Association of sarcopenic obesity with advanced cardiovascular-kidney-metabolic syndrome and incident cardiovascular disease: a longitudinal and cross-sectional study Authors: Roushan Zhang, Dianping Yu, Xin Liu, Hui Xue, Jiajun Xie, Li Wang Published: 2026 ABSTRACT ObjectiveThis study aimed to examine the associations of sarcopenic obesity, obesity only, and sarcopenia only with advanced cardiovascular-kidney-metabolic (CKM) syndrome and incident cardiovascular disease (CVD), and to assess whether sarcopenia and obesity were jointly associated with long-term cardiovascular risk.Research design and methodsWe adopted a dual-cohort design combining longitudinal discovery and cross-sectional validation. The discovery cohort included 7,759 participants from the China Health and Retirement Longitudinal Study (CHARLS, 2011–2018). Body composition phenotypes were classified into four groups: Normal, Obesity Only, Sarcopenia Only, and Sarcopenic Obesity (SO). Sarcopenia was defined by AWGS 2019 criteria. CKM syndrome was staged (0–4) according to the 2023 American Heart Association (AHA) Presidential Advisory. Cox proportional hazards models were used to estimate the association between body composition phenotypes and incident CVD during follow-up. In participants with complete CKM staging data at both 2011 and 2015, a descriptive 4-year observed CKM transition probability matrix was constructed to describe stage movement from 2011 to 2015. External consistency was examined using data from 3,936 participants in NHANES 2011–2018, where muscle mass was measured by dual-energy X-ray absorptiometry (DXA).ResultsIn the CHARLS longitudinal cohort, sarcopenic obesity showed the highest cumulative incidence of CVD events (29.3%). After full adjustment, sarcopenic obesity was associated with a more than two-fold higher risk of incident CVD compared with the Normal group (HR 2.07, 95% CI 1.76–2.45), followed by Obesity Only (HR 1.51, 95% CI 1.31–1.74) and Sarcopenia Only (HR 1.23, 95% CI 1.03–1.45). A positive additive interaction between sarcopenia and obesity was observed (RERI = 0.34, P = 0.027). This association remained consistent in the restricted analysis among participants free of advanced CKM at baseline. In the descriptive four-year transition analysis, early CKM stages frequently moved toward higher-risk stages, whereas advanced stages were more persistent. In NHANES, sarcopenic obesity was associated with the highest likelihood of prevalent advanced CKM syndrome (OR 2.04, 95% CI 1.53–2.70).ConclusionSarcopenic obesity was consistently associated with a greater burden of advanced CKM and a higher risk of incident CVD in the CHARLS longitudinal cohort, with similar cross-sectional associations observed in the NHANES validation cohort. Although causal relationships cannot be established, these findings suggest that sarcopenic obesity may serve as a clinically useful phenotypic marker for identifying individuals at increased cardiometabolic risk.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13098-026-02247-4. Introduction Sarcopenia is recognized as a progressive and generalized skeletal muscle disorder, characterized by the accelerated loss of muscle mass and function, which is strongly associated with adverse outcomes including falls, functional decline, frailty, and mortality [1, 2]. As the global population ages, body composition inevitably shifts. While sarcopenia represents a significant health burden, the prevalence of obesity is simultaneously rising, leading to an increasing overlap of these two conditions [3]. Traditionally, obesity is defined by excess body fat accumulation that impairs health. Since the introduction of the BMI concept, numerous large-scale population studies have demonstrated a J-shaped relationship between BMI and disease risk, with BMI > 30 kg/m² significantly associated with adverse outcomes [4–6]. However, there is a complex paradox: not all older adults with obesity exhibit sarcopenia, as excess body weight can stimulate muscle and bone mechanoreceptors, triggering growth factor synthesis that preserves muscle mass [7]. Nevertheless, the intersection of population aging and the obesity epidemic has made sarcopenic obesity an increasingly prevalent condition [8]. The coexistence of sarcopenia and obesity represents a “double burden” that may be associated with a higher risk of adverse outcomes than either condition alone. Visceral adiposity is strongly associated with metabolic derangements, including hypertension, dyslipidemia, insulin resistance, and diabetes [9, 10]. It is hypothesized that obesity-induced physical inactivity may be linked to increased metabolic disturbances, while altered skeletal muscle energy metabolism further aggravates adiposity, forming a potential feedback loop [11, 12]. Although low muscle mass is a known risk factor for the loss of physical independence, reduced muscle function appears to play a more dominant role. The presence of both deficits signifies a critical risk for disability, comorbidity, and mortality in later life [13, 14]. Consequently, SO has substantial clinical implications, as it has been associated with increased risk of metabolic syndrome, complications, and geriatric syndromes such as frailty [15]. Epidemiological data underscore this burden: in the US population aged over 60, the prevalence of SO is reported at 28.3%, rising to 66.6% among Mexican Americans [16]. Crucially, current evidence suggests that older adults with concomitant obesity and sarcopenia face a higher risk of cardiovascular disease (CVD) and mortality compared to those with either condition alone [17, 18]. SO is closely linked to heightened cardiovascular risk factors, including insulin resistance and abnormal glucose-lipid metabolism [18–21], and has been strongly associated with increased CVD risk in Asian populations, such as Korean adults [22]. Increasing evidence has begun to examine body composition phenotypes within the specific framework of cardiovascular-kidney-metabolic (CKM) syndrome. The 2023 American Heart Association Presidential Advisory defined CKM syndrome as a continuum of cardiovascular, kidney, and metabolic risk and proposed a staging system from stage 0 to stage 4 [23]. A NHANES-based analysis published in JAMA further demonstrated the substantial population burden of CKM syndrome, reporting that nearly 90% of U.S. adults had CKM stage ≥ 1 and approximately 15% had advanced CKM stages [24]. Beyond conventional BMI-defined obesity, adiposity-related indices reflecting central fat distribution, such as body roundness index, have been associated with advanced CKM syndrome and may provide more informative CKM risk stratification than BMI alone [25]. In parallel, sarcopenia and low muscle mass have been linked to adverse outcomes within the CKM framework, including disease progression, mortality, and incident cardiovascular disease among individuals with CKM stages 0–3 [26, 27]. More recently, sarcopenic obesity has also been examined in relation to advanced CKM syndrome and cardiovascular events within CKM stages, supporting the potential value of integrating both adiposity and skeletal muscle phenotypes into CKM risk assessment [28]. To better characterize the systemic interplay between metabolic drivers and organ damage, the American Heart Association (AHA) recently defined Cardiovascular-Kidney-Metabolic (CKM) syndrome. This novel framework integrates obesity, diabetes, chronic kidney disease (CKD), and CVD, introducing a staging system to reflect the progressive nature of the disease [23]. The burden of these interconnected conditions is substantial, disproportionately affecting minority populations and driving high morbidity and mortality [9, 29–31]. The prognostic implications of CKM staging are profound; a study of 50,678 US adults demonstrated that the 15-year survival difference between CKM Stage 0 and Stage 4 was 8.1 months, with an absolute risk difference of 9.6% [32]. However, the association of body composition phenotypes with advanced CKM syndrome, incident CVD, and observed CKM stage movement within the newly defined CKM framework remains under-investigated. Therefore, this study aimed to examine the association of body composition phenotypes, including sarcopenic obesity, obesity only, sarcopenia only, and normal phenotype, with advanced CKM syndrome and incident CVD using longitudinal data from CHARLS and cross-sectional validation from NHANES. Methods Study design and population We adopted a dual-cohort design combining longitudinal discovery and cross-sectional validation to evaluate the associations of body composition phenotypes with advanced CKM syndrome and incident CVD. Data were drawn from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative longitudinal survey of the Chinese population aged 45 years and older. The baseline survey was conducted in 2011 (Wave 1), with follow-up assessments occurring biennially through 2018 (Wave 4) [33]. Further details about CHARLS are available on its official website (http://charls.pku.edu.cn/en). For the present analysis, we included participants who underwent complete anthropometric measurements and blood biomarker assays at baseline. Participants were excluded if they: (1) lacked data necessary to define sarcopenic obesity or CKM staging; (2) had missing covariates; (3) presented with biologically implausible extreme BMI values (≤ 10 or ≥ 60 kg/m²) due to likely raw data entry errors or were lost to follow-up (Fig. 1). The final longitudinal sample comprised 7,759 participants. To examine the external consistency of our findings across different ethnicities and distinct nutritional environments, we utilized data from the National Health and Nutrition Examination Survey (NHANES). We aggregated data from 2011 to 2018 cycles. Applying exclusion criteria consistent with the CHARLS cohort, a total of 3,936 participants were included for cross-sectional validation. Both studies were approved by their respective institutional review boards, and informed consent was obtained from all participants. Fig. 1Flowchart for study population selection Flowchart for study population selection Definition of body composition phenotypes Participants were stratified into four mutually exclusive phenotypes based on the presence of adiposity and skeletal muscle preservation: Normal, Obesity, Sarcopenia, and Sarcopenic Obesity (SO). Body composition phenotypes were classified based on baseline measurements, and reassessment was not performed at each follow-up wave due to data availability in the CHARLS cohort. Sarcopenia and obesity are inherently dynamic conditions that may fluctuate during follow-up. Consequently, utilizing baseline status as a static exposure may introduce potential exposure misclassification. Sarcopenia was assessed based on the criteria recommended by the AWGS 2019, which consists of three components: muscle strength, appendicular skeletal muscle mass (ASM), and physical performance [34]. Handgrip strength (kg) was measured in both the dominant and non-dominant hands using a YuejianTM WL-1000 dynamometer (Nantong Yuejian Physical Measurement Instrument Co., Ltd., Nantong, China), with participants instructed to squeeze the dynamometer as hard as possible. The maximum value of the two measurements was used for subsequent analyses, in accordance with the AWGS 2019 recommendation. The cut-off points for low grip strength for men and women were < 28 and < 18 kg, respectively. Appendicular skeletal muscle mass (ASM) in the CHARLS cohort was calculated using a validated anthropometric equation specifically developed for Chinese adults [35–38]: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}&\mathrm{ASM=0.193}\times \mathrm{weight(kg)}+0.107\times\text{height (cm)}\\&-4.157\times\mathrm{sex}-0.037\times\mathrm{age(years)}-2.631\end{aligned}$$\end{document} This equation was cross-validated using DXA as the reference standard in a large sample of 763 Chinese participants. Importantly, Bland–Altman analysis from the validation study confirmed no significant systematic bias between the estimated and DXA-measured ASM, with over 95% of the observations falling within the limits of agreement, ensuring the reliability of muscle mass estimation in this population [38, 39]. We further evaluated the performance of this anthropometric equation in the NHANES cohort by comparing equation-derived ASM with DXA-measured ASM. Pearson correlation analysis was used to assess the linear association, and Bland–Altman analysis was performed to evaluate agreement between the two methods. The validation showed a strong correlation between equation-derived ASM and DXA-measured ASM (r = 0.944) and no apparent systematic bias. The corresponding scatter plot and Bland–Altman plot are provided in Supplementary Figure S1. The ASM index (ASMI) was calculated as ASM divided by the square of height (kg/m2). Low muscle mass was defined as an ASMI < 7.0 kg/m2 for men and < 5.7 kg/m2 for women. Physical performance was evaluated using the 5-time chair stand test, with low performance defined as a time ≥ 12 s. Consistent with AWGS 2019, “possible sarcopenia” was defined as low muscle strength or low physical performance, while “confirmed sarcopenia” required the presence of low muscle mass combined with either low muscle strength or low physical performance. Central obesity was defined based on waist circumference, with cut-off points set at ≥ 90 cm for men and ≥ 85 cm for women[40]. Waist circumference was used to define obesity in CHARLS because the discovery cohort consisted of Chinese middle-aged and older adults, in whom central adiposity is particularly relevant to cardiometabolic risk. Compared with BMI, waist circumference better captures abdominal fat accumulation and is less influenced by reduced skeletal muscle mass in older adults. In contrast, BMI ≥ 30 kg/m² was used in NHANES because BMI-based obesity remains the conventional definition in U.S. population-based studies. In NHANES, ASM was directly measured via dual-energy X-ray absorptiometry (DXA), which represents the total lean body mass in the arms and legs. However, DXA cannot be utilized for such measurements in pregnant women, individuals with a height exceeding 192.5 cm, or those with a body weight surpassing 136.4 kg. In the NHANES cohort, sarcopenia was defined using the FNIH criteria based on the ALM-to-BMI ratio. A sarcopenia index below 0.512 for females or below 0.789 for males was classified as sarcopenia [41]. The use of disparate definitions reflects the inherent differences in data collection between the two cohorts and acknowledges the ethnic-specific physiological thresholds for sarcopenia. This dual-definition approach has been widely adopted in trans-ethnic comparative research to maintain the sensitivity and specificity of the diagnosis within each distinct cohort [42, 43]. The definition of obesity employs population-specific cutoff values to account for differences in fat distribution across ethnic groups. For NHANES, standard WHO criteria (BMI ≥ 30 kg/m²) were applied. Based on these criteria, participants were stratified into four groups: non-sarcopenic/non-obese (Normal), sarcopenic only, obese only, and sarcopenic obese. Assessment of CKM syndrome stages Staging of Cardiovascular-Kidney-Metabolic (CKM) Syndrome Consistent with the 2023 American Heart Association (AHA) Presidential Advisory [23], CKM syndrome was categorized into five stages (0–4) to reflect the disease continuum. Stage 0 represented no risk factors. Stage 1 indicated excess adiposity, including overweight/obesity or central obesity. Stage 2 comprised metabolic risk factors, including hypertension-related risk, abnormal glucose metabolism, dyslipidemia, or moderate CKD. Stage 3 was defined as subclinical CVD or high predicted risk, including a predicted 10-year CVD risk ≥ 10% or severe CKD [44]. Stage 4 included participants with established clinical CVD. Definitions of CKM-related clinical components and covariates CKM-related clinical and metabolic components were defined according to the 2023 AHA CKM framework and the available measurements in CHARLS and NHANES. For CKM staging, hypertension-related risk was defined as systolic blood pressure ≥ 130 mmHg, diastolic blood pressure ≥ 80 mmHg, self-reported physician-diagnosed hypertension, or current use of antihypertensive medication. Diabetes mellitus was defined as fasting plasma glucose ≥ 126 mg/dL, HbA1c ≥ 6.5%, self-reported physician-diagnosed diabetes, or current use of glucose-lowering medication or insulin. Prediabetes or glucose intolerance was defined as fasting plasma glucose 100–125 mg/dL or HbA1c 5.7–6.4% among participants without diabetes mellitus. In the CKM staging algorithm, abnormal glucose metabolism included either prediabetes/glucose intolerance or diabetes mellitus, whereas diabetes mellitus was reported separately in baseline characteristics and was used as the diabetes input in the PREVENT risk equations. Dyslipidemia was defined as triglycerides ≥ 150 mg/dL, low HDL-C (< 40 mg/dL in men or < 50 mg/dL in women), self-reported physician-diagnosed dyslipidemia, or current use of lipid-lowering medication. Moderate CKD was defined as eGFR < 60 mL/min/1.73 m², and severe CKD was defined as eGFR < 30 mL/min/1.73 m². Clinical CVD was defined as a self-reported history of heart disease or stroke. To account for potential confounding, multivariable models were adjusted for covariates selected a priori based on biological plausibility. Covariates common to both cohorts comprised sociodemographic factors, specifically age, sex, education level, and marital status. Lifestyle behaviors were also included, defined as smoking status and alcohol consumption. Missing data were handled using a complete-case analysis approach. Participants with missing data on variables required to define body composition phenotypes, CKM staging, follow-up outcomes, or covariates included in the multivariable models were excluded from the corresponding analytic dataset. Missing covariate values were not imputed and were not included as separate missing categories. Therefore, all multivariable models were fitted using participants with complete information on the exposure, outcome, and covariates included in each specific model. Variance inflation factors (VIFs) were calculated to ensure the absence of multicollinearity (all VIFs < 2.0). To avoid over-adjustment bias, clinical comorbidities that served as defining components of CKM staging, including hypertension-related risk, abnormal glucose metabolism or diabetes mellitus, dyslipidemia, CKD, and clinical CVD, were not included as covariates in the multivariable adjustment models. Regarding cohort-specific variables, the CHARLS analysis additionally adjusted for area of residence, while this variable was excluded from the NHANES model due to data unavailability. Statistical analysis Baseline characteristics were compared across the four body composition phenotypes using analysis of variance (ANOVA) for continuous variables and χ² tests for categorical variables. Continuous variables are presented as mean and standard deviation, and categorical variables as number and percentage. For cross-sectional analyses in CHARLS, logistic regression models were used to examine the association between body composition phenotypes and prevalent advanced CKM syndrome, defined as CKM Stage 3–4. Clinical cardiovascular disease (CVD), corresponding to CKM Stage 4, was also analyzed as a secondary cross-sectional outcome. The Normal phenotype served as the reference group. For longitudinal analyses, Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for incident CVD during follow-up from 2011 to 2018. Participants with clinical CVD at baseline were excluded from the incident CVD analysis. Incident CVD was defined as newly reported heart disease or stroke during follow-up. The proportional hazards assumption was assessed using Schoenfeld residuals. To reduce potential reverse causation, a restricted longitudinal analysis was additionally performed among participants free of advanced CKM at baseline, defined as CKM Stage 0–2. In this restricted cohort, Cox proportional hazards models were refitted to evaluate the association between baseline body composition phenotypes and incident CVD. Participants with baseline CKM Stage 3 or Stage 4 were excluded from this sensitivity analysis. Because complete biomarker data required for consistent CKM staging were available in both 2011 and 2015, a descriptive four-year CKM stage transition probability matrix was constructed. Transition probabilities were calculated as the observed proportion of participants moving from each CKM stage in 2011 to each CKM stage in 2015, with Stage 4 treated as an absorbing state. Additive interaction between sarcopenia and obesity was assessed using the relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), and synergy index (SI). Multiplicative interaction was evaluated by including a product term between sarcopenia and obesity. Interaction findings were interpreted cautiously because statistical interaction does not necessarily imply biological synergy. Sensitivity analyses included Fine–Gray competing-risk regression treating death before incident CVD as a competing event, and a lagged analysis excluding participants who developed incident CVD within the first 2 years after baseline. For external validation in NHANES, weighted logistic regression models were used to examine the association between body composition phenotypes and prevalent advanced CKM syndrome, accounting for the complex survey design. Three hierarchical models were fitted. Model 1 was unadjusted. Model 2 was adjusted for age, sex, education level, and marital status. Model 3 was further adjusted for smoking status and alcohol consumption. The CHARLS models additionally adjusted for area of residence. Clinical comorbidities that served as defining components of CKM staging were not included as covariates to avoid over-adjustment. Missing data were handled using complete-case analysis. Variance inflation factors were calculated to assess multicollinearity. All tests were two-sided, with P < 0.05 considered statistically significant. Analyses were conducted using R software. Results Baseline characteristics of the study population The baseline characteristics of the 7,759 participants from the CHARLS cohort are summarized in Table 1, stratified by body composition phenotypes. The overall mean age was 57.4 years, and 52.1% were female. The prevalence of the four phenotypes was distributed as follows: Normal (40.9%), Obesity Only (27.6%), Sarcopenia Only (17.9%), and Sarcopenic Obesity (SO) (13.7%). Compared with the other groups, participants with SO were generally older, predominantly female, had lower educational attainment, and showed the highest prevalence of hypertension (69.1%), diabetes mellitus (24.5%), dyslipidemia (68.9%), history of heart disease (18.0%), and stroke (3.9%) (all P < 0.001). The overall prevalence of diabetes mellitus was 16.6% after applying the revised diagnostic criteria. Regarding the CKM syndrome spectrum, advanced CKM (Stage 3–4) was most common in the Sarcopenia Only group (48.7%), followed by the SO group (45.7%), the Obesity Only group (33.4%), and the Normal group (29.5%). Notably, the SO group had the highest burden of clinical CVD (Stage 4, 20.8%), whereas the Sarcopenia Only group had the highest proportion of Stage 3 high-risk disease (33.5%). Because several variables presented in Table 1 are also used to define CKM stages, including hypertension-related risk, abnormal glucose metabolism or diabetes mellitus, dyslipidemia, kidney function, predicted CVD risk, and clinical CVD history, the observed baseline differences across body composition phenotypes should be interpreted with caution. These comparisons are intended to describe the clinical and metabolic profile of each phenotype rather than to establish independent associations with CKM outcomes. Table 1Baseline characteristics of CHARLS participants by body compositionNormal(N = 3173)Obesity Only(N = 2138)Sarcopenia Only(N = 1387)Sarcopenic Obesity(N = 1061)Overall(N = 7759)P-ValueAge (years) Mean (SD)56.0 (8.45)55.7 (8.01)61.4 (9.70)60.0 (8.97)57.4 (8.95)< 0.001Age Group < 65 years2662 (83.9%)1832 (85.7%)883 (63.7%)740 (69.7%)6117 (78.8%)< 0.001 ≥ 65 years511 (16.1%)306 (14.3%)504 (36.3%)321 (30.3%)1642 (21.2%)Sex Male1836 (57.9%)825 (38.6%)757 (54.6%)302 (28.5%)3720 (47.9%)< 0.001 Female1337 (42.1%)1313 (61.4%)630 (45.4%)759 (71.5%)4039 (52.1%)Residence Urban1017 (32.1%)935 (43.7%)371 (26.7%)405 (38.2%)2728 (35.2%)< 0.001 Rural2156 (67.9%)1203 (56.3%)1016 (73.3%)656 (61.8%)5031 (64.8%)Education Level Primary or less2806 (88.4%)1868 (87.4%)1310 (94.4%)1000 (94.3%)6984 (90.0%)< 0.001 Middle/High School326 (10.3%)239 (11.2%)70 (5.0%)57 (5.4%)692 (8.9%) College or above41 (1.3%)31 (1.5%)7 (0.5%)4 (0.4%)83 (1.1%)Marital Status Married3124 (98.5%)2121 (99.2%)1364 (98.3%)1051 (99.1%)7660 (98.7%)0.084 Widowed/Divorced/Single49 (1.5%)17 (0.8%)23 (1.7%)10 (0.9%)99 (1.3%)Smoking Status Current Smoker1518 (47.8%)664 (31.1%)625 (45.1%)266 (25.1%)3073 (39.6%)< 0.001 Non-Smoker1655 (52.2%)1474 (68.9%)762 (54.9%)795 (74.9%)4686 (60.4%)Alcohol consumption Drinker1426 (44.9%)793 (37.1%)580 (41.8%)284 (26.8%)3083 (39.7%)< 0.001 Non-Drinker1747 (55.1%)1345 (62.9%)807 (58.2%)777 (73.2%)4676 (60.3%)Hypertension No1823 (57.5%)774 (36.2%)702 (50.6%)328 (30.9%)3627 (46.7%)< 0.001 Yes1350 (42.5%)1364 (63.8%)685 (49.4%)733 (69.1%)4132 (53.3%)Diabetes Mellitus No2803 (88.3%)1682 (78.7%)1187 (85.6%)801 (75.5%)6473 (83.4%)< 0.001 Yes370 (11.7%)456 (21.3%)200 (14.4%)260 (24.5%)1286 (16.6%)Dyslipidemia No1946 (61.3%)706 (33.0%)865 (62.4%)330 (31.1%)3847 (49.6%)< 0.001 Yes1227 (38.7%)1432 (67.0%)522 (37.6%)731 (68.9%)3912 (50.4%)History of Heart Disease No2930 (92.3%)1861 (87.0%)1205 (86.9%)870 (82.0%)6866 (88.5%)< 0.001 Yes243 (7.7%)277 (13.0%)182 (13.1%)191 (18.0%)893 (11.5%)History of Stroke No3129 (98.6%)2092 (97.8%)1353 (97.5%)1020 (96.1%)7594 (97.9%)< 0.001 Yes44 (1.4%)46 (2.2%)34 (2.5%)41 (3.9%)165 (2.1%)BMI (kg/m2) Mean (SD)21.8 (2.69)26.5 (3.18)21.2 (2.99)26.3 (3.21)23.6 (3.80)< 0.001BMI Categories (China Criteria) Underweight (< 18.5)269 (8.5%)5 (0.2%)200 (14.4%)4 (0.4%)478 (6.2%)< 0.001 Normal (18.5–23.9)2397 (75.5%)389 (18.2%)999 (72.0%)247 (23.3%)4032 (52.0%) Overweight (24-27.9)475 (15.0%)1170 (54.7%)169 (12.2%)535 (50.4%)2349 (30.3%) Obesity (≥ 28)32 (1.0%)574 (26.8%)19 (1.4%)275 (25.9%)900 (11.6%)Waist Circumference (cm) Mean (SD)77.6 (9.96)94.3 (6.35)76.1 (11.8)94.7 (6.58)84.3 (12.5)< 0.001 ASM (kg) Mean (SD)17.1 (3.84)18.6 (4.27)16.0 (3.93)17.3 (4.06)17.3 (4.10)< 0.001Handgrip Strength (kg) Mean (SD)35.7 (9.10)35.4 (9.86)28.8 (10.1)28.5 (9.64)33.4 (10.1)< 0.0015-Time Chair Stand Test (s) Mean (SD)8.61 (1.92)8.86 (1.84)14.3 (3.90)15.2 (5.09)10.6 (4.06)< 0.001Systolic BP (mmHg) Mean (SD)124 (19.3)132 (20.2)127 (21.9)134 (21.9)128 (20.8)< 0.001Fasting Glucose (mg/dL) Mean (SD)105 (28.9)114 (38.0)108 (42.0)117 (45.3)110 (36.9)< 0.001HbA1c (%) Mean (SD)5.15 (0.630)5.37 (0.874)5.20 (0.858)5.44 (0.999)5.26 (0.808)< 0.001LDL-C (mg/dL) Mean (SD)114 (33.8)120 (36.3)112 (32.7)121 (36.2)116 (34.8)< 0.001Triglycerides (mg/dL) Mean (SD)115 (80.8)159 (113)111 (73.6)154 (106)132 (95.8)< 0.001eGFR (mL/min/1.73m2) Mean (SD)94.6 (14.3)93.4 (14.8)91.1 (14.7)91.2 (14.5)93.2 (14.6)< 0.001Predicted 10-Year CVD Risk (%) Mean (SD)7.40 (6.50)7.71 (6.76)10.8 (8.84)9.54 (7.74)8.39 (7.33)< 0.001CKM Stage (Detailed) Stage 0 (No Risk)538 (17.0%)0 (0%)137 (9.9%)0 (0%)675 (8.7%)< 0.001 Stage 1 (Excess Adiposity)23 (0.7%)107 (5.0%)8 (0.6%)41 (3.9%)179 (2.3%) Stage 2 (Metabolic/Kidney)1676 (52.8%)1316 (61.6%)567 (40.9%)535 (50.4%)4094 (52.8%) Stage 3 (Subclinical CVD or High Predicted Risk)657 (20.7%)405 (18.9%)464 (33.5%)264 (24.9%)1790 (23.1%) Stage 4 (CVD)279 (8.8%)310 (14.5%)211 (15.2%)221 (20.8%)1021 (13.2%)CKM Severity (Binary) Stage 0–2 (Early/Intermediate)2237 (70.5%)1423 (66.6%)712 (51.3%)576 (54.3%)4948 (63.8%)< 0.001 Stage 3–4 (Advanced)936 (29.5%)715 (33.4%)675 (48.7%)485 (45.7%)2811 (36.2%)Data are presented as mean (SD) for continuous variables or n (%) for categorical variables. P-values were calculated using ANOVA for continuous variables and Chi-square tests for categorical variables. Abbreviations: ASM, Appendicular Skeletal Muscle mass; BMI, Body Mass Index; BP, Blood Pressure; CKM, Cardiovascular-Kidney-Metabolic; CVD, Cardiovascular Disease; eGFR, estimated Glomerular Filtration Rate; HbA1c, Hemoglobin A1c; LDL-C, Low-Density Lipoprotein Cholesterol; SD, Standard Deviation. Some baseline characteristics, including hypertension, diabetes mellitus, dyslipidemia, eGFR, predicted 10-year CVD risk, and history of heart disease or stroke, are components of CKM staging. Therefore, differences across body composition phenotypes should be interpreted as descriptive baseline comparisons Table 1Baseline characteristics of CHARLS participants by body compositionNormal(N = 3173)Obesity Only(N = 2138)Sarcopenia Only(N = 1387)Sarcopenic Obesity(N = 1061)Overall(N = 7759)P-ValueAge (years) Mean (SD)56.0 (8.45)55.7 (8.01)61.4 (9.70)60.0 (8.97)57.4 (8.95)< 0.001Age Group < 65 years2662 (83.9%)1832 (85.7%)883 (63.7%)740 (69.7%)6117 (78.8%)< 0.001 ≥ 65 years511 (16.1%)306 (14.3%)504 (36.3%)321 (30.3%)1642 (21.2%)Sex Male1836 (57.9%)825 (38.6%)757 (54.6%)302 (28.5%)3720 (47.9%)< 0.001 Female1337 (42.1%)1313 (61.4%)630 (45.4%)759 (71.5%)4039 (52.1%)Residence Urban1017 (32.1%)935 (43.7%)371 (26.7%)405 (38.2%)2728 (35.2%)< 0.001 Rural2156 (67.9%)1203 (56.3%)1016 (73.3%)656 (61.8%)5031 (64.8%)Education Level Primary or less2806 (88.4%)1868 (87.4%)1310 (94.4%)1000 (94.3%)6984 (90.0%)< 0.001 Middle/High School326 (10.3%)239 (11.2%)70 (5.0%)57 (5.4%)692 (8.9%) College or above41 (1.3%)31 (1.5%)7 (0.5%)4 (0.4%)83 (1.1%)Marital Status Married3124 (98.5%)2121 (99.2%)1364 (98.3%)1051 (99.1%)7660 (98.7%)0.084 Widowed/Divorced/Single49 (1.5%)17 (0.8%)23 (1.7%)10 (0.9%)99 (1.3%)Smoking Status Current Smoker1518 (47.8%)664 (31.1%)625 (45.1%)266 (25.1%)3073 (39.6%)< 0.001 Non-Smoker1655 (52.2%)1474 (68.9%)762 (54.9%)795 (74.9%)4686 (60.4%)Alcohol consumption Drinker1426 (44.9%)793 (37.1%)580 (41.8%)284 (26.8%)3083 (39.7%)< 0.001 Non-Drinker1747 (55.1%)1345 (62.9%)807 (58.2%)777 (73.2%)4676 (60.3%)Hypertension No1823 (57.5%)774 (36.2%)702 (50.6%)328 (30.9%)3627 (46.7%)< 0.001 Yes1350 (42.5%)1364 (63.8%)685 (49.4%)733 (69.1%)4132 (53.3%)Diabetes Mellitus No2803 (88.3%)1682 (78.7%)1187 (85.6%)801 (75.5%)6473 (83.4%)< 0.001 Yes370 (11.7%)456 (21.3%)200 (14.4%)260 (24.5%)1286 (16.6%)Dyslipidemia No1946 (61.3%)706 (33.0%)865 (62.4%)330 (31.1%)3847 (49.6%)< 0.001 Yes1227 (38.7%)1432 (67.0%)522 (37.6%)731 (68.9%)3912 (50.4%)History of Heart Disease No2930 (92.3%)1861 (87.0%)1205 (86.9%)870 (82.0%)6866 (88.5%)< 0.001 Yes243 (7.7%)277 (13.0%)182 (13.1%)191 (18.0%)893 (11.5%)History of Stroke No3129 (98.6%)2092 (97.8%)1353 (97.5%)1020 (96.1%)7594 (97.9%)< 0.001 Yes44 (1.4%)46 (2.2%)34 (2.5%)41 (3.9%)165 (2.1%)BMI (kg/m2) Mean (SD)21.8 (2.69)26.5 (3.18)21.2 (2.99)26.3 (3.21)23.6 (3.80)< 0.001BMI Categories (China Criteria) Underweight (< 18.5)269 (8.5%)5 (0.2%)200 (14.4%)4 (0.4%)478 (6.2%)< 0.001 Normal (18.5–23.9)2397 (75.5%)389 (18.2%)999 (72.0%)247 (23.3%)4032 (52.0%) Overweight (24-27.9)475 (15.0%)1170 (54.7%)169 (12.2%)535 (50.4%)2349 (30.3%) Obesity (≥ 28)32 (1.0%)574 (26.8%)19 (1.4%)275 (25.9%)900 (11.6%)Waist Circumference (cm) Mean (SD)77.6 (9.96)94.3 (6.35)76.1 (11.8)94.7 (6.58)84.3 (12.5)< 0.001 ASM (kg) Mean (SD)17.1 (3.84)18.6 (4.27)16.0 (3.93)17.3 (4.06)17.3 (4.10)< 0.001Handgrip Strength (kg) Mean (SD)35.7 (9.10)35.4 (9.86)28.8 (10.1)28.5 (9.64)33.4 (10.1)< 0.0015-Time Chair Stand Test (s) Mean (SD)8.61 (1.92)8.86 (1.84)14.3 (3.90)15.2 (5.09)10.6 (4.06)< 0.001Systolic BP (mmHg) Mean (SD)124 (19.3)132 (20.2)127 (21.9)134 (21.9)128 (20.8)< 0.001Fasting Glucose (mg/dL) Mean (SD)105 (28.9)114 (38.0)108 (42.0)117 (45.3)110 (36.9)< 0.001HbA1c (%) Mean (SD)5.15 (0.630)5.37 (0.874)5.20 (0.858)5.44 (0.999)5.26 (0.808)< 0.001LDL-C (mg/dL) Mean (SD)114 (33.8)120 (36.3)112 (32.7)121 (36.2)116 (34.8)< 0.001Triglycerides (mg/dL) Mean (SD)115 (80.8)159 (113)111 (73.6)154 (106)132 (95.8)< 0.001eGFR (mL/min/1.73m2) Mean (SD)94.6 (14.3)93.4 (14.8)91.1 (14.7)91.2 (14.5)93.2 (14.6)< 0.001Predicted 10-Year CVD Risk (%) Mean (SD)7.40 (6.50)7.71 (6.76)10.8 (8.84)9.54 (7.74)8.39 (7.33)< 0.001CKM Stage (Detailed) Stage 0 (No Risk)538 (17.0%)0 (0%)137 (9.9%)0 (0%)675 (8.7%)< 0.001 Stage 1 (Excess Adiposity)23 (0.7%)107 (5.0%)8 (0.6%)41 (3.9%)179 (2.3%) Stage 2 (Metabolic/Kidney)1676 (52.8%)1316 (61.6%)567 (40.9%)535 (50.4%)4094 (52.8%) Stage 3 (Subclinical CVD or High Predicted Risk)657 (20.7%)405 (18.9%)464 (33.5%)264 (24.9%)1790 (23.1%) Stage 4 (CVD)279 (8.8%)310 (14.5%)211 (15.2%)221 (20.8%)1021 (13.2%)CKM Severity (Binary) Stage 0–2 (Early/Intermediate)2237 (70.5%)1423 (66.6%)712 (51.3%)576 (54.3%)4948 (63.8%)< 0.001 Stage 3–4 (Advanced)936 (29.5%)715 (33.4%)675 (48.7%)485 (45.7%)2811 (36.2%)Data are presented as mean (SD) for continuous variables or n (%) for categorical variables. P-values were calculated using ANOVA for continuous variables and Chi-square tests for categorical variables. Abbreviations: ASM, Appendicular Skeletal Muscle mass; BMI, Body Mass Index; BP, Blood Pressure; CKM, Cardiovascular-Kidney-Metabolic; CVD, Cardiovascular Disease; eGFR, estimated Glomerular Filtration Rate; HbA1c, Hemoglobin A1c; LDL-C, Low-Density Lipoprotein Cholesterol; SD, Standard Deviation. Some baseline characteristics, including hypertension, diabetes mellitus, dyslipidemia, eGFR, predicted 10-year CVD risk, and history of heart disease or stroke, are components of CKM staging. Therefore, differences across body composition phenotypes should be interpreted as descriptive baseline comparisons Baseline characteristics of CHARLS participants by body composition Data are presented as mean (SD) for continuous variables or n (%) for categorical variables. P-values were calculated using ANOVA for continuous variables and Chi-square tests for categorical variables. Abbreviations: ASM, Appendicular Skeletal Muscle mass; BMI, Body Mass Index; BP, Blood Pressure; CKM, Cardiovascular-Kidney-Metabolic; CVD, Cardiovascular Disease; eGFR, estimated Glomerular Filtration Rate; HbA1c, Hemoglobin A1c; LDL-C, Low-Density Lipoprotein Cholesterol; SD, Standard Deviation. Some baseline characteristics, including hypertension, diabetes mellitus, dyslipidemia, eGFR, predicted 10-year CVD risk, and history of heart disease or stroke, are components of CKM staging. Therefore, differences across body composition phenotypes should be interpreted as descriptive baseline comparisons Cross-sectional association with advanced CKM syndrome and clinical CVD We first examined the cross-sectional association between body composition phenotypes and CKM severity at baseline using the CHARLS dataset (Table 2). In the fully adjusted model (Model 3), all three abnormal body composition phenotypes were significantly associated with higher odds of prevalent Advanced CKM (Stage 3–4) compared with the Normal group. The association was strongest for sarcopenic obesity (OR 2.65, 95% CI 2.20–3.21), followed by Obesity Only (OR 2.09, 95% CI 1.79–2.43) and Sarcopenia Only (OR 1.60, 95% CI 1.35–1.89). A similar pattern was observed for prevalent Clinical CVD (Stage 4). In the fully adjusted model, participants with sarcopenic obesity had the highest odds of Clinical CVD (OR 2.29, 95% CI 1.87–2.80), followed by Obesity Only (OR 1.70, 95% CI 1.42–2.02) and Sarcopenia Only (OR 1.59, 95% CI 1.30–1.93). These findings suggest that the coexistence of sarcopenia and obesity was associated with a greater baseline CKM burden than either condition alone. Table 2Cross-sectional association of body composition phenotypes with advanced CKM and clinical CVD in CHARLSOutcomeGroupModel 1Model 2Model 3OR (95% CI)P-ValueOR (95% CI)P-ValueOR (95% CI)P-ValueAdvanced CKM (Stage 3–4)Normal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Advanced CKM (Stage 3–4)Obesity Only1.20 (1.07–1.35)0.0021.97 (1.70–2.29)< 0.0012.09 (1.79–2.43)< 0.001Advanced CKM (Stage 3–4)Sarcopenia Only2.27 (1.99–2.58)< 0.0011.56 (1.32–1.84)< 0.0011.60 (1.35–1.89)< 0.001Advanced CKM (Stage 3–4)Sarcopenic Obesity2.01 (1.74–2.32)< 0.0012.53 (2.10–3.05)< 0.0012.65 (2.20–3.21)< 0.001Clinical CVD (Stage 4)Normal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Clinical CVD (Stage 4)Obesity Only1.76 (1.48–2.09)< 0.0011.68 (1.41-2.00)< 0.0011.70 (1.42–2.02)< 0.001Clinical CVD (Stage 4)Sarcopenia Only1.86 (1.54–2.25)< 0.0011.59 (1.30–1.93)< 0.0011.59 (1.30–1.93)< 0.001Clinical CVD (Stage 4)Sarcopenic Obesity2.73 (2.25–3.31)< 0.0012.28 (1.87–2.79)< 0.0012.29 (1.87–2.80)< 0.001Values are odds ratios (ORs) with 95% confidence intervals (CIs). The “Normal” group serves as the reference category. Model 1: Unadjusted. Model 2: Adjusted for age, sex, residence, education, and marital status. Model 3: Adjusted for Model 2 covariates plus smoking status and alcohol consumption. Advanced CKM is defined as CKM Stage 3 or 4. Clinical CVD corresponds to CKM Stage 4 Table 2Cross-sectional association of body composition phenotypes with advanced CKM and clinical CVD in CHARLSOutcomeGroupModel 1Model 2Model 3OR (95% CI)P-ValueOR (95% CI)P-ValueOR (95% CI)P-ValueAdvanced CKM (Stage 3–4)Normal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Advanced CKM (Stage 3–4)Obesity Only1.20 (1.07–1.35)0.0021.97 (1.70–2.29)< 0.0012.09 (1.79–2.43)< 0.001Advanced CKM (Stage 3–4)Sarcopenia Only2.27 (1.99–2.58)< 0.0011.56 (1.32–1.84)< 0.0011.60 (1.35–1.89)< 0.001Advanced CKM (Stage 3–4)Sarcopenic Obesity2.01 (1.74–2.32)< 0.0012.53 (2.10–3.05)< 0.0012.65 (2.20–3.21)< 0.001Clinical CVD (Stage 4)Normal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Clinical CVD (Stage 4)Obesity Only1.76 (1.48–2.09)< 0.0011.68 (1.41-2.00)< 0.0011.70 (1.42–2.02)< 0.001Clinical CVD (Stage 4)Sarcopenia Only1.86 (1.54–2.25)< 0.0011.59 (1.30–1.93)< 0.0011.59 (1.30–1.93)< 0.001Clinical CVD (Stage 4)Sarcopenic Obesity2.73 (2.25–3.31)< 0.0012.28 (1.87–2.79)< 0.0012.29 (1.87–2.80)< 0.001Values are odds ratios (ORs) with 95% confidence intervals (CIs). The “Normal” group serves as the reference category. Model 1: Unadjusted. Model 2: Adjusted for age, sex, residence, education, and marital status. Model 3: Adjusted for Model 2 covariates plus smoking status and alcohol consumption. Advanced CKM is defined as CKM Stage 3 or 4. Clinical CVD corresponds to CKM Stage 4 Cross-sectional association of body composition phenotypes with advanced CKM and clinical CVD in CHARLS Values are odds ratios (ORs) with 95% confidence intervals (CIs). The “Normal” group serves as the reference category. Model 1: Unadjusted. Model 2: Adjusted for age, sex, residence, education, and marital status. Model 3: Adjusted for Model 2 covariates plus smoking status and alcohol consumption. Advanced CKM is defined as CKM Stage 3 or 4. Clinical CVD corresponds to CKM Stage 4 Longitudinal incidence rates of cardiovascular events classified by body composition phenotype Among the 7,759 baseline analytic participants, 1,021 participants with clinical CVD at baseline were excluded, leaving 6,738 participants for the incident CVD analysis. The restricted longitudinal analysis included 4,948 participants with baseline CKM Stage 0–2. During follow-up from 2011 to 2018, we analyzed the incidence of new-onset CVD events among participants without clinical CVD at baseline. The crude cumulative incidence of CVD events was highest in the sarcopenic obesity group (29.3%), followed by Obesity Only (20.2%), Sarcopenia Only (18.5%), and Normal (13.6%). Because the body composition phenotype groups differed substantially in age and sex distribution, Fig. 2 presents age- and sex-standardized cumulative incidence estimates. After standardization to the age-sex distribution of the overall incident-CVD analytic population, the sarcopenic obesity group remained the group with the highest cumulative incidence of CVD events, followed by Obesity Only, Sarcopenia Only, and Normal. Consistent with these standardized incidence patterns, the Kaplan–Meier survival analysis (Fig. 3) showed that participants with sarcopenic obesity had the lowest cumulative CVD event-free probability over the follow-up period (log-rank P < 0.0001). Fig. 2Age and sex standardized cumulative incidence of CVD events by body composition phenotype. The bar chart displays cumulative incidence percentages of new-onset cardiovascular events over the follow-up period from 2011 to 2018 in the CHARLS cohort, standardized to the age-sex distribution of the overall incident CVD analytic population. Error bars indicate 95% confidence intervals Age and sex standardized cumulative incidence of CVD events by body composition phenotype. The bar chart displays cumulative incidence percentages of new-onset cardiovascular events over the follow-up period from 2011 to 2018 in the CHARLS cohort, standardized to the age-sex distribution of the overall incident CVD analytic population. Error bars indicate 95% confidence intervals Fig. 3Kaplan–Meier estimates of CVD event-free probability by body composition phenotype. Survival curves show the cumulative CVD event-free probability over the follow-up period across the four body composition phenotypes. The number of participants at risk is shown below the curves at the corresponding CHARLS follow-up time points: baseline, 2 years, 4 years, and 7 years. The P-value was calculated using the log-rank test Kaplan–Meier estimates of CVD event-free probability by body composition phenotype. Survival curves show the cumulative CVD event-free probability over the follow-up period across the four body composition phenotypes. The number of participants at risk is shown below the curves at the corresponding CHARLS follow-up time points: baseline, 2 years, 4 years, and 7 years. The P-value was calculated using the log-rank test Longitudinal association between body composition phenotypes and incident CVD Multivariable Cox regression analyses further confirmed the longitudinal association between body composition phenotypes and incident CVD events. As shown in Table 3, compared with the Normal group, the sarcopenic obesity phenotype remained associated with a more than two-fold higher risk of incident CVD events during follow-up from 2011 to 2018 (HR 2.07, 95% CI 1.76–2.45) in the fully adjusted model. We further performed a restricted longitudinal analysis among participants free of advanced CKM at baseline, defined as CKM Stage 0–2. Participants with baseline CKM Stage 3 or Stage 4 were excluded. In this restricted cohort, the associations remained consistent. In the fully adjusted model, sarcopenic obesity was associated with the highest risk of incident CVD (HR 1.93, 95% CI 1.63–2.28), followed by Obesity Only (HR 1.54, 95% CI 1.34–1.76) and Sarcopenia Only (HR 1.28, 95% CI 1.08–1.51) (Supplementary Table S3) (Fig. 4). Table 3Hazard ratios for incident CVD events by body composition phenotype over 7-year follow-upVariable / GroupModel 1Model 2Model 3HR (95% CI)P-ValueHR (95% CI)P-ValueHR (95% CI)P-ValueNormal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Obesity Only1.53 (1.33–1.76)< 0.0011.50 (1.29–1.73)< 0.0011.51 (1.31–1.74)< 0.001Sarcopenia Only1.40 (1.19–1.65)< 0.0011.22 (1.03–1.45)0.0211.23 (1.03–1.45)0.019Sarcopenic Obesity2.36 (2.01–2.77)< 0.0012.06 (1.75–2.43)< 0.0012.07 (1.76–2.45)< 0.001Values are hazard ratios (HRs) with 95% confidence intervals (CIs) derived from Cox proportional hazards models. Model 1 was unadjusted. Model 2 was adjusted for age, sex, residence, education level, and marital status. Model 3 was further adjusted for smoking status and alcohol consumption Table 3Hazard ratios for incident CVD events by body composition phenotype over 7-year follow-upVariable / GroupModel 1Model 2Model 3HR (95% CI)P-ValueHR (95% CI)P-ValueHR (95% CI)P-ValueNormal1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Obesity Only1.53 (1.33–1.76)< 0.0011.50 (1.29–1.73)< 0.0011.51 (1.31–1.74)< 0.001Sarcopenia Only1.40 (1.19–1.65)< 0.0011.22 (1.03–1.45)0.0211.23 (1.03–1.45)0.019Sarcopenic Obesity2.36 (2.01–2.77)< 0.0012.06 (1.75–2.43)< 0.0012.07 (1.76–2.45)< 0.001Values are hazard ratios (HRs) with 95% confidence intervals (CIs) derived from Cox proportional hazards models. Model 1 was unadjusted. Model 2 was adjusted for age, sex, residence, education level, and marital status. Model 3 was further adjusted for smoking status and alcohol consumption Hazard ratios for incident CVD events by body composition phenotype over 7-year follow-up Values are hazard ratios (HRs) with 95% confidence intervals (CIs) derived from Cox proportional hazards models. Model 1 was unadjusted. Model 2 was adjusted for age, sex, residence, education level, and marital status. Model 3 was further adjusted for smoking status and alcohol consumption Fig. 4Restricted longitudinal analysis of incident CVD among participants free of advanced CKM at baseline. Forest plot showing hazard ratios and 95% confidence intervals for incident CVD during follow-up from 2011 to 2018 among participants with baseline CKM Stage 0–2. The Normal group served as the reference category Restricted longitudinal analysis of incident CVD among participants free of advanced CKM at baseline. Forest plot showing hazard ratios and 95% confidence intervals for incident CVD during follow-up from 2011 to 2018 among participants with baseline CKM Stage 0–2. The Normal group served as the reference category Table 4Additive and multiplicative interaction between sarcopenia and obesity for incident CVDMeasure of InteractionHR / Estimate95% CIP-ValueIndependent Associations (Cox Model)Normal1.00-Sarcopenia Only1.231.03–1.450.019Obesity Only1.511.31–1.75< 0.001Sarcopenic Obesity (Observed)2.081.76–2.46< 0.001Interaction IndicesMultiplicative Interaction Term (Ratio)1.120.89–1.420.319RERI (Relative Excess Risk due to Interaction)0.340.06–0.430.027AP (Attributable Proportion)0.160.01–0.330.035SI (Synergy Index)1.461.05–2.060.025Data are derived from multivariable Cox proportional hazards regression models based on the CHARLS longitudinal cohort. The model was fully adjusted for age, sex, residence, education level, marital status, smoking status, and alcohol consumption. Values for the independent associations represent hazard ratios (HRs) compared with the Normal group. Additive interaction between sarcopenia and obesity was assessed using the relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), and synergy index (SI). RERI was calculated as HR11 − HR10 − HR01 + 1, where HR11 represents the joint exposure to sarcopenia and obesity, HR10 represents sarcopenia only, and HR01 represents obesity only. A RERI > 0 indicates positive additive interaction on the risk scale. The multiplicative interaction term was estimated by including a product term between sarcopenia and obesity in the Cox model. Statistical interaction does not necessarily imply biological or mechanistic synergy; therefore, these findings should be interpreted as evidence of a potential joint association rather than definitive biological interaction. Abbreviations: AP, attributable proportion due to interaction; CI, confidence interval; HR, hazard ratio; RERI, relative excess risk due to interaction; SI, synergy index. The estimates in Table 4 were derived from the fully adjusted interaction model including sarcopenia, obesity, and their product term; therefore, minor differences from the categorical phenotype model in Table 3 may occur because of model specification and rounding. Counts are unweighted; odds ratios were estimated using survey-weighted logistic regression models Table 4Additive and multiplicative interaction between sarcopenia and obesity for incident CVDMeasure of InteractionHR / Estimate95% CIP-ValueIndependent Associations (Cox Model)Normal1.00-Sarcopenia Only1.231.03–1.450.019Obesity Only1.511.31–1.75< 0.001Sarcopenic Obesity (Observed)2.081.76–2.46< 0.001Interaction IndicesMultiplicative Interaction Term (Ratio)1.120.89–1.420.319RERI (Relative Excess Risk due to Interaction)0.340.06–0.430.027AP (Attributable Proportion)0.160.01–0.330.035SI (Synergy Index)1.461.05–2.060.025Data are derived from multivariable Cox proportional hazards regression models based on the CHARLS longitudinal cohort. The model was fully adjusted for age, sex, residence, education level, marital status, smoking status, and alcohol consumption. Values for the independent associations represent hazard ratios (HRs) compared with the Normal group. Additive interaction between sarcopenia and obesity was assessed using the relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), and synergy index (SI). RERI was calculated as HR11 − HR10 − HR01 + 1, where HR11 represents the joint exposure to sarcopenia and obesity, HR10 represents sarcopenia only, and HR01 represents obesity only. A RERI > 0 indicates positive additive interaction on the risk scale. The multiplicative interaction term was estimated by including a product term between sarcopenia and obesity in the Cox model. Statistical interaction does not necessarily imply biological or mechanistic synergy; therefore, these findings should be interpreted as evidence of a potential joint association rather than definitive biological interaction. Abbreviations: AP, attributable proportion due to interaction; CI, confidence interval; HR, hazard ratio; RERI, relative excess risk due to interaction; SI, synergy index. The estimates in Table 4 were derived from the fully adjusted interaction model including sarcopenia, obesity, and their product term; therefore, minor differences from the categorical phenotype model in Table 3 may occur because of model specification and rounding. Counts are unweighted; odds ratios were estimated using survey-weighted logistic regression models Additive and multiplicative interaction between sarcopenia and obesity for incident CVD Data are derived from multivariable Cox proportional hazards regression models based on the CHARLS longitudinal cohort. The model was fully adjusted for age, sex, residence, education level, marital status, smoking status, and alcohol consumption. Values for the independent associations represent hazard ratios (HRs) compared with the Normal group. Additive interaction between sarcopenia and obesity was assessed using the relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP), and synergy index (SI). RERI was calculated as HR11 − HR10 − HR01 + 1, where HR11 represents the joint exposure to sarcopenia and obesity, HR10 represents sarcopenia only, and HR01 represents obesity only. A RERI > 0 indicates positive additive interaction on the risk scale. The multiplicative interaction term was estimated by including a product term between sarcopenia and obesity in the Cox model. Statistical interaction does not necessarily imply biological or mechanistic synergy; therefore, these findings should be interpreted as evidence of a potential joint association rather than definitive biological interaction. Abbreviations: AP, attributable proportion due to interaction; CI, confidence interval; HR, hazard ratio; RERI, relative excess risk due to interaction; SI, synergy index. The estimates in Table 4 were derived from the fully adjusted interaction model including sarcopenia, obesity, and their product term; therefore, minor differences from the categorical phenotype model in Table 3 may occur because of model specification and rounding. Counts are unweighted; odds ratios were estimated using survey-weighted logistic regression models Subgroup and interaction analyses for the association between SO and 7-year incident CVD risk Subgroup analyses were performed to assess the robustness of the association between sarcopenic obesity (SO) and 7-year incident CVD across demographic and lifestyle strata (Fig. 5). The elevated risk associated with SO was generally consistent across age, sex, residence, smoking status, and alcohol consumption subgroups, with no statistically significant interactions detected. The association was observed in both participants aged < 65 years (HR 2.20, 95% CI 1.81–2.66) and ≥ 65 years (HR 1.84, 95% CI 1.32–2.57; P for interaction = 0.683), in males (HR 2.11, 95% CI 1.60–2.79) and females (HR 2.02, 95% CI 1.63–2.50; P for interaction = 0.064), and in both urban and rural residents. Estimates in some education and marital-status strata were less precise because of small sample sizes and limited events. Overall, these findings indicate no significant evidence of effect modification and support the robustness of the association between SO and incident CVD. Fig. 5Subgroup analysis of the association between sarcopenic obesity and 7-year incident CVD. Forest plot showing hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between sarcopenic obesity and incident CVD during follow-up from 2011 to 2018 across demographic and lifestyle subgroups. The Normal group served as the reference category. Models were adjusted for age, sex, residence, education level, marital status, smoking status, and alcohol consumption as appropriate, excluding the corresponding stratification variable. P-values for interaction were calculated using likelihood ratio tests comparing models with and without the interaction term. The hazard ratio estimate for the widowed/divorced/single subgroup was not estimable due tosparse events and limited sample size Subgroup analysis of the association between sarcopenic obesity and 7-year incident CVD. Forest plot showing hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between sarcopenic obesity and incident CVD during follow-up from 2011 to 2018 across demographic and lifestyle subgroups. The Normal group served as the reference category. Models were adjusted for age, sex, residence, education level, marital status, smoking status, and alcohol consumption as appropriate, excluding the corresponding stratification variable. P-values for interaction were calculated using likelihood ratio tests comparing models with and without the interaction term. The hazard ratio estimate for the widowed/divorced/single subgroup was not estimable due tosparse events and limited sample size External validation in NHANES 2011–2018 To examine the external consistency of our findings in a comparably aged Western population with gold-standard body composition measurements, we analyzed 3,936 participants from the NHANES 2011–2018 cycles (Table 5). Consistent with the discovery cohort, the prevalence of Advanced CKM exhibited a stepwise increase across phenotypes: 13.6% in the Normal group, rising to 20.7% in Sarcopenia Only, 22.6% in Obesity Only, and reaching a peak of 25.4% in the Sarcopenic Obesity (SO) group. In the fully adjusted weighted logistic regression model, the SO phenotype showed the highest odds of prevalent advanced CKM syndrome. Obesity Only and Sarcopenia Only were also associated with higher odds of prevalent advanced CKM syndrome, although the magnitude of association was greatest in the SO group. These NHANES findings were broadly consistent with the CHARLS results and support the external consistency of the association between sarcopenic obesity and prevalent advanced CKM syndrome. However, because the NHANES analysis was cross-sectional and used a cohort-specific sarcopenia definition, it should be interpreted as conceptual validation rather than direct replication of an identical phenotype. Although the overall pattern in NHANES was broadly consistent with the CHARLS findings, the sarcopenia-only group in NHANES included only 169 participants. Therefore, estimates for this subgroup should be interpreted cautiously because of reduced precision and potential statistical instability. The validation analysis was interpreted primarily in terms of the overall consistency of the association between sarcopenic obesity and advanced CKM syndrome. Table 5External validation of the association between body composition phenotypes and advanced CKM syndrome in NHANES 2011–2018Group (Phenotype)Total NCases (%)Model 1Model 2Model 3OR (95% CI)P-ValueOR (95% CI)P-ValueOR (95% CI)P-ValueNormal2245305 (13.6)1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Obesity Only1199271 (22.6)1.86 (1.55–2.23)< 0.0011.88 (1.57–2.26)< 0.0011.90 (1.58–2.29)< 0.001Sarcopenia Only16935 (20.7)1.66 (1.11–2.43)0.0111.53 (1.02–2.25)0.0341.60 (1.06–2.35)0.021Sarcopenic Obesity32382 (25.4)2.16 (1.63–2.85)< 0.0012.00 (1.50–2.65)< 0.0012.04 (1.53–2.70)< 0.001Body composition in NHANES was assessed using Dual-energy X-ray Absorptiometry (DXA). Model 1: Unadjusted. Model 2: Adjusted for age, sex, education, and marital status. Model 3: Adjusted for Model 2 plus smoking and alcohol consumption. Abbreviations: Ref, Reference; OR, Odds Ratio; CI, Confidence Interval. NHANES analyses accounted for the complex survey design using appropriate sample weights, strata, and primary sampling units Table 5External validation of the association between body composition phenotypes and advanced CKM syndrome in NHANES 2011–2018Group (Phenotype)Total NCases (%)Model 1Model 2Model 3OR (95% CI)P-ValueOR (95% CI)P-ValueOR (95% CI)P-ValueNormal2245305 (13.6)1.00 (Ref)-1.00 (Ref)-1.00 (Ref)-Obesity Only1199271 (22.6)1.86 (1.55–2.23)< 0.0011.88 (1.57–2.26)< 0.0011.90 (1.58–2.29)< 0.001Sarcopenia Only16935 (20.7)1.66 (1.11–2.43)0.0111.53 (1.02–2.25)0.0341.60 (1.06–2.35)0.021Sarcopenic Obesity32382 (25.4)2.16 (1.63–2.85)< 0.0012.00 (1.50–2.65)< 0.0012.04 (1.53–2.70)< 0.001Body composition in NHANES was assessed using Dual-energy X-ray Absorptiometry (DXA). Model 1: Unadjusted. Model 2: Adjusted for age, sex, education, and marital status. Model 3: Adjusted for Model 2 plus smoking and alcohol consumption. Abbreviations: Ref, Reference; OR, Odds Ratio; CI, Confidence Interval. NHANES analyses accounted for the complex survey design using appropriate sample weights, strata, and primary sampling units External validation of the association between body composition phenotypes and advanced CKM syndrome in NHANES 2011–2018 Body composition in NHANES was assessed using Dual-energy X-ray Absorptiometry (DXA). Model 1: Unadjusted. Model 2: Adjusted for age, sex, education, and marital status. Model 3: Adjusted for Model 2 plus smoking and alcohol consumption. Abbreviations: Ref, Reference; OR, Odds Ratio; CI, Confidence Interval. NHANES analyses accounted for the complex survey design using appropriate sample weights, strata, and primary sampling units Descriptive four-year CKM stage transition dynamics To describe changes in CKM syndrome status over time, we constructed a four-year observed transition probability diagram using CKM stages assessed consistently in 2011 and 2015 (Fig. 6). Overall, early CKM stages showed frequent transitions toward more advanced stages, whereas advanced stages were more persistent. Among participants with baseline Stage 0, only 13% remained in Stage 0 after four years, while 54% transitioned to Stage 2 and 25% transitioned to Stage 3. Similarly, only 7% of participants with baseline Stage 1 remained in Stage 1, whereas 67% transitioned to Stage 2 and 16% transitioned to Stage 3. Among participants with baseline Stage 2, 53% remained in Stage 2, while 33% transitioned to Stage 3 and 9% transitioned to Stage 4. Stage 3 showed greater persistence, with 85% remaining in Stage 3 and 12% transitioning to Stage 4. Stage 4, representing established clinical CVD, was treated as an absorbing state. These findings indicate that movement from early CKM stages to metabolic-risk and high-risk stages was common between 2011 and 2015, while advanced CKM stages showed greater stability. Fig. 6Four-year observed CKM stage transition diagram from 2011 to 2015. Nodes represent CKM stages, and the value within each node indicates the probability of remaining in the same stage after four years. Directed arrows indicate observed transition probabilities between CKM stages. Only transitions with probabilities ≥ 0.05 are displayed for visual clarity. Stage 4, representing established clinical cardiovascular disease, was treated as an absorbing state Four-year observed CKM stage transition diagram from 2011 to 2015. Nodes represent CKM stages, and the value within each node indicates the probability of remaining in the same stage after four years. Directed arrows indicate observed transition probabilities between CKM stages. Only transitions with probabilities ≥ 0.05 are displayed for visual clarity. Stage 4, representing established clinical cardiovascular disease, was treated as an absorbing state Sensitivity analysis Sensitivity analyses yielded generally consistent results. In the Fine–Gray competing-risk model accounting for death before incident CVD, sarcopenic obesity remained associated with incident CVD (SHR 2.01, 95% CI 1.71–2.35), followed by Obesity Only and Sarcopenia Only (Supplementary Table S1). After excluding participants who developed CVD within the first 2 years of follow-up, sarcopenic obesity remained associated with a higher risk of incident CVD in the fully adjusted Cox model (HR 2.09, 95% CI 1.75–2.51) (Supplementary Table S2). In the restricted analysis among participants free of advanced CKM at baseline, sarcopenic obesity showed the highest risk of incident CVD (HR 1.93, 95% CI 1.63–2.28) (Supplementary Table S3). Discussion In this study, sarcopenic obesity was consistently associated with a higher burden of advanced CKM syndrome and a higher risk of incident CVD. In the CHARLS cohort, participants with sarcopenic obesity had more than a two-fold higher risk of incident CVD compared with the Normal group after multivariable adjustment. This association was stronger than that observed for obesity only or sarcopenia only. A positive additive interaction was also observed, although this finding should be interpreted cautiously because statistical interaction does not necessarily imply biological synergy. Because body composition phenotypes were assessed only at baseline, changes in sarcopenia or obesity status during follow-up could not be captured. Therefore, our longitudinal estimates should be interpreted as associations with baseline phenotype status rather than persistent or cumulative exposure to sarcopenic obesity. Such exposure misclassification may have influenced the magnitude of the observed associations. Nevertheless, the main findings remained consistent in sensitivity analyses, including the restricted analysis among participants free of advanced CKM at baseline. In the 2023 Presidential Advisory Report, the American Heart Association (AHA) formally introduced the concept of CKM syndrome, integrating obesity, diabetes, chronic kidney disease (CKD), and cardiovascular disease into a continuous staging continuum (Stage 0–4) [45]. However, the current CKM framework primarily uses adiposity-related measures, such as BMI and waist circumference, to define excess adiposity, while skeletal muscle status is not explicitly incorporated. Our findings suggest that muscle-related measures may provide complementary information for cardiometabolic risk stratification in older adults. This aligns with recent evidence from Jiang et al., who identified a robust correlation between SO and CVD progression [46], and a meta-analysis by Liu et al. confirming SO’s association with elevated risks of stroke, myocardial infarction, and heart failure [47]. Consistently, cross-sectional studies underscore sarcopenia as an independent predictor of adverse cardiac events [48–50]. These data suggest that preserved skeletal muscle may provide complementary metabolic reserve in older adults. Beyond its locomotor function, skeletal muscle is a major regulator of whole-body glucose homeostasis and accounts for a substantial proportion of postprandial and insulin-stimulated glucose disposal [51–53]. This glucose-disposal function is mediated in part by insulin- and contraction-regulated glucose transport pathways in skeletal muscle, together with tissue-specific glucose transporters in adipose tissue and liver that collectively maintain systemic glucose homeostasis [54]. Accordingly, impaired skeletal muscle glucose uptake is closely linked to insulin resistance and may represent an early metabolic abnormality before overt hyperglycemia develops [55]. Consistent with this mechanism, previous reviews have suggested that sarcopenia and metabolic syndrome are closely interconnected, and that their coexistence may aggravate frailty, morbidity, and cardiometabolic risk [56]. Therefore, loss of skeletal muscle mass and function may reduce glucose-disposal capacity and may partly explain why individuals with sarcopenic obesity showed a higher risk of advanced CKM burden and incident CVD than those with obesity alone. In the descriptive four-year observed CKM transition analysis, we further characterized stage movement between 2011 and 2015. This analysis suggested that early CKM stages frequently moved toward higher-risk stages, whereas advanced stages were more persistent. This analysis provides context for the dynamic nature of CKM staging but should not be interpreted as evidence that sarcopenic obesity caused CKM stage progression. Separately, baseline sarcopenic obesity was associated with a greater burden of advanced CKM syndrome and a higher risk of incident CVD. Multiple longitudinal studies have investigated the association between sarcopenic or strength-loss obesity and cardiovascular disease [57–60]. The observed association between SO and more advanced CKM stages may be explained by several biological mechanisms. This interpretation is supported by mechanistic evidence indicating that skeletal muscle is the major tissue controlling postprandial glucose disposal and that impaired skeletal muscle glucose uptake is central to insulin resistance. Muscle atrophy diminishes the capacity for postprandial glucose and lipid clearance, leading to chronic elevation of free fatty acids (FFAs), which is associated with endothelial dysfunction, insulin-mediated renal sodium reabsorption, and sympathetic overactivation [61, 62]. The breakdown of muscle-adipose communication mediated by myokines and exosomes may promote inflammation and adipogenesis, creating a self-perpetuating cycle of metabolic deterioration [63–65]. Consequently, SO is associated with a higher likelihood of more advanced CKM stages within the CKM continuum that may warrant closer clinical monitoring compared with obesity alone [11, 66, 67]. A modest positive additive interaction was observed between sarcopenia and obesity. The RERI was 0.34 and the AP was 0.16, suggesting that approximately 16% of the risk among participants with joint exposure may be attributable to additive interaction. However, the multiplicative interaction was not statistically significant. Therefore, these findings should be interpreted as evidence of a potential additive joint association rather than definitive biological or mechanistic synergy. Our results (HR 2.07 for SO) are highly consistent with a recent independent study using CHARLS data by Xin et al. (2025), which reported a 1.73-fold increase in the risk of Major Adverse Cardiovascular Events (MACE) among SO patients [28]. Similarly, Zooravar et al., utilizing the NHANES cohort, found that sarcopenia not only increased all-cause mortality (HR 1.527) but was also significantly associated with advanced CKM stages [26]. While the association between metabolic syndrome (MetS) and CVD has been extensively characterized in prior literature [68], Standard MetScriteria, which are heavily weighted toward glucolipid metabolism and hemodynamic factors, often overlook skeletal muscle as the body’s critical‘functional reserve. Our data suggest that sarcopenia may be linked to poorer prognosis independent of metabolic risk tier. Given that sarcopenia may be associated with a pro-inflammatory milieu, these findings are consistent with prior work suggesting links between systemic inflammation and CKM risk [69]. By bridging the gap between “anatomical phenotypes” and “physiological staging”, we offer a risk stratification framework with additional prognostic information compared to MetS diagnosis alone. Furthermore, current evidence increasingly questions the utility of BMI in geriatric populations, frequently citing the “obesity paradox“ [70]. Our results were broadly consistent with a recent CHARLS-based study by Xin et al. (2025), which reported an increased risk of major adverse cardiovascular events among participants with sarcopenic obesity. Similarly, Zooravar et al. found that sarcopenia was associated with all-cause mortality and advanced CKM stages in NHANES. Compared with the Obesity Only group, the Sarcopenic Obesity group showed a higher risk of incident CVD in our cohort, suggesting that skeletal muscle status may provide additional prognostic information beyond adiposity status alone. These findings highlight the potential limitations of adiposity-only assessment in older adults and align with studies advocating more refined anthropometric metrics, such as the Body Roundness Index, beyond traditional BMI [71]. Several limitations should be acknowledged. First, because this was an observational study, causal inference cannot be established. Although participants with baseline clinical CVD were excluded from the incident CVD analysis and an additional restricted analysis was conducted among participants free of advanced CKM at baseline, reverse causation and bidirectional relationships between sarcopenia, obesity, CKM syndrome, and CVD cannot be fully excluded. Advanced CKM disease may contribute to muscle loss through inflammation, reduced physical activity, and disease-related catabolism. Second, body composition phenotypes were defined using baseline measurements only and were treated as fixed exposures during follow-up. Because repeated measurements required to consistently define sarcopenia and obesity were not uniformly available across follow-up waves, changes in phenotype status over time could not be modeled. This may have introduced exposure misclassification, and our findings should therefore be interpreted as the risk associated with baseline body composition phenotype rather than persistent or time-updated sarcopenic obesity. Third, muscle mass in CHARLS was estimated using anthropometric equations rather than DXA or CT. Although the equation has been validated in Chinese adults and showed good agreement with DXA-measured ASM in the NHANES validation analysis, measurement error remains possible. In addition, sarcopenia and obesity were defined differently across CHARLS and NHANES because of cohort-specific data availability and population-specific standards. Therefore, the NHANES analysis should be interpreted as external conceptual validation rather than direct replication of an identical phenotype. Fourth, partial overlap existed between obesity as an exposure and adiposity-related criteria included in CKM staging, particularly CKM Stage 1. This structural overlap may have contributed to circularity when interpreting associations between obesity-related phenotypes and CKM stage status. We did not remove adiposity-related criteria from the official AHA CKM staging definition because adiposity is an integral component of the CKM framework and excluding it would reduce comparability with existing studies. Fifth, CKM Stage 3 was defined using predicted 10-year CVD risk because direct biomarkers or imaging indicators of subclinical CVD were unavailable in CHARLS. This may have introduced classification bias. In addition, the CKM stage transition analysis was based on two time points and was presented as a descriptive observed transition probability matrix rather than a formal multi-state Markov model. Thus, it should not be interpreted as evidence of causal CKM progression or time-homogeneous disease transition. Sixth, residual confounding remains possible. Although the models adjusted for major sociodemographic and lifestyle factors, several potentially important confounders, including physical activity, dietary intake, medication use, income, inflammatory biomarkers, healthcare access, and kidney-related markers beyond eGFR, were not fully accounted for because of data limitations. Incident CVD in CHARLS was also based on self-reported heart disease or stroke, which may have introduced recall bias or outcome misclassification. Finally, the interaction findings should be interpreted cautiously. Although a modest positive additive interaction between sarcopenia and obesity was observed, the multiplicative interaction was not statistically significant. Statistical interaction does not necessarily indicate biological or mechanistic synergy. Therefore, the joint association of sarcopenia and obesity with incident CVD should be interpreted as a potential additive joint effect or “double burden,” rather than definitive biological synergy. Moreover, the NHANES validation analysis was cross-sectional and included a relatively small sarcopenia-only subgroup, which may limit precision. Future prospective studies with repeated body composition assessments, standardized definitions, objective CVD ascertainment, and mechanistic validation are needed to confirm these findings and determine whether muscle-related measures can improve CKM risk stratification. Conclusion Sarcopenic obesity was consistently associated with a greater burden of advanced CKM syndrome and a higher risk of incident CVD in the CHARLS longitudinal cohort, with similar cross-sectional associations observed in the NHANES validation cohort. Although causal relationships cannot be established, these findings suggest that sarcopenic obesity may serve as a clinically useful phenotypic marker for identifying individuals at increased cardiometabolic risk. Future studies are needed to determine whether muscle mass or muscle function assessment can improve CKM risk prediction beyond existing clinical and adiposity-based measures. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1