Waist-centered triglyceride-glucose markers and inflammation in relation to incident cardiometabolic multimorbidity: a multinational pooled analysis of five aging cohorts

Abstract To compare triglyceride-glucose-derived adiposity indices and evaluate whether C-reactive protein adds prognostic information for incident cardiometabolic multimorbidity across aging cohorts. We pooled 20,509 participants free of baseline CMD from five aging cohorts in China, Costa Rica, England, South Africa, and the United States. Cohort-specific Cox models were combined with random-effects meta-analysis to compare six triglyceride-glucose-related anthropometric indices and log-transformed C-reactive protein. Joint, interaction, and prediction analyses focused on triglyceride-glucose-waist circumference, with triglyceride-glucose-waist-to-height ratio examined in parallel. Over a median follow-up of 6.8 years, 783 participants developed cardiometabolic multimorbidity. All triglyceride-glucose-related indices were associated with higher risk, with the largest pooled estimates for waist-centered markers. Per one-standard-deviation increase, the pooled hazard ratio was 1.48 (95% confidence interval 1.25–1.74) for triglyceride-glucose-waist circumference and 1.21 (95% confidence interval 1.13–1.29) for log-transformed C-reactive protein. Participants with high triglyceride-glucose-waist circumference and high log-transformed C-reactive protein had the highest risk versus the low-marker/low-inflammation group (hazard ratio 2.50, 95% confidence interval 1.65–3.77), without formal interaction. Adding log-transformed C-reactive protein to models already including triglyceride-glucose-waist circumference increased the concordance index by only 0.004. Waist-centered triglyceride-glucose markers showed the largest pooled associations with incident cardiometabolic multimorbidity. C-reactive protein identified a higher-risk subgroup but added little incremental discrimination beyond triglyceride-glucose-waist circumference. These observational findings support further evaluation of TyG-derived markers in aging populations.