Abstract
<title>Abstract</title> <p>Background Educational attainment is a key protective factor against late-life cognitive decline, yet whether its cognitive returns are heightened in resource-deprived rural environments remains controversial. We investigated whether rural-urban residency moderates the association between education and cognitive function among older Chinese adults. Methods Using data from the China Health and Retirement Longitudinal Study (CHARLS; N = 15,646), cognitive function was operationalized via episodic memory (word recall, 0–20). Hierarchical ordinary least squares (OLS) and linear mixed-effects models estimated cross-sectional and longitudinal Education × Rural interactions. Propensity score matching (PSM), E-value sensitivity analysis, and an expanded nine-mediator framework evaluated causal robustness, dimension specificity, and underlying mechanisms. Results The Education × Rural interaction was positive and statistically significant (β = 0.108, p < 0.001), demonstrating larger cognitive returns of education for rural residents. This advantage remained consistent across gender and age subgroups and expanded longitudinally over time (β = 0.054, p < 0.001). PSM confirmed directional stability (β = 0.055, p = 0.037) despite a 48% attenuation. E-value analysis indicated robust main effects (E = 2.21) alongside a more sensitive interaction contrast (E = 1.20). Dimension-specificity tests revealed independent spatial and economic resource substitution, with ~ 95% of the interaction operating via direct cognitive reserve building rather than measurable behavioral or social pathways. Conclusion Educational attainment yields stronger cognitive protective returns in rural contexts, lending empirical support to the resource substitution hypothesis. However, the sensitivity of the interaction contrast to unmeasured confounding warrants careful interpretation when setting policy expectations.</p>