Introduction

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide, with coronary heart disease (CHD) and type 2 diabetes mellitus (T2DM) being two of the most prevalent and interrelated contributors to this global burden1. Epidemiological studies have shown that approximately 30–40% of patients with CHD also have T2DM, and individuals with T2DM face a two- to fourfold increased risk of cardiovascular events compared to those without diabetes2,3. The coexistence of CHD and T2DM not only amplifies the risk of adverse cardiovascular outcomes but also complicates clinical management, leading to significantly worse long-term prognosis4,5.

Chronic systemic inflammation and insulin resistance (IR) are well-established pathophysiological mechanisms driving the progression of atherosclerosis and subsequent cardiovascular events, particularly in patients with both CHD and T2DM6,7. C-reactive protein (CRP), a sensitive marker of systemic inflammation, and the triglyceride-glucose (TyG) index, a surrogate marker of insulin resistance, have each been independently associated with increased cardiovascular risk8,9,10. However, relying on a single biomarker may not fully capture the complex interplay between metabolic dysfunction and inflammatory processes in high-risk cardiometabolic populations.

The C-reactive protein-triglyceride-glucose index (CTI) has recently emerged as a novel composite marker that integrates both inflammatory and metabolic components, providing a more comprehensive assessment of cardiovascular risk11. While previous studies have demonstrated the prognostic value of CTI in populations with hypertension, cerebrovascular disease, and cancer12,13,14, evidence regarding its association with mortality outcomes in patients with coexisting CHD and T2DM remains limited.

Therefore, this study aimed to investigate the association between CTI levels and both all-cause and cardiovascular mortality in patients with CHD and T2DM using data from the National Health and Nutrition Examination Survey (NHANES). NHANES is a nationally representative, population-based cohort with standardized protocols and long-term follow-up, enabling robust estimation of risk in diverse U.S. populations. By leveraging this rich dataset and rigorous analytic strategies—including survey-weighted methods—our objective was to evaluate the prognostic value (rather than causal inference) of CTI in predicting mortality outcomes. This study seeks to provide novel insights into the potential utility of CTI as a clinically applicable biomarker for risk stratification in this high-risk population.

Methods

Study design and population

This study utilized data from the NHANES 1999–2010 cycles, a nationally representative, cross-sectional survey conducted by the Centers for Disease Control and Prevention (CDC) in the United States. NHANES is designed to assess the health and nutritional status of the general U.S. population, with an emphasis on various chronic diseases, demographic data, and lifestyle factors. The study was conducted in accordance with the ethical guidelines of the NHANES Research Ethics Review Board, and all participants provided written informed consent. The study design and reporting were strictly aligned with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Participants aged ≥ 20 years with diagnosed CHD and T2DM at baseline were included. CHD was defined by self-reported physician diagnosis of angina, myocardial infarction, or coronary heart disease. T2DM was defined as fasting plasma glucose (FPG) ≥ 7.0 mmol/L, glycated hemoglobin (HbA1c) ≥ 6.5%, or self-reported diabetes diagnosis with antidiabetic treatment. Individuals were excluded if they had missing data for C-reactive protein, Triglycerides (TG), fasting plasma glucose, or mortality follow-up information.

A total of 759 participants who met both the CHD and T2DM diagnostic criteria were included in the final analysis. During the follow-up period, 520 participants died from all causes, and 213 died from cardiovascular causes. The study design is illustrated in (Fig. 1).

Fig. 1
figure 1

Flowchart of participant selection and follow-up.

Full size image

Assessment of the CTI

The CTI was calculated using the following formula, integrating markers of systemic inflammation and insulin resistance:

CTI = 0.412×Ln(CRP) + Ln(TG [mg/dL]×FPG [mg/dL]/2)15.

The CTI value for each participant was calculated, and participants were categorized into quartiles (Q1-Q4) based on CTI distribution, with Q1 serving as the reference group.

Follow-up and outcomes

Death data were obtained through linkage to the National Death Index (NDI), which provides cause-of-death information for individuals included in the NHANES dataset. The follow-up period for each participant began at the baseline survey and continued until either the date of death or the study cutoff date, December 31, 2019, whichever occurred first. This study assessed both all-cause mortality and cardiovascular mortality. All-cause mortality was defined as death from any cause (010), while cardiovascular mortality specifically referred to death due to heart disease (054–064) or cerebrovascular disease (070).

Covariates

Baseline covariates were selected based on clinical relevance and prior literature regarding cardiovascular risk factors: Demographic factors: Age, Gender, Race, Education level, Marital status, and Poverty-income ratio (PIR). Clinical comorbidities: Hypertension, Heart failure, Stroke, Cancer, Body mass index (BMI), Waist circumference (WC), Systolic and Diastolic blood pressure (SBP, DBP), Drink, Cardiovascular-Kidney-Metabolic syndrome (CKM stage 0–4) and Frailty status (defined using the frailty index [FI] as: FI ≤ 0.10 as non-frail; 0.11–0.21 as pre-frail; >0.21 as frail). Laboratory measurements: HbA1c, Total cholesterol (TC), High-density lipoprotein cholesterol (HDL-C), Serum creatinine (Cr), Blood urea nitrogen (BUN), Uric acid (UA), Albumin (ALB), White blood cell count (WBC), Hemoglobin (HGB), Platelet count (PLT), Mean corpuscular volume (MCV), Lymphocyte count (LYMPH), estimated Glomerular filtration rate (eGFR, calculated using the CKD-EPI equation), triglycerides, fasting plasma glucose, C-reactive protein.

Variables with > 20% missing data were excluded. For variables with less than 20% missing data, we applied random forest imputation (missForest algorithm) to ensure data completeness. This method outperforms traditional approaches such as Multiple Imputation by Chained Equations (MICE) in handling nonlinear relationships, complex interactions, and mixed-type variables16. All variables were obtained through standardized NHANES protocols; definitions and documentation are available at https://www.cdc.gov/nchs/nhanes/.

Statistical analysis

Baseline characteristics were described according to mortality status. Continuous variables were presented as medians with interquartile ranges (IQR) and compared using the Wilcoxon rank-sum test. Categorical variables were expressed as counts and percentages, with comparisons made using the Chi-square test. Kaplan-Meier survival curves and log-rank tests were used to evaluate differences in survival across CTI quartiles. Cox proportional hazards regression models were constructed to assess the association between CTI and mortality outcomes. Three models were developed: Model 1: Unadjusted. Model 2: Adjusted for demographic factors (age, gender, race, education, poverty-income ratio, drink, cardiovascular-kidney-metabolic syndrome, body mass index, waist circumference). Model 3: Further adjusted for clinical comorbidities and laboratory variables (heart failure, frailty status, systolic blood pressure, triglycerides, HbA1c, estimated Glomerular filtration rate, HDL-C). The proportional hazards assumption was verified using Schoenfeld residuals. Restricted cubic spline (RCS) analyses with four knots were conducted to explore potential non-linear relationships between CTI and mortality risk, adjusting for covariates in Model 3. Subgroup analyses were performed stratified by age, gender, body mass index, HbA1c, estimated glomerular filtration rate, and comorbidities. To evaluate potential effect modification, two approaches were used. First, multiplicative interaction terms (e.g., CTI × categorical subgroup variable) were included in the Cox models to test for statistical interaction using Wald statistics, with P values reported. Second, for continuous effect modifiers (age, body mass index, HbA1c, and estimated glomerular filtration rate), additional effect-based interaction analyses were conducted by incorporating CTI × continuous variable terms into multivariable Cox regression models. Marginal effects and hazard ratios of CTI across the continuous variable spectrum were estimated and visualized using model-based marginal plots, enabling a more granular assessment of effect heterogeneity. Given the complex, multistage probability sampling design of NHANES, we incorporated weighted analyses in this study to enhance the national representativeness of our estimates. All model-based analyses accounted for the NHANES survey design, including sample weights, strata, and primary sampling units. Weighted Kaplan–Meier survival curves, Cox proportional hazards models, and subgroup analyses were performed using the survey (version 4.4-2) package in R. All statistical analyses were conducted using R software (version 4.2.2). A two-sided P-value < 0.05 was considered statistically significant.

Results

Baseline characteristics of the participants

A total of 759 participants with coexisting CHD and T2DM were included in the final analysis. During a median follow-up of 109 months, 520 (68.5%) participants died from all causes, and 213 (28.1%) deaths were attributed to cardiovascular causes. As shown in Table 1, compared to survivors, non-survivors of all-cause mortality were older, more likely to be non-Hispanic White, and had a higher prevalence of heart failure and stroke (all P < 0.001). Additionally, they exhibited lower body mass index, diastolic blood pressure, hemoglobin, platelet count, lymphocyte count, and estimated glomerular filtration rate levels, but higher concentrations of blood urea nitrogen, serum creatinine, and C-reactive protein (all P < 0.05). Similar trends were observed among cardiovascular non-survivors. Notably, CTI levels were significantly elevated in both all-cause and cardiovascular death groups (P < 0.001), whereas TyG index levels did not differ significantly between survivors and non-survivors.

Importantly, the distribution of frailty status also varied significantly according to mortality status (P < 0.001 for all-cause and P = 0.049 for cardiovascular mortality), with a higher proportion of frail individuals among non-survivors. Additionally, a trend toward more advanced CKM stages was observed among non-survivors, although the difference was not statistically significant. These findings suggest that both frailty and cardiovascular-kidney-metabolic syndrome progression may contribute to adverse outcomes and warrant further stratified analyses.

Table 1 Baseline characteristics of participants according to all-cause and cardiovascular mortality status.
Full size table

Kaplan-Meier survival curves across CTI quartiles

Kaplan-Meier survival curves (Fig. 2) revealed significant differences in both all-cause (Fig. 2A) and cardiovascular mortality (Fig. 2B) across quartiles of the CTI. Participants in the highest CTI quartile exhibited the lowest cumulative survival probabilities. Log-rank tests confirmed significant trends, with progressively higher mortality rates observed from Q1 to Q4 for both outcomes (P for trend < 0.001). These findings suggest a dose-response relationship between elevated CTI and increased mortality risk in patients with CHD and T2DM.

To enhance national representativeness, weighted Kaplan-Meier analyses accounting for the NHANES complex survey design were additionally conducted. The weighted results demonstrated consistent survival trends with the unweighted analysis, further supporting the robustness of our findings (Supplementary Fig. S1).

Fig. 2
figure 2

Kaplan-Meier survival curves for all-cause (A) and cardiovascular mortality (B) according to CTI quartiles. CTI, C-reactive protein-triglyceride-glucose index.

Full size image

Cox regression analysis of CTI and mortality outcomes

Multivariate Cox proportional hazards regression models were used to evaluate the association between CTI and mortality (Table 2). After adjusting for potential confounders, CTI was independently associated with increased risk of both all-cause and cardiovascular mortality. For each unit increase in CTI, the hazard ratios (HRs) were 1.68 (95% CI: 1.47–1.92, P < 0.001) for all-cause mortality and 1.70 (95% CI: 1.38–2.10, P < 0.001) for cardiovascular mortality in the fully adjusted model. When CTI was analyzed as a categorical variable, participants in the highest quartile (Q4, ≥ 9.24) had a 3.07-fold increased risk of all-cause death (HR: 3.07, 95% CI: 2.17–4.34, P < 0.001) and a 2.75-fold increased risk of cardiovascular death (HR: 2.75, 95% CI: 1.60–4.72, P < 0.001) compared to those in the lowest quartile (Q1, < 8.18). A significant dose-response relationship was observed across CTI quartiles (P for trend < 0.001).

To ensure the national representativeness of our findings, we additionally performed weighted Cox regression analyses that accounted for the complex sampling design of NHANES. The results remained robust and consistent, with similar effect estimates observed for both continuous and categorical CTI measures (Supplementary Table S1).

Table 2 Multivariable Cox proportional hazards regression analysis for the association between CTI and all-cause and cardiovascular mortality.
Full size table

Restricted cubic spline analysis of CTI and mortality.

Restricted cubic spline (RCS) analyses were conducted to explore the potential nonlinear relationship between CTI and mortality risk (Fig. 3). In fully adjusted models, a linear and dose-dependent association was observed between CTI levels and both all-cause (Fig. 3A) and cardiovascular mortality (Fig. 3B). The risk of death increased steadily as CTI rose, with no significant evidence of nonlinearity (P for nonlinearity > 0.05 for both outcomes). These findings suggest a proportional relationship between elevated CTI and increased mortality risk in patients with CHD and T2DM.

Fig. 3
figure 3

Restricted cubic spline analysis showing linear association between CTI and risks of (A) all-cause mortality and (B) cardiovascular mortality. CTI, C-reactive protein-triglyceride-glucose index.

Full size image

Subgroup analyses of the CTI-mortality relationship

Subgroup analyses were conducted to examine the robustness of the association between CTI and mortality across clinically relevant strata (Fig. 4). For all-cause mortality (Fig. 4A), the positive association between higher CTI and increased mortality risk remained significant in nearly all subgroups, including those stratified by age, gender, heart failure, stroke, drinking status, body mass index, HbA1c, and estimated glomerular filtration rate, with no significant effect modifications observed (all P for interaction > 0.05). Notably, the association remained significant across all stages of cardiorenal-metabolic syndrome (CKM stage), as well as across frailty categories (frail, pre-frail, and non-frail), although the interaction terms for CKM stage and frailty status were not statistically significant (P for interaction = 0.357 and 0.505, respectively). These findings suggest that the prognostic value of CTI is broadly consistent regardless of comorbidity burden or frailty level.

For cardiovascular mortality (Fig. 4B), similar trends were observed. The association between CTI and cardiovascular death remained robust across most subgroups. While slightly higher effect sizes were observed in participants with better renal function (eGFR ≥ 60 mL/min/1.73 m²) and lower HbA1c levels, none of the subgroup interaction tests reached statistical significance (all P for interaction > 0.05). The association was also maintained across different CKM stages and frailty status levels, further highlighting the generalizability of CTI as a risk indicator across diverse clinical phenotypes.

To explore potential interactions with continuous clinical variables, multiplicative interaction models were constructed for age, body mass index, HbA1c, and estimated glomerular filtration rate (Fig. 5A–H). Age demonstrated a statistically significant interaction with CTI on all-cause mortality (P for interaction = 0.012), with a stronger association in younger individuals. No significant interactions were found for body mass index, HbA1c, or estimated glomerular filtration rate in either outcome.

Additionally, complex survey-weighted subgroup analyses based on NHANES design (Supplementary Fig. S2–S3) yielded results consistent with the unweighted findings, further confirming the stability and robustness of CTI in predicting mortality risk across key subpopulations.

Fig. 4
figure 4

Subgroup analysis of the association between CTI and all-cause and cardiovascularmortality in different populations. (A) All-cause mortality; (B) Cardiovascular mortality; CTI,C-reactive protein-triglyceride-glucose index; BMI, Body mass index; HbA1c, Glycatedhemoglobin; eGFR, estimated glomerular filtration rate.

Full size image
Fig. 5
figure 5

Interaction effects between CTI and continuous clinical variables on mortality risk. (A-D) Association of CTI with all-cause mortality across levels of age (A), BMI (B), HbA1c (C), and eGFR (D). (E-H) Association of CTI with cardiovascular mortality across levels of age (E), BMI (F), HbA1c (G), and eGFR (H). BMI, Body mass index; HbA1c, Glycated hemoglobin; eGFR, estimated glomerular filtration rate; HR: Hazard Ratio.

Full size image

Correlation analysis of CTI and clinical indicators

A correlation heatmap was constructed to evaluate the associations between CTI and various demographic, clinical, and biochemical parameters (Fig. 6). As expected, CTI showed significant positive correlations with its core components—C-reactive protein, triglycerides, and fasting plasma glucose. In addition, moderate positive correlations were observed between CTI and body mass index, systolic blood pressure, uric acid, and blood urea nitrogen, suggesting potential links with metabolic dysregulation, renal impairment, and systemic inflammation.

Conversely, CTI was negatively correlated with estimated glomerular filtration rate, HDL-C, and albumin, indicating that elevated CTI levels may reflect deteriorated renal function, poor nutritional status, and adverse lipid profiles. These findings suggest that CTI exhibits concordant variation with multiple conventional mortality-related factors, reinforcing its biological plausibility as a clinically meaningful prognostic biomarker.

Fig. 6
figure 6

Pearson correlation matrix among baseline variables. PIR, Poverty-income ratio; BMI, Body mass index; SBP, Systolic blood pressure; DBP, Diastolic blood pressure; TG, Triglycerides; FPG, Fasting plasma glucose; Cr, Serum creatinine; TC, Total cholesterol; HDL-C, High-density lipoprotein cholesterol; ALB, Albumin; BUN, Blood urea nitrogen; HbA1c, Glycated hemoglobin; UA, Uric acid; eGFR, estimated glomerular filtration rate; WBC, White blood cell count; HGB, Hemoglobin; PLT, Platelet count; MCV, Mean corpuscular volume; LYMPH, Lymphocyte count; CRP, C-reactive protein; TyG, Triglyceride-glucose; CTI, C-reactive protein-triglyceride-glucose index.

Full size image

Discussion

In this nationally representative cohort study, we identified a significant and independent association between elevated CTI levels and increased risks of both all-cause and cardiovascular mortality in patients with CHD and T2DM. These associations persisted after comprehensive adjustment for demographic, clinical, and biochemical confounders, and exhibited a clear linear dose-response relationship. Subgroup analyses further confirmed the robustness of these findings across various clinical strata, particularly among individuals with better glycemic control (HbA1c < 7%). Furthermore, the predictive utility of CTI remained consistent across stages of cardiorenal-metabolic syndrome and frailty status, with no significant effect modification observed. Despite the absence of statistical interaction, the consistent direction and magnitude of associations suggest that CTI may serve as a stable and generalizable prognostic marker, even among individuals with advanced cardiometabolic or functional impairment. This highlights CTI’s potential to capture the cumulative burden of systemic inflammation and metabolic dysregulation, particularly in high-risk populations where conventional markers may be insufficient.

The underlying mechanisms linking CTI to adverse outcomes likely stem from its ability to simultaneously capture two pivotal pathophysiological processes: systemic inflammation and insulin resistance17. Chronic low-grade inflammation, as reflected by elevated C-reactive protein, plays a central role in endothelial dysfunction, plaque instability, and thrombogenesis, thereby accelerating atherosclerotic progression18. Concurrently, insulin resistance contributes to metabolic dysregulation, promoting hyperglycemia, dyslipidemia, and hypertension, all of which synergistically exacerbate cardiovascular risk19,20. While C-reactive protein and the TyG index have individually been associated with cardiovascular events21,22,23, our findings support that CTI, as a composite marker, offers superior risk stratification by integrating these interrelated pathways. This is consistent with previous studies demonstrating CTI’s predictive value in hypertensive populations and stroke risk24,25, underscoring its broader applicability beyond single-pathway biomarkers.

Compared with prior research focusing on TyG-related indices and cardiovascular outcomes26,27,28, our study uniquely highlights the prognostic relevance of CTI in a high-risk cardiometabolic population characterized by the coexistence of CHD and T2DM. Notably, in contrast to the non-significant association observed between the TyG index and mortality risk in our cohort, CTI consistently demonstrated a strong relationship with both all-cause and cardiovascular mortality. These findings suggest that incorporating inflammatory status into metabolic risk assessment enhances the identification of vulnerable individuals who may not be flagged by traditional markers alone.

From a clinical perspective, CTI represents a simple, cost-effective, and easily accessible biomarker that could be integrated into routine evaluations to improve risk stratification and guide personalized management strategies in patients with CHD and T2DM. Given its derivation from commonly available laboratory parameters, CTI holds promise for widespread application, particularly in resource-limited settings29,30. Regular monitoring of CTI may facilitate early identification of patients at elevated risk, enabling timely implementation of targeted interventions aimed at mitigating inflammation and metabolic dysfunction.

Despite these strengths, several limitations warrant consideration. First, the observational nature of this study precludes causal inference, and residual confounding cannot be entirely excluded. Second, reliance on self-reported diagnoses for CHD and comorbidities may introduce misclassification bias. Third, CTI was assessed only at baseline, preventing evaluation of longitudinal changes and their potential prognostic implications. Finally, the NHANES cohort primarily represents the U.S. population, which may limit generalizability to other ethnicities or regions.

Conclusions

In summary, elevated CTI is independently associated with higher risks of all-cause and cardiovascular mortality in patients with CHD and T2DM. CTI may serve as a valuable biomarker for improving risk stratification and guiding clinical management in this high-risk group. Future prospective studies are warranted to confirm these findings and explore whether CTI-guided interventions can improve patient outcomes.