Neighborhood-level intergenerational mobility, poverty, and early developmental health in a population birth cohort

• In Event: Poster Session 14
  In Poster Session: PS 14 Topic - Race, Ethnicity, Culture, Context

Abstract

Introduction: Neighborhoods with high levels of intergenerational mobility are hypothesized to be beneficial for child development, independent of neighborhood poverty. There is evidence that neighborhood intergenerational mobility is associated with child cognition and behavior, as well as adult health. It is unclear, however, if such differences originate earlier in development, and thus if neighborhood mobility is uniquely associated with differences in early health. To address this question, we used a population-based cohort to examine whether neighborhood intergenerational mobility, in, is independently associated with low birthweight (LBW), an important early indicator of developmental health.

Methods: Data come from a de-identified dataset of vital birth records for all singleton children born in California in 2017 (n = 426,873). Low birthweight (LBW) was dichotomized to capture clinical risk (< 2.5 kg; 5.3% of births). Maternal residential address was geocoded at the census-tract level. Neighborhood intergenerational mobility was defined as the national rank of mean adult household income, based on tax data, for those who grew up in low-income families in each census tract (from the Opportunity Atlas, Chetty et al, 2018; https://www.opportunityatlas.org). Neighborhood poverty was obtained from the American Communities Survey (2013 – 2017 estimate). We also examined the robustness of results using a concentrated disadvantage factor (Sampson et al, 2008). Figure 1 illustrates the statewide distribution of key variables. We used multilevel logistic modeling to estimate the association between neighborhood predictors (z-score standardized) and LBW, adjusting for individual-level factors (see Figure 2).

Results: Increased neighborhood social mobility was associated with a decreased likelihood of LBW (OR = 0.96, p < .001) after controlling for individual-level covariates. However, after adjusting for neighborhood poverty this effect was no longer significant (OR = 0.99, p = .33). Meanwhile neighborhood poverty was associated with LBW independent of mobility and individual-level covariates (OR = 1.04, p < .001; see Figure 2). The odds of a baby being born LBW was 9.2% higher when children were born in a census tract where neighborhood poverty was at the 90th percentile compared with a tract where neighborhood poverty was at the 10th percentile. There was no evidence that the effects of neighborhood mobility varied by neighborhoods with different poverty levels (interaction OR = 1.00, p = .63) or by race/ethnicity, though the effect of neighborhood poverty was smaller for Latina mothers (interaction OR = 0.91, p < .001). The pattern of results was similar for concentrated disadvantage, and thus robust to measurement approach.

Conclusions: Although children born in neighborhoods defined by greater intergenerational mobility are more likely to be born at a normal birthweight, neighborhood poverty accounts for these effects. Neighborhood poverty and concentrated disadvantage are robust predictors of increased LBW births, which set the stage for increased risk for a host of negative cognitive and health outcomes later in life. This suggests that neighborhood poverty during the prenatal is an important target for prevention efforts focused on early developmental health, while neighborhood intergenerational mobility may be important for development at later stages.

Authors

• Daniel A Hackman, University of Southern California
  Presenting Author

• Himal Suthar, University of Southern California
  Non-Presenting Author

• Abigail Palmer Molina, Erikson Institute
  Non-Presenting Author

• William C. Dawson, University of California, Berkeley
  Non-Presenting Author

• Emily Putnam-Hornstein, University of North Carolina at Chapel Hill
  Non-Presenting Author