Reduced intake of deoxynivalenol in The Netherlands: a risk assessment update.

In 1998 and 1999, wheat contained high levels of deoxynivalenol (DON) in The Netherlands. Eighty percent of the 1-year-old children exceeded the TDI (20% exceeded 2TDI). Assuming a long-term intake a negative effect on body weight might be possible. In 1999 the Dutch government took several risk management measures. We performed a probabilistic exposure and effect assessment and compared current and previous DON exposure and potential health effects for young children. The DON contamination of wheat was reduced for 50%. The estimated DON-intake was reduced to one-third of the previous levels. One-year-old children showed the highest DON-intake with a median of 0.46 microg/kgbw/day (95th percentile: 1.00 microg/kgbw/day). The probabilistic effect assessment showed that at the 95th percentile of DON-intake a reduced body weight was unlikely to exceed 9%. Taking into account the possible (partly) reversibility of this effect we conclude that no clear adverse health effects will be associated to the exposure (February 2000-December 2002) to DON in The Netherlands.

Risk assessment of deoxynivalenol in food: concentration limits, exposure and effects.

The mycotoxin, deoxynivalenol (DON), is produced world-wide by the Fusarium genus in different cereal crops. We derived a provisional TDI of 1.1 microg/kg body weight (bw) and proposed a concentration limit of 129 microg DON/kg wheat based on this TDI and a high wheat consumption of children. In the period September 1998-January 2000, the average DON concentration in wheat was 446 microg/kg (n = 219) in The Netherlands. During this period, the dietary intake of DON exceeded the provisional TDI, especially in children. Eighty percent of the one-year-olds showed a DON intake above the provisional TDI and 20% of these children exceeded twice the provisional TDI. Our probabilistic effect assessment shows that at these exposure levels, health effects may occur. Suppressive effects on body weights and relative liver weight were estimated at 2.2 and 2.7%. However, the large confidence intervals around these estimates indicated that the magnitudes of these effects are uncertain.