Background: Experimental studies in nonhuman primates provide evidence that low-level exposure to persistent organochlorine pollutants such as polychlorinated biphenyls (PCBs) may affect aspects of their menstrual cycle, including cycle length, regularity, and bleeding duration. Few studies have examined these associations in humans.
Methods: We used data from 2314 pregnant women who participated in the Collaborative Perinatal Project, a cohort study that enrolled pregnant women in the 1960s in 12 centers in the United States. Information about usual (prepregnancy) menstrual cycle length, regularity, bleeding duration, and dysmenorrhea was collected at enrollment, and 11 PCBs and p,p'-DDE (1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene (DDE) were measured in stored blood samples collected during the third trimester of pregnancy. We used multivariate linear and logistic regression to examine the association between organochlorine levels and menstrual cycles, adjusting for demographic factors, cholesterol, and triglycerides.
Results: Total PCBs were positively associated with increasing menstrual cycle length (adjusted difference across 5 categories of PCB exposure = 0.7 days, trend-test P value = 0.02). Irregular cycles were slightly more frequent among those in the 2 highest categories of PCB exposure (odds ratio for highest category = 1.5; 95% confidence interval = 0.70-3.3), and there also was some evidence of an association with DDE. The strengths of these associations increased with various exclusions made to decrease potential misclassification of the outcome and the exposures. There was little evidence for associations between DDE or PCBs and bleeding duration, heavy bleeding, or dysmenorrhea.
Conclusions: This study supports experimental studies in monkeys showing an effect of low-dose PCB exposure on menstrual function.