Predicting cardiovascular-kidney-metabolic multimorbidity in Chinese adults with overweight and obesity using machine learning: an internal evaluation

Abstract

Aim: To develop and internally evaluate machine learning (ML) models for predicting incident cardiovascular-kidney-metabolic (CKM) multimorbidity in Chinese adults with overweight or obesity, and to identify key predictors.

Methods: We included 4,244 participants from the China Health and Retirement Longitudinal Study (CHARLS) with overweight/obesity [body mass index (BMI) ≥ 24 kg/m²] and with zero or one CKM disease group at baseline (2015). CKM multimorbidity at follow-up (2018) was defined as coexistence of ≥ 2 disease groups (cardiovascular, kidney, metabolic). A two-stage feature selection [least absolute shrinkage and selection operator (LASSO) with bootstrap stability analysis] identified predictors from 56 sociodemographic, lifestyle, psychological, clinical, and environmental variables. Seven ML algorithms were compared; performance was assessed by area under the curve (AUC), calibration, Brier score, decision curve analysis, and SHapley Additive exPlanations (SHAP) interpretation for internal model evaluation.

Results: During 3-year follow-up, 648 (15.3%) participants developed incident CKM multimorbidity. Nine predictors were selected: age, depression, pain, health expectation, weight change, dyslipidemia, hypertension, heart disease, and kidney disease. Artificial neural network (ANN) and logistic regression showed the best discrimination (AUC: 0.758 and 0.757) and acceptable calibration in internal testing (Brier score: 0.113 and 0.114). SHAP analysis identified hypertension, dyslipidemia, and depression as top contributors. A nomogram was developed for preliminary risk stratification, but external validation is required before any clinical application.

Conclusion: In this internal evaluation, ANN and logistic regression showed moderate discrimination and acceptable calibration for predicting CKM multimorbidity in Chinese adults with overweight/obesity. Logistic regression performed comparably to complex algorithms yet is simpler; however, external validation is still required before use.

Keywords

Cardiovascular-kidney-metabolic multimorbidity / machine learning / prediction model / CHARLS / overweight / obesity / risk factors

References

[1] GBD 2023 Disease and Injury and Risk Factor Collaborators. Burden of 375 diseases and injuries, risk-attributable burden of 88 risk factors, and healthy life expectancy in 204 countries and territories, including 660 subnational locations, 1990-2023: a systematic analysis for the Global Burden of Disease Study 2023 Lancet 2025 406 1873 922 PMC12535840

[2] Gunnarsson S,Vito O,Unwin RJ. Cardiovascular-kidney-metabolic syndrome: prevalence, risks, disease trajectories, and early-stage management Am J Physiol Cell Physiol 2026 330 C1 8

[3] Bharaj IS,Brar A,Kacheria A.et al. Contemporary and emerging therapeutics in cardiovascular-kidney-metabolic (CKM) syndrome: in memory of Professor Akira Endo Biomedicines 2025 13 2192 PMC12467412

[4] Ndumele CE,Neeland IJ,Tuttle KR.et al. ; American Heart Association. 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

[5] Shi P,Lou C,Fang J,Song Y. Adverse life events across the life course and the risk of new-onset cardiovascular-kidney-metabolic multimorbidity in middle-aged and older adults: longitudinal evidence from CHARLS J Am Heart Assoc 2025 14 e045192 PMC12826910

[6] Shen C,Yang R,Zhou Y.et al. Metabolic dysfunction associated steatotic liver disease, cardiometabolic multimorbidity and mortality: evidence from the UK biobank Clin Res Cardiol 2026 115 1378 88

[7] Zhou Y,Dai X,Ni Y.et al. Interventions and management on multimorbidity: an overview of systematic reviews Ageing Res Rev 2023 87 101901

[8] Hall JE,do Carmo JM,da Silva AA,Wang Z,Hall ME. Obesity-induced hypertension: interaction of neurohumoral and renal mechanisms Circ Res 2015 116 991 1006 PMC4363087

[9] Czaja-Stolc S, Potrykus M, Stankiewicz M, Kaska Ł, Małgorzewicz S. Pro-inflammatory profile of adipokines in obesity contributes to pathogenesis, nutritional disorders, and cardiovascular risk in chronic kidney disease Nutrients 2022 14 1457 PMC9002635

[10] Sun X,Yan AF,Shi Z.et al. Health consequences of obesity and projected future obesity health burden in China Obesity 2022 30 1724 51

[11] Peng W,Chen S,Chen X.et al. Trends in major non-communicable diseases and related risk factors in China 2002-2019: an analysis of nationally representative survey data Lancet Reg Health West Pac 2024 43 100809 PMC10920046

[12] Yan Z,Chang X,Liu Z,Liu R,Du X. The association of obesity and lipid-related indicators with all-cause and cardiovascular mortality risks in patients with diabetes or prediabetes: a cross-sectional study based on machine learning algorithms Front Endocrinol 2025 16 1492082 PMC12171389

[13] Wang C,He S,Xie G.et al. Associations of longitudinal trajectories of triglyceride-glucose index combined with classical and novel obesity indices and cardiovascular disease: evidence from a nationwide prospective cohort study in China Cardiovasc Diabetol 2025 24 431 PMC12613409

[14] Zhang H,Zhang Y,Gao W,Mu Y. Identification of risk factors and development of a predictive model for chronic kidney disease in patients with obesity: a four-year cohort study Lipids Health Dis 2024 23 57 PMC10882765

[15] Jiang Y,Zhao B,Wang X.et al. UKB-MDRMF: a multi-disease risk and multimorbidity framework based on UK biobank data Nat Commun 2025 16 3767 PMC12015417

[16] Huang Q,Zou X,Lian Z.et al. Predicting cardiovascular outcomes in Chinese patients with type 2 diabetes by combining risk factor trajectories and machine learning algorithm: a cohort study Cardiovasc Diabetol 2025 24 61 PMC11806858

[17] Zhu J,Shi Z,Ge Z.et al. Comparative analysis of cardiometabolic multimorbidity predictors in China and the USA: a machine learning approach Diabetes Res Clin Pract 2025 229 112938

[18] Liu T,Krentz A,Lu L,Curcin V. Machine learning based prediction models for cardiovascular disease risk using electronic health records data: systematic review and meta-analysis. Eur Heart J Digit Health. 2025;6:7-22 PMC11750195

[19] Xu D,Xu Z. Machine learning applications in preventive healthcare: a systematic literature review on predictive analytics of disease comorbidity from multiple perspectives Artif Intell Med 2024 156 102950

[20] Khalid F,Alsadoun L,Khilji F.et al. Predicting the progression of chronic kidney disease: a systematic review of artificial intelligence and machine learning approaches Cureus 2024 16 e60145 PMC11166249

[21] 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 PMC3937970

[22] Rubino F,Cummings DE,Eckel RH.et al. Definition and diagnostic criteria of clinical obesity Lancet Diabetes Endocrinol 2025 13 221 62 PMC11870235

[23] Gao M,Lv J,Yu C.et al. ; China Kadoorie Biobank (CKB) Collaborative Group. Metabolically healthy obesity, transition to unhealthy metabolic status, and vascular disease in Chinese adults: a cohort study PLoS Med 2020 17 e1003351 PMC7598496

[24] Zhou Y,Kivimäki M,Yan LL.et al. Associations between socioeconomic inequalities and progression to psychological and cognitive multimorbidities after onset of a physical condition: a multicohort study EClinicalMedicine 2024 74 102739 PMC11327438

[25] Song Y,Zhu C,Shi B.et al. Social isolation, loneliness, and incident type 2 diabetes mellitus: results from two large prospective cohorts in Europe and East Asia and Mendelian randomization EClinicalMedicine 2023 64 102236 PMC10520296

[26] Li X,Zhang W,Zhang W.et al. Level of physical activity among middle-aged and older Chinese people: evidence from the China health and retirement longitudinal study BMC Public Health 2020 20 1682 PMC7653852

[27] Jiang C,Sun J,Lv Y.et al. The Chinese version of the 8-item Center for Epidemiologic Studies Depression Scale: Longitudinal psychometric syntheses with 10-year cohort multi-center evidence in an adult sample Gen Hosp Psychiatry 2024 91 204 11

[28] Zeng M,Chen Y,Lobanov-Rostovsky S.et al. Adiposity and dementia among Chinese adults: longitudinal study in the China Health and Retirement Longitudinal Study (CHARLS) Int J Obes 2025 49 706 14 PMC11999862

[29] Whelton PK,Carey RM,Aronow WS.et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines Hypertension 2018 71 1269 324

[30] American Diabetes Association. 2. Classification and diagnosis of diabetes: Standards of Medical Care in Diabetes-2018 Diabetes Care 2018 41 S13 27

[31] Joint committee issued Chinese guideline for the management of dyslipidemia in adults. [2016 Chinese guideline for the management of dyslipidemia in adults] Zhonghua Xin Xue Guan Bing Za Zhi 2016 44 833 53

[32] Pavlou M,Omar RZ,Ambler G. Penalized regression methods with modified cross-validation and bootstrap tuning produce better prediction models Biom J 2024 66 e202300245 PMC12859537

[33] Witte J,Foraita R,Didelez V. Multiple imputation and test-wise deletion for causal discovery with incomplete cohort data Stat Med 2022 41 4716 43

[34] Silva GFS,Fagundes TP,Teixeira BC,Chiavegatto Filho ADP. Machine learning for hypertension prediction: a systematic review Curr Hypertens Rep 2022 24 523 33

[35] Hu J,Szymczak S. A review on longitudinal data analysis with random forest Brief Bioinform 2023 24 bbad002 PMC10025446

[36] Fu C,Zhang Z,Li Y.et al. Association of the estimated glucose disposal rate combined with a body shape index with all-cause and cardiovascular-specific mortality among individuals with cardiovascular-kidney-metabolic syndrome Cardiovasc Diabetol 2026 25 112 PMC13045107

[37] Zhang X,Yao W,Wang D,Hu W,Zhang G,Zhang Y. Development and validation of machine learning models for identifying prediabetes and diabetes in normoglycemia Diabetes Metab Res Rev 2024 40 e70003 PMC11601146

[38] Jia W,Liu X,Wang Y,Pedrycz W,Zhou J. Semisupervised learning via axiomatic fuzzy set theory and SVM IEEE Trans Cybern 2022 52 4661 74

[39] Kriegeskorte N,Golan T. Neural network models and deep learning Curr Biol 2019 29 R231 6

[40] Nohara Y,Matsumoto K,Soejima H,Nakashima N. Explanation of machine learning models using shapley additive explanation and application for real data in hospital Comput Methods Programs Biomed 2022 214 106584

[41] Wang W,Liu Y,Yuan D.et al. Association between metabolic multimorbidity and the risk of cardiovascular disease, kidney disease, and mortality: longitudinal evidence from CHARLS (2011-2020) Public Health 2025 249 106020

[42] Song G. The C-reactive protein-triglyceride glucose index (CTI) predicts mortality in cardiovascular-kidney-metabolic syndrome: a dual-cohort study with machine learning validation Int J Surg 2026 112 1340 52 PMC12825706

[43] He D,Zhang X,Li C.et al. Rising prevalence of cardiovascular-kidney-metabolic syndrome in China, 2010-2019: national cross-sectional surveys J Am Coll Cardiol 2025 86 213 6

[44] Zhao W,Yan Q,Mou C. Physical activity and cardiovascular-metabolic disease risk across cardiovascular-kidney-metabolic syndrome stages: a population-based cohort study BMC Cardiovasc Disord 2025 25 748 PMC12538840

[45] Choi S,Oh M,Lee DH,Jee SH,Jeon JY. Invasive and non-invasive variables prediction models for cardiovascular disease-specific mortality between machine learning vs. traditional statistics Sci Rep 2025 15 35093 PMC12508446

[46] Wang Y,Yang Y,Chen J.et al. Transition of BMI status from childhood to adulthood and cardiovascular-kidney-metabolic syndrome in midlife: a 36-year cohort study Diabetes Care 2025 48 2045 53