Food Insufficiency, Supplemental Nutrition Assistance Program (SNAP) Status, and 9-Year Trajectory of Cognitive Function in Older Adults: The Longitudinal National Health and Aging Trends Study, 2012–2020
Muzi Na, Nan Dou, Monique J. Brown, Lenis P. Chen-Edinboro, Loretta Anderson, Alexandra Wennberg
Abstract
BACKGROUND: Despite findings from cross-sectional studies, how food insecurity experience/Supplemental Nutrition Assistance Program (SNAP) status relates to cognitive decline over time has not been fully understood. OBJECTIVES: We aimed to investigate the longitudinal associations between food insecurity/SNAP status and cognitive function in older adults (≥65 y). METHODS: Longitudinal data from the National Health and Aging Trends Study 2012-2020 were analyzed (n = 4578, median follow-up years = 5 y). Participants reported food insecurity experience (5-item) and were classified as food sufficient (FS, no affirmative answer) and food insufficient (FI, any affirmative answer). The SNAP status was defined as SNAP participants, SNAP eligible nonparticipants (≤200% Federal Poverty Line, FPL), and SNAP ineligible nonparticipants (>200% FPL). Cognitive function was measured via validated tests in 3 domains, and the standardized domain-specific and combined cognitive function z-scores were calculated. Mixed-effect models with a random intercept were used to study how FI or SNAP status was associated with combined and domain-specific cognitive z-scores over time, adjusting for static and time-varying covariates. RESULTS: At baseline, 96.3% of the participants were FS and 3.7% were FI. In a subsample (n = 2832), 10.8% were SNAP participants, 30.7% were SNAP eligible nonparticipants, and 58.6% were SNAP ineligible nonparticipants. Compared with the FS group in the adjusted model (FI vs. FS), FI was associated with faster decline in the combined cognitive function scores [-0.043 (-0.055, -0.032) vs. -0.033 (-0.035, -0.031) z-scores per year, P-interaction = 0.064]. Cognitive decline rates (z-scores per year) in the combined score were similar in SNAP participants (β = -0.030; 95% CI: -0.038, -0.022) and SNAP ineligible nonparticipants (β = -0.028; 95% CI: -0.032, -0.024), both of which were slower than the rate in SNAP eligible nonparticipants (β = -0.043; 95% CI: -0.048, -0.038; P-interaction < 0.0001). CONCLUSIONS: Food sufficiency and SNAP participation may be protective factors preventing accelerated cognitive decline in older adults.