A mycotoxin-dedicated total diet study in the Netherlands in 2013: Part I – Design

Abstract

A mycotoxin-dedicated total diet study (mTDS) allowing assessment of occurrence and dietary exposure to these substances was developed and carried out in the Netherlands in 2013. First, literature was searched to establish the occurrence profile of mycotoxins. Next, foods as consumed according to the Dutch National Food Consumption Surveys (DNFCS) for young children and persons aged 7-69 years, categorised in several food categories, were ranked according to their summed consumption. Subsequently, foods with the highest consumptions were included to cover >85% of the consumption of a particular food category. In some cases, foods other than those contributing to the upper 85% consumption within a food category were included based on their expected high mycotoxin contamination. In some other cases, foods not contributing to mycotoxin exposure were excluded. This resulted in 130 foods to be included in this mTDS. Since a sample size of 12 was established per food, 1,560 food items were purchased for the population aged 7-69 years. Fifty seven additional food items were purchased to take into account the different consumption profile of young children. The 1,617 food items were prepared as consumed based on information available in the food consumption surveys. The prepared food items were combined according to the different consumption forms of the 130 selected foods (e.g. fresh, canned or frozen). This resulted in 213 subsamples, which were proportionally to their consumption further pooled into 88 composite samples. These composite samples covered 87 and 88% of the amount foods consumed by young children and the population aged 7-69 years, respectively. This design allows analysis of mycotoxin occurrence and the subsequent exposure assessment using aggregated food categories reflected by the 88 composite samples, as well as a more refined approach by analysing 213 subsamples.

Strategies for estimating human exposure to mycotoxins via food

“…Data on food consumption of a population are collected via questionnaires from volunteers and, ideally, at the individual level, at a high level of detail, and covering at least two days. At a national level, such databases are widely used to assess the dietary exposure to mycotoxins by the population (Boon et al, 2009;Sprong et al, 2016b). At the international level, various databases are available for exposure assessment.…”

“…So in case of patulin, apple products, being the main source of this mycotoxin (Beretta et al, 2000), should be sampled and combined into a separate composite sample, not containing other fruit products (Sprong et al, 2016b). The same applies to aflatoxin M 1 which is only present in dairy products, while e.g.…”

“…Even if no differentiation is made in the exposure assessment between regions, the sampling has to cover potential geographical differences. For small countries like the Netherlands, this might be less relevant (Sprong et al, 2016b), but for larger countries with different climate zones of very different ethnic groups or large variations in consumption patterns, this must be carefully addressed (Australia New Zealand Food Standards, 2011). For locally produced foods, it is important to ensure that discount shops, supermarkets and farmers' markets are sampled proportionally to their market share.…”

“…In the French study, 1,319 food items were collected, prepared and pooled, resulting in 577 composite samples for mycotoxin analysis. In the study in the Netherlands, 1,617 food items were pooled into 213 sub-samples, which were further pooled proportionally according to the consumption habits of the population in the Netherlands, eventually resulting in 88 composite samples for analysis (Sprong et al, 2016b).…”

Challenges in risk assessment of multiple mycotoxins in food

“…(Sirot et al, 2013;Sprong et al, 2016b). For the exposure assessment studies, food consumption data were mainly obtained from national food consumption surveys, including different data collection methodologies, namely 1-day (Cressey and Reeve, 2013;Han et al, 2014), 2-day (De Boevre et al, 2013;Sprong et al, 2016a;Van de Perre et al, 2015) and 3-day 24-hour recalls (Zhao et al, 2015), 3-day food diary and 7-day food diary (Sirot et al, 2013), all reported in Table 2.…”