Prospective associations between diet quality, frailty and incident cognitive impairment in the health and retirement study

Background Diet quality is a modifiable determinant of healthy aging, potentially influencing physical and cognitive functioning through inflammatory and metabolic pathways. Although healthier dietary patterns have been linked to reduced frailty and cognitive decline, findings remain inconsistent. The interrelationships among diet quality, physical frailty, and incident cognitive impairment, including potential modification by specific dietary components, remain poorly understood. Methods We analyzed 1,636 cognitively normal participants in 2016 from the Health and Retirement Study, a longitudinal study of adults in the United States (U.S.), with diet assessed in 2013, frailty in 2016, and cognitive follow-up through 2022. Cognitive impairment was assessed with cognitive tests and used the Langa–Weir classification; diet quality was measured using the Healthy Eating Index–2015 (HEI-2015); and physical frailty used a modified Fried phenotype. Survey-weighted regression, Cox models, four-way decomposition, competing-risk and multistate models, and additive Bayesian network (ABN)–informed generalized structural equation models (GSEM) were applied. Results Higher HEI-2015 scores were associated with lower frailty (β = −0.13, P < 0.001), but not with incident cognitive impairment. Frailty was associated with higher cognitive risk in minimally adjusted models (Hazard Ratio (HR) = 2.27, 95% confidence interval (CI): 1.56–3.32), with attenuation after full adjustment. Exploratory analyses identified 13 nominally significant food-item associations with composite cognitive impairment, including higher risks associated with selected dairy products, processed meats, and beverages, and an inverse association with red wine (HR = 0.77, 95% CI: 0.66–0.89). None survived FDR correction. Five dietary associations with probable dementia alone survived FDR correction but remained exploratory. Four-way decomposition provided limited evidence of dietary mediation or interaction, while sensitivity analyses supported attenuation of frailty–cognition associations. Conclusions Higher diet quality was associated with lower physical frailty but not independently with incident cognitive impairment. Food-specific associations were heterogeneous and largely did not survive multiple-testing correction. Findings warrant replication before drawing dietary recommendations for cognitive health.