Stress-hyperglycemia ratio, frailty index, and their product in relation to incident cardiometabolic multimorbidity comprising diabetes, heart disease, and stroke: a longitudinal analysis of CHARLS

Cardiometabolic multimorbidity (CMM), encompassing the co-occurrence of major cardiometabolic conditions such as diabetes, heart disease, and stroke, is a major source of disability and premature mortality in ageing populations. The stress-hyperglycemia ratio (SHR) captures glycemic dysregulation relative to long-term glycemic exposure, whereas the frailty index (FI) quantifies accumulated health deficits. We evaluated whether SHR, FI, and their composite product (SHR-FI), assessed at baseline and across two survey waves, were associated with incident CMM. We conducted a prospective analysis of 6,096 participants from the China Health and Retirement Longitudinal Study who were free of CMM at baseline. SHR, FI, and SHR-FI were assessed at Waves 1 and 3. Time-weighted two-wave summaries and data-derived two-time-point longitudinal patterns were evaluated; cumulative and pattern analyses used Wave 3 as a landmark. Cox models estimated hazard ratios (HRs) and 95% confidence intervals (CIs). Restricted cubic spline, subgroup, receiver operating characteristic, decision-curve, joint-model, landmark, and sensitivity analyses were performed. During the Wave 1-anchored follow-up, 527 participants developed CMM. In fully adjusted models, each SD increase in baseline SHR, FI, and SHR-FI was associated with HRs of 1.09 (95% CI, 1.01–1.18), 1.36 (1.26–1.48), and 1.35 (1.26–1.45), respectively. Associations were stronger for cumulative FI and cumulative SHR-FI: the HRs per SD increase were 1.60 (1.48–1.73) and 1.55 (1.44–1.66). The AUCs were 0.68 for baseline FI and SHR-FI and 0.72 and 0.73 for cumulative FI and SHR-FI, respectively. FI accounted for most of the observed discrimination, while SHR-FI showed a modest complementary signal. Landmark analyses beginning at Wave 3 produced directionally consistent cumulative FI and SHR-FI associations. Higher SHR, greater FI, and a higher SHR-FI product were associated with incident CMM. FI provided most of the observed discrimination, and SHR-FI offered limited additional information in this cohort. Two-wave cumulative summaries and longitudinal patterns may support research-oriented risk stratification, but their clinical value requires calibration, internal validation, and external validation.