Association of the c-reactive protein–triglyceride–glucose index and triglyceride–glucose index with incidence of cardiometabolic multimorbidity in middle-aged and older chinese adults: a cohort study

Cardiometabolic multimorbidity (CMM) has emerged as a major public health challenge in aging populations. The C-reactive protein–triglyceride–glucose index (CTI), a novel marker reflecting combined metabolic and inflammatory burden, has shown promising predictive value in various metabolic conditions. This study aimed to evaluate and compare the predictive potential of CTI versus the traditional triglyceride–glucose (TyG) index for incident CMM in middle-aged and older Chinese adults. Data were derived from the China Health and Retirement Longitudinal Study (CHARLS, 2011–2020), including participants aged ≥ 45 years without CMM at baseline. Cox proportional hazards models were applied to assess the associations of CTI and TyG with incident CMM, and restricted cubic spline (RCS) models were used to explore potential nonlinear relationships. Discriminative performance was evaluated using time-dependent receiver operating characteristic (ROC) curves, while incremental predictive value beyond conventional risk factors was assessed using Harrell’s C-statistic, ΔC-statistic, category-free net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Bootstrap internal validation with 1,000 resamples was performed to quantify and correct for optimism in model discrimination. Interaction analyses were conducted across prespecified subgroups to assess robustness. After excluding 3,054 individuals with CMM at baseline, 1,688 incident CMM cases were identified during a 9-year follow-up, with a cumulative incidence of 26.8%. In the fully adjusted Cox model, CTI was strongly associated with CMM risk (HR = 1.658, 95% CI: 1.525, 1.802, P < 0.001) than TyG (HR = 1.534, 95% CI: 1.426, 1.651, P < 0.001). RCS analysis revealed a linear positive relationship between both indices and CMM risk. ROC analyses indicated comparable performance at year 2, while CTI outperformed TyG at years 4, 7, and 9 (all AUCs significantly higher, P < 0.05). Subgroup analyses showed no significant interactions for CTI ( P for interaction > 0.05), suggesting greater consistency across populations; in contrast, TyG showed a significant interaction with dyslipidemia, implying sensitivity to metabolic context. At 9 years, addition of TyG to the conventional risk-factor model yielded an NRI of 0.183 (95% CI: 0.112–0.242) and an IDI of 0.00783 (95% CI: 0.00302–0.01277), whereas addition of CTI yielded an NRI of 0.242 (95% CI: 0.176–0.301) and an IDI of 0.01172 (95% CI: 0.00496–0.01824). Although CTI showed numerically greater incremental improvement, direct CTI-versus-TyG differences in NRI and IDI were not statistically conclusive. Bootstrap internal validation demonstrated limited optimism (approximately 0.004), with optimism-corrected C-statistics of 0.6772, 0.6864, and 0.6890 for the baseline, TyG-extended, and CTI-extended models, respectively. CTI is linearly associated with CMM risk and demonstrates modest but statistically significant improvement in long-term discriminative ability compared to the traditional TyG index. These findings highlight the potential utility of CTI as a complementary biomarker for early identification and stratification of cardiometabolic risk, though it requires integration with other risk factors for clinical application.