TITLE: Frailty, grip strength, blood-based biomarkers of neuropathology and incident dementia: a 17-year longitudinal population study

Background While previous research has demonstrated that frailty is associated with increased dementia risk over time, large multi-decade prospective population-based cohort studies remain limited. Furthermore, the association between lower grip strength and blood-based biomarkers of neuropathology and global cognitive function remains under-investigated. Methods We investigated the relationship between varying levels of baseline frailty (measured using a deficit accumulation Frailty Index) and dementia incidence over 17 years, as well as associations of grip strength (hand, pinch and key, assessed at 8- and 14-year follow-up) with blood-based biomarkers of neuropathology (year 8) and with global cognitive function (year 14) in a population-representative cohort of 9,940 mid-to-late life (50-75 years) individuals. Results In a subsample of 6,357 participants with available dementia information at 17-year follow-up (mean age at baseline [SD], 61.7 [6.6] years; female, 54.3%), 516 received an all-cause dementia diagnosis. Over this time, individuals in the highest frailty index quintile demonstrated a significantly higher risk of all-cause dementia (HR, 1.75; 95% CI, 1.18-2.60) and a vascular dementia diagnosis (HR, 2.44; 95% CI, 1.22-4.91). Cross-sectionally, lower handgrip strength was associated with elevated blood levels of NfL (β=-0.01, p=0.007) and pTau181 (β=-0.004, p=0.017) at 8-year follow-up, as well as poorer cognitive function on the Montreal Cognitive Assessment (MoCA) (β = 0.02, p=0.037) at 14-year follow-up. Conclusions This study strengthens evidence that frailty is a risk factor for dementia in mid-to-late life. It also suggests that it is of considerable importance to assess frailty and grip strength in individuals ‘at risk’ of cognitive decline.