The Sixteenth Conference of the International Society for Environmental Epidemiology (ISEE): Abstracts
Studies have suggested that a history of pesticide poisoning is related to prevalence of neurologic symptoms, but little is known about the effect of chronic low-level pesticide exposure in the absence of poisoning. We explored this issue in a cross-sectional analysis of data from the AHS, a cohort study of licensed pesticide applicators and their spouses. The present analysis was restricted to 18,782 white male private applicators, mostly farmers, who provided complete data on symptoms (97% of eligible applicators). At enrollment, applicators provided information on lifetime pesticide use and exposure and reported frequency during the prior year of 23 symptoms typically associated with pesticide exposure. Symptoms reflected cognitive (e.g., difficulty concentrating), motor (e.g., weakness, tremor), and sensory (e.g., numbness, visual impairment) as well as nonspecific dysfunction (e.g., headache, fatigue). We analyzed the data with logistic regression, adjusting for age, state, education, cigarette smoking and alcohol use. Lifetime days of pesticide use were categorized in three or four levels, depending on the distribution. A summary measure indicating greater symptom prevalence (10+ vs. >10 symptoms) was associated with lifetime days of use of any insecticide, with a significant dose-response; ORs (95% CIs) were 1.51 (1.27−1.80), 1.84 (1.56−2.17), and 2.49 (2.07−2.99) for 1-50, 51-500, and <500 days, respectively. Weaker relationships were seen for use of herbicides [<500 days, 1.28 (0.99−1.65)], fungicides [<50 days, 1.23 (1.07−1.42)], and fumigants [<50 days, 1.44 (1.22−1.70)]. In a model including all four categories of pesticides, associations with insecticides and fumigants were slightly reduced, but relationships with herbicides and fungicides were no longer present. Among chemical classes of insecticides, the strongest relationships were for organophosphates [<500 days, 2.20 (1.86−2.61)] and organochlorines [<50 days, 2.00 (1.80−2.24)]. Excluding applicators with a history of pesticide poisoning did not change these results. Isolated events involving higher personal exposure to pesticides, particularly those involving exposure to the lungs or digestive tract, were associated with greater symptom prevalence [3.04 (2.65−3.49)], as were physician visits for pesticide-related illness [2.25 (1.95−2.59)] and physician diagnoses of pesticide poisoning [2.46 (1.97−3.08)]. Similar associations with pesticide exposure were seen for individual symptoms as well as for groups of symptoms defined a priori and reflecting affect, systemic effects, peripheral neuropathy, motor dysfunction, cognition, or vision. These results suggest that, even in the absence of a history of pesticide poisoning, greater prevalence of self-reported neurologic symptoms is associated with use of pesticides, particularly organophosphate and organochlorine insecticides.