In a comprehensive study of neural tube defects, we calculated pregnancy exposure to nitrates—commonly occurring substances found in foods, medications, drinking water and cigarette smoke. Public water companies provided information on nitrate levels and water sources (groundwater, surface water or both) for each woman's residence.

	Only groundwater contained nitrate exceeding the current allowable standard of 45 milligrams/liter. Exposure above this maximum contaminant level (MCL) was associated with a 4 times higher risk for anencephaly (absence of the brain). There was no increased risk for spina bifida (open spine defects), another type of neural tube defect.
	Women whose drinking water contained nitrate levels below the MCL had a higher risk for anencephaly, but only when the source was groundwater. No increased risk was seen at comparable nitrate levels when drinking water was a mixture of surface and groundwater.
	Similar levels in groundwater and mixed water did not have the same effect. This inconsistency raises the possibility that some other factor, or combination of factors, is responsible for the increased risk noted in groundwater drinkers.
	Other contaminants in the groundwater—such as pesticides—did not provide an alternate explanation for the increased risk.
	It is possible that nitrate exposure in water was misclassified, since it was estimated at the water supplier rather than at the tap. Mixed water sources may alternate between surface and groundwater, defying accurate classification.

SANTA CLARA COUNTY WATER STUDIES

Several early California Birth Defects Monitoring Program studies looked at heart defects and drinking water in births from 1981-1983; this was prompted by the detection in 1981 of solvent contamination in the groundwater drinking supply for a small area of Santa Clara County.

A possible relationship between drinking tap water and heart defects was noted countywide (beyond the area of known contamination), primarily in 1981. Drinking bottled water was a protective factor. However, there were many potential sources of bias in the study—interviews were conducted 3-7 years after birth; intense publicity about water contamination may have influenced mothers' responses.

A follow-up study used county water company records to determine if mothers' homes received chlorinated water during early pregnancy and to estimate exposure to chlorine (added to water as a disinfectant) and trihalomethanes (a disinfectant by-product). Neither substance was associated with higher risks, whether exposure came through drinking or bathing/ showering.