Abstract
The objective of this study was to test whether there was an association between the use of glyphosate when applied by aerial spray for the eradication of illicit crops (cocaine and poppy) and time to pregnancy (TTP) among fertile women. A retrospective cohort study (with an ecological exposure index) of first pregnancies was undertaken in 2592 fertile Colombian women from 5 regions with different uses of glyphosate. Women were interviewed regarding potential reproductive, lifestyle, and work history predictors of TTP, which was measured in months. Fecundability odds ratios (fOR) were estimated using a discrete time analogue of Cox's proportional hazard model. There were differences in TTP between regions. In the final multivariate model, the main predictor was the region adjusted by irregular relationship with partner, maternal age at first pregnancy, and, marginally, coffee consumption and self-perception of water pollution. Boyacá, a region with traditional crops and. recently, illicit crops without glyphosate eradication spraying (manual eradication), displayed minimal risk and was the reference region. Other regions, including Sierra Nevada (control area, organic agriculture), Putumayo and Nariño (illicit crops and intensive eradication spray program), and Valle del Cauca, demonstrated greater risk of longer TTP, with the highest risk for Valle del Cauca (fOR 0.15, 95% CI 0.12, 0.18), a sugar-cane region with a history of use of glyphosate and others chemicals for more than 30 yr. The reduced fecundability in some regions was not associated with the use of glyphosate for eradication spraying. The observed ecological differences remain unexplained and may be produced by varying exposures to environmental factors, history of contraceptive programs in the region, or psychological distress. Future studies examining these or other possible causes are needed.
The authors gratefully acknowledge the contributions of the people of Colombia, Alejandro Rico, and the epidemiologists who participated as field coordinators in the five regions, as well as the interviewers, local authorities, and communities where the study was carried out. We thank Sandra Reza, who conducted much descriptive data analysis, and the Executive Secretariat of CICAD for their help in conducting the work.