(709) Place, not race? Social vulnerability index and preterm birth risk

Social vulnerability index (SVI) is a CDC metric, ranging from 0 (least vulnerable) to 1 (most vulnerable), intended to identify areas that require urgent aid during natural disasters. SVI has been associated with a range of health outcomes, though only a few studies have focused on SVI and preterm birth (PTB). We aimed to investigate SVI, as an objective measure of structural racism or disadvantage, and PTB in a metropolitan area.

Study Design:

LIFECODES participants in Boston, MA (2009-2017) were geocoded to census tract level and linked to SVI. Pregnancy losses and multiple gestations were excluded. Statistical analyses included: t-test, ANOVA, chi square analysis, linear regression and logistic regression associated with SVI continuously and low, medium, and high SVI tertiles. Covariates in adjusted models were age, nulliparity, BMI, and chronic hypertension. Kaplan Meier survival analysis was performed for likelihood of remaining pregnant to 35 weeks. Models were evaluated using Akaike Information Criterion (AIC) to assess relative goodness of fit.

Results:

N=3273 were included in the analyses.  The cohort represented the the entire range of SVI values. We found no significant relationship with SVI and risk of preterm birth < 37 weeks (table 1). Risk of PTB < 34 weeks in the highest vs lowest SVI tertiles were 4.3% vs 2.0% (< .01). Risk of PTB < 28 weeks were 1.4% (high SVI) vs 0.5% (low SVI) (p< .01). In a model with SVI as a continuous predictor, adjusted odds ratio (OR) for PTB < 37, < 34, and < 28 weeks were 1.4 (0.9, 2.0), 2.5 (1.1, 5.5), 6.7 (1.3, 40.0). See image 2 for Kaplan Meier curve. SVI had a lower AIC than a models with race as a predictor, indicating SVI alone has a better fit for predicting PTB.

Conclusion:

Living in a higher SVI neighborhood is associated with increased OR for PTB < 34 and < 28 weeks among a diverse cohort in a Northeast metro area. In our cohort, SVI more accurately predicted PTB than race. Objective measures of structural disadvantage and/or racism may be a useful measure to identify groups at higher risk of early preterm birth compared to race alone.