Race/Ethnicity as a Moderator of the Effects of Maternal Support and Health Behaviors On Low Birth Weight

Purpose: This study tested a structural equation model using maternal health behaviors and two latent maternal support factors to predict low birth weight across different ethnic groups. Infants born at low birth weight are at greater risk for experiencing negative short and long-term health, cognitive, and developmental outcomes as compared with normal weight infants. Further, racial/ethnic disparities in low birth weight are substantial, with African-American mothers having twice the low birth weight rates of non-Hispanic white and Hispanic mothers. While many studies examine race/ethnicity as a predictor of birth outcomes, few have examined whether race/ethnicity moderates the relationship between various risk factors and birth outcomes. Risky behaviors during pregnancy, such as smoking, and alcohol and drug use, can impact birth weight, while family and social support during pregnancy have the potential to limit or discourage these types of behaviors. Results from prior research are not consistent regarding the effects of different types of support on reducing negative health behaviors to impact birth weight. Support from fathers has been shown to be predictive of improved birth outcomes, though support from family is less so. Support from the community or other environmental contexts could also have important positive benefits, though these have been less studied. Further, all these types of support may have varying effects across race/ethnicity.

Methods: Mothers' baseline data from The Fragile Family and Child Wellbeing Study was utilized for this project. These data consist of nearly 4900 live births in 75 hospitals in 20 large U.S. cities, and are representative of all births in U.S. cities with populations >200,000. Baseline interviews were done with mothers within one day of the child's birth and have rich measures of family and individual sociodemographic and psychosocial characteristics. Two latent support variables, environmental support and paternal support, were constructed for this study. Another latent variable was constructed to measure negative maternal health behaviors. Structural Equation Models (SEM) and path analyses were run for each race/ethnic group to model the relationships among the variables and to compare these relationships across groups.

Results: Results indicated that the relationships between support, maternal health behaviors and low birth weight differ between Hispanic (N=1303) and African- American mothers (N=2326). Negative health behaviors significantly impact low birth weight as hypothesized and also were affected by both environmental and paternal support constructs. The findings suggest that both types of support impact low birth weight through their effect on mothers' health behaviors, and that these effects differ by race/ethnicity. Further, support from fathers strongly impacted maternal health behaviors both directly and indirectly through its effect on environmental supports. The effect of father support was strongest among African-American mothers.

Conclusions: As hypothesized, environmental support, support from fathers, and risky behaviors impact the birth weight of urban children. Maternal race/ethnicity may act as a moderator of these latent variables' effects on low birth weight. Additional research is needed to test these findings further and to understand how race/ethnicity moderates the relationships between different types of support, negative health behaviors, and poor birth outcomes.