Total diet studies in Sweden: Monitoring dietary exposure to persistent organic pollutants by a market basket approach

Abstract

In Sweden, market basket studies (MBS) have been conducted in 1999, 2005, and 2010 by the use of similar methods, with the ambition to follow a 5-year scheme. Data on the mean Swedish consumption of common foods on the market were obtained from national food production and trade statistics, and listed food items were purchased from five major Swedish grocery chains and grouped into, at minimum, 12 different food groups (e.g. cereals, meat, fish, and dairy products). By the use of MBS methods, the Swedish mean exposure of both nutrients and potentially hazardous substances could be followed. Among other substances, POPs, such as PCBs, DDT, dioxins, and brominated flame retardants (e.g. PBDEs and HBCD) have been analysed. The data obtained are well suited for time trend studies, and could be used, e.g. in validation of other methods for exposure assessment. The assessment of the Swedish mean POP exposure by MBS methods were based on analyses of several food composites representing mainly foods of animal origin. Fish was normally the single food group with highest POP levels on a fresh weight basis. Comparing POPs, sumPCBs and sumDDT were the two compound groups with the highest levels and the largest exposure by weight, but the dioxin exposure, although much lower on a weight basis, may be of similar or even greater relevance from a risk assessment perspective. When comparing the MBS performed in 1999, 2005, and 2010 a decrease in occurrence levels in food homogenates were seen for most POPs, resulting in a decreased per capita exposure in 2010, in several cases corresponding to 60 % or less of the 1999 levels. However, some of the studied PBDE congeners did not show any significant decrease during the studied time period. The POP exposure levels estimated in the 1999 TDS showed moderate to good agreement with data from a Swedish exposure study based on individual food analyses and a population-based consumption survey, performed at a similar time-point. The pollution situation in the Baltic region, and the relatively high Swedish fish consumption, are most likely important factors to explain the appreciable contribution from fish to the total exposure to many POPs. At the same time, the total Swedish mean dioxin/PCB exposure from all relevant food groups was not found to be higher than for most comparable member states within the European Union. To conclude, the Swedish MBS approach represents a robust and relatively simple and convenient type of total diet study (TDS) method for estimating the average exposure of various compounds, including POPs, present in food. By this method, changes in exposures over time, and differences between regions and countries may be studied, and emerging food hazard hopefully identified. However, as the Swedish MBS design only produces mean exposure values, this does not provide any information on actual individual exposures in a population, and extreme consumers with potentially risk patterns will not be identified. Moreover, although the TDS concept generally includes analyses of food as consumed, this was not done in the Swedish studies, and therefore the effects of food preparation could not be measured. Thus, complementary methods are recommended in monitoring exposure data in high risk groups.