Testing the Weathering Hypothesis of disparities in birth outcomes

Persistent disparities exist in birth outcomes between African-Americans and European-Americans in the US, with low birth weight and prematurity at least two times higher for African-Americans. The Weathering Hypothesis generates great interest as a model for understanding these disparities, but is difficult to test.

South Carolina possesses one of the best data systems for linking socioeconomic data bases with health and vital records, allowing a test of the Weathering Hypothesis by building a model on how preexisting health conditions and socioeconomic status affect birth outcomes across age and race/ethnic groups. We linked birth records for three years (2004-2006; N=168,572) with billing data from Medicaid and one large health plan, and hospital discharge data for three years prior to delivery. Basic demographics are determined from birth records. Outcomes (prematurity and low birth weight), are determined from birth and hospital discharge records. To approximate socioeconomic factors we overlaid a geocode of census tract information then developed statistical models to examine the independent effects of mother's health conditions as well as potential joint effects of mother's health and SES conditions on birth outcomes. We describe data linking and model building; by November we will have full analysis results.

Effective strategies to reduce racial disparities in birth outcomes need to include actions that improve women's health throughout the life-course. This study will determine which factors are most highly related to poor birth outcomes across racial/ethnic groups, and may shed light on the direction intervention strategies should take to reduce these disparities.