Organochlorine pesticides and reproductive endpoints in Hispanic women

Abstract

Organochlorine pesticides (OCPs) and their persistent metabolites are ubiquitous environmental contaminants which bioaccumulate in adipose tissue and are thought to have potential endocrine disrupting effects. This cross-sectional epidemiologic study utilized an existing national data base, the Hispanic Health and Nutrition Examination Survey (HHANES), conducted in 1982-1984, to investigate associations between exposure to seven OCPs/metabolites, ppDDT, DDE, beta-hexachlorocyclohexane (beta- HCH), hexachlorobenzene (HCB), dieldrin, oxychlordane and trans-nonachlor, and the female reproductive endpoints spontaneous abortion, stillbirth, low birth weight (< 2500 gm.), birth defects, age at menarche and age at menopause. Farm work was also examined as an exposure variable. Factors associated with detection of the OCPs/metabolites in serum were investigated. The study sample included 1,599 Hispanic women 12-74 years of age who resided in three regions of the U.S., the southwest (Mexican American), Dade County, Florida (Cuban), and New York City area (Puerto Rican). Concentrations of the OCPs/metabolites were measured in serum. A reproductive history was collected via interview. Demographic and potential confounding variables evaluated were: age and season at the time serum sample was drawn, Hispanic group, serum cholesterol, education, income, country of birth, smoking alcohol consumption, living outside a central city area, Body Mass Index (BMI), parity, oral contraceptive use and diabetes. A history of spontaneous abortion, low birth weight, having a child with a birth defect and age at menarche were not associated with detection in serum of any of the OCPs/metabolites examined. Logistic regression analysis showed a moderately elevated risk for stillbirth among women who had detectable serum dieldrin; however, the risk estimate was imprecise (adjusted OR=2.24; 95% Cl 0.79-6.35). The adjusted risk estimates for stillbirth and low birth weight were elevated among women who had ever done farm work. These effects were modified by parity. The odds ratios for stillbirth and low birth weight and ever having done farm work among women with parity < 3.0 were 3.80 (95% Cl 1.34-11.54) and 2.67 (95% Cl 1.46-4.86) respectively, and for women with parity > 3.0 were 1.18 (95% Cl 0.59-2.93) and 1.47 (95% Cl 0.83-2.59), respectively. Ever having done farm work was also associated with ever having a child with a muscle, bone or joint defect (adjusted OR=2.08; 95% Cl 0.63-6.83) or a mouth defect (adjusted OR=3.12; 95% Cl 0.81-11.97). The sample size for specific types of birth defects was small; therefore, these risk estimates are imprecise. Exposure to ppDDT, beta-HCH, and trans-nonachlor was associated with earlier age at menopause in a dose dependent manner. Analysis of variance showed a significant decrease in mean age at menopause for women in the highest exposure category for ppDDT, beta-HCH and trans-nonachlor. Women with serum DDT level > 6.00 parts per billion (ppb), beta-HCH > 4.0 ppb, or trans-nonachlor ≥ 2.00 ppb, had an adjusted mean age at menopause on average 5.65, 3.35, and 5.18 years earlier, respectively, than women with serum levels of these pesticides below the detection limit. Significant two-way interactions among the three significant OCPs were found and appeared to represent additive effects. Serum levels of dieldrin, hexachlorobenzene, and oxychlordane, or ever having done farm work, were not associated with age at menopause. Alteration of the timing of menopause may be of clinical importance because of the relationship of menopause to chronic conditions such as osteoporosis and cardiovascular disease. There are important limitations of this study which necessitate viewing the study findings conservatively. Since one cannot be certain in a cross-sectional study which has occurred first, the exposure or the outcome, the possibility that higher levels of serum OCPs may be a consequence of early menopause must be considered. Lactation is an important pathway for elimination of OCPs from the body, and lack of information on lactation history may have lead to significant exposure misclassification. Women who lactated may have had lower levels of OCPs at the time of exposure measurement than at the time of the adverse reproductive events which occurred prior to lactation, driving the effect estimate downward. OCPs are long lived in adipose tissue, adipose levels may range from 200-400 times greater than whole weight serum levels, and exposure for many OCPs is considered to be continual, but the temporal stability of OCP serum levels over many years is unknown; therefore, misclassification of exposure status may have occurred due to the varying length of time between the reproductive events and measurement of exposure.