to the exposure are determined either by a high level of contamination in or by high amounts
consumed of that food.
The contribution of each broad food category to total exposure was calculated from the mean
consumption of consumers32 only, collected in each country for the corresponding food category.
The median of the mean values (expressed in g/day) was determined and multiplied by the mean
contamination (expressed in µg/kg). Results are therefore expressed in ng/day and are
summarised in Table 19. Because the consumption surveys in the countries involved used
different methodologies for data collection, the mean consumption values should not be averaged
and thus the median was used.
31 Lower bound: all samples below LOD or LOQ replaced by 0. Upper bound: all samples below LOD or LOQ
replaced by the level of detection or level of quantification.
32 Consumers are defined as subjects consuming the food category under consideration at least once during the
survey duration.

Table 19: Consumer exposure to benzo[a]pyrene (BaP), PAH2, PAH4 and PAH8 for each food
category for which occurrence data are available. The median value of the mean consumption
reported by the Member States for consumers only was calculated.
Category Consumption Exposure
Median BaP PAH2 PAH4 PAH8
g/day ng/day ng/day ng/day ng/day
Cereals and cereal products 257 67 129 257 393
Sugar and sugar products including chocolate 43 5 13 25 39
Fats (vegetable and animal) 38 26 112 177 239
Vegetables, nuts and pulses 194 50 124 221 378
Fruits 153 5 40 75 87
Coffee, tea, cocoa (expressed as liquid) 601 21 55 106 156
Alcoholic beverages 413 4 12 25 74
Meat and meat products and substitutes 132 42 107 195 279
Seafood and seafood products 27 36 140 289 421
Fish and fishery products 41 21 84 170 210
Cheese 42 6 12 20 30
The two highest contributors to the dietary exposure were cereals and cereal products, and
seafood and seafood products. These two contributors have been used to estimate the exposure of
high consumers (see below). It is important to note that because the contributions are estimated
for relevant consumers of foods from each category, the various contributions should not be
summed to estimate an overall exposure. The reason for this is that the same person will not
necessarily be a consumer of all the food groups.
7.2.1 Mean dietary exposure to PAHs

The next step consists of estimating the average dietary exposure to PAHs in the EU. In order to
sum the various contributions from various food categories in respective Member State, the
average dietary exposure for the whole population (including the non-consumers) was used.
Since different methodologies were used to estimate consumption in the various countries the
midpoint (median) of all individual country averages was taken to represent the best estimate of
an overall average. It can be concluded that the overall average dietary exposure across European
countries for which data are available is 235 ng/day (3.9 ng/kg b.w. per day assuming a body
weight of 60 kg) for benzo[a]pyrene (range: 185-255 ng/day) and 1729 ng/day (28.8 ng/kg b.w.
per day assuming a body weight of 60 kg) for PAH8 (range: 1415-2136 ng/day). These figures
are detailed per country in Table 20.

Table 20: Total dietary exposure to benzo[a]pyrene (BaP), PAH2, PAH4 and PAH8 (ng/day) for
average and high consumers across Europe33.
Average exposure High exposure
Whole population Sum of P97.5 for cereals and
seafood + average exposure for
whole population
Country BaP PAH2 PAH4 PAH8 BaP PAH2 PAH4 PAH8
Belgium 232 637 1158 1732 393 1101 2108 3138
Bulgaria 209 560 1020 1526 385 1053 2027 3018
Czech Republic 239 654 1196 1777 426 1207 2328 3449
Denmark 223 617 1135 1690 299 818 1545 2300
Finland 185 535 978 1422 231 623 1155 1693
France 245 655 1220 1814 380 1021 1966 2921
Germany 255 681 1258 1888 422 1194 2311 3439
Hungary 231 647 1168 1716 314 877 1636 2410
Iceland 205 558 1039 1522 694 2232 4486 6568
Ireland 238 646 1188 1793 370 1049 2013 3009
Italy 255 719 1332 1962 487 1502 2943 4322
Netherlands 239 658 1197 1785 535 1687 3318 4886
Norway 252 765 1449 2136 461 1470 2900 4262
Slovakia 244 626 1158 1727 709 1905 3769 5601
Sweden 230 621 1168 1719 364 1003 1949 2876
United Kingdom 188 499 936 1415 315 854 1661 2489
Median EU 235 641 1168 1729 389 1077 2068 3078

EU: European Union
Investigations have shown that in barbecuing of fatty meat pyrolysis of fat that drops into the
flames may lead to PAH levels that are more than a magnitude higher than PAH levels in
products that are barbecued with practices that avoid the pyrolysis of fat. Consequently, high
consumption of certain home barbecued foods may lead to an exposure to PAHs that
considerably exceeds the above estimated background dietary exposure to these compounds.
7.2.2 Dietary exposure to PAHs for high consumers

Finally, the exposure for high consumers was estimated following the guidelines for the use of
the Concise European Food Consumption Database i.e. by summing the 97.5th percentiles
(consumers only) for the two main contributors and the mean exposure (whole population) for the
other food categories (Table 20). The resulting median of the “high exposures” across Europe is
389 ng/day (6.5 ng/kg b.w. per day assuming a body weight of 60 kg) for benzo[a]pyrene alone
33 Due to the high sensitivity of the analytical method there is a limited impact of using either the lower or the upper
bound, thus all the dietary exposure calculations are based on the upper bound value for contamination.

(range 231-709 ng/day) and 3078 ng/day (51.3 ng/kg b.w. per day assuming a body weight of 60
kg) for the sum of PAH8 (range 1693-6568 ng/day).
The JECFA reviewed intake estimates for the 13 PAHs that the Committee considered to be
carcinogenic and genotoxic. Estimated intakes of benzo[a]pyrene ranged from <1000 to 2000
ng/day. In order to provide a likely intake of benzo[a]pyrene covering the main food groups in
the whole diet for the purposes of risk characterization, a separate determination of the range of
intakes was conducted using only those studies that included foods from the range of major food
groups. These studies included foods that were ‘ready to eat’ (e.g. cooked meat), and therefore
included the likely concentrations of PAHs that arise due to cooking of food. From this analysis,
mean intakes of benzo[a]pyrene ranged from 1.4 to 420 ng/day. From this range, the JECFA
selected the value of 240 ng/day (4 ng/kg b.w. per day assuming a body weight of 60 kg) as being
representative of a mean intake for use in the risk characterisation and assumed that high
consumption could be 2 to 2.5 times higher i.e. 8-10 ng/kg b.w. per day. This mean intake value
is the same as calculated above while the JECFA’s assumption for high consumption is higher
than above.
The major contributors to intakes of PAHs identified by JECFA were “cereals and cereal
products” and “vegetable fats and oils”. The exclusion of cocoa butter and pomace oil – not
consumed as such – from the occurrence data of PAHs submitted to EFSA, allows the
identification “seafood and seafood products” as a more important vector for PAHs than “fat and
oils”.
High levels of PAHs were found in some food supplements. Consumption levels of food
supplements are variable depending on their formation and intended use but are often low at
around a level of 2 g/day. Equally spices with high PAH levels detected are consumed in small
quantities. Neither of those foods will influence the overall dietary intake of PAHs very much
and will not be considered in the exposure assessment. The number of samples of cocoa butter
analysed for all 15 PAHs was too small to be included in Table 11. However, high levels found in
the PAH8 group are indicated in Table 9. Cocoa butter is a constituent of chocolate at a
prescribed minimum level of 18% and will thus be accounted for in the exposure assessment in
the chocolate food category. Another high level category from Table 9 is pomace oil. Pomace oil
is not normally consumed directly but used as cooking oil. Pomace oil is not included in the
exposure calculations as a separate item.
7.3 Sensitivity analysis of dietary exposure

An attempt was made to estimate the impact of using the suggested adjustment factors to
disaggregate the cereals and cereal products, sugar and sugar products, fruits, as well as fish and
fish products categories as described in Table 19. The adjusted and unadjusted dietary intake

results are compared in Table 21. It was assumed that fruits other than dried fruit have no PAH
values above the LOD since no test results were reported for fresh fruit. Should fresh fruit carry
the same contamination level as dried fruit on a wet weight basis the absolute percentage
decrease would be about 5% less than indicated in Table 21. Of the adjustments applied, the
change to using processed cereals instead of all cereals and the recalculation of dried fruit had the
biggest impacts accounting for a third of the total effect each on average. The impact of
disaggregating the fish and fishery products category was slightly less at 24% and splitting
chocolate from other sugar and sugar products had the least effect accounting for 8% of the
difference on average.
Table 21: Difference in average exposure in the whole population when disaggregating some
food categories (cereals and cereal products, sugar and sugar products, fruit, fish and fish
products).
Average exposure High exposure
Whole population Sum of P97.5 for cereals and fish +
average exposure for whole
population
Median EU BaP PAH2 PAH4 PAH8 BaP PAH2 PAH4 PAH8
Unadjusted 308 940 1801 2713 576 1972 3927 5279
Adjusted 235 641 1168 1729 389 1077 2068 3078
Difference -24% -32% -35% -36% -32% -45% -47% -42%
The median average dietary exposure across European countries for which data are available
could range from a low estimate of 235 to a high of 308 ng/day when using adjusted or
unadjusted food categories or 3.9 to 5.1 ng/kg b.w. per day assuming a body weight of 60 kg for
benzo[a]pyrene and from 1729 to 2713 ng/day or 28.8 to 45.2 ng/kg b.w. per day assuming a
body weight of 60 kg for PAH8.
The median high dietary exposure across European countries for which data are available could
range from a low estimate of 389 to a high of 576 ng/day for benzo[a]pyrene or 6.5 to 9.6 ng/kg
b.w. per day assuming a body weight of 60 kg and from 3078 to 5279 ng/day for PAH8 or 51.3 to
88.0 ng/kg b.w. per day assuming a body weight of 60 kg.
It should be pointed out that dietary exposure could be increased due to the formation of PAHs
during home food preparation. Only a few samples of barbecued and grilled meat samples have
been included in this survey. Extensive use of such cooking methods could considerably increase
dietary exposure to PAHs. Chapter 5.1 details the impact of home cooking but it is not possible to
calculate the total impact of those practices on dietary exposure to PAHs based on the available
results.

7.4 Estimates of non-dietary exposure to PAHs
7.4.1 Ambient air

There are only a limited number of data that can be used for estimations of human exposure to
PAHs in ambient air. In the Nordic countries the levels of benzo[a]pyrene are seldom expected to
exceed 1.0 ng/m3 and the levels usually recorded at background stations are one to two orders of
magnitude below this indicative level. Measurements of 18 PAHs (the US EPA 16 PAHs plus
benzo[e]pyrene and 1-methylphenanthrene) in particle samples collected at a heavily trafficked
site in Oslo in 2002/2003 show a variation from approximately 100 ng/mg to more than 800
ng/mg particular matter (PM) 10 (Kochbach et al., 2006). This corresponds to a concentration in
air of the sum of these 18 PAHs of approximately 5 to 40 ng/m3 at a particle concentration of 50
µg/m3 and the concentration of benzo[a]pyrene between 0.2 and 1.6 ng/m3.
Assuming a ventilation volume of 20 m3 per day, the intake of benzo[a]pyrene via inhalation
could be between 4 and 32 ng per person. This is in accordance with EC (2002) that estimated the
non-occupational exposure to a number of PAHs. See Table 22.
Table 22: Estimated non-occupational mean daily intake of different PAHs via air and drinking
water for an adult non-smoker (ng/person). Modified from EC (2002).
PAH Aira Drinking waterb
Anthracene 20
Phenanthrene 400
Fluoranthene 100 2-200
Pyrene 100 0.2-200
Benz[a]anthracene 20 0.2-10
Chrysene 20 200c
Benzo[b]fluoranthene 20 0.1-2
Benzo[j]fluoranthene 0.02-0.2
Benzo[k]fluoranthene 20 0.02-2
Benzo[a]pyrene 20 0.2-2
Benzo[e]pyrene 20
Benzo[ghi]perylene 20 0.2-2
Indeno[1,2,3-cd]pyrene 20 0.2-2
Dibenz[a,h]anthracene 2
a
Assuming a ventilation rate of 20 m3/day.
b
Assuming an ingestion of 2 L/day.
c
Chrysene + triphenylene.
7.4.2 Occupational exposure

Coal gasification and its conversion to coke and coal tar and the processing and use of coal-tar
derived products can cause high levels of occupational exposure. Other industries and

occupations that could be connected to elevated exposure to PAH are e.g. paving and roofing that
involve coal tar, aluminium production (including anode manufacturing), carbon electrode
manufacture, occupation as a chimney sweep and other exposures to soot.
The concentration of benzo[a]pyrene in air in these industries can be as high as 100 μg/m3
compared with typical ambient air concentrations of one or a few ng per m3 or lower. Older
reports from petroleum refineries and copper mines state levels up to 470 μg/m3 (WHO/IARC,
1989).
1-Hydroxypyrene is a urinary metabolite of pyrene that has been widely used as a urinary
biomarker of PAH exposure. Pyrene is present in all PAH mixtures at relatively high
concentrations (2-10% of the total PAHs), and in certain environments the pyrene content of the
total PAHs is fairly constant. The levels of 1-hydroxypyrene can reach 100 μmol/mol creatinine
in highly exposed workers but are generally less than 0.1 μmol/mol creatinine in people not
exposed occupationally.
7.4.3 Smoking
The contribution to the PAH exposure from the inhalation route may increase markedly for
smokers and persons exposed to passive smoking. When a mainstream smoke of commercially
available filter cigarettes was analysed for benzo[a]pyrene in Spain (Kayali and Rubio-Barroso,
1995), UK (Evans et al., 1993) and USA (Tomkins et al., 1985) the benzo[a]pyrene content was
found to be in the range of 2-20 ng/cigarette. In a study by Lodovici et al. (2004a) of 13 different
cigarette brands on the Italian market 3.9 to 9.2 ng benzo[a]pyrene per cigarette was reported in
the mainstream. Assuming 6.5 ng as a mean delivery and that 80% of inhaled particle-bound
benzo[a]pyrene from the mainstream smoke is deposited in the respiratory tract (WHO/IARC,
1986), the additional benzo[a]pyrene intake for a person smoking 20 cigarettes/day is 105 ng,
which is below but in the same order of magnitude of the mean intake by ingestion of food.
Personal exposure to passive smoking is also of high concern for PAH intake. The sidestream
smoke has been found to contain about 10 times more PAHs than the mainstream smoke
(Grimmer et al., 1987; cited in WHO/IPCS, 1998; Lodovici et al., 2004a). Therefore, persons in
rooms with many smokers may be exposed to benzo[a]pyrene concentrations in an approximate
range of 5-20 ng/m3 (Grimmer et al., 1977, cited in: Grimmer, 1983; Valerio et al., 1996).
Assuming an exposure to 10 ng/m3 for 5 hours per day, the additional benzo[a]pyrene intake
from passive smoking could be around 40 ng/day.

8. Hazard identification and characterisation
8.1 Summary of key data

There is an enormous body of literature on benzo[a]pyrene and on some other PAHs,
particularly relating to use as a model compound in studies of, for example, mutagenicity and
metabolic activation. As no new key data could be identified that would alter the previous PAH
assessments, the CONTAM Panel decided to use three recent expert reviews as the starting point
for its evaluation. These were: the International Programme on Chemical Safety (IPCS)
Environmental Health Criteria document on selected non-heterocyclic PAHs (WHO/IPCS, 1998),
the opinion of SCF on the risks to human health posed by PAHs (EC, 2002), and the 2005 review
of PAHs by the JECFA (FAO/WHO, 2006).
Both SCF and the JECFA concluded that a Toxic Equivalency Factor (TEF) approach to the
assessment of PAHs was not appropriate due to limitations in the available data and because of
different modes of action amongst different PAHs. Furthermore both concluded that
benzo[a]pyrene could be used as a marker of PAHs in evaluating the risks to human health of
PAHs in food and identified a study of the carcinogenicity of coal tar mixtures and
benzo[a]pyrene in mice (Culp et al., 1998) as critical for the evaluation. SCF concluded that a
conservative estimate implied that the carcinogenic potency of total PAHs in food would be 10
times higher than expected from benzo[a]pyrene alone. The JECFA used a MOE approach based
on dietary exposure to benzo[a]pyrene and the benzo[a]pyrene content of the coal tar mixtures to
reach conclusions on a level of concern for human health.
8.1.1 Toxicokinetics
Except for benzo[a]pyrene, many of the compounds described in the published studies on the fate
of PAHs in animals are not included in the list of the 15 PAHs identified by SCF in 2002.
Notwithstanding, some of these data were taken into account because they provide a broader
knowledge of the toxicokinetics of PAHs than an overview based on a very limited number of
compounds.
8.1.1.1 Absorption
Absorption of PAHs from the diet is determined by the size and the lipophilicity of the molecule,
the presence of bile in the digestive tract, the dose ingested and the lipid content of the diet. In
experimental animals, the gastrointestinal absorption of PAHs, especially benzo[a]pyrene, is well
documented. In rats, oral absorption of benzo[a]pyrene was reported to occur mainly within 2-4
hours and was estimated to be 35-99% following dietary or gavage exposure (ATSDR, 1995;

WHO/IPCS, 1998; EC, 2002; Ramesh et al., 2004). Values between 75 and 87% were reported
for chrysene and between 42 and 99% for pyrene.
In investigating the role of the lymphatic system in the absorption of 7,12-

dimethylbenzanthracene in the rat, Laher and Barrowman (1987) concluded that the portal
venous system was the major route in the absorption.
Blood levels of fluoranthene, pyrene, and benz[a]anthracene were examined after these
hydrocarbons were dissolved in Tween 80/isotonic saline and administered by gavage to rats at a
dose of 20 mg/kg b.w. (Lipniak and Brandys, 1993). Peak blood concentrations of the three
compounds were reached at 1-2 hours after administration. The peak concentration of
fluoranthene (approximately 30 mg/mL) was twice as high as that of pyrene, and 5 times higher
than benz[a]anthracene.
The bioavailability of a mixture of PAHs orally administered to rats was investigated by van
Schooten et al. (1997). Six rats were gavaged with a mixture of 2.1 μg anthracene, 17.5 μg
pyrene, and 7.6 μg benzo[a]pyrene dissolved in 2 mL sunflower oil. The area under the plasma
concentration time curve (AUC), determined for the 7-h plasma concentration time curve, and
expressed as % initial dose/mL/h, was 3.5 ± 2.4, 3.3 ± 1.9, 1.6 ± 1.6, respectively.
In 40 kg castrated pigs fed milk containing radiolabelled benzo[a]pyrene or phenanthrene (235

and 49 µg/L, respectively, corresponding to approximately 5.9 and 1.2 µg/kg b.w.). Cavret et al.
(2003) found absorption rates of 30.5% and 86.1%, respectively.
Following a single oral administration (100 mg per animal for each compound) in lactating goats,
absorption of phenanthrene and pyrene was about 75%, whereas absorption was 12% for
benzo[a]pyrene (Grova et al., 2002; Lapole et al., 2007).
In general in mammals, the oral absorption of the lower-molecular mass, i.e. 3- and 4-ring PAH
(such as anthracene, phenanthrene, fluoranthene or pyrene) was higher compared with very
lipophilic higher-molecular mass PAHs.
In humans, indirect evidence for the gastrointestinal absorption of PAH was provided by Buckley
and Lioy (1992) who detected in the urine the metabolite 1-hydroxypyrene, following ingestion
of diets containing very low levels of benzo[a]pyrene.
8.1.1.2 Distribution
The tissue distribution of PAHs was mainly investigated in rodents after intravenous or oral
administration of radiolabelled or unlabelled individual hydrocarbons. PAHs and/or metabolites
were detected in almost all organs. Highest levels were found in the gastrointestinal tract and in
all lipid-rich tissues (WHO/IPCS, 1998). In rats treated with benzo[a]pyrene (0.150 mg/kg b.w.)

alone or with phenanthrene (4.3 mg/kg b.w.) and pyrene (2.7 mg/kg b.w.) by oral gavage once
per day for 30 days, the highest concentration (34.5 ng/g) of benzo[a]pyrene, was found in
muscle of animals treated with benzo[a]pyrene alone at 20 days of treatment (Kang et al., 2007).
The highest phenanthrene concentration was 47.1 ng/g in muscle and 118.8 ng/g in fat, and for
pyrene it was 29.7 ng/g in muscle and 219.9 ng/g in fat, in rats treated by the mixture.
Saunders et al. (2002) demonstrated that significant amounts of benzo[a]pyrene and metabolites
(benzo[a]pyrene-trans-4,5-dihydrodiol (±), benzo[a]pyrene-trans-7,8-dihydrodiol (±),
benzo[a]pyrene-trans-9,10-dihydrodiol (±), benzo[a]pyrene-3,6-dione, 3-
hydroxybenzo[a]pyrene, and 9-hydroxybenzo[a]pyrene) were found in brain of rats administered
by gavage a single dose of benzo[a]pyrene (12.5, 25, 50 and 100 mg/kg b.w.). The diol
metabolites (4,5; 7,8; 9,10-diols) were predominant in cerebellum, and cortex relative to plasma
across various time points. From the standpoint of time-course distribution of various
metabolites, the diol metabolites dominated in the earlier time points (up to 12 h post exposure)
and the hydroxyl metabolites in the later (from 24 to 96 h post exposure) time points. These
results are in accordance with previous data (Modica et al., 1983; Lipniak and Brandys, 1993)
published on other PAHs (benz[a]anthracene, chrysene, fluoranthene, pyrene, and
benz[a]anthracene) and confirms the ability of these compounds or their metabolites to cross the
blood-brain barrier.
Studies in pregnant mice and rats have shown that PAHs (benzo[a]pyrene, 7,12-
dimethylbenz[a]anthracene and 3-methylcholanthrene) were widely distributed in maternal
tissues and was detected in foetuses, showing that they crossed the placenta (EC, 2002).
8.1.1.3 Metabolism
For PAHs with planar aromatic structures, metabolic activation is required for their toxic,
mutagenic, and carcinogenic action (Miller and Miller, 1966; Miller, 1978; Harvey, 1991). In
most cases, cytochrome P450 (CYP) catalyzed epoxide formation is an initial step in the
activation process, followed by formation of other, highly reactive electrophilic metabolites
capable of binding to cellular macromolecules, in particular proteins and nucleic acids.
In total, three principal metabolic pathways have been described for PAHs (Xue and
Warschawsky, 2005). Historically, the bay region dihydrodiol epoxide pathway was the first
discovered as being crucial for the mutagenicity of many PAHs (Hall and Grover, 1990; Harvey,
1991). Later, the one-electron oxidation pathway leading to reactive PAH radicals was
discovered (Cavalieri and Rogan, 1992; 1995). More recently, the ortho-quinone pathway, also
resulting in reactive eletrophilic PAH metabolites, has been described (Penning et al., 1996;
1999).
Over the last fifty years, the individual steps of the bay region dihydrodiol epoxide pathway have

been investigated. Studies on metabolism, covalent binding to DNA, mutagenicity, and

carcinogenicity have provided good evidence that a number of PAHs are activated by this
pathway. These include benzo[a]pyrene (Dipple et al., 1984; Gelboin, 1980), chrysene,
(Nordquist et al., 1981), 5-methylchrysene (Hecht et al., 1985), benzo[c]phenanthrene (Dipple et
al., 1987); benz[a]anthracene (Levin et al., 1978); 7,12-dimethylbenz[a]anthracene (Slaga et al.,
1979); and dibenzo[a,l]pyrene (Ralston et al., 1994; 1995).
The pathway in principle proceeds in three enzyme-catalyzed steps. First, an aromatic C-C
double bond is oxidized by CYP enzymes to the corresponding arene oxide (epoxide). Then,
microsomal epoxide hydrolase (EH) converts the epoxide to the corresponding trans-dihydrodiol.
Finally, another C-C double bond adjacent to the diol group is oxidized leading to a vicinal diol
epoxide. The bay region or fjord region diol epoxides are particularly electrophilic and can react
with nucleophiles. Their covalent binding to DNA is considered to be a strong pro-mutagenic
event.
In fact, the situation is more complex since most C-C double bonds in a typical PAH molecule
are pro-chiralic, i.e., their reaction products can have diverse spatial conformations. The enzymes
involved (mainly CYPs and EH) catalyze highly stereoselective reactions leading to optically
active products. With benzo[a]pyrene as an example, epoxidation of the 7,8-bond mainly leads to
7R,8S-oxide while the 7S,8R-stereoisomer is a very minor product (Yang, 1988). Upon catalysis
by EH, however, the 7R,8S-epoxide of benzo[a]pyrene is converted to the 7R,8R-diol, while the
7S,8R-oxide is converted to the 7S,8S-diol (Levin et al., 1980), since the addition of the OH-
group derived from the water molecule results in inversion of the configuration at the reacting
carbon centre. A similar stereoselectivity of epoxide and diol formation was reported for other
PAHs (Yang, 1988). PAH diols thus formed are further metabolized by CYPs to the bay region
diol epoxides which are thought to act as ultimate mutagens/carcinogens. The epoxidation can
lead to both a cis and trans orientation of the epoxide oxygen and the adjacent hydroxyl group of
the diol. Again the reaction is catalyzed by CYPs in a more or less stereoselective manner. The
major bay region dihydrodiol derived from benzo[a]pyrene, 7R,8R-benzo[a]pyrene-diol, is
mainly converted to the (+)anti-isomer, (7R, 8S, 9S, 10R)-diol-epoxide-benzo[a]pyrene, the most
potent mutagen among the four 7,8-diol-9,10-epoxide isomers formed from benzo[a]pyrene.
Other PAHs such as benz[a]anthracene, dibenz[a,j]anthracene, 7,12-dimethylbenz[a]anthracene,
chrysene, 7-methylbenz[a]anthracene, benzo[c]phenanthrene, and 5-methylchrysene are also
metabolically converted to diol-epoxides in a stereoselective way (Baird and Ralston, 1997). The
pattern of isomers formed is dependent on the species and CYP enzyme involved (Harvey, 1991).
Studies with recombinant human CYPs revealed that CYP1A1 was the most effective catalyst for
most reactions while CYP1A2 and CYP1B1 were more active catalyzing the formation of
benzo[a]pyrene-7,8-diol (Shimada et al., 1999a; 1999b). In fact, CYP1B1-knockout mice treated
with 7,12-dimethylbenz[a]anthracene or dibenzo[a,l]pyrene showed decreased tumour rates when
compared to wild-type mice (Shimada and Fujii-Kuriyama, 2004).

PAH-derived diols can undergo phase 2 metabolism catalyzed by sulfotransferases and/or

uridine-diphosphate-glucuronyltransferases. Electrophilic epoxides can be detoxified by non-
enzymatic and/or enzyme-catalyzed reaction with glutathione. In spite of extensive knowledge on
genetic polymorphisms of certain phase 1 and phase 2 enzymes, little is know about the
consequences of these polymorphisms with respect to cancer susceptibility in humans (Shimada,
2006).
In the one-electron oxidation pathway PAHs can form radical cations by removal of one of the π-
electrons, i.e., transfer of one electron to an oxidant. Such a reaction could be catalyzed, e.g., by
CYP peroxidase as suggested by Cavalieri and Rogan (1992; 1995). In experiments investigating
this pathway, one-electron oxidation was either initiated by an electrochemical technique or with
peroxidases such as horseradish peroxidase. It was suggested that PAHs with a lower ionization
potential below 7.35 eV such as benzo[a]pyrene, 7,12-dimethylbenz[a]anthracene or 3-
methylcholanthrene are more vulnerable to one-electron oxidation and form more stable radicals
whereas those with higher ionization potential such as benzo[a]pyrene, chrysene or
benzo[e]pyrene are less likely to be activated by this pathway (Cavalieri and Rogan, 1992). With
some PAHs such as 3-methylcholanthrene one-electron oxidation was reported to account for
most of the DNA adducts formed with liver microsomes from 3-methylcholanthrene-induced
rats. All adducts of this type formed with benzo[a]pyrene were removed by depurination, while
the stable adducts were mostly derived from the bay region dihydrodiol epoxide pathway
(Devanesan et al., 1992). In summary, the one-electron pathway results in unstable adducts but
may play a major role in the generation of apurinic sites. Model experiments suggest that the
relative amount of both pathways is strongly influenced by the ionization potential of the
individual PAH.
In the the ortho-quinone pathway the diols formed from PAHs are further oxidized by
dihydrodiol dehydrogenases (DD). DDs belong to the aldo-keto reductase superfamily of genes.
DDs convert the non-K-region diols to the corresponding ketols with nicotinamide adenine
dinucleotide phosphate (NADP+) serving as the typical electron acceptor (Penning et al., 1996;
1999). The ketols rearrange in a spontaneous reaction to form the catechol which, in the presence
of oxygen, can undergo non-enzymatic oxidation to the corresponding ortho-quinones. The PAH-
ortho-quinones are reactive metabolites acting as Michael acceptors which can form both stable
and unstable, depurinating DNA adducts (McCoull et al., 1999; Shou et al., 1993). Furthermore,
the ortho-quinones, upon reduction can form the catechol or semiquinone radical. These products
can transfer electrons to oxygen thus leading to the generation of reactive oxygen species (ROS)
such as superoxide anions or hydroxyl radicals. In fact, PAH ortho-quinones can lead to the
formation of oxidative DNA damage in vitro model systems (Park et al., 2004).
Besides these, other pathways have been described to play a role in the metabolic activation, most
notably the formation of sulfuric acid esters from hydroxymethylated PAH metabolites. Release
of the sulfate anion leads to the formation of an ultimate electrophilic benzylic carbonium ion

(Surh and Miller, 1994). In the liver of rats treated with 7-hydroxymethyl-12-
methylbenz[a]anthracene, sulfation-dependent formation of DNA adducts has been reported
(Flesher et al., 1997).
In addition to being substrates of CYP enzymes, many PAHs act as inducers of drug-
metabolizing enzymes. Induction is thought to act via activation of the aryl hydrocarbon receptor
(AhR) mainly affecting AhR-dependent genes including CYP1A1, CYP 1A2, and CYP1B1
(Shimada, 2006). Furthermore, a variety of phase II enzymes are also induced (Bock et al., 1998;
Elovaara et al., 2007). Comparative analysis revealed a wide range of potencies of PAHs as
inducers of, e.g., CYP1A1 (Till et al., 1999; Shimada, 2006). Induction of AhR-dependent CYPs
has been described in humans after exposure to tobacco smoke or after consumption of charcoal-
grilled meat (Fontana et al., 1999).
Many of the enzymes involved in the metabolism of PAHs have been shown to be polymorphic.
A genetic polymorphism is a genetic difference occurring in at least 1% of the population. A
genetic polymorphism in xenobiotic metabolising enzymes may change the ratio of
activation/deactivation of PAHs. An increased risk of certain cancers has been linked to certain
genetic polymorphisms (Garte and Crosti, 1999). The exact role of genetic polymorphisms on
risk for PAH carcinogenesis has not been completely elucidated, because compensatory
mechanisms or alternative enzymatic pathways may also be involved. Currently, no clear picture
of the effects of genetic polymorphisms of certain CYPs on the carcinogenic risk of humans
exposed to PAHs can be derived. It seems likely that the influence of polymorphisms depends on
the level of exposure being more important at low doses of exposure.
Garte and Crosti (1999) have proposed a nomenclature system for gene polymorphisms in the
CYP1A1 and glutathione-S-transferase-M1 (GSTM1) enzymes (among others). The gene name is
followed by an asterisk, followed by an Arabic number (1 for the wild type) designating the
specific polymorphism in chronological order of first publication. Letters A, B, etc. may follow
the final number when allelic subtypes exist.
An increased risk of cancer and a higher level of PAH-DNA adducts in lymphocytes was
observed in smokers and coke-oven workers that have the CYP1A1*2/ or CYP1A1*3/ genotypes
and do not express the detoxification enzyme GSTM1 (GSTM1*2/2) (Rojas et al., 2000). Similar
results were reported by Alekandrov et al. (2002) and Lodovici et al. (2004b). CYP3A4, the
major cytochrome P-450 in human liver is also involved in PAH metabolism but the
polymorphism detected in this gene has not been linked to altered cancer risk (Agundez, 2004).
Polymorphisms have also been described in several other enzymes involved in PAH metabolism,
such as microsomal epoxide hydrolase (mEH), myeloperoxidase (MPO) and NAD(P)H:quinone
oxidoreductase (NQO1).
In summary, carcinogenic and mutagenic PAHs can undergo at least three major metabolic
pathways which form highly reactive intermediates. The bay region dihydrodiol epoxide pathway

is regarded as the quantitatively most important pathway leading to stable, pro-mutagenic DNA
adducts. It is catalyzed by CYPs and EH, has been described for all strongly
mutagenic/carcinogenic PAH, and finally leads to the formation of electrophilic bay region diol-
epoxides which belong to the most potent chemical mutagens reported so far. The one-electron
oxidation pathway seems to play a relevant role for certain PAHs with lower ionization potential.
Typically, it results in the formation of unstable DNA adducts eventually leading to apurinic
sites. More recently, the oxidation of PAH dihydrodiols by DDs to PAH-derived ortho-quinones,
has been described. The ortho-quinones act as highly reactive Michael acceptors which can form
both stable and unstable DNA adducts, the latter resulting in apurinic sites, and also lead to the
formation of DNA-damaging reactive oxygen species. In addition, other pathways including
sulfate conjugation of hydroxylated meso-methylated PAHs have been described and may, under
certain conditions, play a relevant role in metabolic activation of PAHs. Most but not all phenolic
PAH metabolites can be regarded as detoxification products. Further metabolism of phenols via
phase II enzymes includes glucuronidation and sulphate conjugation, while electrophilic
metabolites can undergo glutathione conjugation. Induction of certain drug metabolizing enzymes
by a number of PAHs occurs via binding and activation of the arylhydrocarbon receptor.
Currently, no judgement on a clear relationship between the chemical structure of a PAH and the
preferred metabolic route appears possible. Considerations on the role of the ionisation potential
(IP) value of an individual PAH suggest that certain N-heterocyclic PAHs are unlikely to undergo
the radical cation pathway (Xue and Warshawsky, 2005). However, all three metabolic pathways
may contribute to the carcinogenicity/genotoxicity of the PAHs discussed here, while animal
species, organ site, expression of activating/inactivating enzymes etcetera may determine the
relative importance of each mechanism.
8.1.1.4 Excretion
After initial metabolism, most of the detoxified portion of PAH compounds are excreted into the
bile as metabolites, then subsequently eliminated in the faeces. A smaller proportion is excreted
in urine. In rats orally administered with benzo[a]pyrene, between 4 and 12% of the dose was
eliminated in urine within 72 hours (Yamazaki and Kakiuchi, 1989). The two major metabolites
of benzo[a]pyrene found in urine are 3-hydroxy- and 9-hydroxybenzo[a]pyrene, part of them
being conjugated to sulphate or glucuronic acid (Bouchard and Viau, 1997; Ramesh et al., 2001).
Oral administration to rats a mixture of 2.1 μg anthracene, 17.5 μg pyrene, and 7.6 μg
benzo[a]pyrene dissolved in 2 mL sunflower oil (van Schooten et al., 1997) resulted in a
relatively low faecal excretion of unchanged compounds (0.4, 0.2 and 0.3% of dose,
respectively). In faeces, 17.0% of the original dose of pyrene was found as 1-hydroxypyrene and
8.8% of the original dose of benzo[a]pyrene was recovered as 3-hydroxybenzo[a]pyrene. In

urine, the total amount of 1-hydroxypyrene excreted is 3.4% of the administered dose, whereas
no detectable levels of 3-hydroxybenzo[a]pyrene could be found.
PAH metabolites and parent compounds were monitored in urine of rats treated with
benzo[a]pyrene (0.150 mg/kg b.w.) alone or with phenanthrene (4.3 mg/kg b.w.) and pyrene (2.7
mg/kg b.w.) by oral gavage once per day for 30 days (Kang et al., 2007). Benzo[a]pyrene and the
3-hydroxybenzo[a]pyrene, a major metabolite of benzo[a]pyrene, was not detected during
treatment or after withdrawal of treatment in both the benzo[a]pyrene alone or benzo[a]pyrene+
phenanthrene+pyrene exposed groups. During the treatment, the major compounds found were 1-
hydoxypyrene (201-263 ng/mL), 3-hydroxyphenanthrene (114-161 ng/mL), unchanged
phenanthrene (41-69 ng/mL) and pyrene (9-17 ng/mL). The levels of pyrene, phenanthrene and
their metabolites in urine were rapidly decreased after withdrawal of treatment.
The process of excretion of PAH compounds into the intestine via the bile, reabsorption, and
return to the liver by the portal circulation (enterohepatic recirculation) has been demonstrated to
occur for 7,12-dimethylbenz[a]anthracene, benzo[a]pyrene, anthracene and pyrene (Ramesh et
al., 2004). Enterohepatic circulation extends the residence time of PAHs in the body and may
lead to long half-lives of reactive PAH metabolites. Though enterohepatic recycling of PAHs has
been studied in animal models, it has not been confirmed in human studies (de Kok and van
Maanen, 2000).
A single dose (approximately 100 µg/animal) of 14C-labelled benzo[a]pyrene, 14C-labelled

phenanthrene or 14C-labelled pyrene was orally administered to lactating goats and the kinetics of
elimination was studied (Grova et al., 2002). Cumulated excretion of radioactivity in milk during
the five days following administration was 0.2, 1.5 and 1.9% of dose, respectively. Lapole et al.
(2007) demonstrated that milk excretion of phenanthrene and pyrene in goat mainly occurred as
3-hydroxy-phenanthrene and 1-hydroxy-pyrene, respectively.
8.1.1.5 Biomarkers of exposure

The most prominent biomarker of exposure to PAHs in humans is the level of one or more PAH
metabolites is urine. 1-Hydroxypyrene, hydroxynaphtalenes and hydroxyphenanthrenes are the
PAH metabolites most widely accepted as reliable biomarkers for internal PAH exposure because
they are excreted in urine at substantial extent, making possible then monitoring by analytical
methods currently available. They have been extensively used for the estimation of PAH
exposure at different workplaces (Jongeneelen et al., 1990; Angerer et al., 1997). While 1-
hydroxypyrene was also suggested to be useful as a marker of PAH exposure in smokers (Scherer
et al., 2000; Yang et al., 2003), there are few studies suggesting a significant influence of dietary
habits. In a study by van Maanen et al. (1994), increased urinary levels of 1-hydroxypyrene and
blood levels (in mononuclear white blood cells) of modified DNA bases co-eluting with

(benzo[a]pyrene-7,8-diol-9,10-epoxide)-deoxyguanosine standards were found in volunteers after

consumption of two grilled hamburgers per day over five days. The benzo[a]pyrene levels in the
consumed hamburgers were in the range of 8.6-26.5 µg/kg. In a subsequent study with grilled
hamburgers with lower benzo[a]pyrene levels, no such changes were found.
Attempts have also been made to analyse the intact glucuronides of PAH metabolites such as 1-
hydroxypyrene glucuronide (1-OHPG) in human urine. A study by Fagundes et al. (2006) in
Southern Brazil revealed that smoking and extensive drinking of Maté (leaves frequently dried
over open fire) were statistically significantly associated with higher urine 1-OHPG
concentrations. However, 1-OHPG was detectable in urine from all 199 subjects tested.
With the development of more sensitive HPLC methods, the analysis of 3-

hydroxybenzo[a]pyrene in human urine with a limit of detection of 0.05 ng/L urine became
possible (Simon et al., 2000). Förster et al. (2008) found strong correlations of 3-
hydroxybenzo[a]pyrene with 1-hydroxypyrene and the sum of hydroxyphenanthrenes in urine for
a variety of workplaces (converter infeed, production of fireproof materials, coking plants,
production of graphite electrodes).
Li et al. (2008) published an extensive analysis of 22 urinary PAH metabolites namely of

naphthalene, fluorene, phenanthrene, pyrene, benzo[a]anthracene, benzo[c]phenanthrene, and
chrysene in 2748 individuals (Table 23). While metabolites of naphthalene, fluorene,
phenanthrene and pyrene were found in almost all samples, only a minority of samples contained
detectable amounts of metabolites of benzo[a]anthracene, benzo[c]phenanthrene and/or chrysene.
22 PAH metabolites were measured in urine samples using enzymatic deconjugation, followed by
automated solid phase extraction, and quantified by gas chromatograpgy-isotope dilution-high
resolution-mass spectrometry (GC-ID-HR-MS). The LODs ranged from 2.0 ng/L (for 3-
hydroxyfluorene) to 10.4 ng/L (for 3-hydroxybenz[a]anthracene). The geometric mean for 1-
hydroxypyrene was 49.6 ng/L urine. Smokers showed significantly higher urinary levels of all
metabolites with a geometric mean level above LOD (Table 24). In children higher levels were
found than in adolescents or adults. Among major metabolites significant correlations were
found. The authors suggest 1-hydroxypyrene as a useful surrogate marker representing PAH
exposure but do not provide detailed data on the diet of the participants.
At relatively high exposure via the diet (charcoal-grilled meat), but not at lower exposure,
increased levels of 1-hydroxypyrene in urine were reported (van Maanen et al., 1994). Viau et al.
(2002), however, found a large variation of 1-hydroxypyrene excretion in urine in volunteers
having consumed food with a defined range of pyrene contamination. Scherer et al. (2000) also
reported that the correlation between PAH exposure via the diet and 1-hydroxypyrene levels in
urine seems to be weak.

Table 23: Urinary PAH metabolites in 2748 individuals in the U.S. population (taken from Li et
al., 2008, modified).
PAH Hydroxy Percent Geometric mean urine Sum
metabolites detection concentration in adults (ng/L)
(ng/L)
Naphthalene (NAP) 1-OH-NAP 100 2190
2-OH-NAP 100 2620 4810
Fluorene (FLUO) 2-OH-FLUO 100 333
3-OH-FLUO 99 138
9-OH-FLUO 99 230 701
Phenanthrene (PHEN) 1-OH-PHEN 99 145
2-OH-PHEN 95 57
3-OH-PHEN 99 105
4-OH-PHEN 74 43
9-OH-PHEN 84 36 386
Pyrene (PYR) 1-OH-PYR 98 47 47
Benzo[c]phenanthrene 1-OH-BcP 9.8 <LOD
(BcP) 2-OH-BcP 6.8
3-OH-BcP 10
Benz[a]anthracene 1-OH-BaA 9.5 <LOD
(BaA) 3-OH-BaA 12
9-OH-BaA
Chrysene (CHR) 1-OH-CHR 28 <LOD
2-OH-CHR 12
3-OH-CHR 24
4-OH-CHR 4.6
6-OH-CHR 28
LOD: Limit of detection
Table 24: Urinary PAH metabolites in 2748 individuals in the U.S. population, stratified by
smoking status (taken from Li et al., 2008, modified).
PAH Hydroxy Geometric mean Geometric mean urine
metabolites urine concentration concentration in adult non-
in adult smokers smokers
(ng/L) (ng/L)
Naphthalene (NAP) 1-OH-NAP 6293 1523
2-OH-NAP 8597 1682
Fluorene (FLUO) 2-OH-FLUO 990 236
3-OH-FLUO 592 90
9-OH-FLUO 342 200
Phenanthrene (PHEN) 1-OH-PHEN 193 132
2-OH-PHEN 88 48
3-OH-PHEN 194 91
4-OH-PHEN 53 39
9-OH-PHEN 88 26
Pyrene (PYR) 1-OH-PYR 104 40
LOD: Limit of detection

8.1.1.6 Biomarkers of effect

In attempts to identify PAH-derived DNA adducts in peripheral blood lymphocytes various
techniques were used. Topinka et al. (2007) used the 32P-postlabelling technique to quantify
‘benzo[a]pyrene–like adducts’ in police officers and found that the adduct levels were related to
genetic polymorphisms in the CYP1A1 and GSTM1 genes. Other methods comprise the analysis
of specific and unspecific defined DNA adducts and the analysis of DNA strand breaks by single-
cell gel electrophoresis (Marczynski et al., 2005). In particular, benzo[a]pyrene-diolepoxide
adducts, analysed by HPLC-FLD, were dependent on CYP1A1, GSTM1, and GSTP1 genotypes
whereas the overall influence of smoking was poor (Lodovici et al., 2004b). In another study,
using the same method, a strong influence of occupational exposure in coke oven workers but a
weak effect of smoking or consumption of charcoal-broiled meat was found (Pavanello et al.,
2004). Pastorelli et al. (1999) found no significant effect of diet or smoking habits on
benzo[a]pyrene-diolepoxide haemoglobin adducts, measured by GC-MS as benzo[a]pyrene-
tetrols released from adducted haemoglobin, in blood samples from the Italian general
population. The levels were also unrelated to 1-hydroxypyrene levels in urine.
In summary, biomarkers of effect such as DNA adducts, DNA strand breaks and haemoglobin
adducts in peripheral blood seem to be in the status of evaluation. The quantitative relationship
between those markers and the levels of exposure seems to be weak. Although available
analytical methods to detect macromolecular adducts are now highly sensitive and may be used
to evaluate occupational exposure to PAHs, they are not adapted to the context of human dietary
exposure to PAH contamination.
8.1.2 Toxicological studies

As indicated in the introduction, this opinion concerns the 15 PAHs identified by SCF in 2002 as
priority PAHs and benzo[c]fluorene as suggested by the JECFA in 2005. Because the focus of
this opinion is the risk of PAHs following oral exposure, only toxicity data on these compounds
via this route of administration are included. For information on other PAHs and other routes of
administration the reader is referred to WHO/IPCS (1998), EC (2002), and FAO/WHO (2005).
8.1.2.1 Acute and short-term toxicity

Limited acute oral toxicity studies indicate that PAHs have moderate to low acute toxicity. An
LD50 value of more than >1600 mg/kg b.w. has been reported for benzo[a]pyrene in mice and
rats (WHO/IPCS, 1998).

The NOEL for benzo[a]pyrene in a 90-day study in rats was 3 mg/kg b.w. per day on the basis of
toxicity in the liver (Kroese et al., 2001). For the other carcinogenic PAHs no information is
available.
8.1.2.2 Reproductive toxicity

There is limited information on the reproductive toxicity of individual PAHs. For benzo[a]pyrene
no effect on reproductive capacity was found in a one-generation study in mice receiving diets
containing doses of up to 133 mg/kg b.w. per day. However, impaired fertility was seen in the
offspring of female mice given benzo[a]pyrene by gavage at doses >10mg/kg b.w. per day.
Developmental toxicity of benzo[a]pyrene was observed in mice of a susceptible genotype
following administration of 120 mg benzo[a]pyrene/kg b.w. per day via the diet. A NOAEL for
reproductive and developmental effects via the oral route has not been established (FAO/WHO,
2005).
8.1.2.3 Immunotoxicity
The immunosuppressive effects of PAHs have mainly been investigated in studies using
parenteral administration. It has been suggested that PAHs exert immune effects via the aryl
hydrocarbon receptor. Observations in CYP1A1 knock-out mice have indicated that CYP1A1
may protect against immunotoxic effects by benzo[a]pyrene. In a study on immunosuppressive
effects in rats administered benzo[a]pyrene by gavage, a NOEL of 3 mg/kg b.w. per day has been
identified (EC, 2002).
8.1.2.4 Carcinogenicity
The carcinogenicity of PAHs administered by dermal, subcutaneous, inhalation or oral routes has
been assessed in a large number of studies. In most studies, the site of tumour development was
related to the route of administration, e.g. gastric tumours after oral administration, skin tumours
after dermal application. However, tumours at sites other than the site of application were also
observed. For the current opinion data on the carcinogenicity of PAHs following oral
administration are most relevant. Several studies on individual PAHs have been evaluated earlier
by SCF (EC, 2002) and the JECFA (FAO/WHO, 2006). Benzo[a]pyrene, when administered by
the oral route, has been reported to produce tumours of the gastrointestinal tract, liver, lungs and
mammary glands of mice and rats. None of the other priority PAHs have been tested for
carcinogenicity after oral administration. In accordance with SCF and the JECFA the CONTAM
Panel considered two studies using oral administration, a study in mice comparing

benzo[a]pyrene with two coal tar mixtues, and a study in rats given benzo[a]pyrene, as relevant
for the assessment of the carcinogenic potential of PAHs in food.
Groups of 48 female B6C3F1 mice were fed diets containing benzo[a]pyrene at a concentration
of 0, 5, 25 or 100 mg/kg of diet (equivalent to doses of 0, 0.7, 3.6 or 14 mg/kg b.w. per day) for 2
years (Culp et al., 1998). Papillomas and carcinomas were observed in the forestomach,
oesophagus, and tongue (see Table 25).
In the same experiment, groups of 48 female B6C3F1 mice were fed diets containing 0, 0.01,
0.03, 0.1, 0.3, 0.6 or 1.0% coal tar mixture I, which contained benzo[a]pyrene at a concentration
of 2240 mg/kg (equivalent to doses 0.03, 0.09, 0.32, 0.96, 1.92 or 3.2 mg/kg b.w. per day), or 0,
0.03, 0.1 or 0.3% of coal tar mixture II, which contained benzo[a]pyrene at a concentration of
3669 mg/kg (equivalent to doses of 0.16, 0.52 or 1.1 mg/kg b.w. per day).
A significantly increased incidence of alveolar and bronchiolar adenomas and carcinomas was
found at 0.3, 0.6 and 1.0% of mixture I and at 0.1 and 0.3% of mixture II (see Table 25). For
tumours of the forestomach, a significant increase was observed at 0.3, 0.6 and 1.0% of mixture I.
The total numbers of tumour-bearing animals were 5/48, 12/48, 14/48, 12/48, 40/48, 42/48 and
43/48 at 0, 0.01, 0.03, 0.1, 0.3, 0.6 and 1.0% of mixture I, and 5/48, 17/48, 23/48 and 44/48 at 0,
0.03, 0.1 and 0.3% of mixture II, respectively. This study indicated that benzo[a]pyrene alone
induced only tumours of the alimentary tract, whereas the coal tar mixtures also induced liver and
lung tumours.

Table 25: Incidence of neoplasms in female B3C6F1 mice fed benzo[a]pyrene or two coal tar mixtures (Culp et al., 1998) based on the JECFA
(FAO/WHO, 2005).
Tumour type Benzo[a]pyrene concentration in diet Coal tar mixture 1 concentration in diet Coal tar mixture 2 in diet
(mg/kg)a (%)b (%)c
0 5 25 100 0 0.01 0.03 0.1 0.3 0.6 1.0 0.03 0.1 0.3
Hepatocellular adenomas and/or 2/48 7/48 5/47 0/45 0/47 4/48 2/46 3/48 14/45 1/42 5/43 7/47 4/47 10/45
carcinomas
Lung alveolar/bronchiolar adenomas 5/48 0/48 4/45 0/48 2/47 3/48 4/48 4/48 27/47 25/45 21/45 4/48 10/48 23/47
and/or carcinomas
Forestomach papillomas and/or 1/48 3/47 36/46 46/47 0/47 2/47 6/45 3/47 14/46 15/45 6/41 3/47 2/47 13/44
carcinomas
Oesophagus papillomas and/or 0/48 0/48 2/45 27/46 −d − − − − − − − − −
carcinomas
Small intestine adenocarcinomas − − − − 0/47 0/46 0/45 0/47 0/42 22/36 36/41 0/47 0/47 1/37
Tongue papillomas and/or 0/48 0/48 2/46 23/48 − − − − − − − − − −
carcinomas
Larynx papillomas and/or carcinomas 0/35 0/35 3/34 5/38 − − − − − − − − − −
Hemangiosarcomase 1/48 2/48 3/47 0/48 1/48 0/48 1/48 1/48 11/48 17/48 1/45 1/48 4/48 17/48
Histiocytic sarcomas 2/48 2/48 1/47 0/48 1/48 0/48 0/48 1/48 7/48 5/48 0/45 3/48 2/48 11/48
Sarcomasf 1/48 2/47 7/47 0/48 1/48 4/48 3/48 2/48 7/48 1/48 2/45 0/48 4/48 5/48
Total tumour-bearing animalsg - - - - 5/48 12/48 14/48 12/48 40/48 42/48 43/48 17/48 23/48 44/48
a
Benzo[a]pyrene intake levels in dose groups were 0, 0.6, 3 and 12 mg/kg b.w. per day respectively (calculated daily intake, corrected for reduced food consumption).
b
Concentrations of benzo[a]pyrene in the diet in these groups were 0.22, 0.66, 2.2, 6.6, 13.4 and 22.0 mg/kg respectively (determined by HPLC).
c
Concentrations of benzo[a]pyrene in the diet in these groups were 1.1, 3.7, and 11.1 mg/kg respectively (determined by HPLC).
d
− no increase reported (actual tumour incidence not given in Culp et al., 1998).
e
Organs involved included skin, liver, mesentery, mesenteric lymph nodes, heart, spleen, urinary bladder, uterus, thoracic cavity, ovary and skeletal muscle .
f
Organs involved included mesentery, forestomach, skin and kidney.
g
Data on tumour-bearing animals were not included in Culp et al., (1998), but were reported by Schneider et al., (2002).

Table 26: Incidence of neoplasms in male and female Wistar rats gavage-dosed with benzo[a]pyrene (Kroese et al., 2001) based on the JECFA
(FAO/WHO, 2005).
Site Benzo[a]pyrene dose (mg/kg body weight)
0 3 10 30
Males Females Males Females Males Females Males Females
Oral cavity (papillomas, carcinomas, keratoacanthomas, carcinomas) 1/24a 1/19 0/24 0/21 7/37 1/9 25/38 23/31
Oesophagus (sarcomas) −b 0/52 − 1/52 − 9/52 − 0/52
Forestomach (papillomas, carcinomas) 0/52 1/52 8/52 6/51 43/52 30/51 52/52 50/52
Duodenum (adenocarcinomas) 0/51 0/49 0/50 0/48 0/51 0/50 1/49 2/51
Jejunum (adenocarcinomas) 0/51 0/50 0/50 0/48 1/51 0/50 8/49 2/51
Liver (adenomas, carcinomas, cholangiomas, cholangiocarcinomas) 0/52 0/52 4/52 2/52 39/52 41/52 50/52 51/52
Kidney (adenomas, carcinomas) 0/52 − 0/52 − 9/52 − 11/52 −
Auditory canal (papillomas, adenomas, carcinomas) 0/1 0/0 0/0 0/1 3/7 0/0 24/33 15/20
Skin and mammary (adenomas, carcinomas, acanthomas, epitheliomas, 4/52 8/52 6/52 15/52 11/52 11/51 32/51 3/52
sarcomas, histiocytomas)
Pituitary (adenomas, carcinomas) 32/51 29/46 31/52 24/52 23/51 12/47 1/50 8/52
Total tumour-bearing animals 6/52 8/52 16/52 20/52 51/52 47/52 52/52 51/52
a
Some tissues were examined histologically only when abnormalities were observed upon macroscopic examination.
b
− no increase reported (actual tumour incidence not given in Kroese et al., 2001).

Administration of oral doses of benzo[a]pyrene at 0, 3, 10 or 30 mg/kg of body weight per day by

gavage to groups of 104 male and female Wistar rats on 5 days per week for 2 years resulted in a
large variety of tumours, most prominent being those in the liver and forestomach (Kroese et al.,
2001). Papilloma and carcinoma were observed in the forestomach, and adenoma and carcinoma
in the liver of both female and male rats. For tumour incidences see Table 25. In addition to these
tumours, treatment with benzo[a]pyrene also induced soft tissue sarcomas (skin, mammary), and
tumours of the auditory canal, oral cavity, small intestine and the kidney (Table 26). The total
numbers of tumour-bearing animals in the different dose groups were 8/52, 20/52, 47/52, 51/52
for females, and 6/52, 16/52, 51/52, 52/52 for males, at 0, 3, 10 or 30 mg/kg of body weight per
day, respectively.
8.1.2.5 Genotoxicity
On the basis of the available information the JECFA concluded that 15 individual PAHs are
clearly genotoxic, both in vitro and in vivo (FAO/WHO, 2006). These genotoxic PAHs are
benz[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene, benzo[ghi]perylene,
benzo[j]fluoranthene, benzo[k]fluoranthene, chrysene, cyclopenta[cd]pyrene,
dibenz[a,h]anthracene, dibenzo[a,e]pyrene, dibenzo[a,h]pyrene, dibenzo[a,i]-pyrene,
dibenzo[a,l]pyrene, indeno[1,2,3-cd]pyrene and 5-methylchrysene.
An important observation was the binding of the active metabolites of PAHs to DNA,
predominantly to amino groups of guanine and adenine. The major stable adduct is formed at the
N2 position of desoxyguanosine. DNA adduct formation is generally regarded as one of the first
steps in carcinogenicity of the mutagenic PAHs. However, there is a poor quantitative
relationship between levels of tissue adduct and tumour formation.
In the lungs of rats fed with coal tar, levels of benzo[c]fluorene-derived adducts were higher than
those of benzo[a]pyrene-derived adducts (Koganti et al., 2000), indicating that benzo[c]fluorene
may contribute to the formation of lung tumours after oral exposure to coal tar (Goldstein, 2001).
The CONTAM Panel noted, however, that no information was available on the amount of
benzo[c]fluorene in the coal tar mixtures.
8.1.2.6 Observations in humans

Most data, both on the effects of PAHs in human populations and on biomarkers of exposure to
PAHs, mainly refer to occupational and environmental exposure. The available evidence
regarding oral exposure to PAHs was indirect and did not include data on quantitative exposure,
and thus was not suitable for use in this opinion.

8.1.3 Dose-response modelling

The benchmark dose (BMD) approach was suggested by Crump (1984) as an alternative to the
NOAEL and LOAEL for non-cancer health effects because it provides a more quantitative
alternative to the first step in the dose-response assessment than the NOAEL/LOAEL. The BMD
is based on a mathematical model being fitted to the experimental data within the observable
range and estimates the dose that causes a low but measurable response (the benchmark response
BMR) typically chosen at a 5 or 10% incidence above the control. The BMD lower limit
(BMDL) refers to the corresponding lower limits of a one-sided 95% confidence interval on the
BMD. Using the lower bound takes into account the uncertainty inherent in a given study, and
assures (with 95% confidence) that the chosen BMR is not exceeded.
For the evaluation of human and experimental animal data EFSA has proposed to use the BMD
methodology to derive a reference point on the dose-response curve (EFSA, 2005a). The use of
the BMDL calculated for a BMR of the confidence limit 10% (BMDL10), is considered an
appropriate reference point for compounds that are both genotoxic and carcinogenic. Such a
value is the lowest statistically significant increased incidence that can be measured in most
studies, and would normally require little or no extrapolation outside the observed experimental
data.
The CONTAM Panel noted that the JECFA in its evaluation of PAHs performed BMD modelling
on both the rat study on benzo[a]pyrene (Kroese et al., 2001) and on the carcinogenicity study by
Culp et al. (1998) on two mixtures of coal tar (Mixture 1 and Mixture 2) in female B6C3F1 mice
using the benzo[a]pyrene content as a marker of the carcinogenic PAHs in the coal tar mixtures.
Overall, the JECFA concluded that the most appropriate end-point for BMD modelling was the
total tumour-bearing animals from the study of the coal tar mixtures in female mice and
consequently calculated BMD10s and BMDL10s for benzo[a]pyrene as a marker using eight
different mathematical models. The calculated BMD10s were from 0.13-0.29 mg
benzo[a]pyrene/kg b.w. per day and the BMDL10s were from 0.10-0.23 mg benzo[a]pyrene/kg
b.w. per day. The JECFA used the lower end of the range of BMDL10s, 0.1 mg
benzo[a]pyrene/kg b.w. per day, to derive Margins of Exposure (MOEs) for its risk assessment of
PAHs in food.
The CONTAM Panel also considered the 2-year carcinogenicity study on coal tar mixtures
performed by Culp et al. (1998) as the most adequate study for dose-response modelling. The
Panel performed BMD modelling on the total number of tumour-bearing animals in that study as
reported by Schneider et al. (2002) (Table 28). As reported by the JECFA the two coal tar
mixtures did not produce significantly different dose-response curves and therefore the data were
combined. However, the results for the animals receiving the two highest doses of Mixture 1
were omitted due to the premature death of all animals in these dose groups.
However, in addition to using only benzo[a]pyrene as a marker for the carcinogenic PAHs in
food, the CONTAM Panel explored additionally the use of:

1) benzo[a]pyrene and chrysene (PAH2),
2) benz[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene and chrysene (PAH4),
3) the sum of the eight carcinogenic PAHs (benz[a]anthracene, benzo[b]fluoranthene,

benzo[k]fluoranthene, benzo[ghi]perylene, benzo[a]pyrene, chrysene,
dibenz[a,h]anthracene, and indeno[1,2,3-cd]pyrene) (PAH8) that were all quantified in
the two coal tar mixtures.
The concentrations of the eight carcinogenic PAH in the coal tar mixtures are given in Table 27.
Table 27: Concentrations (mg/kg) of carcinogenic PAHs in the two coal tar mixtures tested in
mice by Culp et al. (1998).
Compound Mixture 1 Ratio to BaP Mixture 2 Ratio to BaP
Benzo[a]pyrene 1837 1 2760 1
Chrysene 2379 2960
PAH2 4216 2.30 5720 2.07
Benzo[b]fluoranthene 2097 2890
Benz[a]anthracene 2374 3340
PAH4 8687 4.73 11950 4.33
Benzo[k]fluoranthene 699 1010
Benzo[ghi]perylene, 1493 2290
Dibenz[a,h]anthracene 1353 1990
Indeno[1,2,3-cd]pyrene 267 370
PAH8 12499 6.80 17610 6.38
8.1.3.1 Benchmark dose modelling

The US EPA BMD software (BMDS) version 1.4.1 was used (US EPA, 2007) for modelling the
total tumour-bearing animals. For carcinogenicity data, a number of models are available in the
BMDS, and model fitting, determination of goodness-of-fit, and comparing models to decide
which one to use for obtaining the BMDL10 are outlined. The following dose-response models
were fitted to the dose-incidence data:
• Gamma multihit model
• Logistic model
• Log-logistic model

• Multistage model
• Log probit model
• Probit model
• Quantal linear model
• Weibull model
The BMD10 and BMDL10 values for an extra 10% risk compared to the background were
estimated by performing 250 iterations.
The acceptability of a model can be based on several criteria. Some of the models are nested
models (i.e. they are related to each other such that by leaving out a parameter, one model
reduces to the other; this holds for the one, two, and three-stage models). The fit should not be
significantly worse (using the likelihood ratio test) than the fit provided by the “full” model. The
full model is the model that does not assume any dose–response function (its parameters are
simply the frequencies per dose level) (Falk Filipsson et al., 2003).
For instance, when the full model has 8 parameters (i.e. 8 observed incidences available) a fitted
dose-response model with 3 parameters should result in a log-likelihood that is no more than 5.54
lower than the log-likelihood associated with the saturated model.
While the likelihood ratio test can only be applied to nested models, the Akaike information
criterion (AIC) has been proposed as an approximate criterion for comparing the fits of non-
nested models (Falk Filipsson et al., 2003).
In addition, the BMDS provides statistics for the suitability of the fit. The lower the chi-square
value the better the fit and the calculated p-value should be significantly larger than 0.1 which in
this case was chosen to represent a rejection level (Falk Filipsson et al., 2003).

Table 28: PAH doses and number of total tumour-bearing female mice fed two mixtures of coal tar for two years (Culp et al. (1998) as
reported by Schneider et al. (2002)).
Coal tar BaP dose BaP dose PAH2 dose PAH4 dose PAH8 dose Number of Number of
dose (μg/per (mg/kg b.w. per (mg/kg b.w. per (mg/kg b.w. per (mg/kg b.w. per animals per tumour-
(% in feed) day) day)* day)* day)* day)* group bearing
animals
Mixture 1
0 0 0 0 0 0 48 5
0.01 0.8 0.027 0.061 0.126 0.181 48 12
0.03 2.37 0.079 0.181 0.373 0.537 48 14
0.1 8 0.266 0.611 1.26 1.81 48 12
0.3 23.7 0.789 1.81 3.73 5.37 48 40
Mixture 2
0.03 3.63 0.121 0.250 0.523 0.771 48 17
0.1 13.2 0.440 0.910 1.90 2.80 48 23
0.3 36.3 1.21 2.50 5.23 7.71 48 44
* Assuming an average weight of 30 g for the female mice.

The BMD10 and BMDL10 values, as well as the associated statistics for the models used, are
presented in Tables 29, 30, 31, and 32. In addition to using the US EPA BMD software (BMDS)
the Panel also checked the outcome of the BMD10 calculations with the outcome of the PROAST
model applied by the JECFA in its evaluation (FAO/WHO, 2006). The PROAST model produced
the same BMD10 as the probit model in Table 29.
8.1.3.2 Benchmark dose calculations for benzo[a]pyrene

Table 29: BMD10 and BMDL10 calculations for benzo[a]pyrene based on total tumour-bearing
animals in the 2-year carcinogenicity study on coal tar mixtures by Culp et al. (1998).
Model Log likely- p-value AIC Chi- p-value Accept BMD10 BMDL10
hood square (mg/kg (mg/kg
(parameters) value b.w. per b.w. per
day) day)
Full model -198.8 (8)
Gamma multi-hit -204.6 (3) 0.04 415.2 11.0 0.05 No 0.24 0.09
Logistic -204.1 (2) 0.10 412.2 10.1 0.12 Yes 0.14 0.12
Log-logistic -204.1 (3) 0.06 414.2 9.89 0.08 Yes 0.28 0.18
Multi-stage -204.3 (3) 0.05 414.6 10.5 0.06 Yes 0.14 0.07
Log-probit -204.3 (3) 0.05 414.6 10.3 0.07 Yes 0.29 0.20
Probit -204.3 (2) 0.09 412.6 10.4 0.11 Yes 0.13 0.12
Quantal-linear -206.4 (2) 0.02 416.9 14.7 0.02 No 0.07 0.05
Weibull -204.6 (3) 0.04 415.2 11.0 0.05 No 0.18 0.08
Reduced model -262.9 (1) <0.001
The calculated accepted BMD10 values were from 0.13 to 0.29 mg benzo[a]pyrene/kg b.w. per
day in the feed, with the best fit being 0.13 to 0.14 mg/kg b.w. per day. The BMDL10 values
ranged from 0.07 to 0.20 mg/kg b.w. per day with 0.12 mg/kg b.w. per day representing the best
fit. Although based on the same dataset the current calculation resulted in slightly lower BMDL
values (0.07-0.20) than produced by the JECFA (0.10-0.23). This different outcome was due to
the use of different models for the calculation of the BMDL (bootstrap vs. profile likelihood) as
applied by the JECFA and the CONTAM Panel. In addition, the JECFA included a possible
difference between the two coal tar mixtures as a covariate in their analysis. The CONTAM
Panel considered the difference between the mixtures as not important. In order to be prudent the
CONTAM Panel used the lowest BMDL10 value among the accepted ones for the derivation of a
MOE. Thus, the BMDL10 of 0.07 mg/kg b.w. per day was chosen for benzo[a]pyrene as a marker
for the carcinogenic PAHs in food.

8.1.3.3 Benchmark dose calculations for PAH2

Table 30: BMD10 and BMDL10 calculations for the sum of 2 PAHs based on total tumour-
bearing animals in the 2-year carcinogenicity study on coal tar mixtures by Culp et al. (1998).
day) day)
Logistic -203.9 (2) 0.12 411.8 9.81 0.13 Yes 0.30 0.26
Log-logistic -204.1 (3) 0.06 414.2 9.82 0.08 Yes 0.61 0.41
Multi-stage -204.0 (3) 0.06 414.0 9.95 0.08 Yes 0.34 0.17
Log-probit -204.2 (3) 0.06 414.3 9.96 0.08 Yes 0.64 0.45
Probit -203.9 (2) 0.12 411.9 9.84 0.13 Yes 0.29 0.25
Weibull -204.3 (3) 0.05 414.7 10.4 0.06 No 0.44 0.20
Reduced model -262.9 (1) <0.001
The calculated accepted BMD10 values ranged from 0.29 to 0.64 mg for the sum of the 2 PAH/kg
b.w. per day in the feed, with the best fits being from 0.29 to 0.30 mg/kg b.w. per day. The
BMDL10 values ranged from 0.17 to 0.45 mg/kg b.w. per day with 0.25-0.26 mg/kg b.w. per day
representing the best fit. However, in order to be prudent the CONTAM Panel used the lowest
BMDL10 value among the accepted ones for the derivation of a margin of exposure. Thus, the
BMDL10 of 0.17 mg/kg b.w. per day was chosen for the sum of the 2 PAH as a marker for the
carcinogenic PAHs in food.


day) day)
Logistic -203.9 (2) 0.12 411.8 9.81 0.13 Yes 0.61 0.53
Log-logistic -204.1 (3) 0.06 414.1 9.80 0.08 Yes 1.27 0.84
Multi-stage -204.0 (3) 0.06 414.1 9.98 0.08 Yes 0.69 0.34
Log-probit -204.2 (3) 0.06 414.3 9.97 0.08 Yes 1.33 0.93
Probit -204.0 (2) 0.11 411.9 9.87 0.13 Yes 0.60 0.52
Weibull -204.4 (3) 0.05 414.7 10.5 0.06 No 0.91 0.41
Reduced model -262.9 (1) <0.001
The calculated accepted BMD10 values ranged from 0.60 to 1.33 mg for PAH4/kg b.w. per day in
the feed, with the best fits being from 0.60-0.61 mg/kg b.w. per day. The accepted BMDL10
values ranged from 0.34 to 0.93 mg/kg b.w. per day with 0.52-0.53 mg/kg b.w. per day
representing the best fits. However, in order to be prudent the CONTAM Panel used the lowest
BMDL10 value among the accepted ones for the derivation of a MOE. Thus, the BMDL10 of 0.34
mg/kg b.w. per day was chosen for the PAH4 as a marker for the carcinogenic PAHs in food.


day) day)
Logistic -203.9 (2) 0.12 411.8 9.80 0.13 Yes 0.89 0.77
Log-logistic -204.0 (3) 0.06 414.1 9.76 0.08 Yes 1.85 1.22
Multi-stage -204.1 (3) 0.06 414.2 10.1 0.07 Yes 0.97 0.49
Log-probit -204.2 (3) 0.06 414.3 9.99 0.08 Yes 1.93 1.35
Probit -204.0 (2) 0.11 412.0 9.95 0.13 Yes 0.87 0.76
Weibull -204.4 (3) 0.05 414.8 10.6 0.06 No 1.29 0.58
Reduced model -262.90 (1) <0.001
The calculated accepted BMD10 values ranged from 0.87 to 1.93 mg for PAH8/kg b.w. per day in
the feed, with the best fits being from 0.87-0.89 mg/kg b.w. per day. The BMDL10 values ranged
from 0.49 to 1.35 mg/kg b.w. per day with 0.76-0.77 mg/kg b.w. per day representing the best
fits. However, in order to be prudent the CONTAM Panel used the lowest BMDL10 value among
the accepted ones for the derivation of a MOE. Thus, the BMDL10 of 0.49 mg/kg b.w. per day
was chosen for the PAH8 as a marker for the carcinogenic PAHs in food.
8.1.4 Use of a TEF concept in the risk assessment of mixtures of carcinogenic PAHs
The application of the concept of response addition has been suggested by the US EPA (1986) to
determine the cancer risk from mixtures containing carcinogenic compounds. The assumption
was that such compounds show simple dissimilar action with a complete negative correlation of
tolerance. However, as pointed out by Könemann and Pieters (1996) for compounds with
presumed linear dose-response curves, such as genotoxic and carcinogenic compounds for which
it is assumed that a no-effect-level does not exist and for which the mechanism of action may be
regarded as similar, response addition and dose addition will provide identical results. Therefore,
various authors have used different terminology in the assessment of PAHs, relative response
factors, relative potency factors, or toxic equivalency factors (TEFs).
A number of PAHs as well as coal-tar and some occupational exposures to combustion emissions
containing these compounds have shown carcinogenicity in experimental animals and

genotoxicity and mutagenicity in vitro and in vivo (WHO/IPCS, 1998). Several attempts have
been made to derive relative potency factors, often expressed as TEFs for individual PAH
(relative to benzo[a]pyrene, the best studied PAH) with the purpose of summarising the
contributions from individual PAHs in a mixture into a total benzo[a]pyrene equivalent dose,
assuming additivity in their carcinogenic effects (Nisbeth and LaGoy 1992; Rugen et al., 1989;
Thorslund and Farrar 1990; Krewski et al., 1989; Larsen and Larsen, 1998). Because there is a
total lack of adequate data from oral carcinogenicity studies on individual PAHs other than
benzo[a]pyrene, TEF values for PAHs in food have been suggested based on studies using skin
application, pulmonary instillation and subcutaneous or intraperitoneal injections.
There are several problems in using the TEF approach in the risk assessment of PAHs in food.
The use of the TEF approach requires that the compounds in question exert the toxicological
effect by the same mechanism of action, such as is the case for the polychlorinated dibenzo-p-
dioxins and – dibenzofurans, which act through binding to the Ah-receptor. Although a number
of PAHs bind to the Ah receptor, this effect is not the only effect that determines the carcinogenic
potency of PAH. DNA binding and induction of mutations are other significant effects in the
carcinogenesis of PAH, and there is no indication that different PAHs are activated via the same
metabolic route, bind DNA in the same positions, and induce the same types of mutations in the
same organs or tissues. In fact, the study by Culp et al. (1998) showed that a coal-tar mixture
containing PAHs also produced tumours in other tissues and organs than those affected by
benzo[a]pyrene alone, and that the additional PAHs in the mixture did not significantly
contribute to the incidence of stomach tumours observed after benzo[a]pyrene alone.
The limitations in using the TEF approach for the assessment of carcinogenicity of PAHs
following oral administration was illustrated when it was used on the carcinogenicity data and the
analytical data on the PAH composition in the coal tar mixtures used in the study by Culp et al.
(1998). When the TEF values derived by Larsen and Larsen (1998) were used the carcinogenic
potency of both coal tar mixtures was predicted to be only approximately 1.5 times that of the
benzo[a]pyrene content. However, the observed potencies of the coal tar mixtures were up to 5
times that accounted for the benzo[a]pyrene content. In this case, the use of the TEF approach for
PAH carcinogenicity would underestimate it.
Schneider et al. (2002) also examined the use of the TEF approach on the data from the Culp et
al. (1998) study and from several other studies using dermal or lung application of PAH mixtures
of known composition. They used the TEF derived by Brown and Mittelsman (1993) and
concluded that the benzo[a]pyrene equivalency factors do not adequately describe the potency of
PAH mixtures and lead to underestimation of the carcinogenic potencies in most cases.
The CONTAM Panel concluded that the TEF approach to the risk characterisation for PAHs in
food was not considered to be scientifically valid because of the lack of data from oral
carcinogenicity studies on different PAHs, their different modes of action and the evidence of

poor predictability of the carcinogenic potency of PAH mixtures based on the currently proposed
TEF values.
9. Risk characterisation
Margins of exposure (MOEs) were calculated by dividing the lowest BMDL10 values among the
models with acceptable fits by the mean and high level estimates of dietary exposure to
benzo[a]pyrene, PAH2, PAH4 and PAH8 (Tables 33 and 34).
Table 33: Margins of exposure (MOEs) at the median of the EU mean estimates of dietary
exposure.
Marker for the Median EU34 estimated Exposure ratio BMDL10 MOE*
carcinogenic PAH mean dietary exposure to BaP (mg/kg b.w. per day)
in food (ng/kg b.w. per day)
BaP 3.9 1.00 0.07 17900
PAH2 10.7 2.73 0.17 15900
PAH4 19.5 4.97 0.34 17500
PAH8 28.8 7.36 0.49 17000
* Rounded to the nearest 100, EU: European Union
Table 34: Margins of exposure (MOEs) at the median of the EU 97.5th percentile estimates of
dietary exposure.
Marker for the Median EU34 estimated Exposure ratio BMDL10 MOE*
carcinogenic PAH 97.5%ile dietary to BaP (mg/kg b.w. per day)
in food exposure
(ng/kg b.w. per day)
BaP 6.5 1.00 0.07 10800
PAH2 18.0 2.77 0.17 9500
PAH4 34.5 5.32 0.34 9900
PAH8 51.3 7.91 0.49 9600
* Rounded to the nearest 100, EU: European Union
These MOEs indicate a low concern for consumer health at the mean estimated dietary
exposures. This applies to the full range of estimates of average exposures across EU Member
States (3.1-4.3 ng/kg b.w. per day, MOEs: 16,300-22,600 for benzo[a]pyrene alone and 23.6-35.6
ng/kg b.w. per day, MOEs: 13,800-20,800 for PAH8). However, for high level consumers the
MOEs are close to or less than 10,000, which as proposed by the EFSA Scientific Committee
34 Median EU is the median of all individual Member State averages for the dietary exposure.

(EFSA, 2005a) indicates a potential concern for consumer health and a possible need for risk
management action.
Comparison of the MOE values calculated for benzo[a]pyrene, PAH2, PAH4 and PAH8,
indicates that PAH2, PAH4 and PAH8 can be used as alternatives to benzo[a]pyrene alone as
markers of the carcinogenicity of the genotoxic and carcinogenic PAHs, and would be equally
effective.
10. Suitable indicators for occurrence and toxicity of genotoxic and carcinogenic PAHs
Based on the currently available data, the CONTAM Panel concluded that the most relevant
PAHs for use as indicators of the toxicity of the genotoxic and carcinogenic PAHs were those for
which oral carcinogenicity data were available, i.e. benzo[a]pyrene and the other seven genotoxic
and carcinogenic PAHs measured in the coal tar mixtures tested by Culp et al. (1998). However it
was not possible to adopt a TEF approach to the risk assessment of these PAHs because of the
lack of oral carcinogenicity data for any of the PAHs other than benzo[a]pyrene, their different
modes of action and the evidence of poor predictivity of the carcinogenic potency of PAHs
mixtures based on currently proposed TEF values.
From these eight PAHs, groups were identified based on the frequency of their analytical results
above the LOD in the occurrence data submitted by the Member States. Benzo[a]pyrene was not
detected in 28% of the samples analysed for PAH8 when at least one other PAH was measurable.
Therefore benzo[a]pyrene alone is not a suitable indicator for the occurrence and toxicity of
genotoxic and carcinogenic PAHs.
The most frequently detected PAH was chrysene, and therefore the CONTAM Panel investigated
the sum of benzo[a]pyrene plus chrysene as a possible indicator (termed PAH2). The mean
values of benz[a]anthracene and benzo[b]fluoranthene were higher than those of benzo[a]pyrene
and therefore these were included in a grouping with PAH2 (together termed PAH4). The final
group considered as a potential indicator was the complete 8 genotoxic and carcinogenic PAHs,
comprising PAH4 plus benzo[k]fluoranthene, benzo[ghi]perylene, dibenz[a,h]anthracene and
indeno[1,2,3-cd]pyrene (together termed PAH8).
Based upon the occurrence data (Table 12), PAH2 would identify 43% of the benzo[a]pyrene
negative samples in which other PAHs were present, with 2-9% of the samples containing
detectable concentrations of the remaining 13 PAHs. The correlation between PAH2 and PAH4
or PAH8 was 0.92. PAH4 reduces the number of negative samples with detectable levels of the
remaining 11 PAHs to 1-6% with a correlation to PAH8 of 0.99. PAH8 had a minimal impact on
the total number of samples positive for any PAH.
Since the calculated MOE values indicated that benzo[a]pyrene, PAH2, PAH4 and PAH8 were
equally effective as markers of the carcinogenicity, the occurrence data were more influential in

the overall conclusion that PAH4 and PAH8 were better indicators of the occurrence and toxicity
of the genotoxic and carcinogenic PAHs than benzo[a]pyrene or PAH2. Based on the current
pattern of occurrence in different food categories PAH8 did not provide much added value
compared to PAH4.
11. Uncertainty analysis

The evaluation of the inherent uncertainties in the assessment of exposure to PAHs has been
performed following the guidance of the Opinion of the Scientific Committee related to
Uncertainties in Dietary Exposure Assessment (EFSA, 2006). In addition, the draft report on
“Characterizing and Communicating Uncertainty in Exposure Assessment” which is in
preparation to be published as WHO/IPCS monograph, has been considered (WHO/IPCS, 2007).
According to the guidance provided by the EFSA opinion (EFSA, 2006) the following sources of
uncertainties have been considered: assessment objectives, exposure scenario, exposure model,
and model input (parameters).
11.1 Assessment objectives

The objectives of the assessment were clearly specified in the terms of reference. The CONTAM
Panel assessed the new occurrence data that were collected by EFSA, and evaluated whether
benzo[a]pyrene can still be considered as a suitable marker for both occurrence and carcinogenic
effects for the most relevant PAHs, and which food commodities in the different Member States
contribute most to the exposure to the most relevant PAHs. In its assessment the CONTAM Panel
used the margin of exposure approach (MOE) as adopted by the EFSA Scientific Committee in
the Opinion related to substances which are both genotoxic and carcinogenic (EFSA, 2005a). The
uncertainty in the assessment objectives is considered to be negligible.
11.2 Exposure scenario

In response to the Commission request of 2005 for the monitoring of PAHs in food, 18 countries
submitted analytical results covering testing of a variety of different food products mainly in
2005-2007. Germany submitted most of the results (64%). Overall nearly 10,000 data points were
included in the analysis. There is uncertainty in possible regional differences in PAH
contamination of food commodities, and the CONTAM Panel recognised that the data set is not
fully representative of food on the EU market. However, considering that the data set includes a
large number of analytical data from a wide range of European countries and for a number of
food categories, the uneven distribution of occurrence data over the Member States will not add
significantly to the overall uncertainty.

The food products for which data were provided varied considerably between submissions from
the different Member States, but most samples belonged to the fish and seafood category,
followed by meat and meat products, and fats and oils. The samples originated from both targeted
and random sampling but this could only have resulted in an overestimation of exposure.
The type of food processing may have a significant influence on the PAH levels in the prepared
food commodities. The increasing replacement of traditional smoking processes by liquid smoke
flavouring has led to a decrease of PAH levels in commercially smoked food commodities over
the past two decades. On the other hand, investigations on barbecuing practices have
demonstrated that certain home cooking processes may lead to PAH levels that are more than one
magnitude higher compared to PAH concentrations in foods prepared with other practices.
Because the latter was not taken into consideration, this could have led to some underestimation
of the overall PAH exposure.
11.3 Exposure model

The high proportion of samples having levels below the LOD may have introduced uncertainties
to the overall estimate, however calculations made for this opinion show only a limited effect of
using the upper or lower bound. The use of the upper bound in this opinion tends to slightly
overestimate the dietary exposure. Because of the high proportion of samples below the LOD, all
the exposure calculations were based on the mean concentrations. It is generally accepted that the
use of the mean contamination to represent the long term dietary exposure is expected to be an
overestimation compared with the use of the median. Taken together, the uncertainties regarding
the exposure estimates are considered to be small, and tend to overestimate the exposure.
11.4 Model input (parameters)

There are no prescribed fixed official methods for the analysis of PAHs and laboratories can use
any method of analysis, provided it can be demonstrated in a traceable manner that they strictly
fulfil the analytical requirements laid down in the respective legislation. This may have added to
the uncertainty in the analytical results. However, recent inter-laboratory comparison testing have
shown that the analytical results provided are satisfactory but also indicated that instrument
calibration was for a great part responsible for deviations found. The lack of certified reference
materials and proficiency tests are clearly limitations when the method performance for the
analytical methods for analysis of PAHs in foods is assessed, and add thereby to the overall
uncertainty in the analytical results.

11.5 Other uncertainties

The CONTAM Panel applied the margin of exposure approach as has also been used by the
JECFA and obtained similar results for benzo[a]pyrene. Similar margins of exposure were
calculated for the different exposure scenarios applied, indicating a limited impact of exposure to
different groups of PAHs on the overall risk estimate of exposure to PAHs. It should be noted
however, that due to lack of information, benzo[c]fluorene was not included in the risk
assessment.
In Table 35 a summary of the uncertainty evaluation is presented, highlighting the main sources
of uncertainty and indicating an estimate of whether the respective source of uncertainty might
have led to an over- or underestimation of the exposure or the resulting risk.
Table 35: Summary of qualitative evaluation of the impact of uncertainties on the risk
assessment of the dietary exposure to PAHs.
Sources of uncertainty Direction and
Magnitude
Extrapolation of data from a number of Member States to whole Europe +/-a)
Use of analytical data from both targeted and random sampling +
Limited availability of analytical calibration standards +/-
Influence of non-detects on exposure estimate +
Use of adjustment factors for several broad food categories +/-
Processing of food +/-
Derivation of margin of exposure +
a)
+, ++, +++ = uncertainty with potential to cause small, medium or large over-estimation of exposure/risk
-, --, --- = uncertainty with potential to cause small, medium or large under-estimation of exposure/risk (EFSA,
2006).
The CONTAM Panel considered that the impact of the uncertainties on the risk assessment of
exposure to PAHs is limited and concluded that its assessment of the risk is likely to be
conservative – i.e. more likely to overestimate than to underestimate the risk.
CONCLUSIONS
• Polycyclic aromatic hydrocarbons (PAHs) constitute a large class of organic compounds
that are composed of two or more fused aromatic rings. They are primarily formed by
incomplete combustion or pyrolysis and generally occur in complex mixtures.
• Currently the PAHs of greatest concern are those that are genotoxic and carcinogenic. As
no new toxicological data were available to alter the previous conclusions of the Scientific
Committee on Food (SCF) and the Joint FAO/WHO Expert Committee on Food
Additives (JECFA) regarding those PAHs the evaluation focussed on the following 16

PAHs: benz[a]anthracene, benzo[b]fluoranthene, benzo[j]fluoranthene,

benzo[k]fluoranthene, benzo[ghi]perylene, benzo[a]pyrene, chrysene,
cyclopenta[cd]pyrene, dibenz[a,h]anthracene, dibenzo[a,e]pyrene, dibenzo[a,h]pyrene,
dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3-cd]pyrene, 5-methylchrysene and
benzo[c]fluorene.
• A toxic equivalency factor (TEF) approach to the risk characterisation was not considered
to be scientifically valid because of the lack of data from oral carcinogenicity studies for
different PAHs, their different modes of action and the evidence of poor predictivity of
the carcinogenic potency of PAH mixtures based on the currently proposed TEF values.
• The risk characterisation should be based upon the PAHs for which oral carcinogenicity
data were available, i.e. for benzo[a]pyrene and the genotoxic PAHs that were measured
in the two coal tar mixtures used in the carcinogenicity studies of Culp et al. (1998),
which provided the basis of the SCF and JECFA risk assessments. The EFSA Panel on
Contaminants in the Food Chain (CONTAM Panel) concluded that these eight PAHs
(benzo[a]pyrene, benz[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene,
benzo[ghi]perylene, chrysene, dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene), either
individually or in a combination, were currently the only possible indicators of the
carcinogenic potency of PAHs in food.
• In total, results of 9714 food samples analysed for the presence of one or more of the 16
PAHs recommended to be monitored by the European Commission and submitted by the
Member States were considered in this assessment. Results for the full set of 16 PAHs
were reported for 823 food samples, for the 15 SCF priority PAHs in 1375 food samples
and for the set of PAH8 in 4065 samples.
• Benzo[a]pyrene was present in about 50% of all samples analysed with a range from not
detected to 100% dependent on the food group involved. In about 30% of those samples
that were analysed for the 15 SCF priority PAHs other carcinogenic and genotoxic PAHs
were detected despite testing negative for benzo[a]pyrene.
• In samples analysed for the SCF priority PAHs, benzo[a]pyrene comprised about 8% of
the amount of the 15 PAHs. Combination of two (benzo[a]pyrene+chrysene=PAH2), four
(benzo[a]pyrene+chrysene+benz[a]anthracene+benzo[b]fluoranthene=PAH4) and eight
PAHs (benzo[a]pyrene+benz[a]anthracene+benzo[b]fluoranthene+benzo[k]fluoranthene
+benzo[ghi]perylene+chrysene+dibenz[a,h]anthracene+indeno[1,2,3-cd]pyrene=PAH8)
represented about 34%, 60% and 80% of the amount of 15 PAHs, respectively. Mean
concentrations of benzo[a]pyrene and of those three PAH combinations in 4065 food
samples across 11 food categories/subcategories were used for the exposure assessment.
• High amounts of benzo[c]fluorene were found in some food categories, but too few
analytical results were available to include it into the exposure assessment.

• The median dietary exposure across European countries was calculated both for mean and
high dietary consumers and varied between 235 ng/day (3.9 ng/kg b.w. per day) and 389
ng/day (6.5 ng/kg b.w. per day) respectively for benzo[a]pyrene alone, 641 ng/day (10.7
ng/kg b.w. per day) and 1077 ng/day (18.0 ng/kg b.w. per day) respectively for PAH2,
1168 ng/day (19.5 ng/kg b.w. per day) and 2068 ng/day (34.5 ng/kg b.w. per day)
respectively for PAH4 and 1729 ng/day (28.8 ng/kg b.w. per day) and 3078 ng/day (51.3
ng/kg b.w. per day) respectively for PAH8.
• The two highest contributors to the dietary exposure were cereals and cereal products, and
sea food and sea food products. However, the CONTAM Panel noted that the number of
cereal samples analysed was low.
• For non-smokers the major route of exposure to PAHs is consumption of food, for
smokers the contribution from smoking may be significant. Exposure to PAHs via
inhalation generally contributes relatively little to overall PAH exposure.
• Except for benzo[a]pyrene, limited data exist on the toxicokinetics of the 16 carcinogenic
PAHs.
• In mammals, absorption of benzo[a]pyrene varies between 12 and 99% of ingested dose,

depending on the dose and species investigated. Generally a higher absorption was
observed for lower molecular mass PAHs, whereas higher molecular mass PAHs are
poorly absorbed.
• After being absorbed, PAHs are rapidly distributed to almost all organs and are able to
cross the placental barrier.
• PAHs are extensively metabolised in mammals and do not bioaccumulate. Different

metabolic pathways can lead to highly reactive intermediates involved in the
mutagenic/carcinogenic process of PAHs.
• Biomarkers of exposure most widely used for PAHs are the levels of certain urinary PAH
metabolites. These data confirm that humans are exposed internally to a variety of PAHs
including those that are genotoxic/carcinogenic. The data furthermore suggest that food is
one among several routes of exposure, occupation and smoking being other notable
sources. Currently, no quantitative assessment of total PAH exposure and the relative
contribution of the diet can be made based on biomarkers.
• In accordance with the approach of the EFSA Scientific Committee, the CONTAM Panel
calculated BMDL10 values for benzo[a]pyrene, PAH2, PAH4 and PAH8 using a range of
statistical models, and selected the lowest BMDL10 values from the statistical models that
adequately fit the data. These values were 0.07, 0.17, 0.34 and 0.49 mg/kg b.w. per day,
respectively.

• The CONTAM Panel used a MOE approach based on dietary exposure for average and
high level consumers to benzo[a]pyrene, PAH2, PAH4 and PAH8, respectively and their
corresponding BMDL10 values derived from the two coal tar mixtures that were used in
the carcinogenicity studies of Culp et al. (1998). The resulting MOEs for average
consumers were 17,900 for benzo[a]pyrene, 15,900 for PAH2, 17,500 for PAH4 and
17,000 for PAH8. For high level consumers, the respective MOEs were 10,800, 9,500,
9,900 and 9,600. These MOEs indicate a low concern for consumer health at the average
estimated dietary exposures. This applies to the full range of estimates of average
exposures across EU Member States (3.1-4.3 ng/kg b.w. per day, MOEs: 16,300-22,600
for benzo[a]pyrene alone and 23.6-35.6 ng/kg b.w. per day, MOEs: 13,800-20,800 for
PAH8). However, for high level consumers the MOEs were close to or less than 10,000,
which as proposed by the EFSA Scientific Committee indicates a potential concern for
consumer health and a possible need for risk management action.
• The CONTAM Panel concluded that benzo[a]pyrene is not a suitable indicator for the
occurrence of PAHs in food. Based on the currently available data relating to occurrence
and toxicity, the CONTAM Panel concluded that PAH4 and PAH8 are the most suitable
indicators of PAHs in food, with PAH8 not providing much added value compared to
PAH4.
RECOMMENDATIONS (INCL. KNOWLEDGE/DATA GAPS)

• Toxicological data for individual PAHs as well as oral carcinogenicity data with mixtures
relevant for dietary exposure are needed.
• Additional carcinogenicity and occurrence data for benzo[c]fluorene are needed.
REFERENCES
Agundez, J.A. 2004. Cytochrome P450 gene polymorphism and cancer. Curr. Drug Metab. 5:
211-224.
Alekandrov, K., Cascorbi, I., Rojas, M., Bouvier, G., Kriek, E. and Bartsch, H. 2002. CYP1A1
and GSTM1 genotypes affect benzo[a]pyrene DNA adducts in smokers’ lung: comparison
with aromatic/hydrophobic adduct formation. Carcinogenesis 23: 1969-1977.
Anastasio, A., Mercogliano, R., Vollano, L., Pepe, T. and Cortesi, M.L. 2004. Levels of
benzo[a]pyrene (BaP) in “Mozzarella di Bufala Campana” cheese smoked according to
different procedures. J. Agric. Food Chem. 52: 4452-4455.

Angerer, J., Mannschreck, C. and Güdel, J. 1997. Occupational exposure to polycyclic aromatic
hydrocarbons in a graphite-electrode producing plant: biological monitoring of 1-
hydroxypyrene and monohydroxylated metabolites of phenanthrene. Int. Arch. Occup.
Environ. Helath 69: 323-331.
ATSDR (Agency for Toxic Substances and Disease Registry). 1995. Toxicological profile for
polycyclic aromatic hydrocarbons. U.S. Department of Health and Human Services, Public
Health Service, Agency for Toxic Substances and Disease Registry.
http://www.atsdr.cdc.gov/toxprofiles/tp69.pdf.
Baird, W.M. and Ralston, S.L. 1997. In: Comprehensive Toxicology (Bowden, G.T. and Fisher,
S.M., ed.). Elsevier, Oxford, UK, pp. 171-200.
Björseth, A. and Ramdahl, T. 1985. Sources and emissions of PAH. In: Handbook of polycyclic
aromatic hydrocarbons: Volume 2: Emission sources and recent progress in analytical
chemistry (Björseth, A. and Ramdahl, T. ed.). New York, Marcel Dekker, pp 1-20.
Boström, C-E., Gerde, P., Hanberg, A., Jernström, B., Johansson, C., Kyrklund, T., Rannug, A.,
Törnqvist, M., Victorin, K. and Westerholm, R. 1998. Cancer Risk Assessment, Indicators,
and Guidelines for Polycyclic Aromatic Hydrocarbons in the Ambient Air. Environ. Health
Perspect. 110(3): 451-488.
Buckley, T.J. and Lioy, P.J. 1992. An examination of the time course from human dietary
exposure to polycyclic aromatic hydrocarbon to urinary elimination of 1-hydroxypyrene. Br.
J. Ind. Med. 49: 113–124.
Butler, J.D. and Crossley, P. 1981. Reactivity of Polycyclic aromatic hydrocarbons adsorbed on
soot particles. Atmos. Environ. 15: 91-94
Bock, K.W., Gschaidmeier, H., Heel, H., Lehmköster, T., Münzel, P.A., Raschko, F. and Bock
Hennig, B. 1998. Ah receptor-controlled transcriptional regulation and function of rat and
human UDP-glucuronosyltransferase isoforms. Adv. Enzyme Regul. 38: 207-222.
Bouchard, M. and Viau, C. 1997. Urinary excretion of benzo(a)-pyrene metabolites following

intravenous, oral, and cutaneous benzo(a)pyrene administration. Can. J. Physiol. Pharmacol.
75: 185–192.
Brown, R. and Mittelsman, A. 1993. Evaluation of existing methods to rank the relative
carcinogenicity of polycyclic aromatic compounds (PAH), Draft. Technical Resources, Inc.,
Contract No. 68-01-0022, for Office of Emergency and Remedial Response, Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Office of
Pesticides, Pollution Prevention and Toxic Substances (OPPTS).
Cavalieri, E.L. and Rogan, E.G. 1995. Central role of radical cations in metabolic activation of
polyyclic aromatic hydrocarbons. Xenobiotica 25: 677-688.

Cavalieri, E.L. and Rogan, E.G. 1992. The approach to understanding aromatic hydrocarbon
carcinogenesis. The central role of radical cations in metabolic activation. Pharmacol. Ther.
55: 183-194.
Cavret, S., Laurent C., Feidt, C., Laurent, F. and Rychen, G. 2003. Intestinal absorption of 14C
from 14C-phenanthrene, 14C-benzo(a)pyrene and 14Ctetrachlorodibenzo-para-dioxin:
Approaches with the Caco-2 cell line and portal absorption measurements in growing pigs.
Reprod. Nutr. Dev. 43: 145-154.
Chatonnet, P. and Escobessa, J. 2007. Impact of toasting oak barrels on the presence of
polycyclic aromatic hydrocarbons in wine. J. Agric. Food Chem. 55: 10351-10358.
Chen, B.H. and Lin, Y.S. 1997. Formation of polycyclic aromatic hydrocarbons during
processing of duck meat. J. Agric. Food Chem. 45: 1394-1403.
Crump, K.S. 1984. An improved procedure for low-dose carcinogenic risk assessment from
animal data. J. Environ. Pathol. Toxicol. Oncol. 5: 339-348.
Culp, S.J., Gaylor, D.W., Sheldon, W.G., Goldstein, L.S. and Beland, F.A. 1998. A comparison
of the tumours induced by coal tar and benzo[a]pyrene in a 2-year bioassay. Carcinogenesis,
19: 117-124.
de Kok, T.M.C.M. and van Maanen, J.M.S. 2000. Evaluation of fecal mutagenicity and colorectal
cancer risk. Mutat. Res. 463: 53-101.
De Kruijf, N., Schouten, T. and Van der Stegen, G.H.D. 1987. Rapid determination of
benzo[a]pyrene in roasted coffee and coffee brew by high performance liquid
chromatography with fluorescence detection. J. Agric. Food Chem. 35: 545-549.
Devanesan, P.D., Rama Krishna, K.V., Todorovic, R., Rogan, E.G., Cavalieri, E.L., Jeong, H.,
Jankowiak, R. and Small, G.J. 1992. Identification and quantitation of benzo[a]pyrene-DNA
adducts formed by rat liver microsomes in vitro. Chem. Res. Toxicol. 5: 302-309.
DIPP (the Danish Import Promotion Program). 2007. A survey of the market for dried fruit in
Denmark, DIPP, Copenhagen, Denmark. http://www.dipp.eu/en/marketinfo.aspx.
Dipple, A., Moschel, R.C. and Bigger, C.A.H. 1984. Polynulear aromatic carcinogens. In:
Chemical Carinogens, ACS Monograph, vol. 182. (Searle, C.E., ed.). American Chemical
Society Press, Washington DC, pp. 41-163.
Dipple, A., Pigott, P.A., Agarwal, S.K., Yagi, H., Sayer, J.M. and Jerina, D.M. 1987. Optically
active benzo[c]phenanthrene diol-epoxides bind extensively to adenine in DNA. Nature 327:
535-536.

EC (European Commission). 2002. Opinion of the Scientific Committee on Food on the risks to
human health of Polycyclic Aromatic Hydrocarbons in food.
http://ec.europa.eu/food/fs/sc/scf/out153_en.pdf
EC (European Commission). 2001. Outcome of the expert group meeting on 3 October on ways
to prevent contamination of olive residue oil and other oils with polycyclic aromatic
hydrocarbons (PAH). Summary record of the 85th meeting of the Standing Committee on
Foodstuffs, 25th October 2001, agenda item 9.
http://europa.eu.int/comm/food/fs/rc/scfs/rap09_en.pdf
EC-DG-JRC-IRMM/CRL (European Commission, Directorate General, Joint Research Centre,

Institute for Reference Materials and Measurements/Community Reference Laboratory for
polycyclic aromatic hydrocarbons). 2007. Report on the first inter-laboratory comparison test
organised by the Community Reference Laboratory for Polycyclic Aromatic Hydrocarbons,
15+1 EU priority PAHs in acetonitrile. EUR 22696 EN, European Communities, 2007.
EC-DG-JRC-IRMM/CRL (European Commission, Directorate General, Joint Research Centre,

Institute for Reference Materials and Measurements/Community Reference Laboratory for
polycyclic aromatic hydrocarbons). 2008. Report on the second inter-laboratory comparison
test organised by the Community Reference Laboratory for Polycyclic Aromatic
Hydrocarbons, 15+1 EU priority PAHs in edible oil and acetonitrile. EUR 23251 EN,
European Communities, 2008.
EFSA (European Food Safety Authority). 2005a. Opinion of the Scientific Committee on a
request from EFSA related to a harmonized approach for risk assessment of substances which
are both genotoxic and carcinogenic. The EFSA Journal 282: 1-31.
http://www.efsa.europa.eu/EFSA/Scientific_Opinion/sc_op_ej282_gentox_en3,0.pdf
EFSA (European Food Safety Authority). 2005b. European Food Consumption Database:
Current and medium to long-term strategies. EFSA Scientific Colloquium Summary Report.
http://www.efsa.europa.eu/EFSA/Scientific_Document/Sc_Colloque_no3_European_food_co
nsumption_database,0.pdf
EFSA (European Food Safety Authority). 2006. Opinion of the Scientific Committee related to
Uncertainties in Dietary Exposure Assessment.
http://www.efsa.europa.eu/EFSA/Scientific_Opinion/sc_op_uncertainty%20exp_en,5.pdf
EFSA (European Food Safety Authority). 2007. Findings of the EFSA data collection on
polycyclic aromatic hydrocarbons in food. A report from the Unit of data Collection and
Exposure on a Request from the European Commission. EFSA/DATEX/002.
http://www.efsa.europa.eu/EFSA/Scientific_Document/datex_report_pah.pdf.
EFSA (European Food Safety Authority). 2008. Guidance Document for the Use of the Concise
European Food Consumption Database in Exposure Assessment.
The EFSA Journal (2008) 724, 100-114

http://www.efsa.europa.eu/EFSA/General/Coincise_database_guidance_document_and_anne
xes,2.pdf
Elovaara, E., Mikkola, J., Stockmann-Juvala, H., Luukkanen, L., Keski-Hynnilä, H., Kostiainen,
R., Pasanen, M., Pelkonen, O. and Vainio, H. 2007. Polycyclic aromatic hydrocarbon (PAH)
metabolizing enzyme activities in human lung, and their inducibility by exposure to
naphthalene, phenanthrene, pyrene, chrysene, and benzo(a)pyrene as shown in the rat lung
and liver. Arch. Toxicol. 81: 169-182.
EMEP (Co-operative programme for monitoring and evaluation of the long-range transmissions
of air pollutants in Europe). 2001. Heavy metals and POPs within the EMEP region 1999.
EMEP/CCC-Report 9/2001. Norwegian Institute for Air Research PO Box 100, NO-2027
Kjeller, Norway.
Evans, W. H., Thomas, N. C., Boardman, M. C. and Nash, S. J. 1993. Relationships of polycyclic
aromatic hydrocarbon yields with particulate matter (water and nicotine free) yields in
mainstream and sidestream cigarette smoke. Sci. Tot. Environ. 136(1-2): 101-109.
Fagundes, R.B., Abnet, C.C., Strickland, P.T., Kamangar, F., Roth, M.J., Taylor, P.R. and
Dawsey, S.M. 2006. Higher urine 1-hydroxypyrene glucuronide (1-OHPG) is associated with
tobacco smoke exposure and drinking mate in healthy subjects from Rio Grande do Sul,
Brazil. BMC Cancer 6, 139.
Falk Filipsson, A.F., Sand, S., Nilsson, J. and Victorin, K. 2003. The Benchmark dose method-
Review of available models, and recommendations for application in health risk assessment.
Crit. Rev. Toxicol. 33(5): 505-542.
FAO (Food and Agriculture Organization of the United Nations). 2004. The state of world
fisheries and aquaculture. FAO, Rome, Italy.
http://www.fao.org/DOCREP/007/y5600e/y5600e05.htm#P1257_75403.
FAO/WHO (Food and Agriculture Organization of the United Nations/World Health

Organization). 2005. Joint FAO/WHO Expert Committee on Food Additives (JECFA). Sixty-
fourth meeting, Rome, 8-17 February 2005. Summary and Conclusions.
http://www.who.int/ipcs/food/jecfa/summaries/summary_report_64_final.pdf
FAO/WHO (Food and Agriculture Organization of the United Nations/World Health

Organization). 2006. Safety evaluation of certain contaminants in food. Prepared by the
Sixty-fourth meeting of the Joint FAO/WHO Expert Committee on Food Additives (JECFA).
WHO Food Additives Series: 55; FAO Food and Nutrition Paper 82; World Health
Organization WHO, 2006; Food and Agriculture Organization of the United Nations FAO,
2006. http://whqlibdoc.who.int/publications/2006/9241660554_eng.pdf
The EFSA Journal (2008) 724, 101-114

Flesher, J.W., Horn, J. and Lehner, A.F. 1997. 7-Sulfooxymethyl-12-methylbenz[a]anthracene is

an exceptionally reactive electrophilic mutagen and ultimate carcinogen. Biochem. Biophys.
Res. Commun. 231: 144-148.
Fontana, R.J., Lown, K.S., Paine, M.F., Fortlage, L., Santella, R.M., Felton, J.S., Knize, M.G.,
Greenberg, A. and Watkins, P.B. 1999. Effects of a chargrillef meat diet on expression of
CYP3A, CYP1A, and P-glycoprotein levels in healthy volunteers. Gastroenterology 117: 89-
98.
Fontcuberta, M., Arqués, J.F., Martínez, M., Suárez, A., Villalbí, J.R., Centrich, F., Serrahima,
E., Duran, J. and Casas, C. 2006. Polycyclic aromatic hydrocarbons in food samples collected
in Barcelona, Spain. J. Food Prot. 69(8): 2024-2028.
Förster, K., Preuss, R., Rossbach, B., Brüning, T., Angerer, J. and Simon, P. 2008. 3-
Hydroxybenzo[a]pyrene in the urine of workers with occupational exposure to polycyclic
aromatic hydrocarbons in different industries. Occup. Environ. Med. 65: 224-229.
Fromberg, A., Højgård, A. and Duedahl-Olesen, L. 2007. Analysis of polycyclic aromatic

hydrocarbons in vegetable oils combining gel permeation chromatography with solid-phase
extraction clean-up. Food Addit. Contam. 24(7): 758-767.
Galinaro, C.A., Cardoso, D.R. and Franco, D.W. 2007. Profiles of polycyclic aromatic
hydrocarbons in Brazilian sugar cane spirits: Discrimination between cachacas produced from
nonburned and burned sugar cane crops. J. Agric. Food Chem. 55: 3141-3147.
García-Falcón, M.S., Cancho-Grande, B. and Simal-Gándara, J. 2005. Minimal clean-up and

rapid determination of polycyclic aromatic hydrocarbons in instant coffee. Food Chem. 90:
643-647.
García Falcón, M.S., González Amigo, S., Lage Yusty, M. A. and Simal Lozano, J. 1999.
Determination of benzo[a]pyrene in some Spanish commercial smoked products by HPLC-
FL. Food Addit. Contam. 16(1): 9-14.
García Falcón, M. S., López de Alda Villaizán, M. J., González Amigo, S., Simal Lozano, J. and
Lage Yusty, M. A. 1996. Enrichment of benzo[a]pyrene in smoked food products and
determination by HPLC-FL. J. Chromatogr. A 753: 207-215.
García Falcón, M. S. and Simal-Gandara, J. 2005. Determination of polycyclic aromatic

hydrocarbons in alcoholic drinks and the identification of their potential sources. Food Addit.
Contam. 22: 791-797.
Gilbert, J., and Knowles, M.E. 1975. The chemistry of smoked foods: a review. J. Food Technol.
10: 245-261.
The EFSA Journal (2008) 724, 102-114

Garte, S. and Crosti, F. 1999. A nomenclature system for metabolic gene polymorphisms. IARC
Sci. Publ. 148: 5-12.
Gelboin, H.V. 1980. Benzo[alpha]pyrene metabolism, activation and carcinogenesis: role

andregulation of mixed-function oxidases and related enzymes. Physiol. Rev. 60: 1107-1166.
Goldstein, L.S. 2001. To BaP or not to BaP ? That is the question. Environ. Health Perspec.
109(8): A356-357.
Grimmer, G. 1983. Environmental Carcinogens: Polycyclic Aromatic Hydrocarbons. CRC Press,

Boca Raton, FL.
Greenberg, A., Darack, F. and Harkov, R. 1985. Polycyclic aromatic hydrocarbons in New Jersey
(USA): A comparison of winter and summer concentrations over a two-year period. Atmos.
Environ. 19: 1325-1340.
Grova, N., Feidt, C. Laurent, C. and Rychen, G. 2002. [14C] Milk, urine and faeces excretion
kinetics in lactating goats after an oral administration of [14C] polycyclic aromatic
hydrocarbons. Int. Dairy J. 12: 1025-1031.
Guillén, M.D. and Sopelana, P. 2004. Occurrence of polycyclic aromatic hydrocarbons in

smoked cheese. J. Dairy Sci. 87: 556-564.
Guillén, M.D., Sopelana, P. and Partearroyo, M.A. 2000. Determination of polycyclic aromatic
hydrocarbons in commercial liquid smoke flavorings of different compositions by gas
chromatography-mass spectrometry. J. Agric. Food Chem. 48: 126-131.
Guillén, M.D., Sopelana, P., and Partearroyo, M.A. 1997. Food as a source of polycyclic
aromatic carcinogens. Rev. Environ. Health, 12: 133-146.
Hall, M. and Grover, P.L. 1990. Polycyclic aromatic hydrocarbons: metabolisim, activation and
tumor initiation. In: Chemical Carcinogenesis and Mutagenesis Vol. 1. (Cooper, C.S. and
Grover, P.L., ed.). Springer- Verlag, Heidelberg, pp. 327-372.
Harvey, R.G. 1991. Polycyclic aromatic hydrocarbons, Chemistry and Carcinogenicity.

Cambridge University Press, Cambridge, UK, pp. 11-87.
Hecht, S.S., Radok, L., Amin, S., Huie, K., Melikian, A.A., Hoffmann, D., Pataki, J. and Harvey,
R.G. 1985. Tumorigenicity of 5-methylchrysene dihydrodiols and dihydrodiol epoxides in
newborn mice and on mouse skin. Cancer Res. 45: 1449-1452.
Hischenhuber, C. and Stijve, T. 1987. Determination of benzo[a]pyrene in roasted coffee and

coffee brews by HPLC with fluorescence detection. Dtsch. Lebensm. Rundsch. 83: 1-4.
Houessou, J.K., Benac, C., Delteil, C. and Camel, V. 2005. Determination of polycyclic aromatic
hydrocarbons in coffee brew using solid-phase extraction. J. Agric. Food Chem. 53: 871-879.
The EFSA Journal (2008) 724, 103-114

Houessou, J.K., Maloug, S., Leveque, A.-S., Delteil, C., Heyd, B. and Camel, V. 2007. Effect of
Roasting Conditions on the Polycyclic Aromatic Hydrocarbon Content in Ground Arabica
Coffee and Coffee Brew. J. Agric. Food Chem. 55: 9719-9726.
Houessou, J.K., Goujot, D., Heyd, B. and Camel, V. 2008. Modeling the formation of some
polycyclic aromatic hydrocarbons during the roasting of Arabica Coffee samples. J. Agric.
Food Chem. 56: 3648-3656.
ISO (International Organization for Standardization). 2007. Animal and vegetable fats and oils –
Determination of polycyclic aromatic hydrocarbons by on-line donor acceptor complex
chromatography and HPLC with fluorescence detection. Draft International Standard
ISO/DIS 22959. ISO/TC 34/SC 11.
Jahncke, M.L. and Herman, D. 2001. Control of food safety hazards during cold-smoked fish
processing. J. Food Sci. – Supplement to Vol. 66(7): S1104-S1112.
Jongeneelen, F.J., van Leeuwen F.E. and Oosterink, S. 1990. Ambient and biological monitoring
of cokeoven workers: determinants of the internal dose of polycyclic aromatic hydrocarbons.
Br. J. Ind. Med. 47: 454-461.
Kang, H.G., Jeong, S.H., Cho, M.H., and Cho, J.H. 2007. Changes of biomarkers with oral
exposure to benzo(a)pyrene, phenanthrene and pyrene in rats. J. Vet. Sci. 8: 361-368.
Karl, H. and Leinemann, M. 1996. Determination of polycyclic aromatic hydrocarbons in smoked

fishery products from different smoking kilns. Z. Lebensm. Unters. Forsch. 202: 458-464.
Kayali, M. N. and Rubio-Barroso, S. 1995. Determination of benzo(a)pyrene in total particulate

matter of virginia and black tobacco smoke by HPLC with fluorimetric detection. J. Liquid
Chrom. Rel. Technol. 18(8): 1617-1632.
Kayali-Sayadi, M.N., Rubio-Barroso, S., Cuesta-Jimenez, M.P. and Polo-Diez, L.M. 1999. A
new method for the determination of selected PAHs in coffee brew samples by HPLC with
fluorimetric detection and solid-phase extraction. J. Liq. Chrom. Rel. Technol. 22(4): 615-
627.
Kochbach, A., Li, Y., Yttri K.E., Cassee, F.R., Schwarze, P.E. and Namork, E. 2006.
Physicochemical characterisation of combustion particles from vehicle exhaust and
residential wood smoke. http://www.particleandfibretoxicology.com/content/3/1/1.
Koganti, A., Singh, R., Rozett, K., Modi, N., Goldstein, L.S., Roy. A., Zhang F., Harvey, R.G.
and Weynand, E.H. 2000. 7H-benzo[c]fluorene: a major DNA adduct-forming component of
coal tar. Carcinogenesis 21(8): 1601-1609.
Krewski, D., Thorslund, T. and Withey, J. 1989. Carcinogenic risk assessment of complex
mixtures. Toxicol. Ind. Health. 5: 851-867.
The EFSA Journal (2008) 724, 104-114

Kroese, E.D., Muller, J.J.A., Mohn, G.R., Dortant, P.M. and Wester, P.W. 2001. Tumourigenic
effects in Wistar rats orally administered benzo[a]pyrene for two years (gavage studies).
Implications for human cancer risks associated with oral exposure to polycyclic aromatic
hydrocarbons. National Institute of Public Health and the Environment, RIVM Report no.
658603 010, November 2001, Bilthoven.
Könemann, W.H. and Pieters, M.N. 1996. Confusion of concepts in mixture toxicology. Food
Chem. Toxicol. 34: 1025-1031.
Laher, J. M. and Barrowman, J. A. 1987. Intestinal absorption of carcinogenic hydrocarbons.

Ann. N. Y. Acad. Sci. 534:565-574.
Lapole, D., Rychen, G., Grova, N., Monteau, F., Le Bizec, B. and Feidt, C. 2007. Milk and urine
excretion of polycyclic aromatic hydrocarbons and their hydroxylated metabolites after a
single oral administration in ruminants. J. Dairy Sci. 90: 2624-2629.
Larsen, J.C. and Larsen, P.B. 1998. Chemical Carcinogens. In: Air Pollution and Health (Hester,
R.E. and Harrison, R.M., ed.). Issues in Environmental Sciences and Technology, 10, The
Royal Society of Chemical, Cambridge.
Larsson, B.K., Sahlberg, G.P., Eriksson, A.T. and Busk, L.A. 1983. Polycyclic aromatic
hydrocarbons in grilled food. J. Agric. Food Chem. 31: 867-873.
Levin, W., Buening, M.K., Wood, A.W., Chang, R.L., Kedzierski, B., Thakker, D.R., Boyd,
D.R., Gadaginamath, G.S., Armstrong, R.N., Yagi, H., Karle, J.M., Slaga, T.J., Jerina, D.M.
and Conney, A.H. 1980. An enantiomeric interaction in the metabolism and tumorigenicity of
(+)- and (-)-benzo[a]pyrene 7,8-oxide. J. Biol. Chem. 255: 9067-9074.
Levin, W., Thakker, D.R., Wood, A.W., Chang, R.L., Lehr, R.E., Jerina, D.M. and Conney, A.H.
1978. Evidence that benzo(a)anthracene 3,4-diol-1,2-epoxide is an ultimate carcinogen on
mouse skin. Cancer Res. 38: 1705-1710.
Lijinsky, W. and Ross, A.E. 1967. Production of carcinogenic polynuclear hydrocarbons in the
cooking of food. Food Cosmet. Toxicol. 5: 343-347.
Lintas, C., de Matthaeis, M.C. and Merli, F. 1979. Determination of benzo[a]pyrene in smoked,
cooked and toasted food products. Food Cosmet. Toxicol. 17: 325-328.
Lipniak, M. and Brandys, J. 1993. Toxicokinetics of fluoranthene, pyrene and benz(a)anthracene

in the rat. Polycyclic Aromatic Compounds. 3: 111-119.
Li, Z., Sandau, C.D., Romanoff, L.C., Caudill, S.P., Sjodin, A., Needham, L.L. and Patterson, Jr.
D.G. 2008. Concentration and profile of 22 urinary polycyclic aromatic hydrocarbon
metabolites in the US population. Environ. Res. 107(3): 320-331.
The EFSA Journal (2008) 724, 105-114

Llobet, J.M., Falcó, G., Bocio, A. and Domingo, J.L. 2006. Exposure to polycyclic aromatic
hydrocarbons through consumption of edible marine species in Catalonia, Spain. J. Food.
Prot. 69(10): 2493-2499.
Lodovici, M. Akpan, V., Evangelisti, C. and Dolara, P. 2004a. Sidestream tobacco smoke as the
main predictor of exposure to polycyclic aromatic hydrocarbons. J. Appl. Toxicol. 24: 277-
281.
Lodovici, M., Luceri, C., Guglielmi, F., Bacci, C., Akpan, V., Fonnesu, M.L., Boddi, V. and
Dolara, P. 2004b. Benzo[a]pyrene diolepoxide (BPDE)-DNA adduct levels in leukocytes of
smokers in relation to polymorphism of CYP1A1, GSTM1, GSTP1, GSTT1, and mEH.
Cancer Epidemiol. Biomark. Prev. 13: 1342-1348.
Maier, H.G. 1991. Teneur en composés cancérigènes du café en grains. Café, cacao, Thé 35(2) :
133-142.
Marczynski, B., Preuss, R., Mensing, T., Angerer, J., Seidel, A., El Mourabit, A., Wilhelm, M.
and Brüning, T. 2005. Genotoxic risk assessment in white blood cells of occupationally
exposde workers before and after alteration of then poylcyclic aromatic hydrocarbon (PAH)
profile on the production material. Int. Arch. Occup. Environ. Helath 78: 97-108.
McCoull, K.D., Rindgen, D., Blair, I.A. and Penning, T.M. 1999. Synthesis and characterization
of polycyclic aromatic hydrocarbon o-quinones depurinating N7-guanine adducts. Chem.
Res. Toxciol. 12: 237-246.
Miller, E.C. 1978. Some current perspectives on chemical carcinogenesis in humans and
experimental animals: presidential address. Cancer Res. 38: 1479-1486.
Miller, E.C. and Miller, J.A. 1966. Mechanism of chemical carcinogenesis: nature of proximate
carcinogens and interactions with macromolecules. Pharmacol. Rev. 18: 805-838.
Modica, R., Fiume, M., Guaitani, A. and Bartosek, I. 1983. Comparative kinetics of
benzo[a]anthracene, chrysene, triphenylene in rats after oral administration. I. Study with
single compounds. Toxicol. Lett. 18: 103-109.
Mottier, P., Parisod, V. and Turesky, R.J. 2000. Quantitative determination of polycyclic
aromatic hydrocarbons in barbecued meat sausages by gas chromatography coupled to mass
spectrometry. J. Agric. Food Chem. 48: 1160-1166.
Nisbeth, I.C. and LaGoy, P.K. 1992. Toxic equivalency factors (TEFs) for polycyclic aromatic
hydrocarbons (PAH). Regul. Toxicol. Pharmacol. 16: 290-300.
Nordquist, M., Thakker, D.R., Vyas, K.P., Yagi, H., Levin, M., Ryan, D.E., Thomas, D.E.,
Conney, A.H. and Jerina, D.M. 1981. Metabolism of chrysene and phenanthrene to bay-
region diol-epoxide by rat liver enzymes. Mol. Pharmacol. 19: 168-178.
The EFSA Journal (2008) 724, 106-114

OSPAR (Protection of the Marine Environment of the North-East Atlantic). 2008. Background
Concentrations for Metals, PAH and Dioxins in Biota Excerpt from Draft MCWG 2008
report Presented by the Chair of the ICES Marine Chemistry Working Group (MCWG).
ASMO 08/6/5 Add.2-E(L). Oslo, Norway, 21-25 April 2008.
Park, K.S., Sims, R.C., Dupont, R.R., Doucette, W.J. and Matthews, J.E. 1990. Fate of PAH
compounds in two soil types: Influence of volatilization, abiotic loss and biological activity.
Environ. Toxicol. Chem. 9: 187-195
Park, J., Szewczuk, L.M. and Penning, T.M. 2004. PAH-o-quinones produce significant amounts
of 8-oxo-2’-deoxyguanosine (8-oxo-dG) in salmon testis DNA. Abstract # 1934. Ann. Meet.
Soc. Toxicol., Baltimore, MD.
Pastorelli, R., Guanci, M., Restano, J., Berri, A., Micoli, G., Minoia, C., Alcini, D., Carrer, P.,
Negri, E., La Vecchia, C., Fanelli, R. and Airoldi, L. 1999. Seasonal effects on airborne
pyrene, urinary 1-hydroxypyrene, and benzo[a]pyrene diol epoxide hemoglobin adducts in
the general population. Cancer Epidemiol. Biomark. Prev. 8: 561-565.
Pavanello, S., Siwinska, E., Milezynska, D. and Clonfero, E. 2004. GSTM1 mull genotype as a
risk factor for anti-BPDE-DNA adduct formation in mononuclear white blood cells of coke-
oven workers. Mutat. Res. 558: 53-62.
Pham, T., Lum, K. and Lemieux, C. 1993. Sources of PAHs in the St. Lawrence River (Canada)
and their relative importance. Chemosphere 27(7): 1137-1149.
Penning, T.M., Ohnishi, S.T., Ohnishi, T. and Harvey, R.G. 1996. Generation of reactive oxygen
species during the enzymatic oxidation of polycyclic aromatic hydrocarbon dihydrodiols
catalyzed by dihydrodiol dehydrogenase. Chem. Res. Toxicol. 9: 84-92.
Penning, T.M., Burczynski, M.E., Hung, C.-F., McCoull, K.D., Palackal, N.T. and Tsuruda, L.S.
1999. Dihydrodiol dehydrogenase and polycyclic aromatic hydrocarbon activation:
generation of reactive and redox active o-quinones.
Phillips, D.H. 1999. Polycyclic Aromatic Hydrocarbons in the Diet. Mutat. Res., 443: 139- 147.
Ralston, S.L., Lau, H.H., Seidel, A., Luch, A. and Baird, W.M. 1995. Stereoselctive activation of
dibenzo[a,l]pyrene to (-)-anti(11R,12S,13S,14R)- and (+)-syn(11S,12R,13S,14R)-11,12diol-
13,14-epoxides which bind extensively to deoxyadenosine residues of DNA in the human
mammary carcinoma cell line MCF-7. Carcinogenesis 16: 1907-1919.
Ralston, S.L., Lau, H.H., Seidel, A., Luch, A., Platt, K.L. and Baird, W.M. 1994. The potent
carcinogen dibenzo[a,l]pyrene is metabolically activated to fjord-region 11,12-diol-13,14-
epoxides in human mammary carcinoma MCF-7 cell cultures. Cancer Res. 54: 887-890.
The EFSA Journal (2008) 724, 107-114

Ramesh, A., Inyang, F., D. Hood, D.B., Archibong, A.E., Knuckles, M.E. and Nyanda, A.M.
2001. Metabolism, bioavailability, and toxicokinetics of benzo[a]pyrene [B(a)P] in F-344 rats
following oral administration. Exp.Toxic. Pathol. 53: 253–270.
Ramesh, A., Walker, S.A., Hood, D.B., Guillén, M.D., Schneider, K. and Weyand, E.H. 2004.
Bioavailability and risk assessment of orally ingested polycyclic aromatic hydrocarbons. Int.
J. Toxicol. 23: 301-333.
Reinik, M., Tamme, T., Roasto, M., Juhkam, K., Tenno, T. and Kiis, A. 2007. Polycyclic
aromatic hydrocarbons (PAHs) in meat products and estimated PAH intake by children and
the general population in Estonia. Food Addit. Contam. 24(4): 429-437.
Rojas, M., Cascorbi, I., Alexandrov, K., Kriek, E., Auburtin, G., Mayer, L., Kopp-Schneider, A.,
Roots, I. and Bartsch, H. 2000. Modulation of benzo(a)pyrene diolepoxide-DNA adduct
levels in human white blood cells by CYP1A1, GSTM1 and GSTT1 polymorphism.
Carcinogenesis 21: 35-41.
Rose, M., White, S., Macarthur, R., Petch, R.G., Holland, J. and Damant, A.P. 2007. Single-
laboratory validation of a GC/MS method for the determination of 27 polycyclic aromatic
hydrocarbons (PAHs) in oils and fats. Food Addit. Contam. 24(6): 635-651.
Rugen, P.J., Stern, C.D. and Lamm, S.H. 1989. Comparative carcinogenicity of the PAH as a
basis for acceptable exposure levels (AELs) in drinking water. Regul. Toxicol. Pharmacol. 9:
273.
Saint-Aubert, B., Cooper, J.F., Astre, C., Spiliotis, J. and Joyeux, H. 1992. Evaluation of the
induction of polycyclic aromatic hydrocarbons by cooking on two geometrically different
types of barbecue. J. Food Compos. Anal. 5: 257-263.
Saunders, C.R., Ramesh, A. and Shockley, D.C. 2002. Modulation of neurotoxic behavior in F-
344 rats by temporal disposition of benzo(a)pyrene. Toxicol.Lett. 129: 33-45.
Scherer, G., Frank, S., Riedel, K., Meger-Kossien, I. and Renner, T. 2000. Biomonitoring of
exposure to polycyclic aromatic hydrocarbons of non-occupationally exposed persons.
Cancer Epidemiol. Biomark. Prev. 9: 373-380.
Schneider, K., Roller, R., Kalberlah, F. and Schuhmacher-Wolz, U. 2002. Cancer risk assessment
for oral exposure to PAH mixtures. J. Appl. Toxicol. 22: 73-83.
Shimada, T. 2006. Xenobiotix-metabolizing enzymes involved in activation and detoxification of

carcinogenic polycyclic aromatic hydrocarbons. Drug Metab. Pharmacokinet. 21: 257-276.
Shimada, T. and Fujii-Kuriyama, Y. 2004. Metabolic activation of polycyclic aromatic

hydrocarbons to carcinogens by cyctochromes P450 1A1 and 1B1. Cancer Sci. 95: 1-6.
The EFSA Journal (2008) 724, 108-114

Shimada, T., Gillam, E.M., Oda, Y., Tsumura, F., Sutter, T.R., Guengerich, F.P. and Inoue, K.
1999a. Metabolism of benzo[a]pyrene to trans-7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene by
recombinant human cytochrome P450 1B1 and purified liver epoxide hydrolase. Chem. Res.
Toxicol. 12: 623-629.
Shimada, T., Oda, Y., Gillam, E.M.J., Guengerich, F.P. and Inoue, K. 1999b. Metabolic
activation of polycyclic aromatic hydrocarbons and other procarcinogens by cytochromes
P450 1A1 and P450 1B1 allelic variants and other human cytochromes P450 in Salmonalla
typhmurium NM2009. Drug Metab. Dispos. 29: 1176-1182.
Shou, M., Harvey, R.G. and Penning, T.M. 1993. Reactivity of benzo[a]pyrene-7,8-dione with
DNA. Evidence for the formation of deoxyguanosine adducts. Carcinogenesis 14: 475-482.
Simko, P. 2005. Factors affecting elimination of polycyclic aromatic hydrocarbons from smoked
meat foods and liquid smoke flavorings. Mol. Nutr. Food Res. 49: 637-647.
Simko, P., Karovicová, J. and Kubincová, M. 1991. Changes in benzo[a]pyrene content in

fermented salami. Z. Lebensm. Unters. Forsch. 193: 538-540.
Simon, P., Lafontaine, M., Delsaut, P., Morele, Y. and Nicot, T. 2000. Trace determination of
urinary 3-hydroxybenzo[a]pyrene by automated column-switching high-performance liquid
chromatography. J. Chromatogr. B 748: 337-348.
Slaga, T.J., Gleason, G.L., DiGiovanni, J., Sukumaran, K.B. and Harvey, R.G. 1979. Potent
tumor initiating activity of the 3,4-dihydrodiol-7,12-dimethylbenz[a]anthracene in mouse
skin. Cancer Res. 39: 1934-1936.
Speer, K., Steeg, E., Horstmann, P., Kühn, T. and Montag, A. 1990. Determination and
distribution of polycyclic aromatic hydrocarbons in native vegetable oils, smoked fish
products, mussels and oysters, and bream from the river Elbe. J. High Res. Chrom., 13: 104-
111.
Surh, Y.-J. and Miller, J.A. 1994. Roles of electrophilic sulphuric acid ester metabolites in
mutagenesis and carcinogenesis by som polycyclic aromatic hydrocarbons. Chem.-Biol.
Interact. 92: 351-362.
Thorslund, T.W. and Farrar, D. 1990. Development of relative potency estimates for PAH and
hydrocarbon combustion product fractions compared to benzo[a]pyrene and their use in
carcinogenic risk assessment. EPA/600/R-92/134, Dept. Commerce, NTIS.
Till, M., Riebniger, D., Schmitz, H.-J. and Schrenk, D. 1999. Potency of various polycyclic
aromatic hydrocarbons as inducers of CYP1A1 in rat hepatocyte cultures. Chem.-Biol.
Interact. 29: 135-150.
The EFSA Journal (2008) 724, 109-114

Tomkins, B. A., Jenkins, R. A., Griest, W. H., Reagan, R. R. and Holladay, S. K. 1985. Fluid
chromatographic determination of benzo(a)pyrene in total particulate matter of cigarette
smoke. J. AOAC 68: 935-940
Topinka, J., Sevastyanova, O., Binkova, B., Chvatalova, I., Milcova, A., Lnenickiva, Z.,
Novakova, Z., Solansky, I. and Sram, R.J. 2007. Biomarkers of air pollution exposure – a
study of policemen in Prague. Mutat. Res. 624: 9-17.
US EPA (United Sates, Environmental Protection Agency). 1986. Guidance for health risk from
exposure to chemical mixtures. U.S. Environmental Protection Agency, Fed Reg, 51, 34014.
Valerio, F., Ciccarelli, F. and Roggi, C. 1996. Esposizione personale a benzo[a]pirene in aree
urbane e rurali (dati preliminari) [Personal exposure to benzo[a]pyrene in urban and rural
areas (preliminary data)]. In: I valori di riferimento e i valori limite nella prevenzione
ambientale e occupazionale (Aprea, C., Sciarra, G., Fiorentino, M.L. and Minoia, C., ed.),
Morgan Edizioni Tecniche, Milano, Italy, pp. 337-341.
van der Wielen, J.C.A., Jansen J.T.A., Martena, M.J., De Groot, H.N. and In’tVeld, P.H. 2006.
determination of the level of benzo[a]pyrene in fatty foods and food supplements. Food
Addit. Contam. 23(7): 709-714.
van Maanen, J.M., Moonen, E.J., Maas, L.M., Kleinjans, J.C. and van Schooten, F.J. 1994.
Formation of aromatic DNA adducts in white blood cells in relation to urinary excretion of 1-
hydroxypyrene during consumption of grilled meat. Carcinogenesis 15: 2263-2268.
van Schooten, F.J., Moonen, E.J.C., van der Wal, L., Levels, P. and Kleinjans, J.C.S. 1997.
Determination of polycyclic aromatic hydrocarbons (PAH) and their metabolites in blood,
feces and urine of rats orally exposed to PAH contaminated soils. Arch. Environ. Contam.
Toxicol. 33: 317-322.
Varlet, V., Serot, T., Monteau, F., Le Bizec, B. and Prost, C. 2007. Determination of PAH profile
by GC-MS/MS in salmon by processed by four cold-smoking techniques. Food Addit.
Contam. 24(7): 744-757.
Veyrand, B., Brosseaud, A., Sarcher, L., Varlet, V., Monteau, F., Marchand, P., Andre, F. and Le
Bizec, B. 2007. Innovative method for determination of 19 polycyclic aromatic hydrocarbons
in food and oil samples using gas chromatography coupled to tandem mass spectrometry
based on an isotope dilution approach. J. Chromatogr. A: 1149: 333-344.
Viau, C., Diakite, A., Ruzgyte, A., Tuchweber, B., Blasi, C., Bouchard, M. and Vyskocil, A.
2002. Is 1-hydroxypyrene a reliable bioindicator of measured dietary polycyclic aromatic
hydrocarbon under normal conditions? J. Chromatogr. B 778: 165-177.
Visciano, P., Perugini, M., Amorena, M. and Ianieri, A. 2006. Polycyclic aromatic hydrocarbons
in fresh and cold-smoked Atlantic salmon fillets. J. Food Prot. 69: 1134-1138.
The EFSA Journal (2008) 724, 110-114

White, S., Fernandes, A. and Rose, M. 2008. Investigation of the formation of PAHs in foods
prepared in the home and from catering outlets to detremine the effects of frying, grilling,
barbecuing, toasting and roasting. FD 06/13.) Food Standard Agency (FSA)/Central Science
Laboratory (CSL).
WHO/IPCS (World Health Organization/International Programme on Chemical Safety). 2007.

Draft guidance document on characterizing and communicating uncertainty in exposure
assessment. http://www.who.int/ipcs/methods/harmonization/areas/draftundertainty.pdf
WHO/IPCS (World Health Organization/International Programme on Chemical Safety). 1998.

Selected Non-heterocyclic Polycyclic Aromatic Hydrocarbons. Environmental Health Criteria
202. International Programme on Chemical Safety, World Health Organization, Geneva.
WHO/IARC (World Health Organization/The International Agency for Research on Cancer)

1986. Tobacco Smoking. IARC Monographs on the Evaluation of Carcinogenic Risks to
Humans, Vol. 38. International Agency for Research on Cancer, Lyon.
WHO/IARC (World Health Organization/The International Agency for Research on Cancer).

1989. Occupational exposures in petroleum refining: Crude oil and major petroleum fuels.
Lyon, International Agency for Research on Cancer. IARC Monographs on the Evaluation of
the Carcinogenic Risk of Chemicals to Man, Volume 45.
Xue, W. and Warschawsky, D. 2005. Metabolic activation of polycyclic and heterocyclic

aromatic hydrocarbons and DNA damage: A review. Toxicol. Appl. Pharmacol. 206: 73-93.
Yamazaki, H. and Kakiuchi, Y. 1989. The uptake and distribution of benzo[a]pyrene in rat after
continuous oral administration. Toxicol. Environ. Chem. 24(1/2): 95-104.
Yang, S.K. 1988. Stereo selectivity of cytochrome P450 isozymes and epoxide hydrolase in the
metabolism of polycyclic aromatic hyrocarbons. Biochem. Pharmacol. 37: 61-70.
Yang, M., Kim, S., Lee, E., Cheong, H.K., Chang, S.S., Kang, D., Choi, Y., Lee, S.M. and Janh,
J.Y. 2003. Sources of polycyclic aromatic hydrocarbon exposure in non-occupationally
exposed Koreans. Environ. Mol. Mutagen. 42: 250-257.
LIST OF ABBREVIATIONS AND ACRONYMS

AhR Hydrocarbon receptor
AIC Akaike information criterion
ATSDR Agency for Toxic Substances and Disease Registry
AUC Area under the plasma concentration time curve
BaA Benz[a]anthracene
BaP Benzo[a]pyrene
The EFSA Journal (2008) 724, 111-114

BbFA Benzo[b]fluoranthene
BcFL Benzo[c]fluorene
BcP Benzo[c]phenanthrene
BE Belgium
BG Bulgaria
BghiP Benzo[ghi]perylene
BjFA Benzo[j]fluoranthene
BkFA Benzo[k]fluoranthene
BMD Benchmark dose
BMDL Benchmark dose lower confidence limit
BMDS US EPA Benchmark dose software
BMR Benchmark response
CAOBISCO Association of chocolate, biscuit and confectionery industries of the
European Union
CAS Chemical Abstract Service Number
CHR Chrysene
CONTAM Panel EFSA Panel on Contaminants in the Food Chain
CPP Cyclopenta[cd]pyrene
CRL Community Reference Laboratory
CRM Certified reference material
CYP Cytochrome P450
CZ Czech Republic
DBaeP Dibenzo[a,e]pyrene
DBahA Dibenz[a,h]anthracene
DBahP Dibenzo[a,h]pyrene
DBaiP Dibenzo[a,i]pyrene
DBalP Dibenzo[a,l]pyrene
DD Dihydrodiol dehydrogenases
DE Germany
DIPP The Danish Import Promotion Program
DK Denmark
DNA Deoxyribonucleic acid
EC European Commission
EFSA European Food Safety Authority
EH Epoxide hydrolase
EMEP Co-operative programme for monitoring and evaluation of the long-range
transmissions of air pollutants in Europe
EPA Environmental Protection Agency
EPER European Pollutant Emission Register
EU European Union
FAO Food and Agriculture Organization of the United Nations
The EFSA Journal (2008) 724, 112-114
FI Finland
FID Flame ionisation detector
FLD Fluorescence detector
FLUO Fluorene
FR France
GB Great Britain
GC Gas chromatography
GC-ID-HR-MS Gas chromatograpgy-isotope dilution high resolution mass spectrometry
GC-MS Gas chromatography-mass spectrometry
GPC Gel permeation chromatography
GSTM1 Glutathione-S-transferase-M1
HORRAT Horwitz ratio
HPLC High performance liquid chromatography
HPLC-FLD High performance liquid chromatograph-fluorescence detector
HU Hungary
ICES International Council for the Exploration of the Sea
IE Ireland
IP Indeno[1,2,3-cd]pyrene
IP Ionisation potential
IPCS International Programme on Chemical Safety
IS Iceland
ISO International Organization for Standardization
IT Italy
JECFA Joint FAO/WHO Expert Committee on Food Additives
LB Lower bound
LD50 Lethal dose 50%
LDPE Low-density polyethylene
LOAEL Lowest-observed-adverse-effect-level
LOD Limit of detection
LOQ Limit of quantification
LSF Liquid smoke flavouring
MCH 5-methylchrysene
3-MCPD 3-monochloropropane-1,2-diol
mEH Microsomal epoxide hydrolase
ML Maximum level
MOE Margin of exposure
MPO Myeloperoxidase
MS Mass spectrometry
MW Molecular weight
NADP+ Nicotinamide adenine dinucleotide phosphate
The EFSA Journal (2008) 724, 113-114

NAD(P)H Nicotinamide adenine dinucleotide phosphate (reduced form)

NAP Naphthalene
NL The Netherlands
NO Norway
NOAEL No-observed-adverse-effect-level
NOEL No-observed-effect-level
NQO1 NAD(P)H:quinone oxidoreductase
NRL National Reference Laboratory
1-OHPG 1-hydroxypyrene glucuronide
OSPAR Protection of the Marine Environment of the North-East Atlantic
PAH Polycyclic aromatic hydrocarbon
PAH2 Sum of benzo[a]pyrene and chrysene
PAH4 Sum of benz[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene and
chrysene
PAH8 Sum of benz[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene,
benzo[k]fluoranthene, benzo[ghi]perylene, chrysene,
dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene
PDA Photo-diode array
PET Polyethylene terephthalate
PHEN Phenanthrene
PLE Pressurised liquid extraction
PM Particular matter
PTFE Polytetrafluoroethylene
PTV Programmed temperature vaporization
PYR Pyrene
ROS Reactive oxygen species
RSD Relative standard deviation
RSDr Relative standard deviation for repeatability
RSDR Relative standard deviation for reproducibility
SCF Scientific Committee on Food
SCOOP Scientific cooperation
SE Sweden
SFE Supercritical fluid extraction
SK Slovakia
SPE Solid phase extraction
TEF Toxic equivalency factor
UB Upper bound
US United States (of America)
UV Ultraviolet
WHO World Health Organization
The EFSA Journal (2008) 724, 114-114

EFSA Journal - 2008 - Polycyclic Aromatic Hydrocarbons in Food Scientific Opinion of The Panel On Contaminants in The

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

EFSA Journal - 2008 - Polycyclic Aromatic Hydrocarbons in Food Scientific Opinion of The Panel On Contaminants in The

Uploaded by

Copyright:

Available Formats

The EFSA Journal (2008) 724, 1-114

Polycyclic Aromatic Hydrocarbons in Food1

Scientific Opinion of the Panel on Contaminants in the Food Chain

Adopted on 9 June 2008

SCIENTIFIC PANEL MEMBERS

dibenzo[a,h]pyrene, dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3-cd]pyrene and 5-

2 OJ L 34, 8.2.2005, p.43

The EFSA Journal (2008) 724, 2-114

Keywords: Polycyclic aromatic hydrocarbons (PAHs), food, occurrence, indicators, exposure,

The EFSA Journal (2008) 724, 4-114

The EFSA Journal (2008) 724, 5-114

8.1.2.4 Carcinogenicity ..................................................................................................... 76

The EFSA Journal (2008) 724, 6-114

BACKGROUND AS PROVIDED BY THE EUROPEAN COMMISSION

The EFSA Journal (2008) 724, 7-114

The EFSA Journal (2008) 724, 8-114

TERMS OF REFERENCE AS PROVIDED BY THE EUROPEAN COMMISSION

The updated opinion should take into account

The EFSA Journal (2008) 724, 9-114

The EFSA Journal (2008) 724, 10-114

The EFSA Journal (2008) 724, 11-114

Table 1: Polycyclic aromatic hydrocarbons considered in the present opinion.

Benz[a]anthracene BaA 228.3 56-55-3

Benzo[b]fluoranthene BbFA 252.3 205-99-2

Benzo[j]fluoranthene BjFA 252.3 205-82-3

Benzo[k]fluoranthene BkFA 252.3 207-08-9

Benzo[ghi]perylene BghiP 276.3 191-24-2

Benzo[a]pyrene BaP 252.3 50-32-8

The EFSA Journal (2008) 724, 12-114

Chrysene CHR 228.3 218-01-9

Cyclopenta[cd]pyrene CPP 226.3 27208-37-3

Dibenz[a,h]anthracene DBahA 278.3 53-70-3

Dibenzo[a,e]pyrene DBaeP 302.3 192-65-4

Dibenzo[a,h]pyrene DBahP 302.3 189-64-0

Dibenzo[a,i]pyrene DBaiP 302.3 189-55-9

The EFSA Journal (2008) 724, 13-114

Dibenzo[a,l]pyrene DBalP 302.3 191-30-0

Indeno[1,2,3-cd]pyrene IP 276.3 193-39-5

5-methylchrysene MCH 242.3 3697-24-3

Benzo[c]fluorene BcFL 216.3 205-12-9

The EFSA Journal (2008) 724, 14-114

Smoked meats and smoked meat products 5.0

Muscle meat of smoked fish and smoked fishery products, excluding

Muscle meat of fish, other than smoked fish 2.0

Crustaceans, cephalopods, other than smoked. The maximum level applies

Bivalve molluscs 10.0

The EFSA Journal (2008) 724, 15-114

13 OJ L 184, 15.7.1988, p.61

The EFSA Journal (2008) 724, 16-114

3. Sampling and methods of analysis

The EFSA Journal (2008) 724, 17-114

3.2 Methods of analysis

The EFSA Journal (2008) 724, 18-114

A number of PAH compounds, such as benzofluoranthene isomers, have been reported to be

The EFSA Journal (2008) 724, 20-114

4. Sources and environmental fate

4.1 Formation and production

The EFSA Journal (2008) 724, 21-114

The EFSA Journal (2008) 724, 22-114

4.2 Environmental fate

The EFSA Journal (2008) 724, 23-114

The EFSA Journal (2008) 724, 24-114

The EFSA Journal (2008) 724, 25-114