Abstract
<title>Abstract</title> <p>Background Cognition and independence in basic activities of daily living (ADL) often change together, yet their joint social patterning is poorly characterised. We examined educational differences in cognitive-functional states and adjacent-wave transitions among older adults in China and the United States. Methods We analysed adults aged 65 years or older in the China Health and Retirement Longitudinal Study (CHARLS; 2011–2020) and the US Health and Retirement Study (HRS; 2006–2020). Cognitive scores were calibrated against the baseline age- and sex-specific distribution, and the same calibration was applied at later waves. Low cognitive performance was defined as a residual z score <=-1.0; ADL disability was at least one basic ADL difficulty. Four states captured neither condition, either alone, and dual impairment. Discrete-time clustered logistic models estimated education-associated adjacent-wave transitions, adjusting for age at the origin wave, sex, origin state, origin cognition, and wave-pair effects. Sequential adjustment, inverse probability-of-selection weighting, complete-case analysis, continuous cognition, and measurement sensitivities assessed robustness. CHARLS used baseline weights; HRS estimates were unweighted. Results The cohorts included 4,785 CHARLS and 11,355 HRS participants. Observed transitions were available for 1,701 CHARLS (35.5%) and 8,767 HRS (77.2%) participants; contributors were younger and socially advantaged. In primary models, low versus high education was associated with transition to low cognitive performance in CHARLS (odds ratio (OR) 5.16, 95% confidence interval (CI) 2.58–10.32) and HRS (OR 2.30, 95% CI 1.96–2.71). Associations persisted after full adjustment (OR 2.91, 95% CI 1.38–6.14; OR 2.03, 95% CI 1.69–2.44) and in selection-weighted analyses. Low education was also associated with less favourable continuous cognitive change in both cohorts. Associations with ADL and dual-impairment transitions were smaller and less consistent after full adjustment. Conclusions Lower education consistently marked vulnerability to transition into low cognitive performance in both cohorts. Functional transition estimates were more sensitive to adjustment and selection. Education-sensitive geriatric assessment may identify preventive and supportive needs, although these associations are neither causal effects nor nationally representative HRS estimates.</p>