Possible Sarcopenia and Subsequent Cognitive Function in Older Chinese Adults: Limited Evidence for Mediation by...

To examine the longitudinal association between possible sarcopenia and subsequent cognitive function in older Chinese adults and whether a 27-item frailty index accounted for this association.

We analyzed China Health and Retirement Longitudinal Study data using a primary three-stage sequence: possible sarcopenia, physical activity and covariates in 2011; frailty in 2013; and cognition in 2015. Multiple imputation (m = 20), bootstrap estimation with Rubin pooling and independent R 4.5.0 replication were used. Sensitivity analyses included a 16-item non-disability score, exploratory physical activity effect modification, inverse probability of censoring weighting, five additional temporal sequences and reverse-path models.

The primary sequence included 2036 participants. Possible sarcopenia had a negative but imprecise adjusted total association with later cognition (TE = -0.112, 95% CI -0.242 to 0.019; p = 0.094). The direct association was similar (NDE = -0.107, 95% CI -0.238 to 0.024), while the indirect association through frailty was small and included zero (NIE = -0.0047, 95% CI -0.0141 to 0.0046). The non-disability score and all additional sequences showed no precise indirect association. The physical-activity interaction was unsupported (p = 0.192). IPCW strengthened the total, but not the indirect, association.

The primary sequence did not provide precise evidence of an association between possible sarcopenia and later cognition, and the analysis did not provide statistically precise evidence that frailty mediated the association. Findings were time-window-dependent and compatible with bidirectional or shared aging processes rather than a validated early marker or a single frailty-mediated pathway.