Abstract
INTRODUCTION: The association between muscle strength and cognitive function is well-established, yet its influence on working memory and related brain activation remains unclear. This study aimed to investigate the unique contributions of grip strength and 30-s sit-up to working memory performance and prefrontal cortex activation. METHODS: A total of 192 older adults were recruited from nursing homes and communities. Working memory performance (reaction, time, and accuracy) was assessed using an n-back task, while hemodynamic responses were monitored in six prefrontal subregions using a 24-channel fNIRS system. Muscle strength metrics included normalized grip strength and normalized 30-s sit-up. For each dependent variable, four hierarchical regression models were constructed. RESULTS: Normalized grip strength showed robust independent associations with high-load working memory performance (2-back reaction time: β = -0.22, p < 0.001; accuracy: β = 0.19, p = 0.002) and with activation in the dorsolateral prefrontal cortex (L-DLPFC during 2-back: β = 0.24, p < 0.001). When both strength metrics were included simultaneously (Model 3), normalized grip strength remained significantly associated (all p < 0.01), while the association of normalized 30-s sit-up became non-significant (all p > 0.20). The associations exhibited a clear prefrontal functional hierarchy (DLPFC > VLPFC > FPA) and cognitive load modulation pattern (0-back < 1-back < 2-back). CONCLUSION: The findings indicate that normalized grip strength, more so than normalized 30-s sit-up, is associated with the level of prefrontal hemodynamic activation during working memory tasks in older adults. This association is finely modulated by the functional hierarchy of the prefrontal cortex and task complexity. This highlights that standardized grip strength may be a key biomarker linked to the physical-mental health nexus, informing the development of resistance training interventions aimed at preserving cognitive health.