Estimated glucose disposal rate and incident cardiovascular disease across ageing cohorts in China, England, and the United States: cross-national consistency and transportability

The estimated glucose disposal rate (eGDR), an insulin-resistance surrogate derived from routine clinical variables, has been associated with cardiovascular disease (CVD). We examined its association with incident CVD, cross-national transportability, and incremental discrimination across ageing populations in China, England, and the United States. We analysed harmonized prospective cohort data from the China Health and Retirement Longitudinal Study (CHARLS), the English Longitudinal Study of Ageing (ELSA), and the Health and Retirement Study (HRS). Primary analyses included participants free of diabetes, heart disease, and stroke at baseline; incident CVD was defined as new-onset heart disease or stroke. Associations per 1-SD decrease in eGDR were estimated using attained-age Cox models and pooled by random-effects meta-analysis. Frozen-coefficient models assessed transportability across cohorts, and incremental discrimination was evaluated against established risk factors, HbA1c, and eGDR’s constituent variables. The analytic samples comprised 6,059 CHARLS participants (1,219 incident CVD events), 4,061 ELSA participants (865 events), and 4,299 HRS participants (809 events). Hazard ratios per 1-SD lower eGDR were 1.24 (95% CI 1.17–1.32) in CHARLS, 1.23 (1.15–1.33) in ELSA, and 1.23 (1.14–1.33) in HRS; the pooled estimate was 1.24 (1.19–1.29; I 2 =0%). The pooled hazard ratio for the lowest versus highest eGDR quartile was 1.83 (1.64–2.05). Results were similar in sensitivity analyses. Frozen-coefficient models showed modest cross-national discrimination (external Harrell C, 0.570–0.647), while absolute-risk calibration varied across source–target cohort combinations. eGDR did not improve discrimination beyond waist circumference, hypertension, and HbA1c entered separately. Lower eGDR was consistently associated with incident CVD across the three ageing populations. Cross-national discrimination was modest, whereas absolute-risk calibration varied across cohorts. eGDR is best interpreted as a parsimonious summary of routinely measured cardiometabolic information rather than a source of additional discriminative information beyond its constituent variables.