Safety net program participation and churning during the COVID-19 pandemic: a longitudinal analysis of low-income California families.

Objective: Studies of participation in US safety net programs consistently find improvements in health equity, yet few have measured multi-program participation over time, or linked these trajectories to health during the COVID-19 pandemic. This study characterizes longitudinal participation phenotypes in SNAP, WIC, and Medicaid and examines their associations with mental and physical health. Methods: We used longitudinal data from a cohort of low-income families (n = 361, 2019-2023) and applied sequence analysis (Hamming distance) and cluster analysis to identify distinct trajectories of program use. We then characterized sociodemographic profiles across clusters and applied regression analyses to estimate associations between the safety net use phenotypes and self-reported health. Results: We identified six unique trajectories of participation. While program use increased initially (2019-2021), it was followed by declines and "churning" (cyclical loss/gain) in 2022-2023. The most prevalent clusters were "all programs" (25%) and "rapid churning" (22%). Consistent with the hypothesis that administrative burdens harm health, participants with stable, full participation ("all programs") had consistently lower prevalence of adverse health outcomes compared to other clusters. Vulnerable groups (lower-income, racial/ethnic minorities) were disproportionately represented in high-churn clusters. Conclusion: Administrative volatility and "churning" were common, and results suggest that stability in coverage is a key driver of health equity. Policymakers should prioritize specific administrative safeguards, such as cross-program coordination, data sharing, and recruitment strategies to support vulnerable groups.