categories of congenital anomalies (central nervous system, eye, ear, chromosomal, and oral cleft), and two of nine categories of childhood disorders (infections of the kidney and urinary tract and disorders of the respiratory tract). There was no positive association with spontaneous abortions or low birth weight. The wells, which were closed in 1979, were contaminated with trichloroethylene, tetrachloroethylene, and chloroform. The health outcomes evaluated were identified by a telephone sample survey conducted in 1982 of 4936 pregnancies and 5018 residents of Woburn aged 18 or younger. Several of these outcomes were grouped according to “assertions in the literature of potential links with chemicals, pesticides or trace elements,” a grouping that was subsequently criticized (MacMahon, 1986). MacMahon (1986) also commented on the lack of correction for the multiple comparisons made as a result of the health survey and the fact that six of the interviewers were from families of the leukemia patients.
Several studies have reported low birth weight as a consequence of exposure to toxins at Love Canal (Vianna and Polan, 1984; Goldman et al., 1985; Paigen and Goldman, 1987). The timing and extent of the effect differed in these studies. Vianna and Polan (1984) evaluated the cohort of infants born in the Love Canal area from 1940 through 1978. As a surrogate for exposure, they considered whether or not a family resided in an area known to have more water that percolated up to the ground through swales. Such homes arguably could be expected to incur greater exposure to contaminants in these swales and to outgassing emissions from them that could enter basements, especially during times of sewer overflows. Using a five-year moving average, they found a significant excess of low-birth-weight babies occurred during the time when dumping was estimated to be most active in the area, namely 1940 through 1953. No such excess was evident for later years.
Goldman et al. (1985) studied a larger population and a somewhat different cohort of births. Using single family homes in the entire Love Canal neighborhood—an area three times as large as that included in Vianna and Polan—they found an excess of low birth weight between the years 1963 and 1980. The adjusted odds ratio was 3.0 (range, 1.3 to 7.0) for 131 exposed and 357 controls, with a prevalence of low-birth-weight babies of 16 percent for the swale area, 10 percent for the nonswale area, and 4.8 percent for the controls. Paigen and Goldman (1987) defined exposure in two ways: distance of the home from the canal and proximity of the homes to possible paths of chemical migration, inferred from hydrogeological information about the migration of groundwater pollutants from the abandoned site into homes and basements, via swales, or large subterranean, wet