You are here:
TIME TO PREGNANCY IN RELATION TO TOTAL TRIHALOMETHANE LEVELS IN TAP WATER
Citation:
Mendola, P, S. Swan, Windham, Gayle C, L. Fenster, K. Waller, AND C. Ren. TIME TO PREGNANCY IN RELATION TO TOTAL TRIHALOMETHANE LEVELS IN TAP WATER. Presented at 14th Annual Meeting of International Society for Environmental Epidemiology, Vancouver, Canada, August 11-15, 2002.
Description:
Time to pregnancy in relation to total trihalomethane levels in tap water
Shanna H. Swan, Cuirong Ren, Gayle C. Windham, Laura Fenster, Kirsten Waller. (University of Missouri and California Department of Health Services).
We have previously reported increased risks of spontaneous abortion in association with high levels of exposure via ingestion to trihalomethanes in tap water. The mechanism of action remains unclear. To see if exposure to trihalomethanes is associated with other adverse reproductive outcomes, we examined time-to-pregnancy in the California Women's Reproductive Health Study (CWRHS). Participants were married women, enrolled between May 1990 and June 1991. Of 1,092 eligible women, 403 completed the study. Women were eligible even if they used barrier methods of contraception (condoms, diaphragms, caps or foam) but not if they or their partners were sterilized or used oral contraceptives, IUDs or implanted contraceptives. Subjects collected daily urine samples and kept a daily diary until their second missed or until the end of the study (range 1-8 cycles, mean 4.4 cycles). We estimated the day of ovulation and recorded the number of acts of unprotected intercourse the woman reported in a seven-day window around the day of ovulation (from five days prior to the estimated day of ovulation until one day after). This analysis includes only the 440 cycles with unprotected intercourse in the window; these were the only cycles considered to be at risk of pregnancy for this analysis. We identified all clinically confirmed pregnancies that were conceived after the start of urine collection. Of these (n=63), 54 occurred in cycles at risk, the remainder being the result of contraceptive failure or inaccurate reporting. We used survival analysis methods that allowed for time-dependent covariates and right censoring and examined the total trihalomethane level (TTHM) in relation to cycles till pregnancy. We used TTHM levels reported quarterly by utilities in the study area to estimate exposure based on subject residence, as measured during a 90-day time period starting 30 days before the start of urine collection. Adjusting for number of acts of unprotected intercourse, attitude towards becoming pregnant, caffeine and smoking, women whose tap water contained more than 60 g/l of TTHM had a somewhat delayed time to pregnancy, but estimates varied with exposure definition, which will be detailed in this presentation. Estimating exposure by including reported tap water consumption somewhat weakened this association and the amount of time spent showering did not appear to alter these results appreciably. In conclusion, these data, based on small numbers of pregnancies, suggest that elevated levels of TTHMs may affect fertility as measured by cycles-to-pregnancy.
This is an abstract of a proposed presentation and does not necessarily reflect EPA policy.