We generated descriptive statistics and bivariate associations for all variables and examined them for distributional normality. We used multiple linear regression and logistic regression to assess the effects of proximity to the WTC site and stage of pregnancy when the event occurred, on gestational duration, birth weight, birth length, head circumference, ponderal index, and sex-specific small size for gestational age (SGA) among term deliveries (Alexander et al. 1996 (link)). We excluded 10 women who were determined not to have been pregnant on 11 September 2001, based on the gestational age of the baby at birth and the date of delivery. In addition, only women who had completed 36 weeks and 6 days of gestation, or 258 days, and considered full term were included in the analyses, because preterm delivery may have other, complex etiologies. These criteria excluded a total of 29 women, for a final sample of 300 women.
We included race/ethnicity (Asian, African American vs. all others) and Medicaid status (marker for poverty) as covariates in all analyses because of their potentially confounding effects on the relationship between exposure and birth outcomes. Other relevant covariates included infant sex, maternal height, prepregnancy weight, parity (0, ≥1), maternal age in years, cesarean section, and maternal medical complications (including preeclampsia, placental abruption, hypertension, and diagnosis of diabetes during the pregnancy). Smoking was not included because enrollment criteria required that women be nonsmokers during their entire pregnancy. In this cohort, sexually transmitted disease and alcohol use (self-report by interview yielded very low use) were not significantly associated with exposure or birth outcomes (p > 0.05), and their inclusion did not alter effect sizes. Therefore, we excluded them to limit the number of independent variables. All analyses were done using SPSS version 11.5 (SPSS Inc., Chicago, IL).
We included race/ethnicity (Asian, African American vs. all others) and Medicaid status (marker for poverty) as covariates in all analyses because of their potentially confounding effects on the relationship between exposure and birth outcomes. Other relevant covariates included infant sex, maternal height, prepregnancy weight, parity (0, ≥1), maternal age in years, cesarean section, and maternal medical complications (including preeclampsia, placental abruption, hypertension, and diagnosis of diabetes during the pregnancy). Smoking was not included because enrollment criteria required that women be nonsmokers during their entire pregnancy. In this cohort, sexually transmitted disease and alcohol use (self-report by interview yielded very low use) were not significantly associated with exposure or birth outcomes (p > 0.05), and their inclusion did not alter effect sizes. Therefore, we excluded them to limit the number of independent variables. All analyses were done using SPSS version 11.5 (SPSS Inc., Chicago, IL).
Free full text: Click here
