CANCER INCIDENCE AMONG PESTICIDE APPLICATORS EXPOSED TO ALACHLOR OR ATRAZINE IN THE AGRICULTURAL HEALTH STUDY

Abstract

Objectives We evaluated the cancer incidence among over 49,000 pesticide applicators with exposure to either alachlor or atrazine in the Agricultural Health Study (AHS), a prospective cohort study of over 89,000 restricted use pesticide applicators and spouses of farmer applicators in Iowa and North Carolina, USA. Methods Comprehensive questionnaire data concerning lifestyle, occupations, personal and family medical history and use of 50 important agricultural pesticides, were collected prior to cancer incidence. Standardized incidence ratios (SIR) were estimated for those exposed and those not exposed to alachlor and for those exposed and those not exposed to atrazine, two important agricultural herbicides. Comparing the cancer experience of pesticide applicators from each state with that expected based on state specific incidence rates. Poisson regression was used to controlling for the effects of potential confounders (including age, sex, smoking, alcohol, education, state of residence, enrollment year, family history of cancer and the 5 [in the case of alachlor] or [10 in the case of atrazine] other pesticides most correlated with either alachlor or atrazine). Results A total of 1,466 incident malignant neoplasms occurred during the study period, 1993–2000. The SIR analysis showed a moderate deficit for all malignant neoplasm combined among both users and non-users of these herbicides. In Poisson regression analysis of alachlor, we found a significant increasing trend for the incidence of all lymphohematopoietic cancer and cumulative alachlor exposure (p for trend = 0.02). Incidence of leukemia and multiple myeloma was markedly increased among applicators in the highest exposure category. Similar analyses for other cancers did not suggest any association. For atrazine, no significant exposure-response trends were observed for any cancers or for all cancers combined, but non-significant positive trends were observed for cumulative atrazine exposures cancers of the rectum, bladder, and testicles. Conclusion Our findings suggests an association between alachlor exposure and increased incidence of lymphohematopoietic cancer in the Agricultural Health Study. Findings for atrazine are inconclusive at this time. Although the statistical power of the AHS is still low for some cancer evaluations after only 5.3 years of cohort follow-up, the comprehensive prospectively collected data, especially concerning pesticide exposures is an important strength of the study. Further details of the study are provided at our website: www.aghealth.org. Words = 415