Author(s): HE, Z., XIAO, J.W., Institution: PEKING UNIVERSITY, Country: CHINA, Abstract-ID: 1972

INTRODUCTION:
Identifying modifiable factors to mitigate cognitive decline is a public health priority. Although physical activity (PA) protects against cognitive deterioration, underlying mechanisms remain insufficiently understood, with prior research emphasizing physiological pathways. Psychological frailty (PF)—diminished psychological resilience and impaired emotional regulation—has emerged as a risk factor for adverse health outcomes in older populations. This study investigates the association between PA and cognitive decline among middle-aged and older Chinese adults and examines whether PF mediates this relationship.
METHODS:
This cross-sectional study used data from the 2018 China Health and Retirement Longitudinal Study (CHARLS), including 3,657 participants aged ≥45 years. PA was quantified using metabolic equivalents (METs). Cognitive decline was assessed with the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE). PF was measured using a validated Psychological Frailty Index (PFI) integrating emotional status, psychological distress, and functional limitations. Multivariate logistic regression and restricted cubic spline models examined nonlinear PA–cognition associations. Generalized additive models tested PF's mediating role, adjusting for demographics, lifestyle factors, and chronic conditions.
RESULTS:
A significant nonlinear dose–response relationship emerged between PA and cognitive decline (P-non-linear < 0.05). Relative to moderate-to-high PA (Q3), both low PA (Q1: OR=1.27, 95%CI:1.05–1.55) and very high PA (Q4: OR=1.33, 95%CI:1.07–1.65) were associated with elevated cognitive decline risk, suggesting an optimal PA range. This inverse association was more pronounced among males, rural residents, less-educated individuals, and those with poorer function. Mediation analysis revealed a significant indirect pathway: PA was nonlinearly associated with PF (Path a), and higher PF was significantly associated with greater cognitive decline (Path b, P<0.001). Including PF partially attenuated the direct PA–cognition relationship, supporting its mediating role.
CONCLUSION:
PF mediates the PA–cognitive decline link in middle-aged and older adults, confirming a nonlinear association and identifying an optimal activity level. Findings support "vulnerability amplification," wherein PA most benefits disadvantaged groups. This suggests PA benefits cognition both directly via neurobiological mechanisms and indirectly by enhancing psychological resilience—a "body–mind–cognition" pathway integrating lifestyle factors in cognitive aging. Public health strategies should promote sustained, moderate-intensity PA and incorporate psychological support, particularly for vulnerable populations.This study provides novel evidence that PF mediates the PA–cognitive decline relationship in middle-aged and older adults. Public health strategies should promote sustained, moderate-intensity PA and incorporate psychological support interventions, particularly for vulnerable populations.