104
N. Arnich et al.
Food Sampling
Two main criteria were considered for food sampling: (i) the most heavily-consumed
foods and (ii) foods not heavily consumed but likely to be highly contaminated.
A total of 212 different food types were thus selected, covering around 90 % of
dietary consumption in the adult and child populations. Out of these 212 food types,
116 were considered as having no or little inter-regional variability (composition or
contamination). The other 96 foods were the subject of inter-regional lists in order
to take into account potential variability in composition or contamination between
regions (production and/or animal feeding methods, environmental pressure). Eight
inter-regional food lists were thus drawn up.
For each of the 212 food types, a sampling plan was followed taking into account
consumption habits in France, the flavour, the product’s origin, claims such as ‘lowfat’ or ‘organic’ for example, points of purchase (hypermarket or supermarket, retail
shops, markets), the storage method (fresh, deep-frozen, canned), the market shares
of the various brands, and so on. Purchases were made year-round, from June 2007
to January 2009, thus covering seasonal variations in food supply. Lastly, each
sample was purchased twice during the study, in order to cover potential seasonal
variability in composition or contamination. In the end, approximately 20,000 foods
were purchased in thirty-three large towns across mainland France.
Sample Analysis
For each food, only the edible part was used, and then the foods were prepared
‘as consumed’. For example, fruits and vegetables were washed. Vegetables, meat
and seafood products were cooked: braised, pan-fried, grilled, baked, deep-fried,
etc. The foods were then combined into 1,319 composite samples representative
of shopping baskets and consumer purchases for the eight surveyed inter-regions
and analyzed by accredited laboratories. Each composite sample was composed
of 15 subsamples of equal weight of the same food. These analyses led to the
production of over 230,000 analytical results, after relevant substances were tested
in the various samples: each substance was thus tested in those foods that were
known or assumed to contain it according to the scientific literature. Food analyses
were undertaken by around a dozen laboratories chosen for their capacities (national
reference laboratories and accredited laboratories) for most of the tested substances
in the targeted foods.
Exposure Assessment
Individual and population dietary exposure to each compound were assessed by
combining contamination data and national consumption data, as described below:
E i;j D
P n
kD1 C i;k L k;j
BW i
(9.1)
N. Arnich et al.
Food Sampling
Two main criteria were considered for food sampling: (i) the most heavily-consumed
foods and (ii) foods not heavily consumed but likely to be highly contaminated.
A total of 212 different food types were thus selected, covering around 90 % of
dietary consumption in the adult and child populations. Out of these 212 food types,
116 were considered as having no or little inter-regional variability (composition or
contamination). The other 96 foods were the subject of inter-regional lists in order
to take into account potential variability in composition or contamination between
regions (production and/or animal feeding methods, environmental pressure). Eight
inter-regional food lists were thus drawn up.
For each of the 212 food types, a sampling plan was followed taking into account
consumption habits in France, the flavour, the product’s origin, claims such as ‘lowfat’ or ‘organic’ for example, points of purchase (hypermarket or supermarket, retail
shops, markets), the storage method (fresh, deep-frozen, canned), the market shares
of the various brands, and so on. Purchases were made year-round, from June 2007
to January 2009, thus covering seasonal variations in food supply. Lastly, each
sample was purchased twice during the study, in order to cover potential seasonal
variability in composition or contamination. In the end, approximately 20,000 foods
were purchased in thirty-three large towns across mainland France.
Sample Analysis
For each food, only the edible part was used, and then the foods were prepared
‘as consumed’. For example, fruits and vegetables were washed. Vegetables, meat
and seafood products were cooked: braised, pan-fried, grilled, baked, deep-fried,
etc. The foods were then combined into 1,319 composite samples representative
of shopping baskets and consumer purchases for the eight surveyed inter-regions
and analyzed by accredited laboratories. Each composite sample was composed
of 15 subsamples of equal weight of the same food. These analyses led to the
production of over 230,000 analytical results, after relevant substances were tested
in the various samples: each substance was thus tested in those foods that were
known or assumed to contain it according to the scientific literature. Food analyses
were undertaken by around a dozen laboratories chosen for their capacities (national
reference laboratories and accredited laboratories) for most of the tested substances
in the targeted foods.
Exposure Assessment
Individual and population dietary exposure to each compound were assessed by
combining contamination data and national consumption data, as described below:
E i;j D
P n
kD1 C i;k L k;j
BW i
(9.1)