Abstract
Biological age (BA), compared to chronological age, offers a more accurate reflection of aging status. In this prospective UK Biobank study, BA acceleration was measured using the Klemera-Doubal method BA (KDM-BA) and Phenotypic age (PhenoAge). Cox models estimated associations of BA acceleration with incident T2D (n = 271,885) and CAD (n = 270,054). Both additive and multiplicative interactions between BA acceleration and polygenic risk score (PRS) were examined. Predictive performance was assessed by adding BA, PRS, and their interactions to traditional risk models. BA acceleration was positively associated with incident T2D (HRQ4 vs Q1 for KDM-BA: 2.38 [95% CI, 2.22-2.56]; HRQ4 vs Q1 for PhenoAge: 1.85 [95% CI, 1.72-1.99]) and CAD (HRQ4 vs Q1 for KDM-BA: 1.67 [95% CI, 1.58-1.76]; HRQ4 vs Q1 for PhenoAge: 1.33 [95% CI, 1.27-1.39]). Significant multiplicative interactions were observed between BA acceleration and PRS (all P for multiplicative interaction ≤ 0.002). Individuals with highest BA acceleration and PRS had strongest risk elevation for T2D (HR for KDM-BA, 6.89 [95% CI, 6.03-7.87]; HR for PhenoAge, 6.28 [95% CI, 5.28-7.46]) and CAD (HR for KDM-BA, 2.80 [95% CI, 2.59-3.02]; HR for PhenoAge, 2.25 [95% CI, 2.07-2.45]). Additive interactions were observed for T2D with 18-28% of risk attributable to BA-genetic interaction. Adding BA measures and PRS to traditional risk models significantly improved prediction for both diseases (Δ C-statistic 0.024-0.034). In conclusion, BA acceleration was positively associated with incident T2D and CAD, especially in individuals with high genetic predisposition, and improved T2D and CAD prediction beyond traditional risk factors.