Food insecurity increases risk for peripartum cardiomyopathy and early presentation
K Sompel, E Shalowitz, E Kudron, S Son, D KaoAbstract
Background
Peripartum cardiomyopathy (PPCM) is a leading cause of global maternal mortality with worse outcomes associated with African Ancestry and delayed presentation. Recent studies suggest socioeconomic factors also contribute to disparities in PPCM, though the role of food insecurity is unknown.
Purpose
Determine the relationship between food insecurity and timing of PPCM presentation.
Methods
Hospital discharge data were obtained from the states of Arkansas, Arizona, Colorado, Florida, Louisiana, Maryland, New York, South Dakota, Washington, and Wisconsin between 2003-2023. Patient characteristics were determined using ICD-9 and -10 codes. Antepartum presentation was defined as hospitalization with PPCM ≤ 30 days prior to delivery, peripartum presentation as PPCM reported during at time of delivery, and postpartum presentation was defined as rehospitalization with a new diagnosis of PPCM within 150 days following delivery. Social Vulnerability Index (SVI) was determined using Centers for Disease Control postal code data. Using the US Department of Agriculture Food Access Research Atlas, ‘food deserts’ were defined as census tracts with ≥ 20% poverty rate and ≥ 33% with living > 1.5 km from a grocery store in cities, 15 km outside cities, or alternatively as living 0.8 km from a grocery store with no vehicle or > 30 km regardless of vehicle availability. Rate of enrollment in the federal Supplemental Nutrition Assistance Program (SNAP) was also included. Associations between maternal features, and PPCM presentation were determined using multivariable logistic regression.
Results
In total 7,156,032 delivering patients were identified of which 5,638 patients were diagnosed with PPCM. 111 (2.0%) presented antepartum, 1,709 presented peripartum (30.3%) and 3,818 (67.7%) presented postpartum. Total SVI and indices of each individual SVI pillar were higher in PPCM patients indicating a greater degree of vulnerability compared to non-PPCM counterparts (p<0.001 for all). Mothers who developed PPCM were also more likely to live in food deserts (17.9% vs. 14.2%), even when controlled for vehicle access this disparity persisted (p<0.001 for both). Mothers who developed PPCM also had higher rates of enrollment in SNAP, with a stronger association observed in antepartum presentation (p<0.001). Multivariable analysis shown in Figure 1 found that higher overall SVI and residency in a food desert was associated with an increased risk for peripartum and postpartum PPCM. However, higher SVI was most associated with postpartum presentation, while a poverty rate > 30% and residency in a food desert was most associated with peripartum presentation. Controlled for SVI, SNAP enrollment was not associated with increased risk for PPCM.
Conclusion
Food insecurity was associated with increased likelihood of PPCM, particularly for ante- and peripartum presentation when controlled for clinical features and social vulnerability.