Following the EPANET exposure assessment, there were 61 children with congenital anomalies among 1,658 children with some prenatal PCE exposure and 95 children with congenital anomalies among 2,999 children with no prenatal PCE exposure. The corresponding prevalence proportions per 1,000 births were 3.7 and 3.2, respectively. Many characteristics of the exposed and unexposed groups were similar (Table ). For example, mothers in both groups were predominantly white, and comparable proportions had medical conditions, prenatal multivitamin use, and exposure to non-drinking water sources of solvents. However, there were also many differences between the groups. Due to the timing of the PCE contamination, exposed mothers were more likely to give birth in later calendar years. Exposed mothers and fathers were older than unexposed parents, and exposed mothers were less likely to smoke cigarettes during the first trimester. Exposed children with anomalies were also more likely to be male and have mothers with high educational levels and prior pregnancy losses, and alcoholic beverage consumption during the first trimester.
Distribution of Selected Characteristics of Parents and Children by Prenatal PCE Exposure and Outcome Status
Nearly 39% of mothers could not recall their water consumption or bathing patterns during the study pregnancy (data not shown). However, among women who could recall this information, the proportions who drank bottled water (about 22%), consumed more than four glasses of tap water per day (about 51%), and took long showers (about 23%) were similar across the exposed and unexposed groups.
There was a wide distribution of PCE exposure levels encompassing several orders of magnitude in the exposed group. Average monthly PCE exposure levels during the LMP year ranged from 9.6E-05 to 131.8 grams. The 25th
percentiles were 0.1, 0.6, 2.3, and 6.4 grams, respectively. As previously described, the exposure measures were based on the mass of PCE delivered to a home in each calendar year. The annual mass of PCE entering a home was diluted in an estimated 90,000 gallons of water, the annual usage of average households in Massachusetts [25
], and only a small portion of this water was directly consumed by the subjects. Using this annual estimate of household water use, we converted the PCE mass delivered to a home during pregnancy to average annual point concentrations and estimated that the PCE concentrations in the water entering the homes ranged from less than 1 ug/L to 5,197 ug/L. These concentrations are consistent with actual water sampling data from the time period [1
The crude and unadjusted GEE odds ratios for all congenital anomalies combined were 1.2 (95% CI: 0.8, 1.6) and 1.1 (95% CI: 0.8, 1.6), respectively, among children with any prenatal PCE exposure (Table ). These odds ratios were virtually unchanged when maternal and paternal age were controlled simultaneously (multivariate GEE OR: 1.2, 95% CI: 0.8, 1.7, Table ) and when other confounders, including calendar year of birth; mother's educational level, cigarette smoking, alcoholic beverage consumption, and prior pregnancy losses; and child's gender were controlled one at a time (GEE ORs: 1.1-1.2, data not presented in Table ). These results were unchanged when only major malformations were examined.
Frequencies, Crude, Unadjusted and Adjusted GEE Odds Ratios (OR) and 95% Confidence Intervals (CI) for All Congenital Anomalies Combined by Prenatal PCE Exposure
The parental-age adjusted GEE odds ratio for all anomalies was elevated by 40% (95% CI: 0.9-2.2) among children whose average monthly prenatal exposure was greater than 1.136 grams, the cut point corresponding to an average drinking water concentration of 40 ug/L, and elevated by 50% among children whose average monthly prenatal exposure was > = 75th percentile (95% CI: 0.9-2.5) (Table ). The 75th percentile corresponded to an average monthly prenatal exposure of 2.3 grams. No meaningful increases in risk were seen for lower exposure levels. Again, these results were unchanged when only major malformations were included.
When organ system and diagnostic groups were examined (Table ), we found large increases in the odds ratios for neural tube defects (GEE OR 3.5, 95% CI: 0.8-14.0) and oral clefts (GEE OR 3.2, 95% CI 0.7-15.0); and modest increases in the odds ratios for gastrointestinal (GEE OR 1.8, 95% CI: 0.7-4.4) and genitourinary malformations (GEE OR 1.6, 95% CI: 0.6-3.8), including hypospadias (GEE OR: 1.4, 95% CI: 0.4-5.4); and chromosomal malformations (GEE OR: 1.4, 95% CI: 0.3-6.1) among children with any prenatal PCE exposure (Table ). No meaningful increases in odds ratios were seen for cardiac and musculoskeletal malformations, and there were too few exposed cases to estimate odds ratios for eye, ear, respiratory, and other malformations (Table ).
Frequencies, Odds Ratios (ORs), and 95% Confidence Intervals (CI) for Congenital Anomaly Categories by Prenatal PCE Exposure
Among the nine children affected by neural tube defects, there were four exposed cases of anencephaly in three different families (crude prevalence 2.4/1,000) vs. no unexposed cases; one exposed case of spina bifida (crude prevalence 0.6/1,000) vs. three unexposed cases (crude prevalence 1/1,000); and one exposed case of Arnold-Chiari malformation (crude prevalence 0.6/1,000) vs. no unexposed cases.
Adjusted GEE odds ratios were fairly stable when we attempted to control for confounders one at a time among organ system subgroups with at least 20 cases. Adjusted odds ratios ranged from 0.8 to 1.0 for cardiac defects (crude GEE OR: 0.9), 1.6 to 2.0 for gastrointestinal defects (crude GEE OR: 1.8), 1.3 to 2.0 for genitourinary defects (crude GEE OR: 1.6), and 0.9 to 1.1 for musculoskeletal defects (crude GEE OR: 0.9). Confounders controlled in these analyses included those with apparent differences between the exposed and unexposed groups (Table ), including calendar year of birth; maternal and paternal age; maternal cigarette smoking, alcoholic beverage consumption, and prior pregnancy losses; and child's gender.
The small number of affected children limited our ability to examine the organ system and diagnostic group associations for the presence of a dose-response relationship; however, we found that odds ratios for all gastrointestinal defects combined and oral clefts were further increased among children whose average monthly prenatal exposure was greater than 1.136 grams (Table ). No dose-response relationship was observed for neural tube defects.
While we were able to validate only a small number of questionnaire reports against prenatal and obstetric records, we found excellent agreement between the information provided by the mothers and these records. For example, 92% of clinically recognized miscarriages, and 100% of live births noted in survey were reported in the medical record. There was also excellent agreement between the survey and medical record on gestational duration, birth weight, prenatal cigarette smoking, alcohol consumption, and multivitamin use. Furthermore, when we compared questionnaire and birth certificate data from all index children born in Massachusetts (n = 2,490), we found very good agreement on month and year of birth, mother's and father's age at the birth, birth weight, number of prior live births, and number of prior pregnancy terminations (including spontaneous and induced abortions).
In contrast, when we compared the mother's self-assessed exposure status to that derived from the EPANET assessment, we found that only 15% of mothers considered exposed by the EPANET assessment thought that their drinking water was contaminated, whereas 28% of these mothers thought that their water was not contaminated and 57% were unsure. Similarly, we found that 37% of mothers considered unexposed by the EPANET assessment thought that their drinking water was not contaminated while 9% thought that their drinking water was contaminated and 53% were unsure.