Trajectories of muscle strength and physical performance preceding dementia in older US and European populations.

The association between muscle function and dementia risk remains elusive, as studies suggest that impaired muscle function may act as both a risk factor for and a consequence of dementia, hindering causal inference. We aimed to clarify the temporal relationship between muscle function and incident dementia by investigating non-linear trajectories of muscle function in the years preceding dementia onset in older US and European populations. Case-control study. Data were combined from the English Longitudinal Study of Ageing (ELSA, 2004-2018, waves 2-9), Health and Retirement Study (HRS, 2004-2018, waves 7-14), and Survey of Health, Ageing and Retirement in Europe (SHARE, 2004-2017, waves 1-7). For handgrip strength analysis, 18,335 participants aged 60 and older were included from the ELSA, HRS, and SHARE cohorts. For gait speed analysis, 11,690 participants aged 60 and older were included from the ELSA and HRS cohorts. Muscle strength was assessed by handgrip strength using a Smedley dynamometer, and physical performance was evaluated by gait speed using the Timed 8-Foot Walk test, with assessments conducted biennially or quadrennially. Dementia was diagnosed using self-reported physician diagnosis and cognitive-functional assessments. Trajectories of muscle strength and physical performance were analyzed on a backward timescale using latent-process mixed models within a nested case-control design. Significant differences in muscle function trajectories were observed between cases and controls 12 and 13 years prior to dementia onset (handgrip strength: coefficient [SE], -0.23 [0.05], P < 0.001; gait speed: coefficient [SE], -0.24 [0.08], P = 0.003). The pathological trajectories of handgrip strength and gait speed revealed periods of acceleration beginning 6 and 8 years prior to diagnosis, respectively. After adjusting for pre-dementia acceleration, greater handgrip (per 1-kg increment) was associated with a modest reduction in dementia risk (hazard ratio, 0.98; 95 % CI, 0.97-0.99), while faster gait speed (per 1-m/s increment) markedly lowered risk (hazard ratio, 0.35; 95 % CI, 0.23-0.53). These findings highlight muscle function as a cost-effective tool for early detection and dynamic monitoring of dementia risk and identify it as a modifiable target for prevention. Muscle function may also assist in identifying high-risk groups for preferential enrollment into clinical trials for dementia prevention and treatment.