Longitudinal dynamics in frailty and incident all-cause dementia: a multicohort study

• Progression from robust to pre-frail/frail or from pre-frail to frail status is related to an increased risk of dementia. • Recovery from pre-frailty to robustness is associated with a reduced risk of dementia. • Reversion from frailty to pre-frail/robust status shows a trend toward lower dementia risk. • Findings highlight the importance of dynamic frailty monitoring and individualized interventions for dementia prevention. Frailty changes over time, but how these changes relate to the risk of dementia remains uncertain. To explore the temporal relationship between dynamic changes in frailty and incident all-cause dementia. This pooled analysis included adults aged ≥ 60 years from four prospective cohorts. Frailty was assessed using a modified Fried phenotype at two consecutive surveys, allowing the identification of longitudinal frailty transition patterns. The outcome was all-cause dementia, identified through active follow-up assessments or passive data collection. Hazard ratios (HRs) and 95% confidence intervals (CIs) were derived from Cox proportional hazards analyses. Among 15,897 participants (51.3% female, mean age: 66.9 years), 1015 developed incident dementia over 131,621.3 person-years of follow-up. An increased risk of dementia was observed among robust participants who progressed to pre-frail/frail status, compared to those with stable robustness (pooled HR: 1.62, 95% CI: 1.06–2.47; I 2 = 66.8%). Similarly, pre-frail participants who progressed to frail status exhibited a higher dementia risk compared to those who remained pre-frail (pooled HR: 1.85, 95% CI: 1.48–2.34; I 2 = 0.0%). Conversely, participants who recovered from pre-frail to robust status experienced a reduced dementia risk (pooled HR: 0.59, 95% CI: 0.46–0.76; I 2 = 0.0%). A higher cumulative frailty score was related to a higher dementia risk (HR: 1.26, 95% CI: 1.22–1.31; I 2 = 0.0% per score increment). Participants whose frailty scores increased over time had a higher risk of dementia compared to those with stable scores (pooled HR: 1.33, 95% CI: 1.07–1.65; I 2 = 75.5% per score increment), while those with decreasing scores showed a mitigated dementia risk (pooled HR: 0.77, 95% CI: 0.65–0.91; I 2 = 0.0% per score decrement). Progression in frailty was associated with an increased risk of dementia, while reversal of pre-frailty reduced the risk. Our results emphasize the importance of ongoing monitoring and individualized interventions for managing frailty in older adults.