Abstract
BACKGROUND: Biological aging (BA) may influence chronic kidney disease (CKD) development. We evaluated the association of accelerated BA-quantified using Klemera-Doubal method biological age (KDM-BA) and phenotypic age (PhenoAge)-with incident CKD, and assessed its predictive value beyond conventional risk factors (CKD Prediction Consortium [CKD-PC] model) in participants with diabetes.
METHODS: This two-country cohort study included 14,274 participants with diabetes from the UK Biobank and 7,900 from the China Renal Data System (CRDS). KDM-BA and PhenoAge were calculated from clinical biomarkers, and their acceleration (deviation from chronological age) was evaluated. Cox regression assessed associations with incident CKD, while C-index, net reclassification improvement (NRI), and integrated discrimination improvement (IDI) evaluated predictive performance.
RESULTS: Over median follow-ups of 13.3 (UK Biobank) and 3.3 (CRDS) years, 1,676 and 709 incident CKD cases were documented. Each standard deviation increase in KDM-BA acceleration was associated with 31% (95%CI: 23-39%) (UK) and 68% (95%CI: 58-80%) (China) higher CKD hazard. PhenoAge acceleration similarly increased CKD risk (29% and 28%, respectively). In the UK Biobank, adding KDM-BA or PhenoAge acceleration to the CKD-PC model (C-index = 0.770) led to modest but statistically significant improvements in prediction (C-index increase = 0.004 [95%CI: 0.002-0.006] and 0.003 [0.001-0.005], respectively), while leukocyte telomere length (LTL) provided no benefit (0.0001 [-0.0004, 0.0005]). Both BA measures enhanced reclassification (NRI: 0.034-0.076; IDI: 0.002-0.003). CRDS analyses yielded consistent results.
CONCLUSION: Accelerated BA is consistently associated with higher CKD hazard in diabetes across European and Asian populations. KDM-BA and PhenoAge offer practical tools for refining CKD risk stratification.