Frailty and prefrailty were associated with a significantly higher risk of new-onset chronic kidney disease compared to non-frailty, with hazard ratios of 1.606 and 1.143, respectively.
Does the frailty phenotype increase the risk of developing new-onset CKD in CKD-free adults?
Pre-frail and frail phenotypes are associated with a significantly higher risk of developing incident CKD, and a prediction model incorporating frailty showed stable discriminative performance over a decade.
ABSTRACT Background The longitudinal relationship between frailty phenotype and CKD development, and the modifying role of genetic CKD risk in this association, remains unclear. Research on simple, noninvasive and quantifiable CKD prediction models incorporating frailty is limited. Methods We analysed 214 502 CKD‐free participants from the UK Biobank cohort. Frailty was assessed using a modified phenotype with five components: weight loss, exhaustion, low grip strength, low physical activity and slow gait speed, to match UK Biobank data. Self‐reported walking pace and weight change served as proxies where objective measures were unavailable. A polygenic risk score for CKD was calculated based on 258 single nucleotide polymorphisms. Cox proportional hazards models were used to assess the association between the frailty phenotype and new‐onset CKD. The interaction between frailty and genetic risk on CKD outcomes was also examined. A noninvasive CKD prediction model, integrating frailty, age, gender, diabetes, hypertension, BMI and smoking status, was developed and validated internally using the UK Biobank and externally using CHARLS cohorts. Results Among the 214 502 CKD‐free participants with a median age of 57 (49–62) years, 50.3% were female. A total of 109 290 individuals (51.0%) were classified as non‐frail, 96 941 (45.2%) as pre‐frail and 8271 (3.9%) as frail. Over the course of a median follow‐up period of 12.9 years, we documented 8079 (3.8%) cases of CKD. Compared with non‐frailty, the hazard ratio (HR) for new‐onset CKD in prefrailty and frailty was 1.143 (95% CI, 1.090–1.199, p < 0.001) and 1.606 (95% CI, 1.474–1.749, p < 0.001) in the multivariate model, respectively. Each one‐point increase in frailty score was associated with a higher risk of CKD (HR = 1.142; 95% CI, 1.116–1.170, p < 0.001) in the multivariable model. Participants with frailty and high genetic risk had the greatest risk of CKD (HR = 1.981; 95% CI, 1.733–2.266, p < 0.001) compared with those without frailty and low genetic risk. In the simple CKD prediction model incorporating frailty, it demonstrated an AUC of 0.734 at 5 years, 0.745 at 8 years and 0.749 at 10 years in internal testing. External validation also showed consistent discrimination and calibration with an AUC of 0.740. Conclusions Pre‐frail and frail phenotypes were associated with a higher risk of developing CKD, showing a dose–response relationship. A noninvasive prediction model incorporating frailty and clinical parameters exhibited stable discriminative performance over a decade in both European and Asian cohorts.
Ruan et al. (2026) studied this question. Frailty and prefrailty were associated with a significantly higher risk of new-onset chronic kidney disease compared to non-frailty, with hazard ratios of 1.606 and 1.143, respectively.