Meta-analysis on occupational exposure to pesticides--neurobehavioral impact and dose-response relationships.

While the health impact of high exposures to pesticides is acknowledged, the impact of chronic exposures in the absence of acute poisonings is controversial. A systematic analysis of dose-response relationships is still missing. Its absence may provoke alternative explanations for altered performances. Consequently, opportunities for health prevention in the occupational and environmental field may be missed. Objectives were (1) quantification of the neurotoxic impact of pesticides by an analysis of functional alterations in workers measured by neuropsychological performance tests, (2) estimates of dose-response relationships on the basis of exposure duration, and (3) exploration of susceptible subgroups. The meta-analysis employed a random effects model to obtain overall effects for individual performance tests. Twenty-two studies with a total of 1758 exposed and 1260 reference individuals met the inclusion criteria. At least three independent outcomes were available for twenty-six performance variables. Significant performance effects were shown in adults and referred to both cognitive and motor performances. Effect sizes ranging from dRE=-0.14 to dRE=-0.67 showed consistent outcomes for memory and attention. Relationships between effect sizes and exposure duration were indicated for individual performance variables and the total of measured performances. Studies on adolescents had to be analyzed separately due to numerous outliers. The large variation among outcomes hampered the analysis of the susceptibility in this group, while data on female workers was too scant for the analysis. Relationships exist between the impact of pesticides on performances and exposure duration. A change in test paradigms would help to decipher the impact more specifically. The use of biomarkers appropriate for lower exposures would allow a better prevention of neurotoxic effects due to occupational and environmental exposure. Intervention studies in adolescents seem warranted to specify their risk.