Prevalent risk to pregnant women: studying household pesticide exposures. (Science Selections).

Traditionally, risk assessment of pesticide exposure has focused on agricultural and occupational use of pesticides. However, there is growing evidence that pesticide concentrations may be even higher in urban areas--especially within homes--than in rural areas. In this month's issue, a team of researchers led by Gertrud Berkowitz at the Mount Sinai School of Medicine report that pregnant women in New York City face considerable pesticide exposure [EHP 111:79-84]. The findings are part of Mount Sinai's prospective Children's Environmental Health Study, which is examining the effects of indoor pesticide exposure on feral growth and development among these women's babies.

Among the pesticides studied were chlorpyrifos and pentachlorophenol. Chlorpyrifos residues persist up to two weeks after application, exposing people to levels far above those recommended by the U.S. Environmental Protection Agency (EPA)--potential exposure to young infants can reach levels 60-120 times greater than the EPA-recommended reference levels. Although an agreement between the EPA and pesticide manufacturers ended the sale of virtually all household-use products containing chlorpyrifos by late 2001, at the time the study was being set up (1998), chlorpyrifos was the most frequently used pesticide in Manhattan and Brooklyn, and the chemical may still be stored in some homes. Pentachlorophenol is used as a fungicide and herbicide, and was widely used as a wood preservative until the 1970s.

Other studies have quantitated urban use of household pesticides and measured pesticide exposure in adults and children. This is the first, however, to look at urinary pesticide metabolites in pregnant women as a marker of their exposure to these chemicals.

For this part of the study, the cohort included 386 pregnant Hispanic, black, white, and mixed-race women who went on to give birth at Mount Sinai Hospital between May 1998 and July 2001. The researchers collected a urine specimen from each woman during her third trimester and quantified the levels of urinary pesticide metabolites. Each woman also filled out a questionnaire that assessed her exposure to pesticides in her home and in common areas of her apartment building.

When considering reported pesticide use by someone living within the home, exposure was higher among black and Hispanic women, younger women, and single and cohabiting (versus married) women. However, when considering any reported pesticide use, including that by an exterminator or building employee (such as a superintendant), the sociodemographic differences disappeared. Only 46.4% of the women reported that they or a family member had applied pesticides during the woman's pregnancy. However, when pesticides applied by exterminators and building employees were also considered, a total of 72.3% of this pregnant cohort were exposed, a number close to the 80% previously reported for a different pregnant New York cohort.

However, the researchers found that pesticide metabolite levels were higher in this population than in some previously described populations. For example, the median metabolite concentration of 11.3 [micro]g/g for chlorpyrifos was similar to that found in another recent study of children but higher than the median found in the National Health and Nutrition Examination Survey (NHANES) III. Similarly, the median metabolite concentration of 7.3 [micro]g/g for pentachlorophenol was over six times that found in NHANES III but similar to levels among children in a recently reported German study.

To explain the discrepancies between questionnaire and metabolite data, the researchers point out that questionnaires tend to yield only limited information on the specific type and amount of exposure, and are also subject to over- and underreporting. On the other hand, metabolite data reflect not only home-use pesticide exposures but also exposures through food, the workplace, and other sources. While understandable, these limitations point up the difficulty of accurately estimating pesticide exposure and should be considered, the authors say, when interpreting the results of this and similar studies.