Pesticide Exposures and BMI in a Cohort of Pesticide Applicators Open Access
Downloadable ContentDownload PDF
The rising trend in the number of obese and overweight individuals in the United States in recent years has led to much scientific hypothesizing to identify causes beyond the well-known impact of high calorie diet combined with lack of physical exercise. Animal and in vitro studies suggest that exposure to organochlorines, organophosphates, carbamates, PBBs, PCBs, plastics, heavy metals, and solvents may be associated with weight gain through disruption of hormones, neurotransmitters, and metabolic processes, and damage to body tissues, especially nerve and muscle tissue. Several pesticides are among the chemicals that have been hypothesized to induce metabolic changes associated with weight gain in animal studies. Human studies are limited. Some have shown that rural populations have higher mean weights for each age group compared to non-rural populations. Studies have examined the relationship between blood levels of pesticides or their metabolites in relation to current body mass index (BMI) and have found that subjects with higher BMIs had higher levels of these substances in their blood. No study could be identified that obtained BMI and pesticide exposure levels at one point in time and BMI at a later point in time to prospectively assess the effects of exposure on BMI. Our research investigated the hypothesis that exposure to pesticides may be associated with higher BMI using epidemiologic data from the Agricultural Health Study (AHS). This large cohort study of almost 90,000 Iowa and North Carolina pesticide applicators and their spouses was initiated in 1993 (Phase 1: baseline/enrollment) with a follow-up 5 years later (Phase 2). The population available for our analysis included the 9,076 male private applicators that completed Phase 1, Phase 2, and the diet history distributed during Phase 2. No energy input variable was available for baseline and no measure of energy expenditure was available for Phase 1 or Phase 2; thus, we used available data to indirectly estimate and control for these. The calculated cumulative intensity-weighted days of exposure to pesticide classes from the AHS database was our main independent variable. Data were available to assess the association between BMI and exposure to particular pesticide classes for three separate studies: cross-sectional analysis of exposure to pesticide classes and BMI at enrollment; prospective analysis of BMI at 5-year follow-up in relation to pesticide exposure at enrollment; and cross-sectional analysis of BMI and pesticide exposures at follow-up. We found that triazine herbicides were significantly positively associated with BMI across stratified groups and in all three of our studies: enrollment cross-sectional, prospective, and follow-up cross-sectional. While we also saw positive associations between BMI and organochlorines and phenoxy herbicides, both of which were suggested by our literature review, these results were eliminated when we controlled for triazine herbicide exposure in our multiple and logistic regression analyses. Differences in results for the three studies are slight and may be influenced by differences in other predictors included in the models. Further investigation of triazine herbicides as a class and the individual agricultural pesticides included in this class, is warranted.