Abstract
Background
The C-reactive protein–triglyceride glucose index (CTI) has been associated with stroke risk in prior studies. However, its relevance among individuals with cardiovascular-kidney-metabolic (CKM) syndrome remains uncertain.
Methods
This study included 5767 participants with CKM syndrome stages 0–3 from the China Health and Retirement Longitudinal Study. CTI was calculated as: CTI = 0.412 × ln[hs-CRP (mg/L)] + ln[(triglyceride (mg/dL) × fasting glucose (mg/dL)) / 2]. The primary outcome was incident stroke, assessed via self-reported questionnaires, with follow-up spanning 2011 to 2020. Cox proportional hazards models, restricted cubic spline (RCS) analysis, and subgroup analyses were used to examine the association between CTI and stroke risk. Time-dependent receiver operating characteristic (ROC) curve analysis was performed to compare the predictive performance of CTI with the triglyceride–glucose index (TyG) and the metabolic score for insulin resistance. Multiple testing correction was performed using the false discovery rate approach.
Results
After adjustment for potential confounders, higher CTI levels were significantly associated with increased stroke risk (HR 1.33, 95% CI 1.18–1.51). RCS analysis indicated a linear association, with no evidence of nonlinearity (P for nonlinearity = 0.182). Subgroup analyses indicated that elevated CTI was significantly associated with higher stroke risk in individuals at CKM stages 2 (HR 1.27, 95% CI 1.04–1.56; P = 0.019, adjusted P = 0.023) and 3 (HR 1.25, 95% CI 1.04–1.50; P = 0.022, adjusted P = 0.025), but no significant association was observed at stages 0–1. ROC analysis revealed that CTI consistently outperformed TyG in predicting stroke at years 5 and 7, as confirmed by DeLong’s test (adjusted P = 0.028 for both).
Conclusion
CTI is positively and linearly associated with stroke risk in individuals with CKM syndrome, particularly in stages 2 and 3. Furthermore, CTI provides superior predictive accuracy compared to TyG.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-025-05143-3.
Keywords: C-reactive protein–triglyceride glucose index (CTI), Stroke, Cohort study, Cardiovascular-kidney-metabolic syndrome (CKM syndrome), Triglyceride-glucose index (TyG)
Introduction
Stroke remains one of the leading causes of disability and mortality worldwide, placing a substantial burden on both society and the economy [1]. Therefore, preventing its underlying causes is critically important. Major modifiable risk factors include hypertension, diabetes, hyperlipidemia, obesity, cardiovascular disease, metabolic syndrome, and chronic kidney disease, and effective management of these comorbidities can significantly reduce the risk of stroke [1, 2]. In recent years, growing recognition of the complex interplay among cardiovascular, renal, and metabolic disorders has prompted a shift from treating individual diseases to preventing systemic risk. To support this shift, the American Heart Association (AHA) proposed a new diagnostic framework known as Cardiovascular-Kidney-Metabolic (CKM) Syndrome, which aims to improve global prevention and management strategies through early identification and comprehensive intervention [3]. CKM Syndrome is categorized into stages 0 to 4, with stroke identified as a clinical event associated with stage 4. Accordingly, stroke prevention efforts should primarily focus on individuals in stages 0 through 3, where timely interventions can help halt or slow disease progression.
The C-reactive protein–triglyceride glucose index (CTI) combines markers of inflammation (CRP) and insulin resistance (TyG), providing a more comprehensive assessment of both conditions. This integrated measure offers better predictive value for cardiovascular disease and diabetes than either marker alone [4–6]. The triglyceride–glucose index (TyG), first proposed by Simental et al. in 2008 as a surrogate marker for insulin resistance [7] has been linked to various diseases, including diabetes, hypertension, fatty liver, cardiovascular and cerebrovascular diseases, kidney disorders, reproductive system conditions, and malignancies [8]. Cohort studies and meta-analyses have consistently demonstrated a positive association between elevated TyG and stroke risk [9–11]. Moreover, higher TyG levels are associated with early neurological deterioration and worse stroke outcomes [12, 13]. Inflammation is also a well-established contributor to stroke pathogenesis [14] and as a key inflammatory marker, CRP plays an important role in atherosclerosis and stroke development [15, 16]. While CTI, as an integrated marker of both inflammation and insulin resistance, has been studied only sparingly, a recent study reported a positive association between CTI levels and stroke risk [17]. To date, however, no studies have examined the relationship between CTI and stroke in individuals with CKM syndrome, particularly in stages 0 to 3.
Therefore, this study aims to investigate the association between CTI and stroke risk among individuals with stages 0 to 3 CKM syndrome using data from the China Health and Retirement Longitudinal Study (CHARLS) database.
Methods
Study population
This study used data from CHARLS, a nationally representative longitudinal survey employing multistage stratified sampling. The survey targets individuals aged 45 years and older across 450 villages and urban communities in 150 counties and districts in China. Baseline data were collected in 2011, with follow-up waves conducted every 2 to 3 years [18]. To date, five waves have been completed: Wave 1 (2011), Wave 2 (2013), Wave 3 (2015), Wave 4 (2018), and Wave 5 (2020). Peking University’s Ethics Committee approved the CHARLS study, with informed consent from all participants.
We included participants diagnosed with stage 0–3 CKM syndrome in 2011 as the baseline cohort and followed them through 2020. Inclusion criteria required complete demographic information, medical history, physical examination data, and blood test results. A total of 5,767 participants met these criteria and were included in the final analysis (see Fig. 1 for details).
Fig. 1.
Flowchart of the study population. CHARLS China Health and Retirement Longitudinal Study, CKM syndrome cardiovascular-kidney-metabolic syndrome, CTI C-reactive protein–triglyceride glucose index, LDL-C low-density lipoprotein cholesterol
Definitions
The CTI was calculated using the following formula [5]:
 |
The Metabolic Score for Insulin Resistance (METS-IR) was calculated using the following formula [19]:
 |
where BMI = weight (kg) / height² (m²).
The CKM syndrome stages 0–3 classification system, as defined by the AHA, categorizes individuals based on their health status [3]. Stage 0 includes individuals with no identifiable CKM syndrome risk factors. Stage 1 refers to those with excess or dysfunctional adiposity. Stage 2 includes individuals with metabolic risk factors such as elevated triglycerides, hypertension, diabetes, or metabolic syndrome, as well as those with moderate- to high-risk chronic kidney disease (CKD). Stage 3 is defined by the presence of subclinical cardiovascular disease or equivalent risk conditions, including a high predicted risk of cardiovascular events or very high-risk CKD [3].
Follow-up of stroke events
The primary outcome was the first occurrence of stroke, including both hemorrhagic and ischemic types. Stroke events were identified through self-reports obtained via structured questionnaires. Participants were asked whether they had ever received a physician-confirmed diagnosis of stroke [20]. The timing of stroke onset was determined based on responses to the questions: “When were you first diagnosed with a stroke or first became aware of having had one?” and “When did your most recent stroke diagnosis occur?” The cohort was followed from 2011 through five interview waves, continuing until the first reported stroke event or the end of the follow-up period in 2020, whichever came first [20].
Covariates
In the baseline survey of Wave 1 conducted in 2011, data were collected on sociodemographic characteristics, health conditions, and blood test results. Demographic and health-related information was obtained using structured questionnaires administered by trained interviewers. Demographic variables included age, sex, educational level, and place of residence. Health-related variables covered smoking status, alcohol use, hypertension, diabetes, dyslipidemia, use of antihypertensive medications, antidiabetic medications, and antihyperlipidemic medications. Blood test data included measurements of hemoglobin, low-density lipoprotein cholesterol (LDL-C), creatinine, and fasting glucose.
Statistical analysis
Continuous variables with a normal distribution were presented as mean ± standard deviation (SD), and comparisons between groups were performed using the t test. For non-normally distributed continuous variables, data were expressed as median and interquartile range (IQR), and the Mann–Whitney U test was used for group comparisons. Categorical variables were summarized as frequency and percentage, with comparisons conducted using the chi-square test.
To examine the association between the CTI and stroke risk, a Cox proportional hazards regression model was employed. Results were reported as hazard ratios (HRs) with 95% confidence intervals (CIs). Three models were constructed for sensitivity analysis: Model 1 was unadjusted; Model 2 was adjusted for age, sex, education level, alcohol consumption, and smoking status; Model 3 included additional adjustments for hemoglobin level, diabetes, dyslipidemia, LDL-C, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications.
To further evaluate the impact of CTI, we conducted stratified analyses by dividing CTI into tertiles and plotted Kaplan–Meier cumulative incidence curves. Incidence rates (per 1000 person-years) were calculated by dividing the number of stroke events by the total person-years at risk within each tertile. In addition, a restricted cubic spline (RCS) analysis was performed to investigate potential nonlinear associations between CTI and stroke risk.
Subgroup analyses were conducted to evaluate potential effect modification by CKM syndrome stage (0–1, 2, 3), age (< 60 vs. ≥60 years), sex, alcohol consumption, smoking status, and place of residence. The predictive performance of CTI for stroke was assessed and compared with TyG and METS-IR using time-dependent receiver operating characteristic (ROC) curves and the area under the curve (AUC).
Variance inflation factors (VIFs) were used to assess multicollinearity in Model 3, with values < 5 indicating no significant multicollinearity. E-values were calculated to evaluate the potential impact of unmeasured confounding on the association between CTI and stroke [21].
All statistical analyses were conducted using R software (version 4.4.1). Statistical significance was set at a two-sided P-value < 0.05. For multiple comparisons, P values were adjusted using the false discovery rate (FDR) method, and FDR-adjusted P < 0.05 was considered statistically significant where applicable.
Results
Baseline characteristics of the study population
Among 5767 participants followed for 9 years, the mean age was 57.6 ± 8.8 years, and 54.8% were female. Compared to those without stroke, participants who experienced stroke had significantly higher rates of hypertension (36.2% vs. 20%, P < 0.001) and dyslipidemia (16.1% vs. 6.9%, P < 0.001), as well as elevated levels of CRP (median 1.2 vs. 0.9 mg/L, P < 0.001), triglycerides (1.6 vs. 1.4 mmol/L, P < 0.001), and CTI (8.9 vs. 8.7, P < 0.001). The proportion of participants at CKM stage 3 was also higher in the stroke group (44.7% vs. 30.2%, P < 0.001). Detailed characteristics are presented in Table 1.
Table 1.
Baseline characteristics of participants by stroke status (N = 5767)
| Characteristic | Levels | Overall | Non-stroke | Stroke | P value |
|---|---|---|---|---|---|
| N | 5767 | 5250 | 517 | ||
| Age (mean (SD)) | 57.6 (8.8) | 57.4 (8.8) | 59.8 (8.5) | < 0.001 | |
| Sex (%) | Female | 3161 (54.8) | 2890 (55.0) | 271 (52.4) | 0.271 |
| Male | 2606 (45.2) | 2360 (45.0) | 246 (47.6) | ||
| Education (%) | Below secondary | 3979 (69.0) | 3606 (68.7) | 373 (72.1) | 0.116 |
| Secondary or above | 1788 (31.0) | 1644 (31.3) | 144 (27.9) | ||
| Residence (%) | Rural | 3910 (67.8) | 3543 (67.5) | 367 (71.0) | 0.115 |
| Urban | 1857 (32.2) | 1707 (32.5) | 150 (29.0) | ||
| Smoking (%) | Current (still smoking) | 1675 (29.0) | 1518 (28.9) | 157 (30.4) | 0.004 |
| Former (now quit) | 487 (8.4) | 425 (8.1) | 62 (12.0) | ||
| Never | 3605 (62.5) | 3307 (63.0) | 298 (57.6) | ||
| Drinking (%) | No | 3798 (65.9) | 3463 (66.0) | 335 (64.8) | 0.628 |
| Yes | 1969 (34.1) | 1787 (34.0) | 182 (35.2) | ||
| Hypertension (%) | No | 4531 (78.6) | 4201 (80.0) | 330 (63.8) | < 0.001 |
| Yes | 1236 (21.4) | 1049 (20.0) | 187 (36.2) | ||
| Dyslipidemia (%) | No | 5320 (92.2) | 4886 (93.1) | 434 (83.9) | < 0.001 |
| Yes | 447 (7.8) | 364 (6.9) | 83 (16.1) | ||
| Diabetes (%) | No | 5508 (95.5) | 5030 (95.8) | 478 (92.5) | 0.001 |
| Yes | 259 (4.5) | 220 (4.2) | 39 (7.5) | ||
| HB (mean (SD), g/dl) | 14.4 (2.3) | 14.4 (2.2) | 14.6 (2.4) | 0.026 | |
| Glucose (mean (SD), mg/dL) | 108.9 (31.6) | 108.5 (31.1) | 112.3 (36.1) | 0.009 | |
| Creatinine (mean (SD), mg/dL) | 0.8 (0.2) | 0.8 (0.2) | 0.8 (0.2) | 0.009 | |
| Uric acid (mean (SD), mg/dL) | 4.4 (1.2) | 4.4 (1.2) | 4.5 (1.2) | 0.05 | |
| TC (mean (SD), mmol/l) | 5.0 (1.0) | 5.0 (1.0) | 5.2 (1.0) | < 0.001 | |
| HDL-C (mean (SD), mmol/l) | 1.3 (0.4) | 1.3 (0.4) | 1.3 (0.4) | < 0.001 | |
| LDL-C (mean (SD), mmol/l) | 3.0 (0.9) | 3.0 (0.9) | 3.2 (0.9) | 0.001 | |
| TG (mean (SD), mmol/l) | 1.5 (1.1) | 1.4 (1.1) | 1.6 (1.1) | < 0.001 | |
| CTI (mean (SD)) | 8.7 (0.8) | 8.7 (0.8) | 8.9 (0.8) | < 0.001 | |
| TyG (mean (SD)) | 8.7 (0.6) | 8.6 (0.6) | 8.8 (0.7) | < 0.001 | |
| CRP (median [IQR], mg/L) | 0.9 [0.5, 1.9] | 0.9 [0.5, 1.9] | 1.2 [0.7, 2.5] | < 0.001 | |
| CKM (%) | 0 | 448 (7.8) | 433 (8.2) | 15 (2.9) | < 0.001 |
| 1 | 1179 (20.4) | 1114 (21.2) | 65 (12.6) | ||
| 2 | 2322 (40.3) | 2116 (40.3) | 206 (39.8) | ||
| 3 | 1818 (31.5) | 1587 (30.2) | 231 (44.7) |
HB hemoglobin, TC total cholesterol, HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, TG triglycerides, CTI C-reactive protein–triglyceride glucose index, TyG triglyceride–glucose index, CRP C-reactive protein, CKM syndrome cardiovascular–kidney–metabolic syndrome
Associations between CTI and stroke risk
Table 2 presents the association between CTI and stroke incidence. In the unadjusted model (Model 1), a higher CTI was associated with an increased risk of stroke (HR 1.40, 95% CI 1.27–1.54). This association remained statistically significant after adjustment for all covariates in Model 3 (HR 1.33, 95% CI 1.18–1.51).
Table 2.
Sensitivity analyses of the association between CTI and stroke risk
| Variable | Model 1 | Model 2 | Model 3 | Absolute risk (per 1000 person-years) | |||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | ||
| CTI (pre 1-unit) | 1.40 (1.27, 1.54) | < 0.001 | 1.39 (1.26, 1.53) | < 0.001 | 1.33 (1.18, 1.51) | < 0.001 | |
| CTI (tertiles) | |||||||
| Low (5.83 < CTI ≤ 8.29) | Ref | Ref | Ref | 6.39 | |||
| Medium (8.29 < CTI ≤ 8.97) | 1.75 (1.38, 2.22) | < 0.001 | 1.69 (1.33, 2.14) | < 0.001 | 1.56 (1.23, 1.98) | < 0.001 | 11.1 |
| High (8.97 < CTI ≤ 12.79) | 2.09 (1.66, 2.63) | < 0.001 | 2.04 (1.62, 2.57) | < 0.001 | 1.71 (1.34, 2.19) | < 0.001 | 13.2 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
Model 1: Unadjusted. Model 2: Adjusted for age, sex, education level, alcohol consumption, and smoking status. Model 3: Additionally adjusted for hemoglobin level, diabetes, dyslipidemia, LDL-C, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications
CTI C-reactive protein–triglyceride glucose index, HR hazard ratio, CI confidence interval
When CTI was categorized into tertiles, participants in the middle (HR 1.56, 95% CI 1.23–1.98) and highest tertiles (HR 1.71, 95% CI 1.34–2.19) had significantly higher stroke risk compared to the lowest tertile (P for trend < 0.001), with corresponding incidence rates of 6.39, 11.1, and 13.2 per 1000 person-years.
Kaplan–Meier curves further confirmed a higher cumulative incidence of stroke in the highest CTI tertile (log-rank P < 0.001; Fig. 2).
Fig. 2.
Kaplan–Meier analysis of stroke incidence across CTI tertiles. CTI C-reactive protein–triglyceride glucose index
Dose–response relationship
The RCS analysis revealed a linear relationship between CTI and stroke risk (P overall < 0.001), with no evidence of nonlinearity (P for nonlinearity = 0.182; Fig. 3).
Fig. 3.
Dose–response relationship between CTI and stroke risk. The restricted cubic spline curve illustrates the adjusted hazard ratios for stroke across CTI levels. The model was adjusted for age, sex, education level, alcohol consumption, smoking status, hemoglobin level, diabetes, dyslipidemia, low-density lipoprotein cholesterol, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications. CTI C-reactive protein–triglyceride glucose index, HR hazard ratio
Subgroup analyses
Stratified analysis showed that higher CTI was significantly associated with increased stroke risk in individuals at CKM stages 2 (HR 1.27, 95% CI 1.04–1.56; P = 0.019, adjusted P = 0.023) and 3 (HR 1.25, 95% CI 1.04–1.50; P = 0.022, adjusted P = 0.025), but no significant association was observed at stages 0–1 (Fig. 4). The positive association between higher CTI levels and stroke risk remained consistent across subgroups defined by age, sex, education level, place of residence, smoking status, and alcohol consumption. No significant interactions were detected in any of the subgroup analyses (all P > 0.05).
Fig. 4.
Subgroup analyses of stroke risk by CTI. Forest plot showing hazard ratios (HRs) and 95% confidence intervals (CIs) for stroke risk across subgroups defined by CTI. The model was adjusted for age, sex, education level, alcohol consumption, smoking status, hemoglobin level, diabetes, dyslipidemia, low-density lipoprotein cholesterol, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications. P values for multiple comparisons were adjusted using the false discovery rate method. CTI C-reactive protein–triglyceride glucose index, CKM syndrome cardiovascular–kidney–metabolic syndrome, HR hazard ratio, CI confidence interval
Predictive performance of CTI for stroke
Time-dependent ROC analysis showed that CTI consistently had higher AUC values than both TyG and METS-IR at 3, 5, 7, and 8 years (Figs. 5, 6). Detailed AUC values and corresponding 95% confidence intervals are presented in Supplementary Table 1. After applying DeLong’s test with FDR correction, the difference between CTI and TyG reached significance at years 5 and 7 (adjusted P = 0.028 for both). No statistically significant differences were observed between CTI and METS-IR at any time point.
Fig. 5.
Time-dependent ROC curves for CTI and TyG in stroke prediction. A–D illustrate the comparative predictive performance of CTI and TyG models for stroke at 3, 5, 7, and 8 years, respectively. Models were adjusted for age, sex, education level, alcohol consumption, smoking status, hemoglobin level, diabetes, dyslipidemia, low-density lipoprotein cholesterol, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications. P values for multiple comparisons were adjusted using the false discovery rate method. CTI C-reactive protein–triglyceride glucose index, TyG triglyceride–glucose index, ROC receiver operating characteristic, AUC area under the curve, Adj. P adjusted P value
Fig. 6.
Time-dependent ROC curves for CTI and METS-IR in stroke prediction. A–D illustrate the comparative predictive performance of CTI and METS-IR models for stroke at 3, 5, 7, and 8 years, respectively. Models were adjusted for age, sex, education level, alcohol consumption, smoking status, hemoglobin level, diabetes, dyslipidemia, low-density lipoprotein cholesterol, fasting glucose, creatinine, place of residence, hypertension, and the use of antihyperlipidemic, antihypertensive, and antidiabetic medications. P values for multiple comparisons were adjusted using the false discovery rate method. CTI C-reactive protein–triglyceride glucose index, METS-IR metabolic score for insulin resistance, ROC receiver operating characteristic, AUC area under the curve, Adj. P adjusted P value
Sensitivity analysis
To evaluate the robustness of our findings, we assessed multicollinearity and the potential impact of unmeasured confounding. All VIFs in Model 3 were < 5, indicating no significant multicollinearity. E-values were 1.99 for CTI as a continuous variable and 2.81 for the highest vs. lowest tertile comparison, suggesting the observed associations were relatively robust to unmeasured confounding.
Discussion
Using nationally representative cohort data from CHARLS, we found that elevated CTI levels were significantly associated with an increased risk of stroke among individuals at CKM syndrome stages 0 to 3. This association remained robust after adjusting for all relevant covariates. The positive association persisted across most subgroups, except for the subgroup at CKM stages 0–1. RCS analysis revealed a significant linear relationship between CTI levels and stroke incidence. Moreover, ROC analysis showed that CTI exhibited superior predictive performance for stroke risk compared to TyG.
Insulin resistance and inflammation are recognized as risk factors for stroke, with numerous epidemiological studies confirming their association [22–25]. CTI, a novel marker reflecting both inflammation and insulin resistance, was first introduced by Ruan et al. for evaluating cancer prognosis [26]. Since then, CTI has been increasingly utilized in the prediction and prognosis of metabolic and cardiovascular diseases [4–6] and more recently, in stroke research as well [27]. In our study, we observed a positive linear association between CTI and stroke risk among individuals at CKM syndrome stages 0–3, consistent with findings reported by Huo et al. [17]. This linear relationship suggests that lower CTI levels were associated with reduced stroke risk, which may have implications for risk assessment and prevention strategies. For example, according to the CTI formula—CTI = 0.412 × ln[hs-CRP (mg/L)] + ln[(triglyceride (mg/dL) × fasting glucose (mg/dL))/2]—managing fasting glucose, triglycerides, and hs-CRP levels may be helpful in reducing stroke risk, although further interventional studies are needed to confirm this potential.
Previous studies have shown that TyG, as a surrogate for insulin resistance, predicts various cardiometabolic outcomes. Elevated TyG levels are associated with chest pain and cardiovascular mortality, especially in patients with existing cardiovascular disease [28] and have been implicated in the risk of NAFLD/MAFLD [29] glucose metabolism transitions [30] and impaired cardiorespiratory fitness [31]. Longitudinal studies also associate TyG with incident hypertension [32] as well as prognostic value in heart failure [33] and ischemic stroke [9]. These findings support the clinical value of TyG across a wide spectrum of cardiometabolic disorders.
However, TyG’s focus on insulin resistance alone may limit its predictive scope. CTI enhances this by incorporating an inflammatory component, thereby capturing two central and interacting mechanisms of CKM syndrome. Our data demonstrated that CTI consistently outperformed TyG in stroke prediction, emphasizing the added value of inflammation in risk assessment.
The growing burden of CKM syndrome, particularly since the COVID-19 pandemic [34] has drawn attention to its association with frailty [35] reduced serum Klotho levels [36] increased social vulnerability [37] and hormonal dysregulation [38]. CKM syndrome stages differ in their links to mortality and cardiovascular outcomes [39] highlighting the need for integrative biomarkers such as CTI that can inform early prevention and intervention strategies.
The CTI may elevate stroke risk through multiple interrelated mechanisms. Insulin resistance impairs endothelial function and promotes platelet adhesion, activation, and aggregation, thereby increasing thrombosis risk and leading to stroke [40, 41]. It also contributes to the development of metabolic syndrome and accelerates atherosclerosis [42] while simultaneously triggering inflammation and oxidative stress, both linked to increased stroke incidence [40]. In addition to insulin resistance, inflammation independently contributes to endothelial dysfunction, cytokine release, and oxidative stress, further reinforcing atherosclerosis and vascular injury [43, 44]. Moreover, inflammation and insulin resistance interact bidirectionally: inflammation worsens insulin resistance, while insulin resistance induces the release of pro-inflammatory mediators, amplifying systemic inflammation and stroke risk [45, 46]. Cui et al. reported a mutual mediation effect between TyG and CRP in relation to cardiovascular disease [47]. In our study, ROC analysis showed that CTI had a stronger predictive value for stroke than TyG alone, indirectly supporting the hypothesis that inflammation and insulin resistance reinforce each other. Therefore, CTI may serve as a promising biomarker for assessing vascular and metabolic disease risk. Beyond the CTI, other composite indices incorporating inflammatory and lipid-related parameters have also shown prognostic relevance in stroke-related outcomes. For instance, the Atherogenic Plasma Index (API), calculated as a logarithmic ratio of triglycerides to HDL-C, has been associated with increased ischemic stroke risk and poor outcomes. Okşen et al. [48] showed that API was independently associated with one-year mortality in ischemic stroke patients, though the CHA₂DS₂-VASc score demonstrated superior prognostic value. Similarly, Aslan et al. [49] found a significant association between elevated API and ischemic cerebrovascular events due to carotid artery disease. These findings underscore the potential of metabolically integrated indices in vascular risk assessment. Compared with API, CTI may offer broader mechanistic insight by combining insulin resistance and inflammatory status.
We found that CTI was not significantly associated with stroke risk in individuals at CKM syndrome stages 0–1. However, in stages 2 and 3, elevated CTI was significantly associated with an increased risk of stroke, which may be attributed to the higher levels of insulin resistance and inflammation typically present at these stages. Gao et al. reported a positive association between the Systemic Immune-Inflammation Index and the risk of CKM syndrome and its individual components [50]. According to the 2023 AHA official statement, CKM syndrome progresses from stage 0 to stage 4. Stage 2 is marked by metabolic risk factors or moderate CKD, involving mechanisms such as insulin resistance, dyslipidemia, and inflammation. Stage 3 is characterized by subclinical cardiovascular damage—such as coronary artery calcification or elevated high-sensitivity cardiac biomarkers—where inflammation, insulin resistance, and organ injury interact more extensively [51]. These findings underscore the importance of closely monitoring patients at these stages, as elevated CTI levels were associated with increased stroke risk, highlighting its potential value in identifying individuals at higher risk within this population.
CTI demonstrates strong potential for clinical implementation, as it can be readily calculated from routine blood tests, including fasting glucose, triglycerides, and CRP, enhancing its applicability in standard healthcare settings. In our study, CTI consistently showed higher AUC values for stroke prediction than TyG and METS-IR (Figs. 5, 6), with statistically significant differences observed compared to TyG at years 5 and 7. These findings suggest that CTI may offer added predictive value by capturing both inflammatory and metabolic risk components. Therefore, incorporating CTI into traditional risk models, such as the Framingham Stroke Risk Score, may improve predictive performance. However, further studies are required to validate this potential enhancement. Although causal relationships cannot be established due to the observational design, the observed linear association between CTI and stroke risk highlights its potential utility in clinical risk stratification. CTI may serve as a practical and accessible biomarker to aid integrated vascular risk assessment and guide early intervention strategies.
This study has several strengths, including its nationwide cohort design focusing on individuals in CKM syndrome stages 0–3 and a large sample size that provides adequate statistical power. However, several limitations should be noted. First, stroke diagnoses were based on self-reported questionnaires without classification by stroke subtype. Self-reported data may introduce recall bias and misclassification, particularly underdiagnosis of silent strokes, which could potentially attenuate the observed associations. The validity of self-reported stroke has been evaluated in previous studies [52, 53]. Second, despite adjusting for a wide range of covariates, residual confounding from unmeasured or imprecisely measured factors (e.g., lifestyle or inflammatory markers) cannot be ruled out, though our E-value analysis suggests that a strong unmeasured confounder would be needed to nullify the observed association. Third, a substantial number of participants were excluded due to missing data, which may have led to selection bias and potentially affected the magnitude or direction of the association between CTI and stroke risk. Fourth, we did not assess whether CTI improves risk prediction beyond established clinical stroke risk scores, such as the Framingham Stroke Risk Profile, due to limitations in available data; this represents an important direction for future research. Fifth, as the study population consisted exclusively of Chinese individuals, the generalizability of our findings to other ethnic or geographic populations may be limited. External validation in diverse cohorts is needed to confirm the association between CTI and stroke risk. Finally, although our findings indicate the potential utility of CTI as a risk marker, further research is required to determine clinically meaningful cutoff values and to explore its integration into existing stroke risk prediction models. Given the observational nature of our study, establishing a definitive cutoff may not be appropriate at this stage, and future studies are needed to validate thresholds in clinical contexts.
Conclusion
In individuals with CKM syndrome stages 0 to 3, this study found a linear, positive association between the CTI and stroke risk. The association was particularly pronounced in stages 2 and 3, highlighting its potential value in risk assessment. Additionally, CTI outperformed TyG in predicting stroke, underscoring its potential utility as a marker for vascular and metabolic disease assessment.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The data used in this study were derived from CHARLS. We would like to thank the CHARLS research team for their time and effort in conducting the CHARLS project.
Author contributions
YSX and WBQ designed the study; TL and SQC collected the data; TL and SQC performed statistical analyses and interpreted the results; YSX, SQC, JYZ, QYW, WYL, GHP, LT and YHH drafted the manuscript; and WBQ revised the manuscript. The final manuscript was read and approved by all authors.
Funding
No funding was provided.
Data availability
The dataset used in this study was publicly available and can be accessed at http://charls.pku.edu.cn/en. The data generated from the analysis can be obtained from the corresponding author upon request.
Declarations
Ethics approval and consent to participate
The study involving human participants was reviewed and approved by the Ethics Review Committee of Peking University (approval numbers: IRB00001052-11015 for the household survey and IRB00001052-11014 for blood samples). The CHARLS study was conducted in accordance with the Declaration of Helsinki, and written informed consent had been obtained from all participants before their participation in this study.
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.
Yinsong Xu, Shiqin Chen and Jingying Zhu contributed equally to this work.
References
- 1.Boehme AK, Esenwa C, Elkind MS. Stroke risk factors, genetics, and prevention. Circ Res. 2017;120:472–95. 10.1161/CIRCRESAHA.116.308398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kelly DM, Rothwell PM. Does chronic kidney disease predict stroke risk independent of blood pressure? A systematic review and meta-regression. Stroke. 2019;50:3085–92. 10.1161/STROKEAHA.119.025442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ndumele CE, Rangaswami J, Chow SL, Neeland IJ, Tuttle KR, Khan SS, Coresh J, Mathew RO, Baker-Smith CM, Carnethon MR, et al. Cardiovascular-kidney-metabolic health: a presidential advisory from the American heart association. Circulation. 2023;148:1606–35. 10.1161/CIR.0000000000001184. [DOI] [PubMed] [Google Scholar]
- 4.Xu M, Zhang L, Xu D, Shi W, Zhang W. 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. Front Cardiovasc Med. 2024;11:1485538. 10.3389/fcvm.2024.1485538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gao A, Peng B, Gao Y, Yang Z, Li Z, Guo T, Qiu H, Gao R. 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;17:157. 10.1186/s13098-025-01687-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.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 Disord. 2025;25:126. 10.1186/s12902-025-01947-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Simental-Mendia LE, Rodriguez-Moran M, Guerrero-Romero F. The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord. 2008;6:299–304. 10.1089/met.2008.0034. [DOI] [PubMed] [Google Scholar]
- 8.Sun Y, Ji H, Sun W, An X, Lian F. Triglyceride glucose (TyG) index: a promising biomarker for diagnosis and treatment of different diseases. Eur J Intern Med. 2025;131:3–14. 10.1016/j.ejim.2024.08.026. [DOI] [PubMed] [Google Scholar]
- 9.Wu Y, Yang Y, Zhang J, Liu S, Zhuang W. The change of triglyceride-glucose index May predict incidence of stroke in the general population over 45 years old. Cardiovasc Diabetol. 2023;22:132. 10.1186/s12933-023-01870-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Liao C, Xu H, Jin T, Xu K, Xu Z, Zhu L, Liu M. Triglyceride-glucose index and the incidence of stroke: a meta-analysis of cohort studies. Front Neurol. 2022;13:1033385. 10.3389/fneur.2022.1033385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Yang Y, Huang X, Wang Y, Leng L, Xu J, Feng L, Jiang S, Wang J, Yang Y, Pan G, et al. The impact of triglyceride-glucose index on ischemic stroke: a systematic review and meta-analysis. Cardiovasc Diabetol. 2023;22:2. 10.1186/s12933-022-01732-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Liu Y, Wang Z, Zhang Z, Lu Z, Zhang L, Ding W, Fang K, Pan X, Ni M, Liu J. Correlation between triglyceride-glucose index and early neurological deterioration in patients with acute mild ischemic stroke. Front Neurol. 2024;15:1441116. 10.3389/fneur.2024.1441116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sun W, Shen H, Wu X, He A, Yao X, Chen F, Song H, Huang X. Influence of TyG index on large vascular occlusive stroke following endovascular treatment. CNS Neurosci Ther. 2024;30:e70143. 10.1111/cns.70143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chaturvedi S, De Marchis GM. Inflammatory biomarkers and stroke subtype: an important new frontier. Neurology. 2024;102:e208098. 10.1212/WNL.0000000000208098. [DOI] [PubMed] [Google Scholar]
- 15.Libby P, Ridker PM, Hansson GK. Leducq transatlantic network on A. Inflammation in atherosclerosis: from pathophysiology to practice. J Am Coll Cardiol. 2009;54:2129–38. 10.1016/j.jacc.2009.09.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zhou Y, Han W, Gong D, Man C, Fan Y. Hs-CRP in stroke: a meta-analysis. Clin Chim Acta. 2016;453:21–7. 10.1016/j.cca.2015.11.027. [DOI] [PubMed] [Google Scholar]
- 17.Huo G, Tang Y, Liu Z, Cao J, Yao Z, Zhou D. 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;24:142. 10.1186/s12933-025-02686-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43:61–8. 10.1093/ije/dys203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bello-Chavolla OY, Almeda-Valdes P, Gomez-Velasco D, Viveros-Ruiz T, Cruz-Bautista I, Romo-Romo A, Sanchez-Lazaro D, Meza-Oviedo D, Vargas-Vazquez A, Campos OA, et al. METS-IR, a novel score to evaluate insulin sensitivity, is predictive of visceral adiposity and incident type 2 diabetes. Eur J Endocrinol. 2018;178:533–44. 10.1530/EJE-17-0883. [DOI] [PubMed] [Google Scholar]
- 20.Qu L, Fang S, Lan Z, Xu S, Jiang J, Pan Y, Xu Y, Zhu X, Jin J. Association between atherogenic index of plasma and new-onset stroke in individuals with different glucose metabolism status: insights from CHARLS. Cardiovasc Diabetol. 2024;23:215. 10.1186/s12933-024-02314-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the E-Value. Ann Intern Med. 2017;167:268–74. 10.7326/M16-2607. [DOI] [PubMed] [Google Scholar]
- 22.Rundek T, Gardener H, Xu Q, Goldberg RB, Wright CB, Boden-Albala B, Disla N, Paik MC, Elkind MS, Sacco RL. Insulin resistance and risk of ischemic stroke among nondiabetic individuals from the Northern Manhattan study. Arch Neurol. 2010;67:1195–200. 10.1001/archneurol.2010.235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jin A, Wang S, Li J, Wang M, Lin J, Li H, Meng X, Wang Y, Pan Y. Mediation of systemic inflammation on insulin resistance and prognosis of nondiabetic patients with ischemic stroke. Stroke. 2023;54:759–69. 10.1161/STROKEAHA.122.039542. [DOI] [PubMed] [Google Scholar]
- 24.Chen W, Wang S, Lv W, Pan Y. Causal associations of insulin resistance with coronary artery disease and ischemic stroke: a Mendelian randomization analysis. BMJ Open Diabetes Res Care. 2020;8. 10.1136/bmjdrc-2020-001217. [DOI] [PMC free article] [PubMed]
- 25.Chei CL, Yamagishi K, Kitamura A, Kiyama M, Imano H, Ohira T, Cui R, Tanigawa T, Sankai T, Ishikawa Y, et al. C-reactive protein levels and risk of stroke and its subtype in japanese: the circulatory risk in communities study (CIRCS). Atherosclerosis. 2011;217:187–93. 10.1016/j.atherosclerosis.2011.03.001. [DOI] [PubMed] [Google Scholar]
- 26.Ruan GT, Xie HL, Zhang HY, Liu CA, Ge YZ, Zhang Q, Wang ZW, Zhang X, Tang M, Song MM, et al. A novel inflammation and insulin resistance related indicator to predict the survival of patients with cancer. Front Endocrinol (Lausanne). 2022;13:905266. 10.3389/fendo.2022.905266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tang S, Wang H, Li K, Chen Y, Zheng Q, Meng J, Chen X. C-reactive protein-triglyceride glucose index predicts stroke incidence in a hypertensive population: a National cohort study. Diabetol Metab Syndr. 2024;16:277. 10.1186/s13098-024-01529-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Zhao Y, Gu Y, Zhang B. Associations of triglyceride-glucose (TyG) index with chest pain incidence and mortality among the U.S. Population. Cardiovasc Diabetol. 2024;23:111. 10.1186/s12933-024-02209-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Xue Y, Xu J, Li M, Gao Y. Potential screening indicators for early diagnosis of NAFLD/MAFLD and liver fibrosis: triglyceride glucose index-related parameters. Front Endocrinol (Lausanne). 2022;13:951689. 10.3389/fendo.2022.951689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chen X, Liu D, He W, Hu H, Wang W. Predictive performance of triglyceride glucose index (TyG index) to identify glucose status conversion: a 5-year longitudinal cohort study in Chinese pre-diabetes people. J Transl Med. 2023;21:624. 10.1186/s12967-023-04402-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Guo D, Wu Z, Xue F, Chen S, Ran X, Zhang C, Yang J. Association between the triglyceride-glucose index and impaired cardiovascular fitness in non-diabetic young population. Cardiovasc Diabetol. 2024;23:39. 10.1186/s12933-023-02089-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Xin F, He S, Zhou Y, Jia X, Zhao Y, Zhao H. The triglyceride glucose index trajectory is associated with hypertension: a retrospective longitudinal cohort study. Cardiovasc Diabetol. 2023;22:347. 10.1186/s12933-023-02087-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Khalaji A, Behnoush AH, Khanmohammadi S, Ghanbari Mardasi K, Sharifkashani S, Sahebkar A, Vinciguerra C, Cannavo A. Triglyceride-glucose index and heart failure: a systematic review and meta-analysis. Cardiovasc Diabetol. 2023;22:244. 10.1186/s12933-023-01973-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Trimarco V, Izzo R, Pacella D, Virginia Manzi M, Trama U, Lembo M, Piccinocchi R, Gallo P, Esposito G, Morisco C, et al. Increased prevalence of cardiovascular-kidney-metabolic syndrome during COVID-19: a propensity score-matched study. Diabetes Res Clin Pract. 2024;218:111926. 10.1016/j.diabres.2024.111926. [DOI] [PubMed] [Google Scholar]
- 35.Santulli G, Visco V, Varzideh F, Guerra G, Kansakar U, Gasperi M, Marro A, Wilson S, Ferrante MNV, Pansini A, et al. Prediabetes increases the risk of frailty in prefrail older adults with hypertension: beneficial effects of Metformin. Hypertension. 2024;81:1637–43. 10.1161/HYPERTENSIONAHA.124.23087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Tang J, Xu Z, Ren L, Xu J, Chen X, Jin Y, Liang R, Zhang H. Association of serum Klotho with the severity and mortality among adults with cardiovascular-kidney-metabolic syndrome. Lipids Health Dis. 2024;23:408. 10.1186/s12944-024-02400-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Li J, Lei L, Wang W, Ding W, Yu Y, Pu B, Peng Y, Li Y, Zhang L, Guo Y. Social risk profile and cardiovascular-kidney-metabolic syndrome in US adults. J Am Heart Assoc. 2024;13:e034996. 10.1161/JAHA.124.034996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Guldan M, Unlu S, Abdel-Rahman SM, Ozbek L, Gaipov A, Covic A, Soler MJ, Covic A, Kanbay M. Understanding the role of sex hormones in cardiovascular kidney metabolic syndrome: toward personalized therapeutic approaches. J Clin Med. 2024;13. 10.3390/jcm13154354. [DOI] [PMC free article] [PubMed]
- 39.Li N, Li Y, Cui L, Shu R, Song H, Wang J, Chen S, Liu B, Shi H, Gao H, et al. Association between different stages of cardiovascular-kidney-metabolic syndrome and the risk of all-cause mortality. Atherosclerosis. 2024;397:118585. 10.1016/j.atherosclerosis.2024.118585. [DOI] [PubMed] [Google Scholar]
- 40.Ding PF, Zhang HS, Wang J, Gao YY, Mao JN, Hang CH, Li W. Insulin resistance in ischemic stroke: mechanisms and therapeutic approaches. Front Endocrinol (Lausanne). 2022;13:1092431. 10.3389/fendo.2022.1092431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Hierons SJ, Marsh JS, Wu D, Blindauer CA, Stewart AJ. The interplay between non-esterified fatty acids and plasma zinc and its influence on thrombotic risk in obesity and type 2 diabetes. Int J Mol Sci. 2021;22. 10.3390/ijms221810140. [DOI] [PMC free article] [PubMed]
- 42.Di Pino A, DeFronzo RA. Insulin resistance and atherosclerosis: implications for Insulin-Sensitizing agents. Endocr Rev. 2019;40:1447–67. 10.1210/er.2018-00141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Jayaraj RL, Azimullah S, Beiram R, Jalal FY, Rosenberg GA. Neuroinflammation: friend and foe for ischemic stroke. J Neuroinflamm. 2019;16:142. 10.1186/s12974-019-1516-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Parikh NS, Merkler AE, Iadecola C. Inflammation, autoimmunity, infection, and stroke: epidemiology and lessons from therapeutic intervention. Stroke. 2020;51:711–8. 10.1161/STROKEAHA.119.024157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Investig. 2006;116:1793–801. 10.1172/JCI29069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Puschel GP, Klauder J, Henkel J, Macrophages L-G, Inflammation. Insulin resistance and hyperinsulinemia: a mutual ambiguous relationship in the development of metabolic diseases. J Clin Med. 2022;11. 10.3390/jcm11154358. [DOI] [PMC free article] [PubMed]
- 47.Cui C, Liu L, Qi Y, Han N, Xu H, Wang Z, Shang X, Han T, Zha Y, Wei X, et al. Joint association of TyG index and high sensitivity C-reactive protein with cardiovascular disease: a National cohort study. Cardiovasc Diabetol. 2024;23:156. 10.1186/s12933-024-02244-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Okşen D, Aslan M, Yavuz YE. Comparative prognostic value of CHA2DS2-VASc score and atherogenic plasma index for one-year mortality prediction in ischemic stroke patients. Bagcilar Med Bull. 2025;10:8–14. 10.4274/BMB.galenos.2025.57966. [Google Scholar]
- 49.Aslan M, Okşen D, Yavuz YE, Kaynak Ç. Atherogenic plasma index is associated with ischemic cerebrovascular events due to carotid artery disease. Anatol Curr Med J. 2024;6:198–202. 10.38053/acmj.1448539. [Google Scholar]
- 50.Gao C, Gao S, Zhao R, Shen P, Zhu X, Yang Y, Duan C, Wang Y, Ni H, Zhou L, et al. Association between systemic immune-inflammation index and cardiovascular-kidney-metabolic syndrome. Sci Rep. 2024;14:19151. 10.1038/s41598-024-69819-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Ndumele CE, Neeland IJ, Tuttle KR, Chow SL, Mathew RO, Khan SS, Coresh J, Baker-Smith CM, Carnethon MR, Despres JP, et al. A synopsis of the evidence for the science and clinical management of cardiovascular-kidney-metabolic (CKM) syndrome: a scientific statement from the American heart association. Circulation. 2023;148:1636–64. 10.1161/CIR.0000000000001186. [DOI] [PubMed] [Google Scholar]
- 52.Britton A, Milne B, Butler T, Sanchez-Galvez A, Shipley M, Rudd A, Wolfe CD, Bhalla A, Brunner EJ. Validating self-reported strokes in a longitudinal UK cohort study (Whitehall II): extracting information from hospital medical records versus the hospital episode statistics database. BMC Med Res Methodol. 2012;12:83. 10.1186/1471-2288-12-83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Woodfield R, Group UKBSO, Follow-up UKB, Outcomes Working G, Sudlow CL. Accuracy of patient Self-Report of stroke: a systematic review from the UK biobank stroke outcomes group. PLoS One. 2015;10:e0137538. 10.1371/journal.pone.0137538. [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 dataset used in this study was publicly available and can be accessed at http://charls.pku.edu.cn/en. The data generated from the analysis can be obtained from the corresponding author upon request.