Longitudinal association of frailty levels with knee osteoarthritis in middle-aged and elderly chinese: a longitudinal cohort study

Frailty and Knee osteoarthritis (KOA) are highly prevalent in middle-aged and elderly populations. However, evidence on the longitudinal association of frailty with KOA is limited. The aim of our study was to explore the longitudinal effects of frailty levels on KOA, combining phenotypic frailty and frailty index(FI), using Cox regression analysis in a prospective cohort. The data for this study were sourced from the 2011 and 2018 waves of the China Health and Retirement Longitudinal Study (CHARLS). Participants were categorized into three groups based on their total FI concentration: high (≥ 3rd quartile), medium (between the 1st and 3rd quartiles), and low (≤ 1st quartile). A Cox proportional hazards model was used to assess the associations between frailty status, FI, and incident KOA. The predictive value of phenotypic frailty and FI for KOA was evaluated by calculating the area under the receiver operating characteristic curve (AUC). Restricted cubic spline (RCS) analysis was conducted to examine the dose–response relationship between FI and the risk of incident KOA. Additionally, depressive symptoms were included as a mediator to explore their potential mediating effect on the association between frailty and KOA. After 7 years of follow-up, 14,079 participants were included in the analysis of incident KOA. Both phenotypic frailty and FI were significantly associated with increased risk of incident KOA. Compared with non-frail participants, frail individuals had an adjusted hazard ratio (HR) of 2.10 (95% CI: 1.78–2.48; P < 0.05) for KOA. For the FI, a clear dose-response relationship was observed, with HRs for Q2, Q3, and Q4 relative to Q1 being 1.74 (95% CI: 1.36–2.22), 2.62 (95% CI: 2.08–3.30), and 4.17 (95% CI: 3.33–5.22), respectively (P for trend < 0.001). The time-dependent ROC analysis indicated that both phenotypic frailty and FI provided strong predictive value for KOA (AUC for FI = 0.76 at 5 years). Using the causal mediation framework, the mediation effect of social isolation was not statistically significant (p = 0.074), suggesting that the association is primarily driven by direct biological pathways rather than social factors. Frailty had adverse effects on KOA, with social isolation symptoms acting as the mediator.