32 - County-level density of affordable housing units and preterm birth risks

Publication Number: 4031.32

Daria Murosko, The Children's Hospital of Philadelphia, Philadelphia, PA, United States; Matthew Coyle, Childrens Hospital of Philadelphia, Springfield, PA, United States; Anzu Isa, Childrens Hospital of Philadelphia, Montebello, CA, United States; Scott A. Lorch, The Children's Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA, United States

Background: Housing instability and unaffordability are risk factors for preterm birth ( < 37 weeks' gestation). Policies that promote housing stability, such as eviction moratoria and legal aid, are linked to lower preterm birth rates. The Low-Income Housing Tax Credit (LIHTC) is the largest federal program that creates affordable housing units, yet its relationship with birth outcomes is unknown.

Objective: To determine if county-level LIHTC density is associated with individual risk of preterm birth.

Design/Methods: This longitudinal, retrospective cross-sectional study analyzed the National Center for Health Statistics restricted natality files for births in the United States between 2016 and 2023. We determined county-level density of LIHTC units (per 10,000 people) from the National Housing Preservation Database. Density was categorized as 0 (no units in the county), or below or above the national median value. We used modified Poisson regression with robust standard errors to estimate risk ratios for individual preterm birth associated with county-level LIHTC density, accounting for clustering at the county level. Models included state and year fixed effects and adjusted for county- and individual-level covariates. Models were lagged by one year to account for LIHTC availability during pregnancy.

Results: Across 29,654,268 births, median county LIHTC density was 47.7 units per 10,000 residents. There was wide variation in average county-level density across states, ranging from 37.0 to 106.9 units per 10,000 residents (Figure 1). 3,763 of 25,114 county-years had 0 LIHTC units, accounting for 1.2% of births (Table 1). Counties without LIHTC were more rural, had lower housing costs and poverty, and less obstetric access. Maternal characteristics also varied between high- and low-density areas. LIHTC availability was associated with decreased risk of preterm birth in full models (RR 0.95 [95% CI: 0.93-0.98] for low-density, and RR 0.95 [95% CI: 0.92-0.98] for high-density), compared to no LIHTC availability. Coefficients for low vs. high density categories were similar.

Conclusion(s): Compared to counties with no LIHTC units, living in a county with LIHTC units was associated with lower risk of preterm birth, when adjusting for county- and individual-level risk factors. There was not a dose-response with increasing LIHTC density. LIHTC distributions differed across states. Future work should investigate how state-specific LIHTC allocation policies influence where developments are built, acknowledging that drivers of perinatal health and housing needs vary across counties.

Table 1. County- and Individual-Level Characteristics by County-Level LIHTC Unit Density for U.S. Births, 2016-2023 (N = 29,654,268 births in 25,114 county-years)


Average county-year LIHTC unit density (per 10,000 people) and average annual pre-term birth rate, 2016 - 2023.
The left panel shows average annual preterm birth rate by state. The right panel shows average county-year LIHTC density (per 10,000 individuals) by state. Data for Alaska, Hawaii, and Washington, DC are not shown.

Table 2: County-Level LIHTC Unit Density and Risk of Preterm Birth