Parkinson's disease (PD) is a progressive movement disorder affecting over 500,000 US residents. The etiology of PD is still unclear. A recent twin study suggested that PD was not a mongenic disorder. Genetic susceptibility to environmental exposures involving multiple genes may still play a role: PD has been linked to polymorphisms in several genes, typically involving dopamine neurochemistry or metabolism of xenobiotics. The etiology of PD likely has an environmental component. Epidemiologic studies have demonstrated that PD risk is associated with rural living, well water drinking, farming, pesticides, and metals. In particular, several although not all studies have shown that pesticide exposure is associated with increased PD risk. PD risk has also been associated with exposure to metals and with age, diet, and lifestyle factors including cigarette smoking, which is protective. An interesting potential risk factor is the soil pathogen Nocardia asteroides. This mycobacterium causes nigral degeneration and an L-dopa responsive movement disorder in mice and monkeys. Exposure to Nocardia might present an explanation for the risk associated with rural living and farming, but human studies of PD and Nocardia exposure have been inconclusive. Epidemiologic and experimental evidence indicates that the pathophysiology of PD likely involves several interacting mechanisms, including mitochondrial dysfunction, oxidative stress, protein aggregation, and dysfunction of the ubiquitin-proteasome system, and that environmental neurotoxicants work through these pathways. We are conducting a case-control study of PD nested in the Agricultural Health Study (AHS). The parent AHS is a cohort study of ~90,000 licensed pesticide applicators and their spouses, recruited in 1993-97, designed to study cancer and other health outcomes in relation to farming associated exposures.
The specific aims of the nested case-control study of PD are to examine the relationship of PD (i) to pesticide exposure; (ii) to other neurotoxicants, particularly metals; (iii) to Nocardia asteroides; (iv) to lifestyle factors including diet, smoking, and caffeine; (v) to skin melanin, to examine racial/ethnic differences; and (vi) to polymorphisms in genes involved in xenobiotic metabolism, dopaminergic neurotransmission, or xenobiotic-specific membrane transport. Field work for the case-control study is nearly finished; we have enrolled 114 cases and 379 controls. Suspect cases are identified using information from the AHS, and the presence of PD is verified using an in-home neurologic exam and medical records. Controls are a random sample from the remaining cohort, matched to cases by age, sex, and state. Exposure is evaluated using data from three complementary sources. We utilize interview information on pesticide use, other exposures, and lifestyle already collected in the AHS. In addition, we collect blood samples to measure organochlorines, metals, and Nocardia exposure and for DNA banking. We collect samples of house and farm equipment dust to measure certain pesticides and metals. We also conduct additional interviews to obtain information on lifetime use of specific pesticides implicated in PD by case reports or animal research as well as exposure to other neurotoxicants. This study is the first to use prospectively collected exposure information to evaluate the hypothesis that pesticide exposure is related to PD risk. It exploits the unique opportunity provided by the AHS to address this issue in an occupational group defined by pesticide use, combining rigorous methods of case-finding with several complementary methods of exposure assessment. As a pilot data analysis, we have evaluated cross-sectional and prospective data collected in the AHS, using self-reported PD as an outcome. We used data from ~84,000 licensed private pesticide applicators and their spouses enrolled in the Agricultural Health Study (AHS) to evaluate the relationship of self-reported PD to pesticide exposure. Cohort members provided detailed information on lifetime pesticide use at enrollment and reported physician-diagnosed PD both at enrollment (prevalent cases, n=83) and five years later (incident cases, n=78); self-reported PD cases were compared to the remaining cohort. Prevalent cases were more likely to report parkinsonian symptoms including hand tremor; incident cases also reported increased hand tremor at enrollment, five years before reporting PD. Among incident cases, cumulative days of pesticide use was associated with PD risk for both applicators and their spouses (OR 2.3, 95% CI 1.2-4.5 for the highest quartile of use, compared to the lowest; p for trend 0.009). Receiving pesticide-related medical care was associated with increased risk, and using personal protective equipment to reduce potential exposure was associated with decreased risk. Risk was associated with chloroacetanilide, phenoxyacetate, and thiocarbamate herbicides as well as several organophosphate insecticides and some fungicides (ORs > 1.5). There was an apparent inverse association of PD with pesticide exposure among prevalent cases (OR 0.5, 95% CI 0.2-1.1), possibly because of biased reporting of pesticide use or failure of exposed individuals with PD to enroll in the original AHS cohort. This study provides further support for the hypothesis that exposure to certain pesticides increases PD risk. Findings for specific chemicals must be interpreted cautiously but may provide fruitful leads for further investigation. This analysis is limited by its use of self-reported diagnoses of PD, a problem that will be addressed in our case-control study, which will have neurologist-confirmed cases.
Showing the most recent 10 out of 13 publications
