Human Exposure to Malathion during a Possible Vector-Control Intervention against West Nile Virus. II: Evaluation of the Toxicological Risks Using a Probabilistic Approach

ABSTRACT

In order to prevent the propagation of West Nile Virus (WNV), insecticide sprayings have been carried out in several locations in North America since 1999 with the objective of controlling the mosquito populations that transmit this pathogen. An attempt to quantitatively compare the risk of developing a health response to WNV infection with the toxicological risk of insecticides is presented here. First, the acute and subchronic environmental concentrations resulting from repeated spraying events were modeled according to a reasonable worst-case spraying sequence established in an intervention program proposed by the Government of Quebec (Canada). Second, probability density functions (PDF) were established for some exposure parameters according to the data for the concerned population, when feasible. Monte Carlo analyses were performed by incorporating these PDF into the equations used to calculate the daily absorbed doses resulting from the exposure scenarios presented in the companion article (this issue). The results showed that for a significant proportion of the population, aerial and, to a lesser extent, ground sprayings of malathion can generate acute and subchronic exposure that may exceed some levels of toxicological concern based on the USEPA's reference values. Indeed, in the case of acute exposure following aerial spraying for infants, toddlers, and children, these proportions were respectively 37.1%, 59.5%, and 32.8% of the individuals, and 27.3%, 41.3%, and 24.9% following subchronic exposure. For ground spraying, these values were 12.5%, 24.2%, 8.8%; and 9.8%, 16.5%, and 7.4%. These results allowed the comparison of the probability of exceeding a level of toxicological concern for malathion exposure with the probability of developing WNV symptoms. This comparison shows that in some circumstances, the toxicological risk of malathion may exceed the infectious risk of WNV.

Key Words:

• West Nile virus
• malathion exposure
• risk assessment
• Monte Carlo analysis
• probability density functions

ACKNOWLEDGMENTS

The authors thank Onil Samuel from the Institut national de santé publique du Québec, for his help and suggestions during the course of this work, which received financial support from the Ministère de la santé et des services sociaux du Québec. We also thank Helen Fleischauer for grammatical revision of the article.

Notes

^a Mean values for age groups 7–12 months, > 12–18 months, > 18–24 months, > 24–30 months, 3 years and 4 years.

^b To avoid extremes, the distribution was truncated at 100 mg/day.

^c Adult values as reported by Stanek et al. (1997).

^d The chosen values were those in the middle of the range between the values of the neighboring age groups.

^e Geometric mean and 95th percentile. The values proposed by the USEPA (2004) for indoor children, kindergarden children, children playing outdoors, teenage soccer players, and landscapers were taken as a surrogate for infants, toddlers, children, adolescents, and adults, respectively.

^f Maibach et al. (1971) did not obtain values for the legs; therefore the values for the forearm were taken as a surrogate.

^g The mean of the values for the palm and the dorsum was taken as the mean value for the whole hand.

^h A triangular distribution was assigned.

^a RQ_effect is the cumulative daily absorbed dose divided by the corresponding adverse health-effect reference dose (RfD_effect).

^b The probability of presenting an RQ_effect > 1 (Tox) divided by the probability of developing mild flu-like symptoms caused by WNV infection (Inf, i.e., 0.52%).