Int J Cardiol Cardiovasc Risk Prev. 2026 Jun 30;30:200671. doi: 10.1016/j.ijcrp.2026.200671. eCollection 2026 Sep.
ABSTRACT
BACKGROUND: Insulin resistance and frailty are linked and may increase cardiovascular vulnerability. We evaluated whether an aligned index integrating estimated insulin sensitivity and frailty (eGDRFI) predicts incident cardiovascular disease (CVD) and improves risk stratification beyond conventional factors.
METHODS: Two prospective population-based cohorts free of baseline heart disease or stroke were analyzed: CHARLS (China; n = 3859) and ELSA (England; n = 2486). Estimated glucose disposal rate was derived from waist circumference, hypertension, and HbA1c; frailty was measured using a deficit-accumulation frailty index. eGDRFI was prespecified as 100×FI/eGDR, with higher values indicating greater metabolic-frailty burden. Incident composite CVD comprised self-reported physician-diagnosed heart disease or stroke. Cohort-specific Cox models, restricted cubic splines, subgroup and sensitivity analyses, and incremental prediction metrics were applied.
RESULTS: eGDRFI was right-skewed, with median values of 1.00 (IQR, 0.47-1.84) in CHARLS and 1.30 (0.62-2.59) in ELSA. In fully adjusted models, higher eGDRFI quartiles were progressively associated with composite CVD in CHARLS (Q4 vs Q1: HR 2.587, 95% CI 2.030-3.298) and ELSA (HR 1.881, 95% CI 1.398-2.529; both P trend<0.001); continuous associations were significant. Associations were consistent for heart disease in both cohorts and stroke in CHARLS, whereas ELSA stroke estimates were imprecise. Adding eGDRFI modestly improved AUCs and yielded significant NRI and IDI in both cohorts.
CONCLUSIONS: Higher eGDRFI was associated with incident CVD and modestly improved prediction beyond conventional risk factors, supporting further evaluation of metabolic-frailty assessment as a cardiovascular risk marker.
PMID:42437203 | PMC:PMC13355501 | DOI:10.1016/j.ijcrp.2026.200671