Suo · Neurology 2025 · prospective cohort study and bidirectional Mendelian randomization · n=489,573

Association of Frailty With Dementia and the Mediating Role of Brain Structure and Immunometabolic Signatures.

Cited 4 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective non-randomized cohort study with bidirectional Mendelian randomization

PubMed 40961394 · doi:10.1212/WNL.0000000000214199 · record verified 2026-08-29

What was done

A prospective cohort study of 489,573 dementia-free UK Biobank participants enrolled between 2006 and 2010 (mean age 57.03 years, 54.4% female). Physical frailty was measured using five phenotype criteria: weight loss, exhaustion, physical inactivity, slow walking speed, and low grip strength. Incident dementia was ascertained via linked hospital admission records and death registries using ICD-10 diagnostic codes over a median follow-up of 13.58 years. Analyses included Cox proportional hazards regression, bidirectional Mendelian randomization (MR) to evaluate causality, and structural equation modeling to explore mediation by genetic risk, brain structure, and biological biomarkers.

What was found

Over follow-up, 8,900 incident dementia cases occurred. Dementia risk was higher in prefrailty (hazard ratio [HR]: 1.50, 95% CI 1.44–1.57) and frailty (HR: 2.82, 95% CI 2.61–3.04) compared to nonfrailty. Frail individuals with high genetic risk showed the highest risk compared to nonfrail individuals with low genetic risk (HR: 3.87, 95% CI 3.30–4.55 for high polygenic risk score; HR: 8.45, 95% CI 7.51–9.51 for APOE-ε4 carriers). Forward MR indicated physical frailty was causally associated with dementia (odds ratio [OR]: 1.79, 95% CI 1.03–3.12), whereas reverse MR showed no effect (OR: 1.00, 95% CI 0.98–1.01). Specific numeric effect sizes for structural equation modeling of brain structures and immunometabolic biomarkers were not reported in the abstract.

Why it matters

These findings suggest that physical frailty is not merely a correlative marker but a potentially causal contributor to incident dementia that compounds genetic risk, highlighting frailty as a potential target for dementia prevention.

Limits

Dementia cases were identified solely through hospital admission and mortality records, which may miss milder outpatient cases. The UK Biobank cohort is subject to healthy volunteer selection bias, and specific quantitative mediation metrics for brain imaging and immunometabolic measures were not detailed in the abstract.

Cited by