UC Irvine

While wide-scale public health prevention efforts have significantly decreased national teen birth rates, disparities in teen births persist at local levels. The aims of this dissertation were to conduct a spatial analysis of maternal teen birth status in all births surrounding the New Bedford Harbor Superfund site (MA, USA) over two distinct time periods (1992-1998 and 2002-2008); to assess whether joint exposures of modeled prenatal chemical exposures elevate risk of subsequent teen birth in infant females born between 1992-1998 near NBH (MA Birth Record Cohort); and, to ascertain whether similar or different combinations of joint exposures affect subsequent teen birth in New Bedford Cohort (NBC) females. The spatial analyses in Chapter 2 demonstrated a statistically significant hot spot of elevated risk of teen birth west of the NBH only for the later time period (2002-2008). Chapter 3 employed predictive exposure models built from measured biomarkers in the NBC to estimate prenatal exposures for cord serum DDE, cord serum HCB, maternal hair Hg, cord blood Pb, and cord serum PCB4 for all births in four towns surrounding the NBH from 1992-1998. Epidemiological models, using an innovative extension of generalized additive models, for both MBRC (Chapter 3) and the NBC (Chapter 4) demonstrated higher risk for subsequent teen birth across low levels of DDE and higher levels of the remaining chemical exposures. In Chapter 4, the apparent protective effect of DDE persisted even after adjustment for maternal dietary factors. Although this research has its limitations, it provides a novel approach to analyzing mixtures of chemical and non-chemical exposures and makes a significant contribution to the literature on the effects of joint chemical exposures on maternal, child, and adolescent health.