Association Between Housing Affordability and Severe Maternal Morbidity

Abstract

Importance: The number of people living in unaffordable housing (relative to income) is projected to continue increasing as housing cost inflation outpaces incomes in the US. Although reproductive-aged women have disproportionately high housing costs, particularly around the time of childbirth, data on associations between housing costs and maternal health and the role of publicly supported affordable housing programs in mitigating those associations are lacking.

Objective: To estimate associations between area-level rental housing costs and severe maternal morbidity (SMM) and assess the potential mitigating role of publicly supported affordable housing.

Design, setting, and participants: This cross-sectional study linked New Jersey birth files from January 1, 2008, to December 31, 2018, to maternal hospital discharge records and municipal-level housing and demographic data from the state of New Jersey and the US Census Bureau. Data were analyzed from January to September 2022. The birth files contained records for all births in New Jersey, and the hospital discharge records contained information from all inpatient hospitalizations over the study period. A total of 1 004 000 birth records were matched to maternal discharge records and municipal-level data.

Exposures: Municipal-level rental costs relative to income (housing cost burden), availability of publicly supported affordable housing, and housing subsidy per person with an income lower than the federal poverty level.

Main outcomes and measures: Severe maternal morbidity was identified using diagnosis and procedure codes developed by the US Centers for Disease Control and Prevention to measure SMM.

Results: Of 1 004 000 mothers (mean [SD] age at birth, 29.8 [5.9] years; 44.7% White), 20 022 (2.0%) experienced SMM. Higher municipal rental housing costs were associated with greater odds of SMM (odds ratio [OR], 1.27; 95% CI, 1.01-1.60), particularly among mothers with less than a high school education (OR, 1.81; 95% CI, 1.06-3.10), and the positive associations decreased at higher levels of affordable housing availability. Among mothers with less than a high school education, the risk of SMM was 8.0% lower (risk ratio, 0.92; 95% CI, 0.85-1.00) for each additional $1000 annual municipal-level housing subsidy per person with an income lower than poverty level after controlling for rental costs and other characteristics, which translated to a 20.7% lower educational disparity in SMM.

Conclusions and relevance: In this cross-sectional study, living in a municipality with higher rental housing costs was associated with higher odds of SMM, except when there was high availability of publicly supported affordable housing. These results suggest that greater availability of publicly supported affordable housing has the potential to mitigate the association between rental housing costs and SMM and reduce socioeconomic disparities in SMM.

Figure 1.. Percentage of Renter Households With Housing Cost Burden in New Jersey Municipalities

Housing cost burden was defined as spending more than 30% of household income on rent. Data were obtained from American Community Survey 5-year estimates. Municipalities were self-governing administrative divisions (ie, cities, towns, townships, boroughs, or villages) incorporated under state law.

Figure 2.. Marginal Effects Associated With Housing Cost Burden in Municipalities by Availability of Publicly Supported Affordable Housing Among Individuals With Less Than High School Education

Estimates were derived from multilevel logistic regression models of the association between the percentage of renter households with housing cost burden, the availability of publicly supported affordable housing, and educational attainment with severe maternal morbidity (including analysis of 2-way and 3-way interactions between the 3 exposures). The analysis controlled for individual-level factors (age, race and ethnicity, parity, and year of birth) and municipal-level factors (median rent, percentage of residents with income lower than the poverty level, and population size). Whiskers indicate 95% CIs.