Abstract
BACKGROUND: Limited evidence exists regarding the association of frailty with fractures of different types and the potential mediating factors. This study investigated the prospective associations of frailty status with incident fracture risk and potential mediating role of inflammatory biomarkers.
METHODS: This prospective cohort study utilized the UK Biobank data (median follow-up: 13.6 years). Frailty status was assessed using the frailty index (FI) and Fried’s phenotype (FP). The outcomes of interest were any, vertebral, hip, and non-hip non-vertebral (NHNV) fractures. Cox regression models were employed to estimate association between frailty and fracture outcomes. Subgroup analyses by age, sex, and body mass index were conducted. Additionally, mediation analyses were performed to explore whether and quantify the extent to which inflammation may mediate the frailty-fracture association.
RESULTS: Among 418,700 participants aged 38-72 years, frailty significantly increased risk of incident fracture, taking the frailty index as an example, with hazard ratios (HRs) and 95 % confidence intervals (CIs) of 1.53 (1.45-1.60) for any fractures, 2.33 (2.04-2.66) for vertebral fractures, 1.74 (1.55-1.95) for hip fractures, and 1.43 (1.35-1.52) for NHNV fractures. The FI-any fracture association was stronger in participants aged ≥60 years compared to younger participants (Pinteraction = 0.0414); FP exhibited more pronounced associations with any fractures in men than women (Pinteraction = 0.0001). Moreover, three inflammatory biomarkers – C-reactive protein, neutrophils, and platelets – significantly mediated 0.86-2.20 % of the frailty-fracture relationships.
CONCLUSION: Frailty was independently linked to increased risks of incident fracture, partially mediated through inflammatory biomarkers. These findings underscore the clinical importance of frailty screening in fracture prevention settings.