Академический Документы
Профессиональный Документы
Культура Документы
Environment International
journal homepage: www.elsevier.com/locate/envint
a r t i c l e
i n f o
Article history:
Received 12 October 2012
Accepted 22 December 2012
Available online 31 January 2013
Keywords:
Total diet study
Polycyclic aromatic hydrocarbons
PAH4
Human exposure
a b s t r a c t
In the frame of the second French Total Diet Study (TDS), the 15 + 1 EU priority polycyclic aromatics hydrocarbons (PAHs) were analyzed in 725 foodstuffs habitually consumed by the French population, using gas
chromatography coupled to tandem mass spectrometry, after pressurized liquid extraction and purication
on PS-DVB stationary phase. The highest PAH concentrations recovered in foodstuffs corresponded to the
following contributors: chrysene (25.7%), benzo[b]uoranthene (15.0%) and benz[a]anthracene (9.0%)
whereas the lowest concentrations were those of dibenz[a,h]anthracene, 5 methylchrysene and dibenzo[a,
h]pyrene (below 2.0%). By food groups, the current highest levels of total PAH were detected in mollusks
and crustaceans, followed by the different oil based products. To estimate French population's exposure,
contamination data were combined with national individual food consumption data. Mean daily exposure
to the sum of benzo[a]pyrene, benz[a]anthracene, chrysene and benzo[b]uoranthene (PAH4) was estimated
to be 1.48 ng/kg bw/day in adults and 2.26 ng/kg bw/day in children. The main contributors to PAH exposure
for adults are fats, bread and dried bread products followed by crustaceans and mollusks. The margin of
exposure (MOE) approach indicates that exposure to PAHs through food is not a major health problem for
French consumers.
2013 Elsevier Ltd. All rights reserved.
1. Introduction
Polycyclic aromatic hydrocarbons (PAHs) are a large group of
organic compounds containing two or more aromatic rings and
belonging to the Food and Environment Contaminants. They are
produced by natural and anthropogenic processes. Among these
sources, pyrogenic (incomplete burning of coal, oil, gas, wood,
garbage, or other organic substances) and petrogenic inputs are the
two main sources of PAHs (Hodgeson, 1990; Wang et al., 2001).
Non smokers are mainly exposed to PAHs through food and air.
PAHs are signicantly present in food due to heat processes such
as smoking, grilling and smoke-drying. Food can also be naturally
Corresponding author.
E-mail address: laberca@oniris-nantes.fr (B. Veyrand).
1
Alexandra Tard is now employed with the European Food Safety Authority (EFSA)
in its ANS Unit that provides scientic and administrative support to the ANS Panel.
The present article is published under the sole responsibility of the authors and may
not be considered as an EFSA scientic output. The positions and opinions presented
in this article are those of the authors alone and are not intended to represent the
views or scientic works of EFSA. To know about the views or scientic outputs of
EFSA please consult its website under http://www.efsa.europa.eu.
0160-4120/$ see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.envint.2012.12.011
12
monitoring benzo[a]pyrene alone is not a suitable indicator of the occurrence of total PAHs in food while the sum of benzo[a]pyrene, chrysene,
benzo[b]uoranthene and benz[a]anthracene (PAH4) and/or the sum of
PAH4, benzo[k]uoranthene, benzo[ghi]perylene, dibenz[a,h]anthracene
and indeno[1,2,3-c,d]pyrene (PAH8) are most suitable indicators of
PAHs in food, although considering PAH8 do not provide much added
value compared to PAH4. As a consequence, regulation 1881/2006/EC
has been amended (European Union (EU), 2011), for application starting
1 September 2012, to introduce new maximum levels for the 4 PAHs of
interest whilst maintaining a separate maximum level for benzo[a]
pyrene, thus ensuring a more efcient associated risk assessment.
Based on the relative carcinogenic potential of the 11 most toxic PAHs
that were most representative of food contamination (AFSSA, 2003), toxic
equivalency factors (TEFs) used to be adopted for risk characterization, and
an extra risk factor of 106 for cancer could be calculated on the basis of a
virtually safe dose of 5 ng/kg bw/day (RIVM, National Institute of Public
Health and the Environment, 2001). EFSA however, no longer recommends using TEFs due to a lack of data from oral carcinogenicity studies
for PAHs and the low predictivity associated to this approach. Risk characterization is now estimated by calculating margins of exposure on the
basis of the lower limit of the condence interval of the benchmark dose
associated with an incidence of 10% of induced tumors, i.e. BMDL10 of
0.34 mg/kg bw/day for the sum of the PAH4 (EFSA, 2008).
Performed at the national level, total diet studies (TDSs) aim at
assessing the dietary exposure of the population to chemicals, using
contamination data from food samples representative of the whole
diet and prepared as consumed by the population (EFSA, 2011;
WHO, 1985). In 2006, the French Agency for Food, Environmental
and Occupational Health and Safety (ANSES), started the second
TDS to estimate dietary exposure to different substances including
PAHs. The National Reference Laboratory for PAHs (LABERCA, Nantes,
France) was appointed to analyze the 15 + 1 EU priority PAHs by an
in-house developed and accredited method (Veyrand et al., 2007).
In the present study, the occurrence data collected during the second
French TDS for a number PAHs and their distribution in the French diet
are presented together with the evaluation of the dietary intake by the
French population for corresponding risk assessment purposes. As mentioned in the scientic opinion of EFSA (2008), PAH4 is the most suitable
indicators of PAHs in food and therefore, results of this study have mainly
focused on this indicator.
2. Experimental
2.1. Sample selection and preparation
Consumption data by the general French population were taken from
the second French Individual and National Food Consumption Survey
(INCA2), conducted between 2006 and 2007 (Dubuisson et al., 2009;
Lioret et al., 2010), which described the dietary habits (in terms of quality,
quantity, and food preparation) of children and adults representative of
the French population. A food sampling has been made on the basis of
the INCA2 study (Sirot et al., 2009) taking into account the most consumed products and also some foodstuffs less consumed but suspected
to be contaminated at high levels. Samples of foodstuffs were purchased
from June 2007 to January 2009 in eight great regions of the French metropolitan territory (when consumed in the region) and each food sample
was collected, when possible, in winterautumn and in springsummer
in order to take into account potential seasonal variations in the contamination. The corresponding sampling methodology has already been described in detail (Sirot et al., 2009). Food samples included: meat and
meat products, sh, crustaceans and mollusks, vegetables, eggs and egg
products, dairy products, cereals, oils and fats, industrial bakery, and
drinks (Table 1). A total of 1319 composite samples have been selected
for the TDS, covering approximately 90% of the total diet of adults and
children. For each sample, 15 sub-samples of equal weight of the same
food item were cooked in stainless steel, aluminum or cast iron cookware
and mixed (using stainless steel knives and plastic recipients) to obtain
a homogenized individual composite sample. Among the 1319 composite samples prepared, 725 samples have been selected for PAH analysis,
corresponding to 114 different food items.
2.2. Reagents and chemicals
All solvents used (e.g. dichloromethane, hexane, acetone, ethanol, cyclohexane, ethyl acetate and toluene) were of picograde quality and
obtained from Promochem (Wesel, Germany). Florisil (100200
mesh) was from Promochem (Wesel, Germany). SPE Envi Chrom-P cartridges (80160 m spherical particles) were provided by Sigma-Aldrich
(St. Quentin Fallavier, France). The isotopic labeled internal standard compounds 13C-PAHs and 12C-PAHs were purchased from Promochem.
Fluorinated PAHs were purchased from Chiron (Trondheim, Norway).
2.3. Sample preparation
Preparations of the samples were based on a previous method
described by Veyrand et al. (2007).
For solid materials, samples were freeze-dried. The dry residue
was weighed in order to determine its water content. One gram of
the dry residue was taken and spiked with a mixture of 13C-labeled
internal standard (IS, n = 14).
Extraction of PAHs was performed using pressurized liquid extraction
on an ASE 300 (Accelerated Solvent Extractor), from Dionex Corp. A
cellulose lter was placed at the bottom of the ASE cell (66 mL) and lled
up with 15.0 g of Florisil. The phase was pre-washed in the ASE system
with CH2Cl2. One gram of the dry residue sample was introduced into
the cell, and extraction was performed with a mixture hexane/acetone
(50/50, v/v).
For liquid matrices, 10 mL was transferred into a 50 mL tube and
spiked with the 13C-labeled internal standard mixture. PAHs were
extracted two times by liquid/liquid extraction with a cyclohexane/
ethyl acetate (50/50, v/v) mixture.
The extracts of liquid and solid matrices were puried onto a SPE
cartridge (Envi Chrom-P) as follows: after conditioning, the sample
was loaded. The cartridge was washed with a mixture of cyclohexane/
ethanol (70/30, v/v) and target compounds were eluted using 12 mL
of cyclohexane/ethyl acetate (40/60, v/v) as mobile phase. Fluorinated
PAHs were used as external standards. Two microliters of the nal extract (in toluene) were analyzed by GCMS/MS.
2.4. Gas chromatography coupled to tandem mass spectrometry (GCMS/
MS) measurement
For GCMS/MS analysis, a gas chromatograph (Agilent, 6890 Series)
and a programmable oven reaching up to 350 C was coupled to a
Quattromicro GC triple quadrupole analyzer (Waters, Micromass) operating in electron ionization mode. The sample extracts were injected in the
splitless mode. The injector temperature was kept at 300 C. The transfer
line temperature was set at 320 C. Helium was used as the carrier
gas at a ow rate of 1.0 mL min 1. Separation was performed on a
column Zebron ZB-50MS (30 m 0.25 mm 0.25 m) purchased
from Phenomenex (Le Pecq, France). The column temperature program was as follows: 110 C (1 min), 20 C min1 to 240 C (0 min),
5 C min1 to 320 C (10 min). Electron ionization (EI) was operated at
70 eV in any case. Temperature of the source was kept at 230 C. Argon
was used as collision gas at a pressure of approximately 3104 mbar.
PAHs were detected and quantied by monitoring two specic
transitions.
2.5. Performances
The described method has been validated and performances have
been found t-for-purpose: limits of detection (LODs) and limits of
13
Table 1
Number of analysis (n), mean concentrations expressed in g/kg of fresh matter (medium bound values), and percentage of detection in the 725 analysis.
BaP
BaA
BbF
CHR
PAH4
Food category
Conc (g/kg)
% > LOD
Conc (g/kg)
% > LOD
Conc (g/kg)
% > LOD
Conc (g/kg)
% > LOD
Conc (g/kg)
37
6
4
3
80
4
8
7
3
8
46
18
61
8
30
3
6
64
38
16
22
32
68
9
29
75
38
2
0.234
0.241
0.151
0.088
0.045
0.021
0.024
0.044
0.034
0.024
0.025
0.019
0.010
0.034
0.009
0.011
0.012
0.012
0.068
0.011
0.014
0.015
0.005
0.003
0.002
0.002
0.002
0.001
76
100
100
67
88
50
56
67
50
61
61
41
33
83
27
34
50
30
100
5
13
34
28
56
62
47
32
0
0.490
0.408
0.363
0.184
0.055
0.067
0.076
0.047
0.037
0.038
0.029
0.036
0.021
0.012
0.022
0.013
0.019
0.014
0.053
0.011
0.009
0.009
0.007
0.003
0.002
0.002
0.001
0.002
92
100
100
100
88
100
88
100
75
100
72
92
100
17
40
66
63
73
100
27
50
10
75
78
35
27
13
0
1.553
0.326
0.181
0.147
0.080
0.056
0.037
0.056
0.037
0.036
0.026
0.027
0.013
0.012
0.013
0.017
0.014
0.015
0.053
0.010
0.012
0.010
0.009
0.002
0.002
0.002
0.001
0.001
89
100
100
100
100
100
88
100
100
100
76
93
100
50
57
71
100
67
100
64
63
50
74
56
62
60
34
0
2.009
0.946
0.649
0.076
0.138
0.171
0.143
0.105
0.078
0.064
0.071
0.061
0.047
0.031
0.043
0.041
0.032
0.029
0.093
0.036
0.028
0.028
0.022
0.004
0.004
0.003
0.002
0.002
95
100
100
100
100
100
91
100
100
100
96
95
67
100
80
76
100
69
71
73
63
63
82
67
72
59
34
0
4.285
1.921
1.344
0.495
0.319
0.315
0.28
0.251
0.187
0.162
0.151
0.142
0.091
0.089
0.087
0.082
0.076
0.071
0.068
0.068
0.064
0.058
0.042
0.012
0.011
0.009
0.006
0.006
quantication (LOQs) for solid matrices were evaluated and ranged from
0.008 to 0.017 g kg1 and from 0.026 to 0.055 g kg1 respectively. For
liquid matrices, LODs were between 0.001 and 0.080 g/L and LOQs between 0.003 and 0.240 g/L. The linearity was assessed on seven calibration levels for each analyte over the respective range of 0.150 g kg1 of
dry matter. Correlation coefcient (R2) was better than 0.990 for all
analytes. Recoveries were found between 50% and 70%.
2.6. Internal quality controls and statistical analysis
Accuracy of the method has been veried yearly through the participation to European prociency tests organized by the European
Union Reference Laboratory (JRC-IRMM, Geel, Belgium). A quality
control sample and a blank sample have been systematically analyzed
within every batch, and performances have been checked through a
quality control chart over the whole study.
For each congener, two hypotheses have been made to manage
the left-censored data (i.e. concentration under the analytical limits).
Under the lower bound hypothesis (LB), censored data were assigned
a zero, and under the upper bound hypothesis (UB), censored data
were assigned the analytical limit for the considered couple sample/
PAH congener. Ordinary statistical methods were used to calculate
the arithmetical mean and 95th percentile (P95) on number (n) of
samples of general food groups or product groups.
where Ei,k is the exposure of the subject i to the congener k, Ci,j is the
consumption level of the food j by the subject i, Kj,k is the contamination level of congener k in the food j, BWi is the individual body
weight of the subject i and n is the total number of foods in the diet
of the subject i. Each food item consumed by each subject was
assigned the mean contamination level of the food samples collected
during seasons in the region of the subject, or the mean of the other
regions when the food we not collected in the region.
Means and 95th percentiles of exposure were then calculated for
adults and children. Exposure levels were calculated for the sum of
PAH4 in weight.
For the congeners that were taken into account in the calculation of
PAH4, the medium bound was applied (MB), i.e. when a congener was
not detected in the food matrix, then concentration was assumed to be
0.5 LOD.
To evaluate the risk of the exposure to PAHs, margins of exposure
were calculated for adults and children as the ratio between EFSA's
BMDL10 for PAH4 (0.34 mg/kg bw/day) and exposure (mean and P95).
Percentage of contribution of the different food groups considered
in the present study was calculated for adults and children, as the rate
of the exposure through the food group consumption in the total
mean dietary exposure.
3. Results
14
30%
General distribution of PAH4 analytes (%)
25.7%
BaP
8%
BbF
28%
25%
CHR
47%
Contribution (%)
BaA
17%
20%
15.0%
15%
10%
9.0%
7.2%
6.8%
5.6%
5%
4.5%
5.0%
4.4%
2.0%
1.6%
2.1%
2.0%
5.4%
1.9%
1.7%
0%
BaP
CHR
BaA
BbF
BkF
IP
DBahA BghiP
BjF
BcF
CPP
5-MCH
Fig. 1. Relative proportions for the mean concentrations (medium bound values) of the 15 + 1 EU priority PAHs in the 725 food samples analyzed. General distribution of PAH4
analytes are illustrated in the top right-hand corner.
interest were reported below 0.5 LOD (0.07 g/kg) attesting for low
contamination of this particular food product (Martorell et al., 2010).
Oil-based products were found to be the second most contaminated matrices, with a mean PAH4 concentration of 1.92 g/kg for oils
and 1.34 g/kg for margarine. Similarly, Martorell et al. (2010)
reported a mean PAH4 concentration of 1.96 g/kg for oils and fats.
Sub-ppb BaP levels have also been reported for various fat products
in the US (Kazerouni et al., 2001). These levels of contaminations
are 2 to 3 times lower than those described in the EFSA scientic
opinion, which mentioned a mean PAH4 concentration between
4.27 g/kg and 4.65 g/kg in 899 fats and oils (excluding cocoa butter
and pomace oil). Perell et al. (2009) reported a PAH4 concentration
of 19.58 g/kg for raw olive oil.
Concentrations observed in the brewed coffee samples analyzed in
the present study (Table 1) were very low. PAHs, found in coffee
powder as a result of the drying process as reported in the EFSA
report (mean lower bound concentration of 2.32 g/kg in 28 coffee
samples), were not extracted in the present study during the sample
35.0
30.0
R = 0.991
25.0
R = 0.995
20.0
15.0
10.0
R = 0.992
5.0
0.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
OILS
BbF
17%
MOLLUSCS
BaP
13%
15
BbF
36%
BaP
5%
BaA
21%
CHR
47%
CHR
49%
BaA
12%
Fig. 3. Mean distribution of the 4 analytes of PAH4 in (a) 6 oil samples and (b) 21 mollusk samples.
Children (n = 1444)
Analyte
Mean
P95
Mean
P95
BaA
BaP
BbF
BcFL
BghiP
BjF
BkF
CHR
CPP
DBahA
DbaeP
DbahP
DbaiP
DbalP
IP
MCH
PAH4
PAH8
0.300
0.191
0.301
0.260
0.417
0.157
0.132
0.688
0.424
0.071
0.124
0.110
0.111
0.129
0.185
0.115
1.478
2.281
0.593
0.350
0.715
0.448
0.788
0.333
0.299
1.383
0.756
0.117
0.217
0.195
0.197
0.223
0.337
0.180
2.998
4.454
0.466
0.304
0.439
0.445
0.631
0.243
0.194
1.053
0.709
0.118
0.202
0.188
0.190
0.208
0.291
0.189
2.259
3.493
0.921
0.619
0.943
0.867
1.327
0.516
0.417
2.250
1.447
0.223
0.395
0.366
0.367
0.395
0.592
0.344
4.694
7.322
16
Table 3
Exposure (mean and 95th percentile) to PAH4 through the different food groups of French adults and children (ng/kg bw/day) and contribution of the food groups to the total mean
exposure.
Food group
Adults (n = 1918)
Children (n = 1444)
Mean
P95
Contribution (%)
Mean
P95
Contribution (%)
0.193
0.240
0.107
0.015
0.093
0.095
0.047
0.186
0.035
0.060
0.03
0.029
0.065
0.058
0.021
0.006
0.021
0.040
0.029
0.001
0.020
0.023
0.016
0.002
0.028
0.011
0.008
0.000
1.974
0.787
0.799
0.153
0.172
0.572
0.256
0.321
0.203
0.202
0.101
0.253
0.294
0.114
0.096
0.196
0.052
0.066
0.089
0.031
0.106
0.036
0.057
0.035
0.111
0.021
0.046
0.030
13.1
16.2
7.3
1.0
6.3
6.4
3.2
12.6
2.4
4.1
2.0
2.0
4.4
3.9
1.4
0.4
1.4
2.7
2.0
0.1
1.4
1.5
1.1
0.1
1.9
0.7
0.6
0.0
0.098
0.342
0.111
0.027
0.156
0.166
0.196
0.161
0.117
0.146
0.07
0.043
0.140
0.123
0.025
0.018
0.027
0.063
0.041
0.001
0.070
0.032
0.018
0.006
0.001
0.023
0.039
0.000
2.365
1.457
0.874
0.199
0.288
0.794
0.575
0.280
0.417
0.423
0.220
0.299
0.311
0.214
0.133
0.091
0.081
0.108
0.128
0.047
0.290
0.056
0.078
0.044
0.044
0.038
0.132
0.093
4.3
15.2
4.9
1.2
6.9
7.4
8.7
7.1
5.2
6.5
3.1
1.9
6.2
5.4
1.1
0.8
1.2
2.8
1.8
0.0
3.1
1.4
0.8
0.3
0.0
1.0
1.7
0.0
4. Conclusions
In summary, the results of the present TDS allowed estimating the
general contamination of the French diet in PAHs: mollusks and
oil-based products were found to be the most contaminated samples;
chrysene, benzo[b]uoranthene and benz[a]anthracene were quantied in foodstuffs as the most important contributors. PAH4 was furthermore conrmed as suitable indicator of PAHs in food. This study
also allowed calculating the exposure of the French population to
PAHs through the whole diet. MOEs approach indicates that exposure
of this class of compounds through the diet does not represent a food
safety issue. Comparison of the present data with the last exposure
estimation performed in 2003 in France indicates a three times decrease in the exposure level. Similar trends have been reported in
other countries and related mainly to environmental causes (general
reduction of industrial and trafc emissions (Wang et al., 2009)
resulting in a decrease in atmospheric PAH levels (Meijer et al.,
2008)). Aside from these environmental conditions that may lead to
a decrease in the food contamination, the establishment of temporal
trends must carefully take into account the TDS methodology and associated analytical performances. Indeed, the continuous enhancement of analytical performances allows decreasing the limits of
quantication, which consequently leads to decreasing upper bound
values of exposure.
References
AFSSA. Avis de l'Agence franaise de scurit sanitaire des aliments relatif a une
demande d'avis sur l'valuation des risques prsentes par le Benzo(a)pyrne
(B(a)P) et par d'autres hydrocarbures aromatiques polycycliques (HAP), prsents
dans diverses denres ou dans certaines huiles vgtales, ainsi que sur les niveaux
de concentration en HAP dans les denres au-del desquels des problmes de
sant risquent de se poser. Saisine n 2000-SA-0005 - AFSSA, Maisons Alfort,
France; 2003.
Bartle KD. In: Creaser C, Purchase R, editors. Food contaminants, sources and surveillance.
Cambridge: The Royal Society of Chemistry; 1991. p. 4160.
17