Examining seasonal patterns of birth defects may help to identify environmental risk factors. Because the teratogenic window for most birth defects is during gestational weeks 3 to 8, investigating exposures closer to the timing of conception is important. However, studies are usually based on month of birth, which is not the biologically relevant exposure period and does not account for differences in gestational length. We aimed to determine whether the occurrence of birth defects varied by month of conception using the population-based New York State Congenital Malformations Registry (CMR). We merged live birth certificates (n = 2,044,091) with CMR records for mothers residing in New York State, excluding New York City, for the years 1992 through 2006. We categorized birth defects according to the National Birth Defects Prevention Network guidelines and performed Cochran-Armitage trend, Hewitt-Rogerson, and Walter-Elwood tests on month of conception and chi-square tests on season of conception. We graphed seasonal distributions and seasonality test results. We performed stratified analyses by maternal and infant characteristics. Of 42 groups examined in the 15-year period, 24 (57%) had at least one statistically significant test result, suggesting a trend or seasonal variation: Cochran-Armitage (18), Hewitt-Rogerson (17), Walter-Elwood (4), and chi-square (5). Ventricular septal defect showed the most consistent results: Cochran-Armitage (p = 0.0006), Hewitt-Rogerson (December to May; p = 0.0130), Walter-Elwood (March 14; p = 0.0027), and chi-square (winter; p = 0.0046). Congenital cataract, pulmonary valve atresia/stenosis, coarctation of aorta, biliary atresia, and renal agenesis or hypoplasia had at least three significant tests. These results may help to generate hypotheses about environmental factors that vary by season for further studies.
Read full abstract