Abstract
Background
Metabolic syndrome, characterized by a clustering of cardiometabolic risk factors, is a key driver of chronic diseases. The Cardiometabolic Index (CMI) is a useful metric, but its exclusion of hyperglycemia—a cornerstone of metabolic dysregulation—limits its scope. A Modified Cardiometabolic Index (MCMI) incorporating glucose has been proposed, but its long-term predictive value for incident type 2 diabetes, cardiovascular disease, and subsequent multimorbidity has not been prospectively validated. We aimed to evaluate the MCMI as a predictor for these outcomes and compare its performance against the original CMI.
Methods
This study included 8,251 participants aged ≥ 45 years from the China Health and Retirement Longitudinal Study. The MCMI was calculated at baseline (2011) using waist-to-height ratio, the triglyceride to HDL-C ratio, and fasting blood glucose. Multivariable Cox proportional hazards models were used to assess the associations between the MCMI (as both a continuous variable and in quartiles) and the incidence of 13 chronic diseases over a median follow-up of 7.0 years. The predictive accuracy of the MCMI and CMI was compared using the area under the receiver operating characteristic curve (AUC) and the DeLong test.
Results
A higher baseline MCMI was significantly associated with an increased risk of developing type 2 diabetes (HR 1.15, 95% CI 1.12–1.18), hypertension (HR 1.12, 95% CI 1.09–1.15), dyslipidemia (HR 1.11, 95% CI 1.08–1.14), heart disease (HR 1.09, 95% CI 1.03–1.16), and stroke (HR 1.13, 95% CI 1.07–1.19), with significant dose-response relationships observed across quartiles. The MCMI was also associated with progression to multimorbidity, showing a stronger association with the development of a second chronic disease (HR 1.22, 95% CI 1.13–1.31) than a first (HR 1.09, 95% CI 1.03–1.16). Compared to the CMI, the MCMI demonstrated significantly superior predictive accuracy for type 2 diabetes (AUC 0.663 vs. 0.647), hypertension (AUC 0.747 vs. 0.742), stroke (AUC 0.692 vs. 0.686), and dyslipidemia (AUC 0.657 vs. 0.650) (all P < 0.001).
Conclusions
In this large, nationally representative cohort, the MCMI was a robust and independent predictor of incident cardiometabolic diseases and the development of multimorbidity. Its superior predictive accuracy over the original CMI supports its utility as a simple, low-cost, and more effective tool for stratifying cardiometabolic risk and guiding early prevention, particularly for type 2 diabetes and related cardiovascular events, in clinical and public health settings.
Supplementary Information
The online version contains supplementary material available at 10.1186/s13098-026-02090-7.
Keywords: Modified cardiometabolic index, Multimorbidity, Prospective cohort study, Risk stratification
Introduction
The global epidemic of chronic non-communicable diseases (NCDs), largely driven by widespread metabolic dysregulation and an aging population, represents a primary public health challenge. These conditions impose a substantial health and economic burden worldwide [1]. In this context, multimorbidity—the coexistence of two or more chronic diseases in an individual—is increasingly prevalent. Multimorbidity is associated with a reduced quality of life, an increased risk of premature mortality, and significant challenges to healthcare system sustainability [2, 3]. Therefore, developing and validating tools to effectively identify future disease risk is crucial for proactive health management and precision prevention.
Metabolic syndrome, defined by a clustering of cardiometabolic risk factors, is a common pathophysiological basis for many chronic diseases, most notably type 2 diabetes and cardiovascular disease [4]. The original Cardiometabolic Index (CMI), a metric combining the waist-to-height ratio and the triglyceride-to-high-density lipoprotein cholesterol (TG/HDL-C) ratio, was developed to quantify this risk [5]. However, a key limitation of the CMI is its exclusion of blood glucose, a core component of metabolic dysregulation and a cornerstone in the diagnosis of prediabetes and type 2 diabetes. This omission limits a comprehensive assessment of cardiometabolic risk. To address this limitation, a Modified Cardiometabolic Index (MCMI) that incorporates blood glucose was recently proposed [6]. Yet, the association between the MCMI and the risk of incident chronic diseases has not been prospectively evaluated. Furthermore, whether the MCMI has superior predictive ability over the original CMI remains unvalidated.
Therefore, this study used data from the China Health and Retirement Longitudinal Study (CHARLS) to prospectively investigate the associations between the MCMI and the risk of incident chronic diseases, with a focus on cardiometabolic outcomes, and multimorbidity. We also compared the predictive ability of the MCMI with that of the original CMI to validate its utility as a tool for early risk stratification.
Methods
Study population
This study used data from the CHARLS, a nationally representative survey of Chinese adults aged 45 years and older and their spouses. CHARLS employs a multi-stage stratified probability-proportional-to-size sampling method, covering 150 counties and districts across 28 provinces. We used the 2011 wave as the baseline and linked these data to follow-up waves in 2013, 2015, and 2018. The study protocol was approved by the Biomedical Ethics Review Committee of Peking University (IRB 000010052–11015) and followed the principles of the Declaration of Helsinki. All participants provided written informed consent. Trained staff conducted face-to-face interviews using a standardised questionnaire [7].
The participant selection process is shown in Fig. 1. First, participants were excluded if they were aged < 45 years or had missing data required for the MCMI calculation. For the analysis of each incident chronic disease, a separate analytical sub-cohort was created. The chronic diseases for this analysis were selected based on their high prevalence and significant public health burden among the middle-aged and older population in China [8]. From each sub-cohort, we further excluded individuals with the specific disease at baseline (2011) or those with missing outcome data during any follow-up wave.
Fig. 1.
Flowchart of the study population selection. The final analytical cohort for each of the 13 chronic diseases was established by separately excluding participants from the eligible baseline sample (n = 8,251) who had the corresponding prevalent disease in 2011. The numbers shown for each disease represent the final sample size for that specific analysis
Data assessment
Baseline data included sociodemographic characteristics, lifestyle factors, health status, anthropometric measurements, and biochemical markers. Data on sociodemographics, lifestyle, and health status were collected using a standardised questionnaire. Sociodemographic variables included age, sex, residence (urban or rural), educational attainment, and marital status. Lifestyle and health status variables included smoking status, alcohol consumption, physical activity (Metabolic Equivalents of Task), Activities of Daily Living (ADL), self-reported health, and the number of chronic comorbidities (categorised as 0, 1, or ≥ 2).Trained staff performed anthropometric measurements (height, weight, waist circumference [WC]) and measured blood pressure (systolic and diastolic) according to standard protocols. Fasting blood samples were analysed for lipids (total cholesterol, triglycerides, low-density lipoprotein cholesterol), fasting blood glucose, and high-sensitivity C-reactive protein.
Assessment of the modified cardiometabolic index
The MCMI and the original CMI were calculated using data from the 2011 baseline survey.The MCMI was calculated with the following formula: MCMI = ln [(triglycerides (mg/dL) × fasting blood glucose (mg/dL))/high-density lipoprotein cholesterol (mg/dL)] × (waist (cm)/height (cm)) [6].For comparison, the CMI was calculated as follows: CMI = (triglycerides (mmol/L)/high-density lipoprotein cholesterol (mmol/L)) × (waist (cm)/height (cm)) [9].
Ascertainment of chronic disease outcomes
For participants free of a specific chronic disease at baseline, incident cases were identified during the 2013, 2015, and 2018 follow-up waves. An incident case was defined by a self-reported physician diagnosis in response to the question, “Have you been diagnosed with [disease name]?” [10].
Statistical analysis
Baseline characteristics of the study population were summarised across quartiles of the MCMI. Continuous variables were presented as mean ± standard deviation (SD) or median (interquartile range, IQR), and categorical variables as counts (n) and percentages (%). We used one-way analysis of variance (ANOVA) or the Kruskal-Wallis test to compare continuous variables and the χ² test for categorical variables. We used multiple imputation by chained equations to handle missing covariate data. Five complete datasets were generated, and results were pooled using Rubin’s rules.
We used the Kaplan-Meier method and the log-rank test to estimate and compare cumulative incidence across MCMI quartiles. Multivariable Cox proportional hazards models were used to assess the association between MCMI and the risk of incident chronic diseases. MCMI was analysed both as a continuous variable and as a categorical variable (quartiles), with the lowest quartile (Q1) as the reference group.
We fitted three nested models:
Model 1: Unadjusted.
Model 2: Adjusted for sociodemographic variables (age, sex, educational attainment, marital status, and residence).
Model 3: Additionally adjusted for lifestyle factors (smoking, alcohol consumption, physical activity), clinical variables (systolic blood pressure, creatinine, Activities of Daily Living, self-reported health, and number of comorbidities), and high-sensitivity C-reactive protein.
Results were presented as hazard ratios (HRs) with 95% confidence intervals (CIs). We assessed multicollinearity using the variance inflation factor (VIF), with a threshold of < 5.
We used Cox proportional hazards models with restricted cubic splines (RCS) to examine the potential non-linear dose-response relationship between the MCMI and the risk of incident chronic diseases. Stratified analyses were conducted to assess these associations within subgroups defined by age, sex, smoking status, and alcohol consumption. We tested for potential interactions by including product terms between the MCMI and these stratification variables in the models.
Furthermore, we investigated the association between the MCMI and the development of multimorbidity. Using Cox models, we separately assessed two outcomes: the risk of developing a first chronic disease among participants who were disease-free at baseline, and the risk of developing a second chronic disease among those with a single prevalent condition at baseline.
To compare the predictive performance of the MCMI and CMI, we assessed their ability to predict incident chronic diseases using Receiver Operating Characteristic curves. The predictive performance was assessed by the area under the receiver operating characteristic curve, which was calculated from the risk scores predicted by the fully adjusted multivariate Cox model (Model 3). The statistical significance of the difference between the AUCs was determined using the DeLong test.
To assess the robustness of our findings, we conducted three sensitivity analyses. The primary analysis was repeated by: (1) using complete cases only, without multiple imputation; (2) excluding self-rated health from the fully adjusted model; and (3) excluding participants who developed an outcome within the first two years of follow-up to account for potential reverse causality. All statistical analyses were performed using R software (version 4.2.2), and a two-sided P value of less than 0.05 was considered statistically significant.
Results
Baseline characteristics of participants
The baseline characteristics of the 8,251 participants, stratified by quartiles of the MCMI, are shown in Table 1. Compared with individuals in the lowest MCMI quartile, those in the highest quartile were more likely to be female and reside in urban areas (P < 0.001) but had lower rates of current smoking and drinking (P < 0.001). A higher MCMI was also associated with lower physical activity, higher systolic blood pressure, and higher C-reactive protein levels (all P < 0.001). Furthermore, participants with a higher MCMI had a greater prevalence of limitations in ADL (P = 0.032). No significant differences across MCMI quartiles were observed for age, educational attainment, marital status, serum creatinine, or self-reported health.
Table 1.
Baseline characteristics of participants according to quartiles of the modified cardiometabolic Index(MCMI)
| Characteristic | Quartile 1 (N = 2,063) | Quartile 2 (N = 2,063) | Quartile 3 (N = 2,062) | Quartile 4 (N = 2,063) | P-value |
|---|---|---|---|---|---|
| Age, years, mean ± SD | 59.77 ± 9.32 | 59.62 ± 9.24 | 59.58 ± 9.03 | 60.16 ± 8.93 | 0.066 |
| Gender, n (%) | < 0.001 | ||||
| Female | 790 (38%) | 1,045 (51%) | 1,183 (57%) | 1,377 (67%) | |
| Male | 1,273 (62%) | 1,018 (49%) | 879 (43%) | 686 (33%) | |
| Education, n (%) | 0.125 | ||||
| Less than lower secondary education | 1,870 (91%) | 1,869 (91%) | 1,831 (89%) | 1,869 (91%) | |
| Secondary or above | 193 (9.4%) | 194 (9.4%) | 231 (11%) | 194 (9.4%) | |
| Marital status, n (%) | 0.681 | ||||
| Married | 1,793 (87%) | 1,800 (87%) | 1,797 (87%) | 1,776 (86%) | |
| Non-married | 270 (13%) | 263 (13%) | 265 (13%) | 287 (14%) | |
| Residence, n (%) | < 0.001 | ||||
| Rural | 1,484 (72%) | 1,433 (69%) | 1,258 (61%) | 1,131 (55%) | |
| Urban | 579 (28%) | 630 (31%) | 804 (39%) | 932 (45%) | |
| Smoking, n (%) | 1,064 (52%) | 853 (41%) | 723 (35%) | 608 (29%) | < 0.001 |
| Drinking, n (%) | 1,011 (49%) | 827 (40%) | 752 (37%) | 649 (31%) | < 0.001 |
| Physical activity, MET-h/week, median (IQR) | 6720.0 (1582.0–14958.0) | 5520.0 (462.0–13398.0) | 4800.0 (462.0–10920.0) | 2562.0 (0.0–6762.0) | < 0.001 |
| Systolic blood pressure (SBP), mmHg, mean ± SD | 124.35 ± 20.28 | 127.30 ± 21.12 | 130.81 ± 20.52 | 136.28 ± 21.69 | < 0.001 |
| Creatinine (Cr), mg/dL, median (IQR) | 0.77 (0.66–0.88) | 0.76 (0.66–0.88) | 0.75 (0.64–0.89) | 0.76 (0.66–0.88) | 0.436 |
| C-reactive protein (CRP), mg/L, median (IQR) | 0.75 (0.44–1.67) | 0.85 (0.49–1.84) | 1.08 (0.60–2.13) | 1.59 (0.88–3.20) | < 0.001 |
| ADL disability, n (%) | 0.032 | ||||
| No disability (0) | 1,727 (85%) | 1,729 (85%) | 1,688 (83%) | 1,643 (80%) | |
| Any disability (≥ 1) | 313 (15%) | 312 (15%) | 356 (17%) | 411 (20%) | |
| Self-reported health, n (%) | 0.316 | ||||
| Excellent | 9 (0.6%) | 10 (0.7%) | 13 (0.9%) | 8 (0.6%) | |
| Very good | 120 (7.7%) | 109 (7.4%) | 110 (7.7%) | 112 (8.1%) | |
| Good | 275 (18%) | 235 (16%) | 239 (17%) | 234 (17%) | |
| Fair | 766 (49%) | 749 (51%) | 686 (48%) | 634 (46%) | |
| Poor | 381 (25%) | 372 (25%) | 384 (27%) | 401 (29%) |
Data are presented as mean ± standard deviation (SD), median (interquartile range, IQR), or n (%). P-values were calculated using ANOVA for normally distributed variables, Kruskal-Wallis test for skewed variables, and chi-square test for categorical variables. Percentages are based on non-missing values for each variable
ADL, Activities of Daily Living; Cr, Creatinine; CRP, C-reactive protein; IQR, Interquartile Range; MET, Metabolic Equivalent of Task
Association of the modified cardiometabolic index and the risk of new-onset chronic diseases
No significant multicollinearity was detected among the model covariates; all VIFs were < 5 (Supplementary Table 1). During a median follow-up of 7.0 years, Kaplan-Meier analysis was used to assess the association between baseline MCMI and the cumulative incidence of 13 chronic diseases (Fig. 2). A significant dose-response relationship was observed for several cardiometabolic conditions. Specifically, the cumulative incidence of hypertension, diabetes, heart disease, stroke, and dyslipidemia increased progressively across higher MCMI quartiles (log-rank P < 0.05 for all). In contrast, no significant associations were found between MCMI and the incidence of the other chronic conditions.
Fig. 2.
Kaplan-Meier curves for the cumulative incidence of 13 chronic diseases according to quartiles of the Modified Cardiometabolic Index (MCMI). Log-rank tests showed statistically significant differences (P < 0.05) among the four MCMI quartiles for the incidence of Hypertension, Diabetes, Heart Disease, Stroke and Dyslipidemia. No significant association was found for Cancer, Lung Disease, Liver Disease, Kidney Disease, Digestive Disease, Psychological Problem, Asthma or Arthritis
Multivariable Cox proportional hazards models were used to assess the associations between the MCMI and incident chronic diseases (Table 2). In the fully adjusted model (Model 3), when analyzed as a continuous variable, a higher MCMI was significantly associated with an increased risk of hypertension (HR 1.12, 95% CI 1.09–1.15), diabetes (HR 1.15, 95% CI 1.12–1.18), heart disease (HR 1.09, 95% CI 1.03–1.16), stroke (HR 1.13, 95% CI 1.07–1.19), and dyslipidemia (HR 1.11, 95% CI 1.08–1.14). Conversely, a higher MCMI was associated with a lower risk of digestive system diseases (HR 0.90, 95% CI 0.82–0.98). No significant associations with the MCMI were observed for the other chronic diseases.
Table 2.
Hazard ratios for incident chronic diseases associated with a one-unit increase in the continuous modified cardiometabolic index (MCMI)
| Outcome | Events | Model 1 h (95% CI) | P-value | Model 2 h (95% CI) | P-value | Model 3 h (95% CI) | P-value |
|---|---|---|---|---|---|---|---|
| Hypertension | 1505 | 1.13 (1.11–1.16) | < 0.001 | 1.15 (1.12–1.17) | < 0.001 | 1.12 (1.09–1.15) | < 0.001 |
| Diabetes | 650 | 1.16 (1.14–1.19) | < 0.001 | 1.16 (1.13–1.19) | < 0.001 | 1.15 (1.12–1.18) | < 0.001 |
| Cancer | 156 | 1.11 (1.02–1.22) | 0.019 | 1.10 (0.99–1.22) | 0.082 | 1.09 (0.97–1.22) | 0.136 |
| Lung Disease | 857 | 0.98 (0.90–1.08) | 0.723 | 1.02 (0.93–1.11) | 0.683 | 0.96 (0.88–1.06) | 0.418 |
| Heart Disease | 979 | 1.13 (1.09–1.18) | < 0.001 | 1.13 (1.07–1.18) | < 0.001 | 1.09 (1.03–1.16) | 0.002 |
| Stroke | 520 | 1.13 (1.09–1.18) | < 0.001 | 1.15 (1.10–1.20) | < 0.001 | 1.13 (1.07–1.19) | < 0.001 |
| Psychological Problem | 213 | 0.97 (0.81–1.15) | 0.713 | 0.89 (0.74–1.07) | 0.201 | 0.85 (0.70–1.03) | 0.104 |
| Arthritis | 1432 | 1.00 (0.94–1.06) | 0.919 | 0.99 (0.93–1.05) | 0.763 | 0.98 (0.92–1.05) | 0.591 |
| Dyslipidemia | 1152 | 1.14 (1.11–1.16) | < 0.001 | 1.13 (1.10–1.16) | < 0.001 | 1.11 (1.08–1.14) | < 0.001 |
| Liver Disease | 467 | 1.08 (1.01–1.16) | 0.026 | 1.09 (1.02–1.16) | 0.014 | 1.06 (0.98–1.14) | 0.166 |
| Kidney Disease | 609 | 1.03 (0.95–1.12) | 0.495 | 1.05 (0.98–1.14) | 0.185 | 0.99 (0.89–1.09) | 0.773 |
| Digestive Disease | 1117 | 0.93 (0.86–1.00.86.00) | 0.064 | 0.91 (0.83–0.98) | 0.016 | 0.90 (0.82–0.98) | 0.012 |
| Asthma | 291 | 0.98 (0.84–1.14) | 0.785 | 1.01 (0.87–1.18) | 0.861 | 0.96 (0.81–1.12) | 0.582 |
P-values < 0.05 are considered statistically significant. Model 1: Adjusted for age and gender. Model 2: Adjusted for Model 1 variables plus education, marital status, and residence. Model 3: Adjusted for Model 2 variables plus smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, self-reported health, and chronic disease count at baseline
ADL, Activities of Daily Living; CI, Confidence Interval; HR, Hazard Ratio; MET, Metabolic Equivalent of Task
When the MCMI was analyzed by quartiles, a clear dose-response relationship was observed for several outcomes (Table 3). In the fully adjusted model, participants in the highest quartile (Q4) had a significantly higher risk of hypertension (HR 2.41, 95% CI 2.08–2.80), diabetes (HR 4.06, 95% CI 3.16–5.22), stroke (HR 2.51, 95% CI 1.90–3.32), and dyslipidemia (HR 2.84, 95% CI 2.38–3.40) compared with those in the lowest quartile (Q1). A significantly increased risk for heart disease was also found in Q4 (HR 1.37, 95% CI 1.13–1.66). The association with a lower risk of digestive system diseases was consistent, with a significant risk reduction observed in the highest quartile (HR 0.74, 95% CI 0.62–0.88). No significant trends across MCMI quartiles were observed for the other chronic conditions.
Table 3.
Hazard ratios for incident chronic diseases according to quartiles of the modified cardiometabolic index (MCMI)
| Outcome | MCMI Quartile | Model 1 h (95% CI) | P-value | Model 2 h (95% CI) | P-value | Model 3 h (95% CI) | P-value |
|---|---|---|---|---|---|---|---|
| Hypertension | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.17 (1.01–1.37) | 0.036 | 1.21 (1.04–1.40) | 0.015 | 1.22 (1.04–1.41) | 0.012 | |
| Q3 | 1.48 (1.28–1.72) | < 0.001 | 1.58 (1.36–1.84) | < 0.001 | 1.59 (1.37–1.85) | < 0.001 | |
| Q4 | 2.25 (1.95–2.60) | < 0.001 | 2.40 (2.07–2.78) | < 0.001 | 2.41 (2.08–2.80) | < 0.001 | |
| Diabetes | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.27 (0.95–1.69) | 0.104 | 1.25 (0.94–1.66) | 0.130 | 1.23 (0.92–1.64) | 0.154 | |
| Q3 | 2.08 (1.60–2.71) | < 0.001 | 2.05 (1.57–2.67) | < 0.001 | 1.96 (1.50–2.56) | < 0.001 | |
| Q4 | 4.45 (3.49–5.67) | < 0.001 | 4.32 (3.37–5.54) | < 0.001 | 4.06 (3.16–5.22) | < 0.001 | |
| Cancer | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 0.77 (0.49–1.22) | 0.261 | 0.73 (0.46–1.15) | 0.175 | 0.72 (0.45–1.14) | 0.159 | |
| Q3 | 0.86 (0.55–1.35) | 0.521 | 0.79 (0.50–1.24) | 0.298 | 0.75 (0.48–1.19) | 0.221 | |
| Q4 | 1.04 (0.68–1.59) | 0.864 | 0.88 (0.57–1.37) | 0.574 | 0.81 (0.52–1.27) | 0.362 | |
| Lung Disease | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.03 (0.85–1.24) | 0.795 | 1.06 (0.88–1.28) | 0.534 | 1.04 (0.86–1.26) | 0.667 | |
| Q3 | 0.97 (0.81–1.18) | 0.786 | 1.05 (0.86–1.27) | 0.649 | 0.97 (0.80–1.18) | 0.790 | |
| Q4 | 0.97 (0.80–1.17) | 0.725 | 1.05 (0.87–1.28) | 0.596 | 0.92 (0.76–1.13) | 0.435 | |
| Heart Disease | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.38 (1.15–1.66) | 0.001 | 1.32 (1.10–1.59) | 0.003 | 1.30 (1.08–1.57) | 0.006 | |
| Q3 | 1.37 (1.13–1.65) | 0.001 | 1.28 (1.06–1.55) | 0.010 | 1.22 (1.01–1.48) | 0.039 | |
| Q4 | 1.64 (1.37–1.97) | < 0.001 | 1.49 (1.23–1.79) | < 0.001 | 1.37 (1.13–1.66) | 0.001 | |
| Stroke | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.62 (1.21–2.15) | 0.001 | 1.66 (1.24–2.21) | 0.001 | 1.64 (1.23–2.19) | 0.001 | |
| Q3 | 1.83 (1.38–2.43) | < 0.001 | 1.94 (1.46–2.57) | < 0.001 | 1.86 (1.40–2.47) | < 0.001 | |
| Q4 | 2.57 (1.97–3.36) | < 0.001 | 2.73 (2.08–3.60) | < 0.001 | 2.51 (1.90–3.32) | < 0.001 | |
| Psychological Problem | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.38 (0.95–2.03) | 0.094 | 1.28 (0.88–1.88) | 0.199 | 1.27 (0.86–1.86) | 0.228 | |
| Q3 | 1.14 (0.77–1.70) | 0.517 | 1.01 (0.67–1.50) | 0.974 | 0.90 (0.60–1.35) | 0.605 | |
| Q4 | 1.07 (0.72–1.60) | 0.735 | 0.87 (0.58–1.32) | 0.518 | 0.76 (0.50–1.16) | 0.202 | |
| Arthritis | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.19 (1.03–1.38) | 0.017 | 1.17 (1.01–1.35) | 0.034 | 1.16 (1.01–1.35) | 0.042 | |
| Q3 | 1.09 (0.94–1.27) | 0.245 | 1.10 (0.94–1.27) | 0.226 | 1.08 (0.93–1.26) | 0.322 | |
| Q4 | 1.05 (0.90–1.22) | 0.516 | 1.03 (0.88–1.20) | 0.733 | 1.00 (0.85–1.17) | 0.997 | |
| Dyslipidemia | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 1.29 (1.07–1.57) | 0.008 | 1.29 (1.07–1.57) | 0.008 | 1.30 (1.07–1.58) | 0.008 | |
| Q3 | 1.88 (1.57–2.25) | < 0.001 | 1.84 (1.53–2.20) | < 0.001 | 1.77 (1.47–2.12) | < 0.001 | |
| Q4 | 2.99 (2.52–3.56) | < 0.001 | 2.94 (2.46–3.51) | < 0.001 | 2.84 (2.38–3.40) | < 0.001 | |
| Liver Disease | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 0.86 (0.65–1.13) | 0.271 | 0.88 (0.67–1.15) | 0.346 | 0.86 (0.66–1.13) | 0.286 | |
| Q3 | 1.06 (0.82–1.37) | 0.678 | 1.08 (0.83–1.41) | 0.547 | 1.01 (0.77–1.31) | 0.969 | |
| Q4 | 1.20 (0.93–1.54) | 0.163 | 1.25 (0.96–1.62) | 0.094 | 1.11 (0.85–1.44) | 0.439 | |
| Kidney Disease | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 0.87 (0.68–1.10) | 0.236 | 0.90 (0.71–1.15) | 0.406 | 0.87 (0.69–1.11) | 0.264 | |
| Q3 | 1.12 (0.90–1.40) | 0.321 | 1.19 (0.95–1.50) | 0.126 | 1.10 (0.88–1.39) | 0.394 | |
| Q4 | 1.11 (0.89–1.39) | 0.357 | 1.22 (0.97–1.53) | 0.091 | 1.06 (0.84–1.34) | 0.606 | |
| Digestive Disease | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 0.88 (0.75–1.04) | 0.143 | 0.86 (0.72–1.01) | 0.066 | 0.85 (0.72–1.00.72.00) | 0.053 | |
| Q3 | 0.95 (0.81–1.12) | 0.540 | 0.91 (0.77–1.08) | 0.287 | 0.88 (0.75–1.04) | 0.135 | |
| Q4 | 0.84 (0.72–1.00.72.00) | 0.046 | 0.79 (0.67–0.94) | 0.008 | 0.74 (0.62–0.88) | 0.001 | |
| Asthma | Q1 | Ref. | — | Ref. | — | Ref. | — |
| Q2 | 0.88 (0.63–1.23) | 0.456 | 0.91 (0.66–1.27) | 0.593 | 0.90 (0.64–1.26) | 0.531 | |
| Q3 | 0.90 (0.65–1.25) | 0.540 | 0.99 (0.71–1.38) | 0.949 | 0.92 (0.66–1.29) | 0.643 | |
| Q4 | 1.07 (0.78–1.46) | 0.686 | 1.18 (0.85–1.64) | 0.325 | 1.03 (0.74–1.44) | 0.863 |
P-values < 0.001 are displayed as < 0.001. Model 1: Adjusted for age and gender. Model 2: Adjusted for Model 1 variables plus education, marital status, and residence. Model 3: Adjusted for Model 2 variables plus smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, self-reported health, and chronic disease count at baseline
ADL, Activities of Daily Living; CI, Confidence Interval; HR, Hazard Ratio; MET, Metabolic Equivalent of Task
RCS models, fully adjusted as in Model 3, showed significant non-linear dose-response relationships between the MCMI and the risk of hypertension, diabetes, stroke, and dyslipidemia (all P for non-linearity < 0.05) (Fig. 3). In contrast, the associations for heart disease and digestive system diseases were approximately linear (both P for non-linearity > 0.05).
Fig. 3.
Dose-response relationship between the Modified Cardiometabolic Index (MCMI) and the risk of incident chronic diseases. Hazard ratios were estimated using restricted cubic spline Cox regression models with four knots placed at the 5th, 35th, 65th, and 95th percentiles of the MCMI distribution. The reference value for the hazard ratio was set at the median MCMI value. The solid lines represent the estimated hazard ratios, and the shaded areas represent the 95% confidence intervals. All models were adjusted for age, gender, education, marital status, residence, smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, self-reported health, and chronic disease count at baseline. The P-values for overall association and for nonlinearity are presented for each outcome. ADL, Activities of Daily Living; CI, Confidence Interval; HR, Hazard Ratio; MET, Metabolic Equivalent of Task
Significant interactions were found between the MCMI and age, sex, smoking status, and alcohol consumption for the risk of hypertension, diabetes, and dyslipidemia (all P for interaction < 0.05) (Table 4). These associations were stronger among participants aged ≥ 60 years, men, smokers, and those who consumed alcohol. Conversely, no significant interactions with the MCMI were observed for heart disease or stroke.
Table 4.
Stratified analysis of the association between continuous MCMI and incident chronic diseases
| Disease | Subgroup | Level | HR (95% CI) | P-value | P for interaction |
|---|---|---|---|---|---|
| Hypertension | Age group | < 60 | 1.10 (1.06–1.15) | < 0.001 | 0.047 |
| ≥ 60 | 1.24 (1.13–1.35) | < 0.001 | |||
| Gender | Male | 1.38 (1.23–1.54) | < 0.001 | < 0.001 | |
| Female | 1.10 (1.05–1.14) | < 0.001 | |||
| Smoking status | Yes | 1.26 (1.15–1.38) | < 0.001 | 0.008 | |
| No | 1.10 (1.06–1.14) | < 0.001 | |||
| Drinking status | Yes | 1.34 (1.20–1.50) | < 0.001 | 0.001 | |
| No | 1.10 (1.06–1.14) | < 0.001 | |||
| Diabetes | Age group | < 60 | 1.13 (1.09–1.17) | < 0.001 | < 0.001 |
| ≥ 60 | 1.37 (1.26–1.48) | < 0.001 | |||
| Gender | Male | 1.43 (1.32–1.56) | < 0.001 | < 0.001 | |
| Female | 1.13 (1.09–1.17) | < 0.001 | |||
| Smoking status | Yes | 1.42 (1.31–1.53) | < 0.001 | < 0.001 | |
| No | 1.12 (1.08–1.16) | < 0.001 | |||
| Drinking status | Yes | 1.43 (1.31–1.55) | < 0.001 | < 0.001 | |
| No | 1.13 (1.09–1.17) | < 0.001 | |||
| Cancer | Age group | < 60 | 1.08 (0.95–1.22) | 0.249 | 0.899 |
| ≥ 60 | 1.13 (0.86–1.49) | 0.368 | |||
| Gender | Male | 1.16 (0.85–1.58) | 0.363 | 0.999 | |
| Female | 1.08 (0.96–1.22) | 0.203 | |||
| Smoking status | Yes | 1.11 (0.82–1.52) | 0.493 | 0.987 | |
| No | 1.09 (0.97–1.23) | 0.154 | |||
| Drinking status | Yes | 1.27 (0.98–1.64) | 0.074 | 0.446 | |
| No | 1.06 (0.91–1.23) | 0.463 | |||
| Lung Disease | Age group | < 60 | 1.07 (0.93–1.22) | 0.342 | 0.010 |
| ≥ 60 | 0.88 (0.77–1.01) | 0.066 | |||
| Gender | Male | 0.88 (0.77–1.02) | 0.080 | 0.119 | |
| Female | 1.05 (0.93–1.19) | 0.451 | |||
| Smoking status | Yes | 0.90 (0.78–1.04) | 0.145 | 0.354 | |
| No | 1.02 (0.90–1.16) | 0.755 | |||
| Drinking status | Yes | 0.87 (0.76–1.01) | 0.072 | 0.142 | |
| No | 1.04 (0.92–1.18) | 0.520 | |||
| Heart Disease | Age group | < 60 | 1.11 (1.05–1.17) | < 0.001 | 0.066 |
| ≥ 60 | 1.03 (0.92–1.16) | 0.591 | |||
| Gender | Male | 1.12 (0.99–1.26) | 0.070 | 0.551 | |
| Female | 1.09 (1.02–1.16) | 0.011 | |||
| Smoking status | Yes | 1.07 (0.94–1.21) | 0.307 | 0.758 | |
| No | 1.10 (1.04–1.17) | 0.002 | |||
| Drinking status | Yes | 1.05 (0.92–1.21) | 0.467 | 0.502 | |
| No | 1.10 (1.04–1.16) | < 0.001 | |||
| Stroke | Age group | < 60 | 1.10 (1.03–1.18) | 0.005 | 0.440 |
| ≥ 60 | 1.21 (1.09–1.36) | < 0.001 | |||
| Gender | Male | 1.25 (1.12–1.39) | < 0.001 | 0.063 | |
| Female | 1.10 (1.02–1.18) | 0.010 | |||
| Smoking status | Yes | 1.24 (1.11–1.39) | < 0.001 | 0.162 | |
| No | 1.10 (1.03–1.18) | 0.004 | |||
| Drinking status | Yes | 1.18 (1.03–1.35) | 0.019 | 0.485 | |
| No | 1.11 (1.05–1.18) | < 0.001 | |||
| Psychological Problem | Age group | < 60 | 0.80 (0.61–1.04) | 0.094 | 0.678 |
| ≥ 60 | 0.91 (0.69–1.22) | 0.536 | |||
| Gender | Male | 1.01 (0.74–1.39) | 0.932 | 0.272 | |
| Female | 0.78 (0.62–0.99) | 0.041 | |||
| Smoking status | Yes | 0.97 (0.68–1.38) | 0.860 | 0.304 | |
| No | 0.81 (0.65–1.01) | 0.066 | |||
| Drinking status | Yes | 0.75 (0.53–1.07) | 0.117 | 0.641 | |
| No | 0.89 (0.71–1.12) | 0.328 | |||
| Arthritis | Age group | < 60 | 1.00 (0.92–1.08) | 0.949 | 0.620 |
| ≥ 60 | 0.96 (0.85–1.07) | 0.428 | |||
| Gender | Male | 0.98 (0.87–1.11) | 0.779 | 0.642 | |
| Female | 0.98 (0.91–1.07) | 0.698 | |||
| Smoking status | Yes | 1.01 (0.90–1.14) | 0.859 | 0.607 | |
| No | 0.97 (0.90–1.06) | 0.538 | |||
| Drinking status | Yes | 1.02 (0.90–1.16) | 0.757 | 0.441 | |
| No | 0.97 (0.89–1.05) | 0.437 | |||
| Dyslipidemia | Age group | < 60 | 1.10 (1.07–1.13) | < 0.001 | 0.037 |
| ≥ 60 | 1.22 (1.14–1.31) | < 0.001 | |||
| Gender | Male | 1.32 (1.24–1.42) | < 0.001 | < 0.001 | |
| Female | 1.09 (1.05–1.12) | < 0.001 | |||
| Smoking status | Yes | 1.28 (1.20–1.37) | < 0.001 | < 0.001 | |
| No | 1.09 (1.06–1.13) | < 0.001 | |||
| Drinking status | Yes | 1.31 (1.22–1.41) | < 0.001 | < 0.001 | |
| No | 1.09 (1.06–1.13) | < 0.001 | |||
| Liver Disease | Age group | < 60 | 1.05 (0.96–1.15) | 0.251 | 0.956 |
| ≥ 60 | 1.09 (0.92–1.28) | 0.311 | |||
| Gender | Male | 1.13 (0.97–1.32) | 0.109 | 0.270 | |
| Female | 1.04 (0.93–1.16) | 0.497 | |||
| Smoking status | Yes | 1.14 (0.98–1.32) | 0.086 | 0.291 | |
| No | 1.03 (0.92–1.15) | 0.606 | |||
| Drinking status | Yes | 1.12 (0.95–1.32) | 0.195 | 0.633 | |
| No | 1.04 (0.94–1.15) | 0.450 | |||
| Kidney Disease | Age group | < 60 | 0.99 (0.86–1.13) | 0.848 | 0.819 |
| ≥ 60 | 0.98 (0.85–1.15) | 0.843 | |||
| Gender | Male | 0.93 (0.79–1.09) | 0.352 | 0.395 | |
| Female | 1.02 (0.91–1.13) | 0.768 | |||
| Smoking status | Yes | 1.06 (0.91–1.23) | 0.455 | 0.204 | |
| No | 0.93 (0.81–1.08) | 0.336 | |||
| Drinking status | Yes | 1.03 (0.87–1.22) | 0.708 | 0.620 | |
| No | 0.94 (0.82–1.09) | 0.417 | |||
| Digestive Disease | Age group | < 60 | 0.95 (0.85–1.05) | 0.301 | 0.084 |
| ≥ 60 | 0.83 (0.73–0.95) | 0.006 | |||
| Gender | Male | 0.80 (0.69–0.92) | 0.002 | 0.054 | |
| Female | 0.96 (0.87–1.05) | 0.369 | |||
| Smoking status | Yes | 0.81 (0.70–0.94) | 0.005 | 0.211 | |
| No | 0.95 (0.86–1.05) | 0.314 | |||
| Drinking status | Yes | 0.79 (0.69–0.92) | 0.002 | 0.073 | |
| No | 0.95 (0.86–1.05) | 0.321 | |||
| Asthma | Age group | < 60 | 1.03 (0.80–1.33) | 0.806 | 0.412 |
| ≥ 60 | 0.91 (0.74–1.13) | 0.400 | |||
| Gender | Male | 0.89 (0.70–1.13) | 0.339 | 0.351 | |
| Female | 1.03 (0.83–1.28) | 0.791 | |||
| Smoking status | Yes | 0.90 (0.71–1.14) | 0.369 | 0.434 | |
| No | 1.03 (0.82–1.29) | 0.784 | |||
| Drinking status | Yes | 0.79 (0.61–1.02) | 0.075 | 0.049 | |
| No | 1.11 (0.91–1.36) | 0.282 |
All hazard ratios were adjusted for age, gender, education, marital status, residence, smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, self-reported health, and chronic disease count at baseline, except for the stratification variable itself. P for interaction was derived from the coefficient of the cross-product term between the continuous MCMI and the dichotomized subgroup variable
ADL, Activities of Daily Living; CI, Confidence Interval; HR, Hazard Ratio; MET, Metabolic Equivalent of Task
We also assessed the association between the MCMI and the development of multimorbidity (Table 5). In the fully adjusted model, a higher MCMI was associated with an increased risk of developing a first chronic disease among initially healthy participants (HR 1.09, 95% CI 1.03–1.16). The association was stronger for the development of a second chronic disease among those with one prevalent condition at baseline (HR 1.22, 95% CI 1.13–1.31).
Table 5.
Association between continuous MCMI and the incidence of Multimorbidity
| Outcome | Model 1 h (95% CI) | P-value | Model 2 h (95% CI) | P-value | Model 3 h (95% CI) | P-value |
|---|---|---|---|---|---|---|
| Incidence of the first chronic disease (from 0 at baseline) | 1.11 (1.05–1.17) | 0.001 | 1.11 (1.05–1.17) | 0.001 | 1.09 (1.03–1.16) | 0.005 |
| Incidence of a second chronic disease (from 1 at baseline) | 1.23 (1.14–1.32) | < 0.001 | 1.23 (1.15–1.33) | < 0.001 | 1.22 (1.13–1.31) | < 0.001 |
Model 1: Adjusted for age and gender. Model 2: Adjusted for Model 1 variables plus education, marital status, and residence. Model 3: Adjusted for Model 2 variables plus smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, and self-reported health
ADL, Activities of Daily Living; CI, Confidence Interval; HR, Hazard Ratio; MET, Metabolic Equivalent of Task
The MCMI showed superior predictive performance for several chronic diseases compared with the CMI (Fig. 4). In fully adjusted models, the AUC for the MCMI was significantly higher than for the CMI in predicting hypertension (AUC 0.747 vs. 0.742, P < 0.001), stroke (AUC 0.692 vs. 0.686, P < 0.001), diabetes (AUC 0.663 vs. 0.647, P < 0.001), and dyslipidemia (AUC 0.657 vs. 0.650, P < 0.001).
Fig. 4.
Comparison of predictive accuracy between the Modified Cardiometabolic Index (MCMI) and the traditional Cardiometabolic Index (CMI) for incident chronic diseases. The figure displays the AUC and 95% confidence intervals from receiver operating characteristic (ROC) analyses. AUCs were derived from fully adjusted Cox proportional hazards models predicting the risk of 13 different chronic diseases, using either MCMI or CMI as the primary predictor. P-values assess the statistical significance of the difference between the AUC of the MCMI model and that of the CMI model, as determined by the DeLong test. Note that P-values displayed as 0 in the source data represent a statistically significant value of P < 0.001. All models were adjusted for age, gender, education, marital status, residence, smoking, drinking, physical activity (METs), C-reactive protein (CRP), ADL disability, self-reported health, and chronic disease count at baseline. ADL, Activities of Daily Living; AUC, Area Under the Curve; CMI, Cardiometabolic Index; MCMI, Modified Cardiometabolic Index; MET, Metabolic Equivalent of Task
The primary findings remained robust across three sensitivity analyses (Supplementary Table 2). The associations were consistent when repeating the analysis with complete cases only, were unchanged after excluding self-rated health from the model, and remained significant after excluding participants who developed an outcome within the first two years of follow-up to account for potential reverse causality.
Discussion
In this large, nationally representative prospective cohort, we provide the first systematic evaluation of the MCMI in relation to the incidence of 13 chronic diseases. Our principal finding is that a higher MCMI is significantly associated with an increased risk of hypertension, diabetes, heart disease, stroke, and dyslipidemia, with clear dose-response relationships. Importantly, the MCMI showed superior predictive accuracy for these key cardiometabolic diseases compared with the original CMI, supporting its clinical utility as an improved risk stratification tool.
It is important to interpret the magnitude of our findings with caution. While the per-unit risk increase for the continuous MCMI appeared modest, the clear dose-response relationship, with the highest quartile facing a substantially elevated risk for key outcomes like diabetes (HR 4.06), underscores its clinical relevance. From a public health standpoint, this suggests that even small shifts in this widely applicable index can translate into a significant disease burden at the population level, especially in the context of multimorbidity progression.
The strong association we observed between the MCMI and multiple cardiometabolic diseases has important clinical implications. The MCMI provides a more comprehensive risk assessment by integrating waist-to-height ratio, lipids, and blood glucose [4]. Its inclusion of glucose addresses a key limitation of the original CMI, which omits a core component of metabolic dysregulation [11].
While composite indices like TyG-BMI/WC and criteria such as MAFLD are valuable, MCMI’s primary strength lies in its composition [12–14]. It incorporates the TG/HDL-C ratio and WHtR, offering a more robust assessment of atherogenic dyslipidemia and central obesity compared to the less specific components of TyG-based indices [15]. Similarly, while MAFLD is essential for clinical diagnosis, its limitation for population screening is its reliance on imaging and a binary outcome. In contrast, MCMI operates as a continuous, low-cost risk score, making it a more practical and scalable tool for granular risk stratification [16–18].
The observed association between the MCMI and cardiometabolic disease risk is biologically plausible, reflecting the synergistic effects of its components [19]. Waist-to-height ratio is a marker of central obesity, a key driver of insulin resistance, chronic inflammation, and endothelial dysfunction [20]. The triglyceride to high-density lipoprotein cholesterol ratio indicates atherogenic dyslipidemia, characterized by a higher burden of small, dense low-density lipoprotein particles [15]. The inclusion of blood glucose further strengthens the index by directly capturing hyperglycemia, which promotes oxidative stress, the formation of advanced glycation end-products, and vascular damage [21]. The interplay of these components likely accelerates atherosclerosis and elevates cardiometabolic risk through these interconnected pathological pathways.
An unexpected finding was that a higher MCMI was associated with a lower risk of digestive system diseases. This observation, while seemingly counterintuitive, may have several explanations. It could reflect the “obesity paradox,” where a higher body mass might be protective in certain conditions [22]. Alternatively, this association might be confounded by unmeasured lifestyle or dietary factors [23]. The heterogeneity of conditions included under the broad category of digestive system diseases may also have contributed to this inverse association. This novel finding requires validation and further investigation to elucidate the potential underlying mechanisms.
The predictive value of the MCMI was particularly strong in high-risk subgroups, including older adults, men, smokers, and alcohol drinkers, highlighting its utility for targeted screening in these populations. Furthermore, the MCMI was associated not only with the incidence of a first chronic disease but also, more strongly, with the progression to multimorbidity. This finding is of considerable public health relevance, suggesting the MCMI could help identify individuals who may benefit most from comprehensive health management strategies aimed at preventing the accumulation of chronic conditions [24].
This study has several limitations. First, the reliance on self-reported diagnoses for chronic diseases is subject to information bias. Second, despite extensive covariate adjustment, residual confounding cannot be entirely excluded. For instance, the stronger association observed for the incidence of a second chronic disease might be partially confounded by the type of the initial disease.Third, our findings are derived from a single cohort of middle-aged and older Chinese adults. Consequently, a key limitation is the lack of external validation in an independent population. Future studies are essential to confirm our findings and test the performance of MCMI across diverse ethnicities and age groups. Finally, while our lagged analysis mitigated the issue, the possibility of reverse causality for diseases with long subclinical phases remains.
Notwithstanding these limitations, our findings strongly support the clinical utility of the MCMI. As a composite index of routinely measured, low-cost biomarkers, the MCMI is a simple and practical tool for risk stratification. In primary care and public health, it could be used to identify individuals at high risk for cardiometabolic diseases and guide personalized prevention strategies. We recommend that individuals in the highest MCMI quartile be targeted for intensive lifestyle interventions, including diet, exercise, smoking cessation, and alcohol reduction, alongside enhanced clinical surveillance. For patients with an existing chronic condition, a high MCMI should prompt comprehensive management strategies focused on preventing the development of multimorbidity.
Conclusions
In conclusion, this large prospective cohort study establishes the MCMI as a robust, independent predictor of incident hypertension, type 2 diabetes, heart disease, stroke, and dyslipidemia in middle-aged and older Chinese adults. We demonstrate a clear dose-response relationship for these outcomes and show that the MCMI has significantly greater predictive accuracy than the original CMI. Furthermore, the MCMI is a strong predictor of the progression to multimorbidity, particularly the transition from a single chronic disease to multiple conditions. As a simple, low-cost index of central obesity, dyslipidemia, and hyperglycemia, the MCMI is a valuable and readily applicable tool for early risk stratification and for guiding targeted preventive strategies in both clinical and public health settings.
Supplementary Information
Acknowledgements
The authors would like to express their gratitude to the CHARLS team for providing access to the data.
Abbreviations
- ADL
Activities of daily living
- AUC
Area Under the Curve
- CHARLS
China health and retirement longitudinal study
- CI
Confidence interval
- CMI
Cardiometabolic index
- HR
Hazard ratio
- IQR
Interquartile range
- MCMI
Modified cardiometabolic index
- RCS
Restricted cubic splines
- SD
Standard deviation
- VIF
Variance inflation factor
Author contributions
Yilin Pan : Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Visualization, Writing – original draft, Writing – review & editing; Jingru Bi : Formal analysis, Visualization, Writing – review & editing; Long Feng : Investigation, Data curation; Junru Qu : Investigation, Data curation; Zhiqiang Wang : Investigation, Data curation; Beibei Du : Writing – review & editing, Supervision, Funding acquisition, Project administration.
Funding
This study was supported by National Natural Science Foundation of China (No. 82470327, 82100337), Jilin Provincial Natural Science Foundation (No. YDZJ202201ZYTS097) and the Project of Jilin Provincial Department of Finance (No. 2022SCZ40).
Data availability
The data that support the findings of this study are available from the China Health and Retirement Longitudinal Study repository ([http://charls.pku.edu.cn/index/en.html](http:/charls.pku.edu.cn/index/en.html)).
Declarations
Ethics approval and consent to participate
The CHARLS study was approved by the Biomedical Ethics Committee of Peking University (approval number: IRB00001052-11015). All participants provided written informed consent before participating in the survey.
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.
YilinPan and JingruBi contributed equally to this work.
References
- 1.Global burden. Of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396(10258):1204–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hajat C, Stein E. The global burden of multiple chronic conditions: a narrative review. Prev Med Rep. 2018;12:284–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pati S, Swain S, Knottnerus JA, Metsemakers JFM, van den Akker M. Health related quality of life in multimorbidity: a primary-care based study from Odisha, India. Health Qual Life Outcomes. 2019;17(1):116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Saklayen MG. The global epidemic of the metabolic syndrome. Curr Hypertens Rep. 2018;20(2):12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Luo Y, Yin Z, Li X, Sheng C, Zhang P, Wang D, et al. Cardiometabolic index predicts cardiovascular events in aging population: a machine learning-based risk prediction framework from a large-scale longitudinal study. Front Endocrinol (Lausanne). 2025;16:1551779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Guo Y, Su W, Tao L, Zhang G, Wang K. The association between modified cardiometabolic index with non-alcoholic fatty liver disease and liver fibrosis: a cross-sectional study. BMC Gastroenterol. 2025;25(1):265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhou M, Wang H, Zeng X, Yin P, Zhu J, Chen W, et al. Mortality, morbidity, and risk factors in China and its provinces, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019;394(10204):1145–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.He L, Lin C, Tu Y, Yang Y, Lin M, Tu H, et al. Correlation of cardiometabolic index and sarcopenia with cardiometabolic multimorbidity in middle-aged and older adult: a prospective study. Front Endocrinol (Lausanne). 2024;15:1387374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zhuo L, Lai M, Wan L, Zhang X, Chen R. Cardiometabolic index and the risk of new-onset chronic diseases: results of a national prospective longitudinal study. Front Endocrinol (Lausanne). 2024;15:1446276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wakabayashi I, Daimon T. The “cardiometabolic index” as a new marker determined by adiposity and blood lipids for discrimination of diabetes mellitus. Clin Chim Acta. 2015;438:274–8. [DOI] [PubMed] [Google Scholar]
- 12.Zhou Z, Liu Q, Zheng M, Zuo Z, Zhang G, Shi R, et al. Comparative study on the predictive value of TG/HDL-C, TyG and TyG-BMI indices for 5-year mortality in critically ill patients with chronic heart failure: a retrospective study. Cardiovasc Diabetol. 2024;23(1):213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chen Q, Hu P, Hou X, Sun Y, Jiao M, Peng L, et al. Association between triglyceride-glucose related indices and mortality among individuals with non-alcoholic fatty liver disease or metabolic dysfunction-associated steatotic liver disease. Cardiovasc Diabetol. 2024;23(1):232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tang X, Zhang K, He R. The association of triglyceride-glucose and triglyceride-glucose related indices with the risk of heart disease in a national. Cardiovasc Diabetol. 2025;24(1):54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kosmas CE, Rodriguez Polanco S, Bousvarou MD, Papakonstantinou EJ, Peña Genao E, Guzman E, et al. The triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio as a risk marker for metabolic syndrome and cardiovascular disease. Diagnostics (Basel). 2023;13(5):929. 10.3390/diagnostics13050929 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Pipitone RM, Ciccioli C, Infantino G, La Mantia C, Parisi S, Tulone A, et al. MAFLD: a multisystem disease. Ther Adv Endocrinol Metab. 2023;14:20420188221145549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ramandi A, George J, Merat S, Jafari E, Sharafkhah M, Radmard AR, et al. Polypill protects MAFLD patients from cardiovascular events and mortality: a prospective trial. Hepatol Int. 2023;17(4):882–8. [DOI] [PubMed] [Google Scholar]
- 18.Eslam M, Sanyal AJ, George J. MAFLD: a consensus-driven proposed nomenclature for metabolic associated fatty liver disease. Gastroenterology. 2020;158(7):1999-2014.e1. [DOI] [PubMed] [Google Scholar]
- 19.Tune JD, Goodwill AG, Sassoon DJ, Mather KJ. Cardiovascular consequences of metabolic syndrome. Transl Res. 2017;183:57–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ross R, Neeland IJ, Yamashita S, Shai I, Seidell J, Magni P, et al. Waist circumference as a vital sign in clinical practice: a consensus statement from the IAS and ICCR working group on visceral obesity. Nat Rev Endocrinol. 2020;16(3):177–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shah MS, Brownlee M. Molecular and cellular mechanisms of cardiovascular disorders in diabetes. Circ Res. 2016;118(11):1808–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Horwich TB, Fonarow GC, Clark AL. Obesity and the obesity paradox in heart failure. Prog Cardiovasc Dis. 2018;61(2):151–6. [DOI] [PubMed] [Google Scholar]
- 23.Aune D, Giovannucci E, Boffetta P, Fadnes LT, Keum N, Norat T, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality-a systematic review and dose-response meta-analysis of prospective studies. Int J Epidemiol. 2017;46(3):1029–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Smith SM, Wallace E, O’Dowd T, Fortin M. Interventions for improving outcomes in patients with multimorbidity in primary care and community settings. Cochrane Database Syst Rev. 2016;3(3):Cd006560. [DOI] [PMC free article] [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 that support the findings of this study are available from the China Health and Retirement Longitudinal Study repository ([http://charls.pku.edu.cn/index/en.html](http:/charls.pku.edu.cn/index/en.html)).