Abstract
Objective
Chronic inflammation and metabolic dysregulation are key drivers of cardiovascular disease (CVD). Although the triglyceride-glucose (TyG) index and C-reactive protein (CRP) have each been associated with cardiovascular risk, their combined use as a C-reactive protein-triglyceride-glucose index (CTI) may improve risk prediction. This study aimed to investigate the association between the CTI and CVD, with particular focus on various population subgroups, and to evaluate its predictive capacity for incident CVD.
Methods
This investigation utilized data from the China Health and Retirement Longitudinal Study (CHARLS) cohort, which included 8084 middle-aged and older participants. The association between the CTI and the incidence of CVD was examined using Kaplan-Meier survival analysis, multivariable Cox proportional hazards regression models, and restricted cubic splines (RCS). The predictive performance of CTI was evaluated by receiver operating characteristic (ROC) curve analysis, along with net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Furthermore, subgroup and sensitivity analyses were performed to test the robustness of the results.
Results
During follow-up, 2,133 participants (26.39%) developed CVD. After adjustment for potential confounders, each one-unit increase in the CTI was associated with a 44% higher risk of CVD (HR = 1.44, 95% CI: 1.17–1.78). Compared with the lowest quartile, higher quartiles of CTI were associated with a stepwise increase in risk (HR = 1.23, 1.29, and 1.40; P for trend < 0.001). Subgroup analyses indicated stronger associations among participants aged 45–60 years and those who were married. RCS analyses further supported an overall linear relationship between CTI and CVD risk. Moreover, the CTI index improved reclassification metrics (NRI/IDI) and discriminative ability (AUC). Sensitivity analyses corroborated these primary findings.
Conclusion
The CTI, as a composite biomarker reflecting integrated metabolic and inflammatory status, demonstrated a significant positive association with CVD risk. Furthermore, this association exhibited variations across age groups and marital status, indicating the potential utility of CTI as a novel and clinically relevant biomarker for CVD risk stratification.
Graphical abstract
Supplementary Information
The online version contains supplementary material available at 10.1186/s12933-025-02960-w.
Keywords: C-reactive protein-triglyceride glucose index, Cardiovascular diseases, Inflammation, Insulin resistance
Research insights
What is currently known about this topic?
Chronic inflammation and metabolic dysfunction contribute to CVD. The TyG index and CRP are individually linked to CVD risk, but their combined predictive value has not been fully established and validated.
What is the key research question?
Investigate the relationship between CTI and CVD, focusing on different subgroups of individuals with a newly diagnosed cardiovascular disease, and assess its ability to predict CVD.
What is new?
This study provides large-scale, prospective evidence that the CTI is independently associated with incident CVD. It demonstrates a dose-response relationship, confirms the predictive value through reclassification and discrimination metrics (NRI, IDI, AUC), and identifies population subgroups (middle-aged adults and married individuals) in which the association is particularly pronounced.
How might this study influence clinical practice?
The CTI may serve as a integrative biomarker combining metabolic and inflammatory status, potentially enhancing individualized cardiovascular risk assessment.It may complement existing risk prediction models by capturing both metabolic and inflammatory dimensions of CVD risk.
Introduction
The global population is aging, posing a growing public health challenge, particularly in developing countries where this process is accelerating [1–3]. Within this context of demographic transition, CVD has emerged as the leading cause of functional impairment and mortality among elderly populations [4, 5]. As a chronic non-communicable disease, CVD exhibits a strong positive association with advancing age and places a substantial economic burden on society [6–9]. Epidemiological data indicate a 62.6% increase in global cardiovascular disease-related mortality from 1990 to 2022, with deaths reaching 19.8 million in 2022, solidifying CVD’s position as the predominant cause of death worldwide [10, 11]. Notably, current epidemiological models project a 90.0% increase in global CVD prevalence from 2025 to 2050. Concurrently, the estimated number of Cardiovascular Disease-related mortality is projected to rise from 20.5 million in 2025 to 35.6 million by 2050 [12]. Given this escalating public health burden, the early identification of high-risk populations and the timely implementation of interventions represent a crucial preventive strategy for controlling CVD risk factors and preventing disease progression [4, 13–15].
IR is a metabolic disorder characterized by reduced insulin sensitivity and responsiveness [16, 17]. It contributes significantly to CVD through mechanisms including disruption of glucose metabolism, increased free radical generation, activation of the mitochondrial electron transport chain, and elevated production of reactive oxygen species (ROS) [18]. Evidence establishes the TyG index as a well-validated biomarker for assessing IR with high sensitivity and specificity, demonstrating significant predictive value for CVD [1, 19–21]. Inflammatory responses mediate pathological processes underlying CVD, including atherosclerosis, myocardial infarction, and heart failure [22, 23]. Accumulating evidence over the past decade implicates inflammatory vesicle activation in CVD pathogenesis [24–26]. The centrality of inflammation in CVD is well-established, as is the clinical value of CRP as a sensitive systemic inflammatory marker [27–29]. Given the established pathophysiological links between IR, inflammation, and CVD, a composite index integrating these pathways offers substantial clinical value. Ruan et al. proposed the CTI, integrating CRP (inflammation) and TyG (IR) indices, which demonstrated utility in predicting cancer patient survival [30]. Nevertheless, the mechanistic relationship between CTI and CVD risk remains unelucidated. Recent studies have indicated that the CTI, a composite marker reflecting inflammation and insulin resistance, may serve as an effective tool for predicting CVD outcomes. For instance, Sun et al. conducted a retrospective cohort study using data from the National Health and Nutrition Examination Survey (NHANES) and reported a strong association between CTI and both CVD incidence and mortality [31]. Xu and colleagues further confirmed its utility in detecting coronary heart disease prevalence within the NHANES adolescent population [32]. Chen et al. demonstrated the predictive value of CTI for incident CVD in metabolically heterogeneous populations [33], while other studies highlighted its prognostic significance in patients with cardio-renal-metabolic syndrome [34]. Ma and colleagues investigated cumulative CTI exposure and longitudinal changes in a middle-aged and older Chinese cohort [35], and Shi et al. validated its predictive relevance for mortality among individuals with diabetes or prediabetes [36].
Although these studies have provided valuable insights, most were based on retrospective or cross-sectional analyses, primarily focused on mortality rather than incident CVD, and did not comprehensively evaluate population subgroups. Moreover, the predictive value of the CTI across different age groups and glycemic states in nationally representative community cohorts remains poorly understood. To address this gap, we conducted a prospective cohort study using data from the CHARLS, which included over 8,000 middle-aged and older adults with nearly nine years of follow-up. This study aimed to assess the association between baseline CTI and CVD risk and to explore whether this association varies according to age, marital status, gender, and glycemic status.
Methods
Study design
Data for this study were derived from the CHARLS, a nationally representative longitudinal cohort of Chinese adults aged ≥ 45 years. Conducted in multiple waves between 2011 and 2020, CHARLS employed multistage stratified probability sampling to recruit participants across 150 county-level administrative units (districts/counties) within 28 Chinese provinces. Comprehensive descriptions of the study protocol and cohort characteristics are documented in prior publications. The study protocol received ethical approval from the Peking University Institutional Review Board and conformed to the Declaration of Helsinki. All participants provided written informed consent prior to enrollment. Within the CHARLS framework, field interviewers received standardized training; data collection occurred through face-to-face interviews using structured questionnaires.
This investigation utilized the 2011 baseline survey with longitudinal follow-up through 2013, 2015, 2018and 2020. Figure 1 delineates the participant selection workflow. From an initial cohort of 17,708 participants, exclusions comprised: 777 individuals aged < 45 years; 5,692 with missing baseline CRP measurements; 9 lacking triglyceride (TG) data; 16 without fasting blood glucose (FBG) values; 100 with unavailable glycated hemoglobin (HbA1c) data; 1,807 exhibiting pre-existing CVD at baseline; and 1223 lost to follow-up. Consequently, the analytical cohort comprised 8084 participants.
Fig. 1.
Flow chart of the study population
Calculation of CTI
The CTI index was obtained by using the following formula [30]: CTI = 0.412 × lg (CRP [mg/L]) + lg (TG [mg/dl] × FPG [mg/dl])/2.
Assessment of incident CVD
The primary outcome measures comprised incident CVD and time-to-event duration during longitudinal follow-up (Waves 2–5). Referring to prior research [37], information regarding CVD diagnosis was obtained using standardized questions: (1) “Have you ever been diagnosed by a physician with heart disease (e.g., myocardial infarction, coronary heart disease, angina pectoris, congestive heart failure, or other cardiac conditions) or stroke?” and (2) “Are you currently taking any prescribed medication for the management of heart disease or stroke?” Participants who answered affirmatively to either question during the follow-up period were classified as having experienced a CVD event. The CHARLS research team implemented rigorous data recording and validation protocols to ensure data reliability.
Assessments of covariates
Baseline data were systematically collected by trained interviewers using structured questionnaires, encompassing four domains: (1) Demographic and lifestyle characteristics: characteristics, including gender, age, geographic region, educational attainment, marital status, tobacco use, and alcohol consumption; (2) Anthropometric assessments assessments, which height, body weight, waist circumference, body mass index (BMI), and resting systolic and diastolic blood pressure (SBP and DBP); (3) Disease and Medication medication history, covering medical conditions, conditions as hypertension, diabetes, dyslipidemia, liver disease, kidney disease, lung disease, and gastrointestinal disorders, as well as the used to treat hypertension, dyslipidemia, and diabetes; diabetes; (4) Laboratory parameters: CRP, parameters, including CRP, glucose (FPG), total cholesterol (TC), TG, high-density triglycerides (TG), cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and serum uric acid (UA).
Definitions
Hypertension is defined by the fulfillment of one or more of the following criteria: (1) systolic blood pressure ≥ 140 mmHg; (2) diastolic blood pressure ≥ 90 mmHg; (3) self-reported hypertension diagnosed by a physician; or (4) current use of antihypertensive medication. Diabetes mellitus (DM) is characterized by meeting at least one of the following: (1) fasting plasma glucose (FPG) ≥ 126 mg/dL; (2) glycated hemoglobin (HbA1c) ≥ 6.5%; (3) current use of antidiabetic medication; or (4) self-reported physician-diagnosed diabetes. Hyperglycemia is defined as an FPG level ≥ 100 mg/dL or an HbA1c level ≥ 5.7%. Conversely, normal blood glucose is defined as an FPG < 100 mg/dL and an HbA1c < 5.7%. Dyslipidemia is identified by the presence of one or more of the following: triglycerides (TG) ≥ 150 mg/dL, total cholesterol (TC) ≥ 240 mg/dL, high-density lipoprotein cholesterol (HDL-C) < 40 mg/dL, low-density lipoprotein cholesterol (LDL-C) ≥ 160 mg/dL, current use of lipid-lowering medications, or self-reported dyslipidemia diagnosed by a physician. Liver, kidney, lung, and gastrointestinal diseases are defined based on either a documented physician diagnosis or current use of medications specifically prescribed for these conditions.
Statistical analysis
The extent of missing data in this study is detailed in Table S1. To address missing values and mitigate potential bias, multiple imputation was employed (5 imputation cycles, 100 iterations). Normally distributed quantitative variables are presented as mean ± standard error and were compared using analysis of variance (ANOVA). Non-normally distributed quantitative variables are characterized by median and interquartile range, with between-group comparisons conducted via the Kruskal-Wallis test. Categorical variables are expressed as frequency (proportion) and were analyzed using the chi-square test.
The cumulative incidence of CVD was estimated using Kaplan-Meier survival curves, with differences assessed by the log-rank test. Multicollinearity among covariates was examined using variance inflation factors (VIF), and variables with VIF ≥ 5 were excluded from the multivariable models (Supplementary Table S2). To investigate the association between CTI and CVD, three Cox proportional hazards models were constructed: Model 1 was unadjusted; Model 2 was adjusted for demographic and lifestyle factors, including gender, age, residence, marital status, educational level, smoking status, and alcohol consumption; and Model 3 was further adjusted for clinical and metabolic risk factors, including body mass index, waist circumference, systolic and diastolic blood pressure, total cholesterol, LDL cholesterol, HDL cholesterol, uric acid, hypertension, diabetes, dyslipidemia, liver disease, kidney disease, lung disease, gastrointestinal disease, antihypertensive medication use, antidiabetic medication use, dyslipidemia, glycemic status, and geographic region. We also conducted fully adjusted Restricted Cubic Spline (RCS) analyses to explore the dose-response relationship between CTI and CVD risk. If the analysis revealed a significant non-linear relationship, the data were divided into two groups based on the inflection point for separate examination of the association between CTI and CVD. Furthermore, the Likelihood Ratio Test (LRT) was applied to evaluate potential interactions between CTI and covariates. To further assess the relative explanatory power, separate Cox regression models were constructed for the TyG index and CRP, enabling direct comparison with CTI in relation to CVD risk. ROC curves were generated to examine the predictive performance of CTI, TyG, and CRP for incident CVD, and the NRI and the comprehensive IDI was calculated to quantify the incremental predictive value of CTI.
Subgroup analyses were performed according to glycemic status (hyperglycemia and normoglycemia), age (45–60 years and ≥ 60 years), gender, residence, marital status, smoking, alcohol consumption, body mass index (BMI; <18.5, 18.5–25, and ≥ 25 kg/m²), waist circumference (abdominal obesity defined as ≥ 85 cm in men and ≥ 80 cm in women; normal defined as < 85 cm in men and < 80 cm in women), hypertension, diabetes, dyslipidemia, liver disease, kidney disease, lung disease, gastrointestinal disease, use of antihypertensive agents, antidiabetic agents, lipid-lowering agents, and geographic region. In addition, stratified analyses were conducted by age, gender, glycemic status, and variables that exhibited significant interactions in the subgroup analyses. To further assess the robustness of our findings, three sensitivity analyses were performed: (1) exclusion of participants with missing values; (2) exclusion of participants who died during follow-up; and (3) reassessment of the associations using logistic regression models.
All data in this study were processed and analyzed using R software (version 4.5.1). P < 0.05 was considered statistically significant.
Results
Baseline characteristics of participants
This study included 8,084 participants from the CHARLS, with a mean age of 58.05 ± 8.95 years; 4,282 participants (52.97%) were female. Participants in the highest quartile of CTI were more likely to be female, urban residents, individuals with a university education or higher, and former smokers. They also had a higher prevalence of hypertension, diabetes, dyslipidemia, and pulmonary diseases. In contrast, participants in the lowest quartile of CTI demonstrated a higher prevalence of gastrointestinal diseases. With respect to anthropometric and laboratory parameters, individuals in the upper quartile of CTI exhibited higher levels of blood glucose, BMI, body weight, waist circumference, systolic and diastolic blood pressure, C-reactive protein, fasting plasma glucose, total cholesterol, triglycerides, and uric acid. Conversely, higher CTI was associated with lower levels of high-density lipoprotein cholesterol (HDL-C). The demographic and clinical characteristics of the study population are summarized in Table 1.
Table 1.
Patient demographics and baseline characteristics
| Characteristic | Overall N = 8,084 | Q1 N = 2,021 | Q2 N = 2,021 | Q3 N = 2,021 | Q4 N = 2,021 | p-value |
|---|---|---|---|---|---|---|
| Age | 58.05 ± 8.95 | 56.69 ± 8.75 | 58.09 ± 8.86 | 58.61 ± 9.00 | 58.80 ± 9.05 | < 0.001 |
| Gender | 0.5 | |||||
| Female | 4,282 (52.97%) | 1,057 (52.30%) | 1,055 (52.20%) | 1,099 (54.38%) | 1,071 (52.99%) | |
| Male | 3,802 (47.03%) | 964 (47.70%) | 966 (47.80%) | 922 (45.62%) | 950 (47.01%) | |
| Residence | < 0.001 | |||||
| Rural | 6,828 (84.46%) | 1,762 (87.18%) | 1,732 (85.70%) | 1,690 (83.62%) | 1,644 (81.35%) | |
| Urban | 1,256 (15.54%) | 259 (12.82%) | 289 (14.30%) | 331 (16.38%) | 377 (18.65%) | |
| Marital status | 0.12 | |||||
| Married | 7,221 (89.32%) | 1,821 (90.10%) | 1,822 (90.15%) | 1,794 (88.77%) | 1,784 (88.27%) | |
| Other | 863 (10.68%) | 200 (9.90%) | 199 (9.85%) | 227 (11.23%) | 237 (11.73%) | |
| Education | 0.7 | |||||
| College or above | 90 (1.11%) | 19 (0.94%) | 20 (0.99%) | 25 (1.24%) | 26 (1.29%) | |
| High/Vocational school | 759 (9.39%) | 196 (9.70%) | 195 (9.65%) | 175 (8.66%) | 193 (9.55%) | |
| Less than primary school | 1,501 (18.57%) | 361 (17.86%) | 390 (19.30%) | 396 (19.59%) | 354 (17.52%) | |
| Middle school | 1,653 (20.45%) | 437 (21.62%) | 384 (19.00%) | 423 (20.93%) | 409 (20.24%) | |
| No formal education | 2,343 (28.98%) | 573 (28.35%) | 599 (29.64%) | 575 (28.45%) | 596 (29.49%) | |
| Primary school | 1,738 (21.50%) | 435 (21.52%) | 433 (21.43%) | 427 (21.13%) | 443 (21.92%) | |
| Smoking status | 0.018 | |||||
| Current | 2,498 (30.90%) | 630 (31.17%) | 648 (32.06%) | 596 (29.49%) | 624 (30.88%) | |
| Former | 611 (7.56%) | 126 (6.23%) | 139 (6.88%) | 166 (8.21%) | 180 (8.91%) | |
| Never | 4,975 (61.54%) | 1,265 (62.59%) | 1,234 (61.06%) | 1,259 (62.30%) | 1,217 (60.22%) | |
| Drinking status | 0.051 | |||||
| Current | 2,580 (31.91%) | 684 (33.84%) | 665 (32.90%) | 593 (29.34%) | 638 (31.57%) | |
| Former | 610 (7.55%) | 138 (6.83%) | 147 (7.27%) | 170 (8.41%) | 155 (7.67%) | |
| Never | 4,894 (60.54%) | 1,199 (59.33%) | 1,209 (59.82%) | 1,258 (62.25%) | 1,228 (60.76%) | |
| BMI | 23.68 ± 11.18 | 22.23 ± 3.33 | 23.40 ± 16.29 | 24.38 ± 14.23 | 24.72 ± 4.14 | < 0.001 |
| Height | 1.58 ± 0.09 | 1.58 ± 0.09 | 1.58 ± 0.09 | 1.58 ± 0.10 | 1.58 ± 0.09 | 0.3 |
| Weight | 58.75 ± 11.50 | 55.56 ± 10.07 | 57.22 ± 11.00 | 59.98 ± 11.14 | 62.24 ± 12.50 | < 0.001 |
| WC | 83.88 ± 12.39 | 79.93 ± 10.80 | 82.02 ± 11.66 | 85.47 ± 12.58 | 88.11 ± 12.81 | < 0.001 |
| SBP | 128.01 ± 20.76 | 124.11 ± 20.18 | 125.65 ± 19.63 | 129.89 ± 20.89 | 132.38 ± 21.27 | < 0.001 |
| DBP | 74.79 ± 11.93 | 72.73 ± 11.81 | 73.51 ± 11.63 | 75.76 ± 11.87 | 77.15 ± 11.88 | < 0.001 |
| CRP | 2.63 ± 7.29 | 0.46 ± 0.26 | 0.87 ± 0.49 | 1.68 ± 1.11 | 7.51 ± 13.36 | < 0.001 |
| FPG | 109.50 ± 35.59 | 98.03 ± 15.39 | 102.92 ± 19.27 | 107.53 ± 23.28 | 129.53 ± 57.78 | < 0.001 |
| HbA1c | 5.25 ± 0.78 | 5.07 ± 0.46 | 5.14 ± 0.52 | 5.22 ± 0.66 | 5.57 ± 1.17 | < 0.001 |
| TG | 133.19 ± 109.93 | 75.99 ± 28.35 | 105.40 ± 42.43 | 136.41 ± 65.92 | 214.96 ± 175.09 | < 0.001 |
| TC | 193.13 ± 38.25 | 184.08 ± 33.55 | 190.93 ± 35.54 | 195.91 ± 36.74 | 201.59 ± 44.14 | < 0.001 |
| HDL | 51.26 ± 15.32 | 59.58 ± 15.02 | 53.66 ± 14.11 | 48.93 ± 13.38 | 42.86 ± 13.60 | < 0.001 |
| LDL | 115.90 ± 34.61 | 111.58 ± 29.53 | 118.22 ± 32.26 | 120.48 ± 34.33 | 113.33 ± 40.62 | < 0.001 |
| Uric Acid | 4.43 ± 1.23 | 4.10 ± 1.11 | 4.27 ± 1.14 | 4.56 ± 1.22 | 4.77 ± 1.33 | < 0.001 |
| Blood glucose status | < 0.001 | |||||
| Hyperglycemia | 4,761 (58.89%) | 839 (41.51%) | 1,074 (53.14%) | 1,277 (63.19%) | 1,571 (77.73%) | |
| Normal Blood Glucose | 3,323 (41.11%) | 1,182 (58.49%) | 947 (46.86%) | 744 (36.81%) | 450 (22.27%) | |
| Hypertension | 2,884 (35.68%) | 514 (25.43%) | 623 (30.83%) | 789 (39.04%) | 958 (47.40%) | < 0.001 |
| Diabetes | 1,275 (15.77%) | 115 (5.69%) | 205 (10.14%) | 291 (14.40%) | 664 (32.86%) | < 0.001 |
| Dyslipidemia | 7,559 (93.51%) | 1,794 (88.77%) | 1,886 (93.32%) | 1,930 (95.50%) | 1,949 (96.44%) | < 0.001 |
| Liver disease | 267 (3.30%) | 78 (3.86%) | 67 (3.32%) | 63 (3.12%) | 59 (2.92%) | 0.4 |
| Kidney disease | 465 (5.75%) | 121 (5.99%) | 116 (5.74%) | 108 (5.34%) | 120 (5.94%) | 0.8 |
| Lung disease | 806 (9.97%) | 169 (8.36%) | 204 (10.09%) | 199 (9.85%) | 234 (11.58%) | 0.008 |
| Digestive disease | 1,973 (24.41%) | 579 (28.65%) | 494 (24.44%) | 458 (22.66%) | 442 (21.87%) | < 0.001 |
| Meds for hypertension | 1,264 (15.64%) | 180 (8.91%) | 237 (11.73%) | 365 (18.06%) | 482 (23.85%) | < 0.001 |
| Meds for diabetes | 240 (2.97%) | 23 (1.14%) | 29 (1.43%) | 54 (2.67%) | 134 (6.63%) | < 0.001 |
| Meds for dyslipidemia | 276 (3.41%) | 29 (1.43%) | 45 (2.23%) | 80 (3.96%) | 122 (6.04%) | < 0.001 |
| Region of China | 0.024 | |||||
| Central China | 2,423 (29.97%) | 547 (27.07%) | 633 (31.32%) | 636 (31.47%) | 607 (30.03%) | |
| East China | 2,504 (30.97%) | 679 (33.60%) | 608 (30.08%) | 610 (30.18%) | 607 (30.03%) | |
| Northeastern China | 454 (5.62%) | 123 (6.09%) | 98 (4.85%) | 121 (5.99%) | 112 (5.54%) | |
| Western China | 2,703 (33.44%) | 672 (33.25%) | 682 (33.75%) | 654 (32.36%) | 695 (34.39%) | |
| CVD | 2,133 (26.39%) | 406 (20.09%) | 516 (25.53%) | 584 (28.90%) | 627 (31.02%) | < 0.001 |
Association between the CTI and the CVD incidence
During a mean follow-up of nine years, 2,133 participants (26.39%) developed incident CVD, with regional distributions as follows: 176 cases (38.77%) in northeastern China, 687 cases (28.35%) in central China, 683 cases (25.27%) in western China, and 587 cases (23.44%) in eastern China (Fig. 2). Across CTI quartiles, the incidence of CVD increased progressively, with 406 cases (20.09%) in Q1, 516 cases (25.53%) in Q2, 584 cases (28.90%) in Q3, and 627 cases (31.02%) in Q4. Kaplan-Meier survival analysis further demonstrated a significant increase in cumulative CVD incidence with higher CTI quartiles (log-rank test, P < 0.001; Fig. 3). To evaluate the association between CTI and CVD risk, three Cox proportional hazards models were constructed. After adjusting for potential confounders in Model 3, a one-unit increase in CTI was associated with a 44% higher risk of CVD (HR = 1.44, 95% CI: 1.17–1.78). This risk magnitude exceeded that of the TyG index (HR = 1.36, 95% CI: 1.14–1.62) and was greater than that of CRP, for which no significant association with increased CVD risk was observed (HR = 1.00, 95% CI: 1.00-1.01) (Table 2; Supplementary Table S3). Compared with Q1, participants in Q2, Q3, and Q4 had progressively higher risks of CVD, with HRs of 1.23 (95% CI: 1.08–1.41), 1.29 (95% CI: 1.13–1.48), and 1.40 (95% CI: 1.20–1.62), respectively (P for trend < 0.001; Table 2). ROC curve analysis indicated that CTI exhibited the highest predictive accuracy for incident CVD, with AUC values of 0.61 at 2 years, 0.56 at 4 years, 0.56 at 7 years, and 0.57 at 9 years. In comparison, the AUC values for CRP were 0.59, 0.55, 0.55, and 0.55, and those for the TyG index were 0.57, 0.54, 0.54, and 0.56 at the corresponding time points (Supplementary Figure S1). NRI and IDI analyses further confirmed that CTI provided incremental predictive value over CRP and TyG (NRI > 0; IDI>0;Supplementary Figure S2). Finally, RCS analysis demonstrated a significant positive linear association between CTI and CVD risk (P for nonlinearity = 0.108; Fig. 4). These results collectively suggest that CTI is a more effective predictor of CVD risk than either CRP or the TyG index.
Fig. 2.
Prevalence of CVD by province
Fig. 3.
Kaplan-Meier curve analysis demonstrating the cumulative incidence of CVD in different CTI quartile groups
Table 2.
Association between CTI and CVD incidence
| Characteristic | Event(%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | HR(95%CI) | p | ||
| CTI(per 1 unit) | 2133(26.39%) | 1.97(1.68–2.31) | < 0.001 | 1.79(1.52–2.10) | < 0.001 | 1.44(1.17–1.78) | < 0.001 |
| CTI quartile | |||||||
| Q1 | 406(20.09%) | Ref | Ref | Ref | |||
| Q2 | 516(25.53%) | 1.32(1.15–1.50) | < 0.001 | 1.26(1.10–1.43) | < 0.001 | 1.23(1.08–1.41) | 0.002 |
| Q3 | 584(28.90%) | 1.52(1.34–1.72) | < 0.001 | 1.41(1.24–1.60) | < 0.001 | 1.29(1.13–1.48) | < 0.001 |
| Q4 | 627(31.02%) | 1.72(1.52–1.95) | < 0.001 | 1.59(1.41–1.81) | < 0.001 | 1.40(1.20–1.62) | < 0.001 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
Model 1: unadjusted for any covariates; Model 2: adjusted for gender, age, residence, marital status, education level, smoking status, and drinking status; Model 3: based on model 2, BMI, WC, SBP, DBP, TC, LDL, HDL, uric acid, hypertension, diabetes, dyslipidemia, liver diseases, kidney diseases, lung diseases, digestive diseases, hypertension drugs, diabetes drugs, dyslipidemia drugs, blood sugar status, and geographical area were adjusted
HR Hazard Ratio, CI Confidenlce Interval
Fig. 4.
RCS Analysis of CTI and CVD Incidence
Subgroup analysis
To further investigate the association between the CTI and CVD risk, subgroup analyses were performed. The results demonstrated that elevated CTI was consistently associated with an increased risk of CVD across various strata, including gender, glycemic status, hypertension, kidney disease, smoking status, alcohol consumption, body mass index (BMI ≥ 18.5 kg/m²), abdominal obesity, dyslipidemia, absence of diabetes, liver disease, lung disease, or digestive system diseases, as well as non-use of antihypertensive, antidiabetic, or lipid-lowering medications. Importantly, significant interactions were observed between CTI and age (interaction P = 0.005) and marital status (interaction P = 0.040), indicating that these factors may modify the effect of CTI on CVD risk (Fig. 5). The likelihood ratio test (LRT) revealed significant interactions between CTI and age (χ²= 14.54, P < 0.01) as well as between CTI and marital status (χ² = 4.30, p = 0.038) (Supplementary Figure S4).
Fig. 5.
Subgroup risk forest plot
Association between CTI and CVD incidence by glycemic status
During follow-up, 1,327 participants (27.87%) with hyperglycemia and 806 participants (24.26%) with normal blood glucose developed incident CVD. Kaplan-Meier analyses demonstrated significant differences in cumulative CVD incidence across the four CTI quartiles within both glycemic status groups (log-rank P < 0.05; Fig. 6). In the fully adjusted Model 3, each one-unit increase in CTI was associated with a 34% higher risk of CVD among participants with hyperglycemia (HR = 1.34, 95% CI: 1.03–1.74) and a 69% higher risk among participants with normal blood glucose (HR = 1.69, 95% CI: 1.17–2.45) (Table 3). RCS analyses indicated a significant positive linear association between CTI and CVD incidence in both groups (P < 0.05; P for nonlinearity > 0.05; Fig. 7).
Fig. 6.
Kaplan-Meier curve analysis demonstrating the cumulative incidence of CVD in different CTI quartile groups with different glycemic states. A participants with DM; B participants with NGR
Table 3.
Association between CTI and CVD incidence by glycemic status
| Blood glucose status | Characteristic | Event(%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | HR(95%CI) | p | |||
| Hyperglycemia | CTI(per 1 unit) | 1327(27.87%) | 1.80(1.48–2.20) | < 0.001 | 1.68(1.37–2.05) | < 0.001 | 1.34(1.03–1.74) | 0.027 |
| Hyperglycemia | CTI quartile | |||||||
| Hyperglycemia | Q1 | 165(19.67%) | Ref | Ref | Ref | |||
| Hyperglycemia | Q2 | 285(26.54%) | 1.38(1.14–1.68) | < 0.001 | 1.31(1.08–1.59) | 0.005 | 1.31(1.08–1.60) | 0.006 |
| Hyperglycemia | Q3 | 389(30.46%) | 1.63(1.36–1.96) | < 0.001 | 1.51(1.26–1.82) | < 0.001 | 1.42(1.17–1.72) | < 0.001 |
| Hyperglycemia | Q4 | 488(31.06%) | 1.75(1.46–2.08) | < 0.001 | 1.62(1.36–1.94) | < 0.001 | 1.42(1.16–1.75) | < 0.001 |
| Hyperglycemia | P for trend | < 0.001 | < 0.001 | 0.006 | ||||
| Normal Blood Glucose | CTI(per 1 unit) | 806(24.26%) | 2.24(1.64–3.05) | < 0.001 | 1.94(1.42–2.66) | < 0.001 | 1.69(1.17–2.45) | 0.005 |
| Normal Blood Glucose | CTI quartile | |||||||
| Normal Blood Glucose | Q1 | 241(20.39%) | Ref | Ref | Ref | |||
| Normal Blood Glucose | Q2 | 231(24.39%) | 1.25(1.04–1.49) | 0.016 | 1.20(1.00-1.44) | 0.049 | 1.17(0.97–1.42) | 0.093 |
| Normal Blood Glucose | Q3 | 195(26.21%) | 1.35(1.12–1.63) | 0.002 | 1.25(1.04–1.51) | 0.020 | 1.17(0.95–1.44) | 0.145 |
| Normal Blood Glucose | Q4 | 139(30.89%) | 1.72(1.40–2.12) | < 0.001 | 1.60(1.30–1.98) | < 0.001 | 1.53(1.20–1.94) | < 0.001 |
| Normal Blood Glucose | P for trend | < 0.001 | < 0.001 | 0.001 | ||||
Model 1: unadjusted for any covariates; Model 2: adjusted for gender, age, residence, marital status, education level, smoking status, and drinking status; Model 3: based on model 2, BMI, WC, SBP, DBP, TC, LDL, HDL, uric acid, hypertension, diabetes, dyslipidemia, liver diseases, kidney diseases, lung diseases, digestive diseases, hypertension drugs, diabetes drugs, dyslipidemia drugs, and geographical area were adjusted
HR Hazard Ratio, CI Confidence Interval
Fig. 7.
RCS analysis between CTI and CVD incidence by glycemic status. A participants with DM; B participants with NGR
Association between CTI and CVD incidence by gender
During follow-up, 939 male participants (24.70%) and 1,194 female participants (27.88%) developed incident CVD. Kaplan-Meier analyses demonstrated significant differences in cumulative CVD incidence across the four CTI quartiles for both sexes (log-rank P < 0.01; Fig. 8). In the fully adjusted Model 3, CTI was significantly associated with CVD risk in both male and female participants. Specifically, each one-unit increase in CTI was associated with a 48% higher risk of CVD in males (HR = 1.48, 95% CI: 1.08–2.03) and a 46% higher risk in females (HR = 1.46, 95% CI: 1.10–1.95) (Table 4). RCS analyses revealed a positive, albeit non-significant, association between CTI and CVD in males (P > 0.05; P for nonlinearity > 0.05). In contrast, a significant nonlinear relationship was observed in females (P < 0.05; P for nonlinearity < 0.05; Fig. 9). Using a cutoff value of 2.038 in females, participants were stratified into two subgroups to further examine the relationship between CTI and CVD risk. The results indicated a significant association in participants below the cutoff (P < 0.05), whereas no significant association was observed in those at or above the cutoff (P > 0.05) (Supplementary Table S5).
Fig. 8.
Kaplan-Meier curve analysis demonstrating the cumulative incidence of CVD in different CTI quartile groups by gender. (A) male; (B) female
Table 4.
Association between CTI and incidence of CVD by gender
| Gender | Characteristic | Event(%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | HR(95%CI) | p | |||
| Male | CTI(per 1 unit) | 939(24.70%) | 1.88(1.48–2.40) | < 0.001 | 1.91(1.49–2.44) | < 0.001 | 1.48(1.08–2.03) | 0.016 |
| Male | CTI quartile | |||||||
| Male | Q1 | 184(19.09%) | Ref | Ref | Ref | |||
| Male | Q2 | 240(24.84%) | 1.35(1.11–1.63) | 0.002 | 1.31(1.08–1.58) | 0.007 | 1.31(1.07–1.59) | 0.007 |
| Male | Q3 | 241(26.14%) | 1.45(1.20–1.76) | < 0.001 | 1.40(1.15–1.69) | < 0.001 | 1.23(1.00-1.51) | 0.045 |
| Male | Q4 | 274(28.84%) | 1.70(1.41–2.05) | < 0.001 | 1.68(1.39–2.03) | < 0.001 | 1.43(1.15–1.78) | 0.002 |
| Male | P for trend | < 0.001 | < 0.001 | 0.005 | ||||
| Female | CTI(per 1 unit) | 1194(27.88%) | 2.03(1.65–2.51) | < 0.001 | 1.72(1.39–2.14) | < 0.001 | 1.46(1.10–1.95) | 0.010 |
| Female | CTI quartile | |||||||
| Female | Q1 | 222(21.00%) | Ref | Ref | Ref | |||
| Female | Q2 | 276(26.16%) | 1.29(1.08–1.54) | 0.005 | 1.22(1.03–1.46) | 0.025 | 1.19(0.99–1.42) | 0.067 |
| Female | Q3 | 343(31.21%) | 1.55(1.31–1.84) | < 0.001 | 1.43(1.20–1.69) | < 0.001 | 1.36(1.13–1.64) | 0.001 |
| Female | Q4 | 353(32.96%) | 1.73(1.46–2.04) | < 0.001 | 1.53(1.29–1.81) | < 0.001 | 1.40(1.14–1.73) | 0.001 |
| Female | P for trend | < 0.001 | < 0.001 | 0.002 | ||||
Model 1: unadjusted for any covariates; Model 2: adjusted for age, residence, marital status, education level, smoking status, and drinking status; Model 3: based on model 2, BMI, WC, SBP, DBP, TC, LDL, HDL, uric acid, hypertension, diabetes, dyslipidemia, liver diseases, kidney diseases, lung diseases, digestive diseases, hypertension drugs, diabetes drugs, dyslipidemia drugs, blood sugar status, and geographical area were adjusted
HR Hazard Ratio, CI Confidence Interval
Fig. 9.
RCS analysis between CTI and CVD incidence by gender. A male; B female
The association between CTI and the CVD incidence by age
During follow-up, 1,107 participants (22.42%) aged 45–60 and 1,026 participants (32.60%) aged 60 years or older developed incident CVD. Kaplan-Meier analyses demonstrated significant differences in cumulative CVD incidence across the four CTI quartiles within the different age groups (log-rank P < 0.01; Fig. 10). In the fully adjusted Model 3, a significant association between CTI and CVD risk was observed only among participants aged 45–60 years (P < 0.001). Specifically, each one-unit increase in CTI was associated with a 64% higher risk of CVD in this age group (HR = 1.64, 95% CI: 1.22–2.21; Table 5). RCS analyses confirmed a significant positive linear association between CTI and CVD occurrence in participants aged 45–60 years (P < 0.05; P for nonlinearity > 0.05; Fig. 11).
Fig. 10.
Kaplan-Meier curve analysis shows the cumulative incidence of CVD in different age groups and CTI quartile group. A age 45–60; B age ≥ 60
Table 5.
Association between CTI and incidence of CVD by age
| Age | Characteristic | Event(%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | HR(95%CI) | p | |||
| 45–60 | CTI(per 1 unit) | 1107(22.42%) | 2.24(1.80–2.78) | < 0.001 | 2.25(1.81–2.79) | < 0.001 | 1.64(1.22–2.21) | < 0.001 |
| 45–60 | CTI quartile | |||||||
| 45–60 | Q1 | 219(16.27%) | Ref | Ref | Ref | |||
| 45–60 | Q2 | 269(21.94%) | 1.39(1.17–1.67) | < 0.001 | 1.39(1.16–1.66) | < 0.001 | 1.30(1.08–1.56) | 0.005 |
| 45–60 | Q3 | 289(24.53%) | 1.57(1.32–1.87) | < 0.001 | 1.56(1.31–1.86) | < 0.001 | 1.34(1.11–1.62) | 0.003 |
| 45–60 | Q4 | 330(27.80%) | 1.87(1.58–2.22) | < 0.001 | 1.87(1.57–2.22) | < 0.001 | 1.49(1.21–1.84) | < 0.001 |
| 45–60 | P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
| ≥ 60 | CTI(per 1 unit) | 1026(32.60%) | 1.56(1.23–1.98) | < 0.001 | 1.46(1.15–1.86) | 0.002 | 1.27(0.94–1.74) | 0.123 |
| ≥ 60 | CTI quartile | |||||||
| ≥ 60 | Q1 | 187(27.70%) | Ref | Ref | Ref | |||
| ≥ 60 | Q2 | 247(31.07%) | 1.15(0.95–1.39) | 0.157 | 1.13(0.93–1.36) | 0.218 | 1.16(0.96–1.42) | 0.126 |
| ≥ 60 | Q3 | 295(34.99%) | 1.32(1.10–1.59) | 0.003 | 1.27(1.06–1.53) | 0.010 | 1.24(1.02–1.51) | 0.032 |
| ≥ 60 | Q4 | 297(35.61%) | 1.44(1.20–1.73) | < 0.001 | 1.37(1.14–1.65) | < 0.001 | 1.30(1.04–1.61) | 0.019 |
| ≥ 60 | P for trend | < 0.001 | < 0.001 | 0.025 | ||||
Model 1: unadjusted for any covariates; Model 2: adjusted for gender, residence, marital status, education level, smoking status, and drinking status; Model 3: based on model 2, BMI, WC, SBP, DBP, TC, LDL, HDL, uric acid, hypertension, diabetes, dyslipidemia, liver diseases, kidney diseases, lung diseases, digestive diseases, hypertension drugs, diabetes drugs, dyslipidemia drugs, blood sugar status, and geographical area were adjusted
HR Hazard Ratio, CI Confidence Interval
Fig. 11.
RCS analysis between CTI and CVD incidence at different ages. A age 45–60; B age ≥ 60
Association between CTI and CVD incidence according to marital status
During follow-up, 1,870 married participants (25.90%) and 263 unmarried or divorced participants (30.48%) developed incident CVD. Kaplan-Meier analyses demonstrated significant differences in cumulative CVD incidence across the four CTI quartiles only among married participants (log-rank P < 0.05; Fig. 12). In the fully adjusted Model 3, a significant association between CTI and CVD risk was observed exclusively in married participants. Specifically, each one-unit increase in CTI was associated with a 51% higher risk of CVD in this group (HR = 1.51, 95% CI: 1.20–1.90; Table 6). RCS analyses confirmed a significant positive linear association between CTI and CVD incidence among married participants (P < 0.05; P for nonlinearity > 0.05; Fig. 13).
Fig. 12.
Kaplan-Meier analysis of cumulative CVD incidence across CTI quartiles stratified by marital status. Panel (A) represents married participants, and Panel (B) represents other participants
Table 6.
Association between CTI and incidence of cardiovascular disease stratified by marital status
| Marital status | Characteristic | Event(%) | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | HR(95%CI) | p | |||
| Married | CTI(per 1 unit) | 1870(25.90%) | 2.09(1.76–2.47) | < 0.001 | 1.96(1.65–2.32) | < 0.001 | 1.51(1.20–1.90) | < 0.001 |
| Married | CTI quartile | |||||||
| Married | Q1 | 355(19.49%) | Ref | Ref | Ref | |||
| Married | Q2 | 449(24.64%) | 1.30(1.13–1.50) | < 0.001 | 1.24(1.08–1.43) | 0.002 | 1.20(1.04–1.38) | 0.014 |
| Married | Q3 | 509(28.37%) | 1.53(1.33–1.75) | < 0.001 | 1.43(1.25–1.64) | < 0.001 | 1.28(1.11–1.48) | < 0.001 |
| Married | Q4 | 557(31.22%) | 1.77(1.55–2.02) | < 0.001 | 1.67(1.46–1.91) | < 0.001 | 1.41(1.20–1.65) | < 0.001 |
| Married | P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
| Other | CTI(per 1 unit) | 263(30.48%) | 1.27(0.80–2.02) | 0.315 | 1.18(0.73–1.92) | 0.493 | 1.00(0.55–1.82) | 0.990 |
| Other | CTI quartile | |||||||
| Other | Q1 | 51(25.50%) | Ref | Ref | Ref | |||
| Other | Q2 | 67(33.67%) | 1.42(0.99–2.04) | 0.059 | 1.43(0.99–2.07) | 0.058 | 1.50(1.03–2.20) | 0.036 |
| Other | Q3 | 75(33.04%) | 1.41(0.99–2.01) | 0.060 | 1.33(0.93–1.92) | 0.120 | 1.35(0.92–1.99) | 0.126 |
| Other | Q4 | 70(29.54%) | 1.38(0.96–1.99) | 0.079 | 1.33(0.92–1.92) | 0.134 | 1.30(0.85-2.00) | 0.231 |
| Other | P for trend | 0.135 | 0.257 | 0.419 | ||||
Model 1: unadjusted for any covariates; Model 2: adjusted for gender, age, residence, education level, smoking status, and drinking status; Model 3: based on model 2, BMI, WC, SBP, DBP, TC, LDL, HDL, uric acid, hypertension, diabetes, dyslipidemia, liver diseases, kidney diseases, lung diseases, digestive diseases, hypertension drugs, diabetes drugs, dyslipidemia drugs, blood sugar status, and geographical area were adjusted
HR Hazard Ratio, CI Confidence Interval
Fig. 13.
Analysis of the relationship between RCS between CTI and CVD incidence by marital status. Panel (A) represents married participants, and Panel (B) represents other participants
Additional stratified analyses
We conducted stratified analyses based on the presence of chronic diseases, and the results are presented in Supplementary Tables S6-S12 and Supplementary Figures S3-S16. Additionally, we performed combined stratified analyses of glycemic status with gender, age, and marital status, and the corresponding results are provided in Supplementary Tables S13-S15 and Supplementary Figures S17-S22.
Sensitivity analyses
To evaluate the robustness of our findings, several sensitivity analyses were performed. First, after excluding all participants with missing data, the results remained essentially unchanged (Supplementary Table S16). Second, the exclusion of deceased participants did not materially alter the findings (Supplementary Table S17). Third, analyses using a logistic regression model yielded consistent results (Supplementary Table S18).
Discussion
In this prospective cohort study of 8,084 participants from the CHARLS, we observed that the CTI, a composite marker reflecting inflammatory status and metabolic dysfunction, was significantly associated with incident CVD over approximately nine years of follow-up. After comprehensive adjustment for demographic characteristics, lifestyle factors, and clinical risk factors (Model 3), each one-unit increase in CTI was associated with a 44% higher risk of CVD. Notably, the predictive performance of CTI exceeded that of both the TyG index and CRP alone. Subgroup analyses indicated that this association was particularly pronounced among individuals aged 45–60 years, married participants, women, and those without pre-existing diabetes or other chronic conditions. Sensitivity analyses, including exclusion of participants with missing data or death and application of logistic regression, confirmed the robustness of these findings. Furthermore, restricted cubic spline analyses generally demonstrated a linear dose-response relationship between CTI and CVD risk across most strata; however, nonlinear trends were observed in female, suggesting the presence of potential effect thresholds. Collectively, these findings indicate that CTI may serve as a potential biomarker for cardiovascular risk stratification.
Our results are consistent with and extend current theoretical frameworks concerning the joint roles of inflammation and metabolic dysregulation in the development of atherosclerosis and CVD. The TyG index is well known as a surrogate for IR, and CRP is one of the most studied biomarkers for chronic low-grade inflammation; individually, both have been repeatedly associated with cardiovascular risk [38–40]. For instance, recent meta-analysis confirms that elevated CRP, IL-6, Galectin-3, and fibrinogen are strongly associated with the incidence of CVD in general populations [41]. Liang et al.‘s cross-sectional study of an elderly U.S. population further revealed a non-linear relationship between the TyG index and multiple cardiovascular outcomes, including coronary heart disease (CHD) and acute myocardial infarction [42]. Moreover, an earlier study using NHANES showed that elevated CTI levels were associated with both CVD incidence and mortality. Our findings largely agree with these associations and demonstrate that CTI, by integrating both metabolic and inflammatory components, may provide stronger or more consistent predictive value than either component alone [43, 44]. The potential mechanisms underlying CTI’s predictive advantages over the TyG index or CRP may include the following: first, CTI integrates the TyG index, which reflects insulin resistance, and CRP, which indicates systemic inflammation, to provide a more comprehensive assessment of metabolic-inflammatory status [1, 45–48]. Secondly, a single biomarker (e.g. TyG or CRP) only reflects one dimension of a pathophysiological process; however, a combined analysis of multiple indices can reveal synergistic or antagonistic effects between them [48, 49]. Finally, given the significant heterogeneity in biomarker levels between individuals, TyG and CRP provide complementary biological information regarding insulin sensitivity and inflammatory status, respectively [2, 49]. This combined strategy significantly improves the identification and stratification of high-risk individuals.
The stronger CTI-CVD association among participants aged 45–60 suggests that metabolic-inflammatory risk factors exert their greatest incremental harm in midlife-possibly prior to or at the early onset of vascular damage [44]. This supports the life-course hypothesis of CVD, where risk accumulates over years of exposure and is more modifiable before irreversible damage occurs [50]. We observed stronger or nonlinear associations in certain subgroups such as females or married individuals. Possible explanations include gender differences in hormonal regulation, fat distribution (visceral vs. subcutaneous), or immune responses. Married individuals may have different psychosocial stress, support, health behaviors or access to medical care, which may modify the effect of CTI [51]. Our reclassification analyses further demonstrated the incremental predictive value of CTI beyond the TyG index and CRP. The NRI indicated that CTI provided a more pronounced reclassification benefit compared with CRP than with the TyG index, suggesting that the integration of metabolic and inflammatory information captured by CTI yields superior risk stratification over CRP alone. The IDI values were modest but consistently positive, reinforcing the notion that CTI offers incremental gains in discrimination compared with both TyG and CRP. The interaction effects suggest that the same CTI level might have differential risk depending on socio-demographic contexts. Subgroup analysis revealed significant interaction effects between CTI and both age and marital status. Furthermore, elevated CTI was consistently associated with increased cardiovascular risk across various strata, including gender, blood glucose level, hypertension, kidney disease, smoking status, alcohol consumption, body mass index, abdominal obesity, dyslipidemia, absence of diabetes, liver disease, pulmonary disease, gastrointestinal disorders, and non-use of antihypertensive, glucose-lowering, or lipid-lowering medications. These findings suggest that these factors may moderate the effect of CTI on cardiovascular risk.
This study has several significant advantages. First, it is the first to elucidate the characteristics of the association between CTI and CVD incidence within a framework that stratifies by gender, age, and glycemic status. This addresses the lack of research into population heterogeneity of this indicator. Secondly, the prospective cohort design based on a nationally representative sample and the large sample size significantly enhance the robustness of the study findings. Thirdly, multivariate group analysis and interaction effect tests confirmed the consistency of the study results across different population characteristics, providing an empirical basis for clinical risk stratification. Finally, CTI’s low cost and high accessibility as a composite index based on conventional biochemical indicators (CRP, TG, FPG) endow it with significant clinical translational potential.
This study has several limitations that should be acknowledged. First, CTI was assessed only at a single time point (baseline), thus precluding the capture of temporal variations in this metric. Repeated measurements would have allowed for the establishment of within-person trajectories and reduced regression dilution bias. Furthermore, although a wide range of confounders was adjusted for, residual confounding may persist due to unmeasured variables such as dietary habits, physical activity, genetic predisposition, or subclinical conditions, which could influence both CTI and cardiovascular risk. It should also be noted that CVD outcomes were ascertained based on self-reported physician diagnoses. The potential for misclassification-either under or over-reporting-of these diagnoses may bias effect estimates toward the null or exaggerate certain associations. Additionally, the generalizability of our findings may be limited, as the study population consisted exclusively of middle-aged and older Chinese adults; thus, the results may not extend to younger individuals or populations with different ethnic or socioeconomic backgrounds. Although CTI demonstrated superior predictive performance compared to both the TyG index and CRP, its discriminative ability as measured by the AUC remained modest. Therefore, CTI alone is insufficient for clinical decision-making and should be integrated with established risk scores and biomarkers to enhance predictive accuracy. Finally, the observed nonlinear associations in certain subgroups (e.g., gender and married individuals) may be influenced by extreme values or limited sample sizes, underscoring the need for more precise modeling or larger studies in these populations.
Several avenues for future investigation merit consideration. First, longitudinal assessments of CTI with repeated measurements are needed to capture trajectories over time and to clarify the temporal relationship between changes in CTI and incident CVD, thereby strengthening causal inference. Second, mechanistic studies should be undertaken to explore biological pathways underlying the CTI-CVD association, including endothelial dysfunction, oxidative stress, immune cell activation, and potentially the gut microbiome, given emerging evidence linking gut-inflammation interactions with cardiometabolic risk. Third, mediation analyses are warranted to disentangle the contribution of potential intermediates such as blood pressure, renal function, adiposity, and subclinical inflammation. In addition, interventional studies-either randomized controlled trials or quasi-experimental designs-could evaluate whether strategies aimed at lowering CTI (e.g., lifestyle modification, dietary interventions, anti-inflammatory therapy, or metabolic control) reduce subclinical vascular damage or clinical CVD outcomes. Fourth, risk prediction research should incorporate CTI into multivariable models alongside demographic and lifestyle factors, to evaluate calibration, discrimination, and net benefit across diverse cohorts. Furthermore, subgroup analyses and threshold refinement are needed to examine effect heterogeneity and identify clinically relevant cutoffs, particularly across gender, marital status, age strata, and chronic disease status. Finally, cross-population validation in ethnically and geographically diverse cohorts will be essential to confirm external validity and to understand how socio-economic, dietary, and environmental factors may modify the association between CTI and CVD.
Conclusion
In conclusion, this study underscores that CTI, as a composite biomarker reflecting both metabolic dysfunction and systemic inflammation, is strongly associated with incident cardiovascular disease in a middle-aged and older Chinese cohort. Its predictive value exceeds that of TyG and CRP, especially among certain subgroups. These findings offer theoretical insight into how inflammation and metabolic dysregulation jointly contribute to cardiovascular pathogenesis, and practical potential for improving risk stratification and prevention strategies. Further research is warranted to refine CTI-based prediction models, explore mechanisms, and test interventions aimed at lowering CTI to reduce CVD incidence.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This study utilized data from the CHARLS database. The authors express their gratitude to the CHARLS research team and all individuals who participated in the study. The schematic diagram of the graphical abstract was drawn by Figdraw.
Author contributions
LZ, WM, and SL were responsible for the design of the experimental protocol, preparation of graphs, and data analysis. DL and JJG performed the literature search. SL screened the studies and wrote the main paper. QW and JH were responsible for checking and revising the paper. WM and LZ provided critical guidance throughout the study design, data analysis, and interpretation. All authors participated in the finalization of the paper and signed the acknowledgement.
Funding
This work was supported by the Key Project of Chongqing Natural Science Foundation (Science and Technology Development Fund) (CSTB2023NSCQ-ZDJ0001).
Data availability
The data supporting the findings of this study are available on the CHARLS website (http://charls.pku.edu.cn/).
Declarations
Ethics approval and consent to participate
CHARLS was approved by the Institutional Review Board of Peking University (approval number: IRB00001052-11015 for the household survey and IRB00001052-11014 for blood samples), and all participants provided written informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Lin Zhang and Shan Li have contributed equally to this work and share first authorship.
Contributor Information
Jing Hu, Email: hujing820715@tmmu.edu.cn.
Qian Wang, Email: wangqian411@tmmu.edu.cn.
Wei Mao, Email: 395762865@qq.com.
References
- 1.Huo G, Tang Y, Liu Z, et al. Association between C-reactive protein-triglyceride glucose index and stroke risk in different glycemic status: insights from the China health and retirement longitudinal study (CHARLS). Cardiovasc Diabetol. 2025;24(1):142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Khan HTA, Addo KM, Findlay H. Public Health Challenges and responses to the growing ageing populations. Public Health Challenges. 2024;3(3):e213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Huang Z. A study on the causes and solutions to the global aging Population. Advances in Economics, management and political sciences. 2024; 142:1–5.
- 4.Roth GA, Mensah GA, Johnson CO, et al. Global burden of cardiovascular diseases and risk Factors, 1990–2019: update from the GBD 2019 study. J Am Coll Cardiol. 2020;76(25):2982–3021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Vaduganathan M, Mensah GA, Turco JV, et al. The global burden of cardiovascular diseases and risk: a compass for future health. J Am Coll Cardiol. 2022;80(25):2361–71. [DOI] [PubMed] [Google Scholar]
- 6.Yang M, Liu J, Shen Q, et al. Body roundness index trajectories and the incidence of cardiovascular disease: evidence from the China health and retirement longitudinal study. J Am Heart Association. 2024;13(19):e034768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhao D, Wang Y, Wong ND, et al. Impact of aging on cardiovascular diseases: from chronological observation to biological insights: JACC family series. JACC Asia. 2024;4(5):345–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chong B, Jayabaskaran J, Jauhari SM, et al. Global burden of cardiovascular diseases: projections from 2025 to 2050. Eur J Prev Cardiol. 2025;32(11):1001–15. [DOI] [PubMed] [Google Scholar]
- 9.Tan SCW, Zheng BB, Tang ML, et al. Global burden of cardiovascular diseases and its risk Factors, 1990–2021: a systematic analysis for the global burden of disease study 2021. QJM: Int J Med. 2025;118(6):411–22. [DOI] [PubMed] [Google Scholar]
- 10.Liu S, Li Y, Zeng X, et al. Burden of cardiovascular diseases in China, 1990–2016: findings from the 2016 global burden of disease study. JAMA Cardiol. 2019;4(4):342–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ren J, Zhang Y. Targeting autophagy in aging and aging-Related cardiovascular Diseases. Trends Pharmacol Sci 2018; 39(12):1064–76. [DOI] [PMC free article] [PubMed]
- 12.Chong B, Jayabaskaran J, Jauhari SM et al. Global burden of cardiovascular diseases: projections from 2025 to 2050. Eur J Prev Cardiol. 2024. [DOI] [PubMed]
- 13.Gao K, Cao LF, Ma WZ, et al. Association between sarcopenia and cardiovascular disease among middle-aged and older adults: findings from the China health and retirement longitudinal study. EClinicalMedicine. 2022;44:101264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gupta R, Wood DA. Primary prevention of ischaemic heart disease: populations, individuals, and health professionals. Lancet. 2019;394(10199):685–96. [DOI] [PubMed] [Google Scholar]
- 15.Ran J, Zhou P, Wang J, et al. Global, regional, and National burden of heart failure and its underlying causes, 1990–2021: results from the global burden of disease study 2021. Biomark Res. 2025;13(1):16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sun Y, Ji H, Sun W, et al. Triglyceride glucose (TyG) index: A promising biomarker for diagnosis and treatment of different diseases. Eur J Intern Med. 2025;131:3–14. [DOI] [PubMed] [Google Scholar]
- 17.Tao LC, Xu JN, Wang TT, et al. Triglyceride-glucose index as a marker in cardiovascular diseases: landscape and limitations. Cardiovasc Diabetol. 2022;21(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Newsholme P, Cruzat VF, Keane KN, et al. Molecular mechanisms of ROS production and oxidative stress in diabetes. Biochem J. 2016;473(24):4527–50. [DOI] [PubMed] [Google Scholar]
- 19.Ren Q, Huang Y, Liu Q, et al. Association between triglyceride glucose-waist height ratio index and cardiovascular disease in middle-aged and older Chinese individuals: a nationwide cohort study. Cardiovasc Diabetol. 2024;23(1):247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sánchez-Íñigo L, Navarro-González D, Fernández-Montero A, et al. The TyG index May predict the development of cardiovascular events. Eur J Clin Invest. 2016;46(2):189–97. [DOI] [PubMed] [Google Scholar]
- 21.Cui H, Liu Q, Wu Y, et al. Cumulative triglyceride-glucose index is a risk for CVD: a prospective cohort study. Cardiovasc Diabetol. 2022;21(1):22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Speer T, Dimmeler S, Schunk SJ, et al. Targeting innate immunity-driven inflammation in CKD and cardiovascular disease. Nat Rev Nephrol. 2022;18(12):762–78. [DOI] [PubMed] [Google Scholar]
- 23.Ajoolabady A, Pratico D, Lin L, et al. Inflammation in atherosclerosis: pathophysiology and mechanisms. Cell Death Dis. 2024;15(11):817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Olsen MB, Gregersen I, Sandanger et al. Targeting the inflammasome in cardiovascular Disease. JACC basic to translational science. 2022; 7(1):84–98. [DOI] [PMC free article] [PubMed]
- 25.Wang L, Cheng CK, Yi M, et al. Targeting endothelial dysfunction and inflammation. J Mol Cell Cardiol. 2022;168:58–67. [DOI] [PubMed] [Google Scholar]
- 26.Ajoolabady A, Pratico D, Ren J. Endothelial dysfunction: mechanisms and contribution to diseases. Acta Pharmacol Sin. 2024;45(10):2023–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Rizo-Téllez SA, Sekheri M, Filep JG. C-reactive protein: a target for therapy to reduce inflammation. Front Immunol. 2023;14:1237729. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Moore KJ. Targeting inflammation in CVD: advances and challenges. Nat Reviews Cardiol. 2019;16(2):74–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Luan YY, Yao YM. The clinical significance and potential role of C-Reactive protein in chronic inflammatory and neurodegenerative diseases. Front Immunol. 2018;9:1302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ruan GT, Xie HL, Zhang HY, et al. A novel inflammation and insulin resistance related indicator to predict the survival of patients with cancer. Front Endocrinol. 2022;13:905266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sun Y, Guo Y, Ma S, et al. Association of C-reactive protein-triglyceride glucose index with the incidence and mortality of cardiovascular disease: a retrospective cohort study. Cardiovasc Diabetol. 2025;24(1):313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Xu M, Zhang L, Xu D et al. Usefulness of C-reactive protein-triglyceride glucose index in detecting prevalent coronary heart disease: findings from the National health and nutrition examination survey 1999–2018. Frontiers in cardiovascular medicine. 2024; 11:1485538. [DOI] [PMC free article] [PubMed]
- 33.Chen Y, Jia W, Guo J, et al. Association between the C-reactive protein-triglyceride glucose index and new-onset coronary heart disease among metabolically heterogeneous individuals. Cardiovasc Diabetol. 2025;24(1):316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chen J, Yan L, He L, et al. Association between different dimensions of the C-reactive protein-triglyceride-glucose index and future cardiovascular disease risk in individuals with cardiovascular-kidney-metabolic syndrome stages 0–3: a nationwide cohort study. Diabetol Metab Syndr. 2025;17(1):299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ma X, Ma X, Wang Y, et al. Associations of cumulative exposure and dynamic trajectories of the C-reactive protein-triglyceride-glucose index with incident cardiovascular disease in middle-aged and older Chinese adults: a nationwide cohort study. Cardiovasc Diabetol. 2025;24(1):303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Shi W, Shi Y, Park J et al. C-reactive protein-triglyceride glucose index as a risk factor for all-cause and cardiovascular mortality among individuals with diabetes or prediabetes: an observational study based on evidence from NHANES 2003–2008. 2025; 1(1):9–17.
- 37.Li H, Zheng D, Li Z, et al. Association of depressive symptoms with incident cardiovascular diseases in Middle-Aged and older Chinese adults. JAMA Netw Open. 2019;2(12):e1916591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wang A, Wang G, Liu Q, et al. Triglyceride-glucose index and the risk of stroke and its subtypes in the general population: an 11-year follow-up. Cardiovasc Diabetol. 2021;20(1):46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Wang TJ. Assessing the role of circulating, genetic, and imaging biomarkers in cardiovascular risk prediction. Circulation. 2011;123(5):551–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Fan J, Bai J, Liu W, et al. Effects of intensive vs. standard blood pressure control on cognitive function: Post-hoc analysis of the STEP randomized controlled trial. Front Neurol. 2023;14:1042637. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Liu Y, Guan S, Xu H, et al. Inflammation biomarkers are associated with the incidence of cardiovascular disease: a meta-analysis. Front Cardiovasc Med. 2023;10:1175174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Liang D, Liu C, Wang Y. The association between triglyceride-glucose index and the likelihood of cardiovascular disease in the U.S. Population of older adults aged ≥ 60 years: a Population-based study. Cardiovasc Diabetol. 2024;23(1):151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Dang K, Wang X, Hu J, et al. The association between triglyceride-glucose index and its combination with obesity indicators and cardiovascular disease: NHANES 2003–2018. Cardiovasc Diabetol. 2024;23(1):8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Liu C, Liang D. The association between the triglyceride-glucose index and the risk of cardiovascular disease in US population aged ≤ 65 years with prediabetes or diabetes: a population-based study. Cardiovasc Diabetol. 2024;23(1):168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Ruan G-T, Deng L, Xie H-L, et al. Systemic inflammation and insulin resistance-related indicator predicts poor outcome in patients with cancer cachexia. Cancer Metabolism. 2024;12(1):3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Gao A, Peng B, Gao Y, et al. Evaluation and comparison of inflammatory and insulin resistance indicators on recurrent cardiovascular events in patients undergoing percutaneous coronary intervention: a single center retrospective observational study. Diabetol Metab Syndr. 2025;17(1):157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Shan Y, Liu Q, Gao T. Application of the C-reactive protein-triglyceride glucose index in predicting the risk of new-onset diabetes in the general population aged 45 years and older: a National prospective cohort study. BMC Endocr Disorders. 2025;25(1):126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Cui C, Liu L, Qi Y, et al. Joint association of TyG index and high sensitivity C-reactive protein with cardiovascular disease: a National cohort study. Cardiovasc Diabetol. 2024;23(1):156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Wan B, Wang S, Hu S, et al. The comprehensive effects of high-sensitivity C-reactive protein and triglyceride glucose index on cardiometabolic Multimorbidity. Front Endocrinol. 2025;16:1511319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Fuster V, Ibanez B. Address cardiovascular health in middle age: time to remove the blindfold. J Am Coll Cardiol. 2023;81(7):705–7. [DOI] [PubMed] [Google Scholar]
- 51.Dhindsa DS, Khambhati J, Schultz WM, et al. Marital status and outcomes in patients with cardiovascular disease. Trends Cardiovasc Med. 2020;30(4):215–20. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data supporting the findings of this study are available on the CHARLS website (http://charls.pku.edu.cn/).