Development and Validation of an Interpretable Machine Learning Model for Predicting 5-year Major Adverse Cardiovascular Events in Patients with Coronary Artery Disease

Development and Validation of an Interpretable Machine Learning Model for Predicting 5-year Major Adverse Cardiovascular Events in Patients with Coronary Artery Disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Development and Validation of an Interpretable Machine Learning Model for Predicting 5-year Major Adverse Cardiovascular Events in Patients with Coronary Artery Disease Zhongxing Jiang, Haofeng Zhou, Yindu Liu, Han Yin, Junshuo Zhu, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8616117/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 15 You are reading this latest preprint version Abstract Background Coronary artery disease (CAD) remains a major contributor to global cardiovascular mortality. The accurate prediction of prognosis is critical for guide clinical decision-making. This study aimed to develop and validate interpretable machine learning (ML) models for predicting 5-year major adverse cardiovascular events (MACE) in hospitalized CAD patients. Methods A prospective cohort of 705 CAD patients was included and randomly divided into training (n = 564) and validation (n = 141) sets. Eleven key predictors were selected using Least Absolute Shrinkage and Selection Operator (LASSO) regression. Six ML algorithms were developed, and model performance was assessed using discrimination, calibration, and decision curve analysis. Shapley Additive Explanations (SHAP) were applied to enhance model interpretability. Results Key predictors identified by LASSO regression included left ventricular ejection fraction (LVEF), N-terminal pro-B-type natriuretic peptide (NT-proBNP), nitrate use, CAD duration, depressive symptoms, and age. The random forest (RF) model demonstrated superior performance, achieving the highest Area Under the Curve (AUC) in both training (0.887, 95% CI: 0.859–0.915) and validation (0.753, 95% CI: 0.656–0.849) cohorts, along with optimal balance of sensitivity (0.856), F1 score (0.708), and Brier score (0.152). The LASSO method revealed that LVEF, NT-proBNP, and nitrate use were the top 3 predictors of 5-year mace. Depressive symptoms were also associated with increased MACE risk. Conclusions This interpretable RF-based model provides accurate and interpretable 5-year MACE prediction in CAD patients. By integrating clinical and psychosocial features, it supports personalized secondary prevention. External validation is warranted to assess real-world applicability. Coronary artery disease Major adverse cardiovascular events Machine learning SHAP value Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Coronary artery disease (CAD) remains a leading global cause of morbidity and mortality, posing a significant public health burden(1,2). According to the Global Burden of Disease study, ischemic heart disease, the most common manifestation of CAD, consistently ranks among the top causes of death and disability-adjusted life years globally(3). Despite advances in medical therapy, revascularization techniques, and secondary prevention strategies, patients with CAD continue to face substantial long-term risks(4). Critically, the post-discharge period represents a high risk phase rather than an endpoint of vulnerability(5,6). These events can significantly impact long-term survival, quality of life, and healthcare utilization. Consequently, effective long-term risk stratification is essential—not only to identify high-risk individuals but also to personalize post-discharge secondary prevention strategies. Traditional prognostic models for CAD patients, such as simple integer risk score(7) and the biomarker-based ABC-CHD model(8), rely on a relatively limited set of clinical or biomarker variables and were developed primarily in stable outpatient populations, which may limit their applicability to broader hospitalized CAD cohorts. Their performance in real-world clinical settings, particularly for long-term outcomes, remains suboptimal(9). Machine learning (ML) approaches offer the ability to model complex, nonlinear relationships and integrate a broader range of routinely collected features, including advanced imaging, detailed laboratory values, and granular clinical parameters(10,11). This expanded feature space enables improved risk stratification and predictive performance. While ML has demonstrated improved predictive capabilities in CAD, many prior studies have either concentrated on short-term outcomes or been restricted to specific clinical contexts—such as percutaneous coronary intervention (PCI)-treated patients or those with suspected CAD—limiting their generalizability to broader cohorts and long-term secondary prevention(9,12). In this study, we aimed to develop and validate ML-based models for predicting 5-year MACE in hospitalized CAD patients. By leveraging routinely collected clinical data, we sought to improve long-term risk stratification and support individualized care in secondary prevention. 2. Methods 2.1. Design and participants This prospective cohort study included 705 consecutive inpatients diagnosed with CAD at Guangdong Provincial People's Hospital between October 2017 and January 2018. Baseline clinical data were extracted from electronic medical records, including routine examinations, coronary angiography (CAG) results, and discharge diagnoses. Participants were randomly assigned to a training cohort (n = 564) and a validation cohort (n = 141) in an 8:2 ratio. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Medical Ethics Committee of Guangdong Provincial People's Hospital. Written informed consent was obtained from all participants. 2.2. Study variables 2.2.1. Outcome variables The primary outcome was 5-year MACE defined as a composite of cardiac death, unplanned revascularization, cardiac rehospitalization, non-fatal myocardial infarction, and stroke. Outcome data were extracted from electronic medical records and independently verified through structured annual telephone interviews with patients or their families. Participants were prospectively followed on an annual basis for up to five years. Follow-up was censored at the earliest occurrence of a MACE event, all-cause death, loss to follow-up, or study termination on January 31, 2023. 2.2.2. Predictors A total of 54 candidate predictors were selected for model development based on clinical relevance, routine availability, and evidence from prior literature. These variables included five categories: (1) demographic and lifestyle factors, such as age, sex, and body mass index (BMI); (2) psychological factors, such as depressive symptoms and anxiety symptoms, assessed using two standardized self-report questionnaires—the Patient Health Questionnaire-9 (PHQ-9)(13) and the Generalized Anxiety Disorder-7 (GAD-7)(14)—administered by the same researcher to all patients on the night before surgery during hospitalization. Depressive symptoms and anxiety symptoms were defined as PHQ-9 and GAD-7 scores ≥ 9, respectively; (3) clinical history and comorbidities, such as CAD duration, hypertension, and diabetes mellitus; (4) medication and intervention history, such as aspirin use, statin therapy, and previous PCI; and (5) imaging and laboratory findings, such as left ventricular ejection fraction (LVEF), fasting glucose, and N-terminal pro-B-type natriuretic peptide (NT-proBNP). Detailed definitions, measurement methods, and coding schemes for all predictors are provided in Supplementary Table S1 . 2.3. Statistical analysis 2.3.1. Data preparation Missing data were handled via multiple imputation by chained equations among candidate predictors under the assumption of missing at random. Baseline characteristics were compared between patients with and without MACE, as well as between the training and validation cohorts. Student’s t-test was used for continuous variables, and the chi-square test was used for categorical variables. A two-sided p-value < 0.05 was considered statistically significant. All analyses were conducted using R software (version 4.4.3). 2.3.2. Variable selection To reduce model complexity and identify the most informative predictors of 5-year MACE, the least absolute shrinkage and selection operator (LASSO) regression was used. This penalized regression technique applies an L1 penalty to shrink less relevant coefficients toward zero. All predictors were standardized before modeling. The optimal regularization parameter was determined via 10-fold cross-validation in the training cohort, using the area under the receiver operating characteristic curve (AUC) as the selection criterion. The final set of variables corresponding to the lambda value that achieved the highest mean AUC was selected for subsequent model construction. 2.3.3. Model development Six classification algorithms were used to develop prediction models for 5-year MACE: logistic regression (LR), k-nearest neighbors (KNN), support vector machine (SVM), random forest (RF), gradient boosting machine (GBM), and LightGBM. Hyperparameters for each model were tuned using grid search with 10-fold cross-validation in the training cohort. Final model parameters were selected based on a comprehensive assessment of the highest mean AUC, lowest Brier score, and optimal F1 score within the training data. 2.3.4. Performance evaluation Model performance was assessed in terms of discrimination, calibration, and clinical utility. Discrimination was evaluated using AUC derived from receiver operating characteristic (ROC) curves, and AUC comparisons among models were conducted using the DeLong test. Calibration was assessed via Brier scores and visualized with calibration plots. Clinical utility was evaluated using decision curve analysis (DCA). To enhance model interpretability, Shapley Additive Explanations (SHAP) were computed to quantify the marginal contribution of each predictor to model output, both at the global and individual levels. 3. Results 3.1. Participant characteristics Among 705 hospitalized CAD patients, 221 (31.3%) developed a MACE within 5 years. The baseline characteristics of patients with and without MACE are summarized in Table 1 . Compared with non-MACE patients, those who experienced MACE were more likely to be older, have a longer CAD duration, lower LVEF, higher NT-proBNP levels, a history of stroke, advanced New York Heart Association (NYHA) functional class, and elevated blood urea nitrogen (BUN) and creatinine levels. They were also more likely to have received nitrate therapy, undergone multi-stent PCI, and present with extensive coronary artery disease. Baseline characteristics of the training and validation cohorts are shown in Supplementary Table S2, and the two cohorts were generally comparable in demographic, clinical, and biochemical profiles. Table 1 Characteristics of subjects with or without MACE Predictors Non-Mace (n = 484) Mace (n = 221) x 2 /t a p - value Age, Mean ± SD 62.32 ± 10.14 64.98 ± 10.84 -3.07 0.002* Sex, n(%) Female 129(26.7) 65(29.4) 0.45 0.503 Male 355(73.3) 156(70.6) Body mass index(kg/m2), Mean ± SD 24.55 ± 3.08 24.17 ± 3.64 1.33 0.184 Marriage, n(%) Divorced or Widowed or Single 32(6.6) 9(4.1%) 1.35 0.245 Married 452(93.4) 212(95.9) Education, n(%) Less than 6 years 131(27.1) 75(33.9) 4.59 0.204 7–9 years 149(30.8) 54(24.4) 10–12 years 101(20.9) 45(20.4) More than 12 years 103(21.3) 47(21.3) Sleep quality, n(%) Good 104(21.5) 50(22.6) 5.8 0.122 Fair 202(41.7) 78(35.3) Poor 120(24.8) 53(24%) Very poor 58(12) 40(18.1) Smoking status, n(%) Previous/current 210(43.4) 104(47.1) 0.69 0.408 Depressive symptoms , n(%) 56 (12) 38 (17) 3.68 0.055 Anxiety symptoms, n(%) 32 (7) 20 (9) 0.99 0.320 CAD duration(mouth), Mean ± SD 31.06 ± 48.93 40.29 ± 51.33 -2.25 0.025* NYHA, n(%) Class I 186 (38) 68 (31) 17.4 < 0.001* Class II 248 (51) 107 (48) Class III 42 (9) 33 (15) Class III-IV 8 (2) 13 (6) Medical history Hypertension, n(%) 283(58.5) 133(60.2) 0.12 0.730 Diabetes mellitus, n(%) 149(30.8) 74(33.5) 0.39 0.530 Stroke 22 (5) 21 (10) 5.67 0.017* Medication use Aspirin,n(%) 397 (82) 188 (85) 0.79 0.374 Clopidogrel or Ticagrelor,n(%) 360 (74) 179 (81) 3.33 0.068 Anticoagulants,n(%) 23 (5) 14 (6) 0.48 0.489 Statin, n(%) 436(90.1) 208(94.1) 2.64 0.104 βBlocker, n(%) 389(80.4) 187(84.6) 1.55 0.212 ACEI or ARB, n(%) 325(67.1) 158(71.5) 1.13 0.287 Nitrates, n(%) 50(10.3) 36(16.3) 4.49 0.034* CCB, n(%) 108(22.3) 53(24) 0.15 0.695 Previous PCI, n(%) Never 359(74) 146(66) 14.37 0.039* One stent 60(12) 27(12) Two stents 35(7) 23(10) Three or more stents 30(6) 25(11) Number of diseased vessels, n(%) No vessel 87 (18) 32 (14) 13.5 0.004* Single vessel 99 (20) 25 (11) Two vessels 90 (19) 41 (19) Three or more vessels 208 (43) 123 (56) Left main CAD, n(%) No 396 (82) 179 (81) 0.02 0.876 Yes 88 (18) 42 (19) HbA1c(%), Mean ± SD 6.52 ± 1.41 6.76 ± 1.61 -1.94 0.054 Fast glucose(mmol/L), Mean ± SD 6.28 ± 2.72 6.66 ± 2.87 -1.65 0.099 CREA(µmol/L), Mean ± SD 94.87 ± 31.72 105.8 ± 64.91 -2.38 0.018* BUN(mmol/L), Mean ± SD 6.17 ± 2.04 7.17 ± 3.65 -3.83 < 0.001* TC(mmol/L), Mean ± SD 4.46 ± 1.32 4.34 ± 1.18 1.25 0.212 HDL-C(mmol/L), Mean ± SD 1 ± 0.24 0.96 ± 0.23 2.18 0.030* LDL-C(mmol/L), Mean ± SD 2.89 ± 0.92 2.83 ± 0.88 0.84 0.403 TG(mmol/L), Mean ± SD 1.74 ± 1.7 1.77 ± 1.27 -0.25 0.804 ApoA(g/L), Mean ± SD 1.14 ± 0.18 1.13 ± 0.18 1.09 0.277 ApoB(g/L), Mean ± SD 0.91 ± 0.27 0.9 ± 0.26 0.71 0.479 Lpa(nmol/L), Mean ± SD 303.63 ± 359.67 315.45 ± 367.39 -0.4 0.690 FT3(pmol/L), Mean ± SD 4.86 ± 0.8 4.77 ± 0.86 1.31 0.189 FT4(pmol/L), Mean ± SD 12.5 ± 9.5 13.37 ± 12.44 -0.92 0.360 TSH(mIU/L), Mean ± SD 1.92 ± 3.51 1.82 ± 1.65 0.56 0.578 CK(U/L), Mean ± SD 115.39 ± 98.87 128.77 ± 146.84 -1.23 0.218 CKMB(U/L), Mean ± SD 11.86 ± 7.61 12.72 ± 8.84 -1.26 0.208 NT-proBNP(ng/L), Mean ± SD 743.61 ± 2067.21 1537.15 ± 3514.26 -3.12 0.002* PT(s), Mean ± SD 13.65 ± 1.15 13.86 ± 1.19 -2.17 0.031* APTT(s), Mean ± SD 37.99 ± 4.3 38.04 ± 4.63 -0.15 0.883 FIB(g/L), Mean ± SD 3.94 ± 1.1 4.08 ± 1.15 -1.47 0.143 TT(s), Mean ± SD 17.14 ± 8.06 17.57 ± 8.43 -0.63 0.527 INR, Mean ± SD 1.05 ± 0.12 1.06 ± 0.11 -1.56 0.120 D-dimer(µg/mL), Mean ± SD 674.61 ± 940.16 676.23 ± 612.41 -0.03 0.978 hs-CRP(mg/L), Mean ± SD 8.83 ± 17.13 8.53 ± 13.82 0.25 0.806 LVEF(%), Mean ± SD 59.74 ± 9.69 54.81 ± 13.58 4.86 < 0.001* Abbreviations: CAD, coronary artery disease; NYHA, New York Heart Association functional classification; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; PCI, percutaneous coronary intervention; HbA1c, glycated hemoglobin A1c; Fast glucose, fasting blood glucose; CREA, creatinine; BUN, blood urea nitrogen; TC, total cholesterol; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; TG, triglyceride; ApoA, apolipoprotein A1; ApoB, apolipoprotein B; Lpa, lipoprotein(a); FT3, free triiodothyronine; FT4, free thyroxine; TSH, thyroid-stimulating hormone; CK, creatine kinase; CKMB, creatine kinase MB isoenzyme; NT-proBNP, N-terminal pro-brain natriuretic peptide; PT, prothrombin time; APTT, activated partial thromboplastin time; FIB, fibrinogen; TT, thrombin time; INR, international normalized ratio; hs-CRP, high sensitivity C-reactive protein; LVEF, left ventricular ejection fraction. a The comparative analyses for continuous variables and categorical variables were conducted using Student’s t test and Chi-square test, respectively. * p < 0.05 3.2. Variable selection results To identify the most relevant predictors of 5-year MACE, we employed LASSO regression using the glmnet package (version 4.4.3) in R. All 54 candidate variables were z-score normalized before modeling. A 10-fold cross-validation procedure was used to determine the optimal regularization parameter, and the value corresponding to the minimum mean cross-validated deviance (lambda.min) was selected for final variable selection (Supplementary Figure S1 ). At this lambda value, 11 variables with non-zero coefficients were retained (Supplementary Table S1 ), including age, stroke, NYHA functional class, previous PCI, number of diseased vessels, CAD duration, depressive symptoms, nitrates use, NT-proBNP, BUN, and LVEF. Most of these predictors have well-established associations with adverse cardiovascular outcomes. The selected predictors were subsequently used for model development across six classification algorithms. 3.3. Prediction performance for 5-year MACE The discriminatory performance of all six models is illustrated in Fig. 1 , and detailed performance metrics are provided in Table 2 . In the training cohort, the RF model achieved the highest AUC of 0.887 (95% CI: 0.859–0.915), followed by GBM (AUC: 0.805) and LightGBM (AUC: 0.777). The SVM, KNN, and LR models showed moderate discrimination, with AUCs of 0.731, 0.679, and 0.660, respectively. Pairwise comparisons confirmed that RF significantly outperformed all other models in terms of AUC. Table 2 Performance comparison of 6 MACE prediction models with 11 variables Model AUC(95%CI) Threshold* Sensitivity Specificity PPV NPV F1 Brier score LR 0.660(0.611–0.708) 0.363 0.470 0.775 0.497 0.756 0.483 0.201 KNN 0.679(0.632–0.725) 0.260 0.657 0.606 0.441 0.789 0.528 0.201 SVM 0.731(0.685–0.777) 0.296 0.751 0.634 0.493 0.844 0.595 0.201 RF 0.887(0.859–0.915) 0.185 0.856 0.734 0.603 0.915 0.708 0.152 GBM 0.805(0.768–0.843) 0.326 0.707 0.773 0.595 0.848 0.646 0.181 LightGBM 0.777(0.737–0.816) 0.317 0.713 0.705 0.533 0.839 0.610 0.196 LR 0.697(0.604–0.790) 0.345 0.550 0.802 0.524 0.818 0.537 0.185 KNN 0.649(0.538–0.760) 0.300 0.475 0.812 0.500 0.796 0.487 0.190 SVM 0.644(0.533–0.755) 0.303 0.525 0.782 0.488 0.806 0.506 0.199 RF 0.753(0.656–0.849) 0.215 0.675 0.782 0.551 0.859 0.607 0.177 GBM 0.746(0.647–0.845) 0.355 0.575 0.881 0.657 0.840 0.613 0.175 LightGBM 0.735(0.631–0.840) 0.352 0.575 0.891 0.676 0.841 0.622 0.184 Abbreviations: AUC, area under the receiver operating characteristic curve; PPV, positive predictive value; NPV, negative predictive value; F1, F1 score; LR, logistic regression; KNN, k-nearest neighbors; SVM, support vector machine; RF, random forest; GBM, gradient boosting machine; LightGBM, Light Gradient Boosting Machine. *The optimal threshold for binary classification was determined using the maximum Youden index from the ROC curve. In addition to AUC, RF also demonstrated the best overall predictive performance in the training cohort, with the highest sensitivity (0.856), and F1 score (0.708), as well as the lowest Brier score (0.152). GBM and LightGBM also performed well, with F1 score of 0.646 and 0.610, respectively. Model calibration was evaluated using calibration plots in Supplementary Fig. 2 − 1 (training cohort) and Supplementary Fig. 2–2 (validation cohort). In the training cohort, the RF model showed the best calibration, with predicted risks closely aligning with actual outcomes. In the validation cohort, GBM and RF exhibited superior calibration performance compared with other models, with calibration curves closer to the ideal diagonal line. In the validation cohort, RF and GBM remained the top performers, with AUCs of 0.753 (95% CI: 0.656–0.849) and 0.746 (95% CI: 0.647–0.845), outperforming the remaining models (AUC range: 0.644–0.735). GBM and LightGBM achieved the highest specificity (0.881 and 0.891, respectively), while RF offered the best balance between sensitivity (0.675) and negative predictive value (NPV = 0.859). Among the traditional models, LR showed moderate performance (AUC = 0.697), whereas SVM and KNN performed relatively poorly. Decision curve analysis (Fig. 2 ) further confirmed the clinical utility of the RF, GBM, and LightGBM models. These approaches consistently provided greater net benefit across a wide range of decision thresholds compared with the “treat-all” or “treat-none” strategies, supporting their potential role in individualized risk assessment and clinical decision-making. 3.4. Importance of predictive variables for 5-year MACE Variable importance was evaluated based on the random forest model trained with the 11 LASSO-selected predictors. As shown in Fig. 3 , LVEF, NT-proBNP, and nitrates use were identified as the top three predictors contributing most to model performance. Figure 3 also depicts several clinically relevant and directionally consistent feature–outcome relationships through distinct SHAP value distributions. NT-proBNP shows a strong positive association, with higher values—represented by red data points—mainly concentrated in the positive SHAP region, highlighting its adverse impact on 5-year MACE. Furthermore, the SHAP dependence plot illustrates the effect of each feature on individual risk prediction (Fig. 4 ). As illustrated in the SHAP dependence plots, LVEF and depressive symptoms showed distinct associations with the predicted 5-year MACE risk. For LVEF, a U-shaped pattern was observed: SHAP values below 50% or above 70% were associated with increased SHAP values, suggesting elevated risk at both ends of left ventricular function. Patients identified as having depressive symptoms exhibited higher SHAP values, suggesting a possible association between depressive symptoms and elevated MACE risk. Other variables associated with higher SHAP values included age above 70 years, elevated NT-proBNP (> 1100 pg/ml), nitrate use, higher NYHA functional class, history of PCI, elevated BUN, longer CAD duration, and greater number of diseased coronary vessels. These features were all linked to an increased likelihood of adverse cardiovascular events over the 5-year follow-up period. 4. Discussion CAD remains a major global contributor to long-term cardiovascular morbidity and mortality. Accurate post-discharge risk prediction is essential for optimizing secondary prevention. In this prospective study, we developed six ML models to predict 5-year MACE in hospitalized CAD patients, identifying RF as the optimal model. ML offers distinct advantages over traditional statistical models in prognostic prediction, particularly for diseases with multiple and complex risk factors such as coronary heart disease. Unlike conventional linear models that rely on predefined assumptions, ML algorithms can automatically learn nonlinear relationships and interactions from large-scale, high-dimensional clinical data, thereby improving prediction accuracy and enabling personalized risk stratification. In cardiovascular medicine, ML has been increasingly applied to diagnosis, classification, and outcome prediction, showing potential to refine prognostic assessments and guide more targeted interventions for patients undergoing PCI(9). Previous models for predicting cardiovascular risk in coronary artery disease patients have shown varying approaches and performance. Ganz et al. developed a 9-protein risk score with moderate discrimination (C-statistics 0.64–0.75) outperforming the Framingham model for 4-year outcomes(15). Lindholm et al. created the “ABC-CHD” model combining biomarkers and clinical factors, achieving high discrimination for cardiovascular death (c-index 0.78–0.81)(8). Liu et al. applied machine learning to PCI patients, with a random forest model predicting 5-year all-cause mortality with an AUC of 0.71(9). Our findings similarly support the superior predictive value of ML methods like random forest for long-term prognosis in CAD, underscoring the potential of integrating complex clinical and biomarker data to enhance risk stratification. RF—an ensemble method aggregating multiple decision trees—outperforms traditional logistic regression by capturing complex nonlinear relationships without strict parametric assumptions(16–18). In our cohort, RF achieved superior discrimination (training AUC: 0.887, 95% CI: 0.859–0.915; validation AUC: 0.753), sensitivity (0.856), F1-score (0.708), and calibration (Brier score: 0.152). Importantly, we improved the interpretability of the RF model by applying SHAP. Notably, our model revealed a U-shaped relationship between LVEF and 5-year MACE risk, with increased risk observed both below 50% and above 70%. This aligns with prior studies: Liu et al. reported that LVEF below 55% was associated with increased all-cause mortality and major adverse cardiac and cerebrovascular events in PCI patients(19); Chang et al. demonstrated that LVEF exceeding 70% conferred higher mortality risk, potentially linked to left ventricular concentric remodeling, especially in women(20). These findings underscore the importance of monitoring both impaired and supranormal LVEF values in CAD prognostication. While the adverse prognostic impact of reduced LVEF is well established, the mechanisms underlying the risk associated with supranormal LVEF are less widely appreciated. One explanation is that supranormal LVEF often reflects concentric left ventricular remodeling, which can impair diastolic filling and elevate filling pressures, particularly in women(21). In addition, excessive contractility may be accompanied by coronary microvascular dysfunction, where increased myocardial oxygen demand induces microvascular ischemia, myocardial injury, and interstitial fibrosis, ultimately compromising cardiac performance(22). Furthermore, patients with supranormal LVEF frequently exhibit diffuse interstitial myocardial fibrosis—manifested by increased extracellular volume—rather than focal scarring, a pathophysiological substrate closely linked to heart failure with preserved ejection fraction and adverse cardiovascular outcomes(23). Depressive symptoms were also significantly associated with higher MACE risk, highlighting the crucial role of psychosocial factors in CAD prognosis. The interplay between depressive symptoms and cardiovascular outcomes may involve behavioral and biological pathways, including systemic inflammation, autonomic dysfunction, endothelial impairment, hypothalamic–pituitary–adrenal axis dysregulation, platelet hyperactivation, and genetic predisposition(24,25). Clinically, patients with emotional distress often present with concomitant somatic complaints such as chest tightness, chest pain, palpitations, and dyspnea(26,27). These symptoms are frequently misattributed by patients to recurrent cardiac events, prompting more aggressive medical consultation and intervention(28). This heightened vigilance may, in turn, increase the likelihood of subsequent healthcare utilization and rehospitalization. Future research should further explore the contribution of this symptom-perception–driven pathway to long-term outcomes in CAD patients with emotional comorbidities. NT-proBNP was identified as the second most important predictor, reflecting myocardial wall stress and volume overload often resulting from ischemia or myocardial injury. Elevated NT-proBNP (> 1100 pg/ml in our cohort) predicted increased long-term MACE risk, corroborating previous research demonstrating its prognostic value in chronic cardiac conditions(29,30). Vergaro et al. analyzed 12,763 patients with stable heart failure (mean LVEF 33%) in the BIOS consortium and reported optimal NT-proBNP cutoffs across BMI categories for 5-year all-cause mortality ranging from 3,785 pg/ml in underweight to 755-1,554 pg/ml in overweight or obese patients(31). In a prospective study of 355 patients presenting to the emergency department with atrial fibrillation, Holl et al. reported that NT-proBNP > 500 pmol/L was associated with increased risk of death and MACE over 2 years(32). The cutoff identified in our study (> 1,100 pg/ml) is lower than those in acute AF cohorts but close to the range observed in overweight/obese HF patients in Vergaro et al. Possible explanations include differences in underlying disease type and severity, BMI distribution, outcome definitions (MACE vs. mortality), follow-up duration, and measurement conditions. These factors should be considered when applying NT-proBNP thresholds across different clinical settings. The strengths of this study include its prospective design with complete 5-year follow-up and clinically verified endpoints, comprehensive variable selection encompassing psychosocial factors, and rigorous validation through multiple performance metrics coupled with SHAP-enhanced interpretability. However, several limitations warrant consideration. First, the single-center data may limit generalizability to diverse healthcare settings. Second, external validation in multicenter cohorts is needed to confirm robustness despite internal validation. Finally, while SHAP improves transparency, it does not fully resolve the inherent 'black box' nature of ensemble ML models. 5. Conclusion In this prospective cohort study of hospitalized CAD patients, we developed and validated ML-based models for predicting 5-year MACE, with RF demonstrating the best performance across discrimination, calibration, and clinical utility. By incorporating routinely available clinical, laboratory, and psychosocial variables, our models achieved accurate long-term risk stratification and enhanced interpretability through SHAP analysis. These findings support the integration of ML-based risk prediction into secondary prevention strategies to improve individualized patient management after hospital discharge. External validation in diverse populations is warranted to confirm generalizability. Abbreviations CAD = Coronary artery disease MACE = major adverse cardiovascular events ML = Machine learning PCI = percutaneous coronary intervention CAG = coronary angiography PHQ-9 = Patient Health Questionnaire-9 GAD-7 = Generalized Anxiety Disorder-7 LASSO = least absolute shrinkage and selection operator AUC = area under the receiver operating characteristic curve LR = logistic regression KNN = k-nearest neighbors SVM = support vector machine RF = random forest GBM = gradient boosting machine ROC = receiver operating characteristic DCA = decision curve analysis SHAP = Shapley Additive Explanations NT-proBNP = N-terminal pro-B-type natriuretic peptide NYHA = New York Heart Association BUN = blood urea nitrogen Declarations Ethics approval and consent to participate Ethical approval was given by the medical ethics committee of Guangdong General Hospital with the following reference number: No.GDREC2017203H. All participants gave written informed consent. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests Funding This project is supported by grant from the Guangzhou Municipal Science and Technology Program key projects (2023B03J1249), China Heart House-Chinese Cardiovascular Association TCM fund (CCA-TCM-032; 202342) and Su Ke'an Pharmaceutical Research and Development Project (202460). Author Contribution Z.X.J., H.F.Z., Y.D.L., H.Y. and H.M. contributed to the conception and design of the work. Y.D.L., H.Y., J.S.Z., X.Y.X., J.N.C. and R.W. were responsible for the acquisition of data. Z.X.J. and H.F.Z. performed all data analyses. Z.X.J., H.F.Z. and Y.D.L. drafted the manuscript. All the authors gave comments and revised the manuscript. All the authors approved the final version to be submitted. Data Availability The datasets generated and/or analyzed during the current study are not publicly available due to patient privacy and ethical restrictions but are available from the corresponding author upon reasonable request. Requests for access to the data should be directed to Dr. Ma (email: [ [email protected] ]). References Wang Z, Du A, Liu H, Wang Z, Hu J. Systematic Analysis of the Global, Regional and National Burden of Cardiovascular Diseases from 1990 to 2017. J Epidemiol Glob Health. 2022 Mar;12(1):92–103. Neto MG, Saquetto MB, Roever L, Carvalho VO. The Effect of Yoga Intervention on Psychological Symptoms, Health-Related Quality of Life, and Cardiovascular Risk Factors in People with Coronary Artery Disease: A Systematic Review and Meta-Analysis. Heart and Mind. 2024 Oct;8(4):300–9. Vos T, Lim SS, Abbafati C, Abbas KM, Abbasi M, Abbasifard M, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. The Lancet. 2020 Oct;396(10258):1204–22. Freites A, Hernando L, Salinas P, Cánovas E, De La Rosa A, Alonso J, et al. Incidence and prognosis of late readmission after percutaneous coronary intervention. Cardiol J. 2023 Oct 27;30(5):696–704. Vallejo-Vaz AJ, Dharmayat KI, Nzeakor N, Carrasco CP, Fatoba ST, Fonseca MJ, et al. Recurrent cardiovascular and limb events in 294,428 patients with coronary or peripheral artery disease or ischemic stroke on antiplatelet monotherapy: The RESRISK cohort study. Atherosclerosis. 2024 Nov;398:118589. Brown TM, Bittner V, Colantonio LD, Deng L, Farkouh ME, Limdi N, et al. Residual risk for coronary heart disease events and mortality despite intensive medical management after myocardial infarction. Journal of Clinical Lipidology. 2020 Mar;14(2):260–70. Bavry AA, Kumbhani DJ, Gong Y, Handberg EM, Cooper-DeHoff RM, Pepine CJ. Simple Integer Risk Score to Determine Prognosis of Patients With Hypertension and Chronic Stable Coronary Artery Disease. JAHA. 2013 Aug 22;2(4):e000205. Lindholm D, Lindbäck J, Armstrong PW, Budaj A, Cannon CP, Granger CB, et al. Biomarker-Based Risk Model to Predict Cardiovascular Mortality in Patients With Stable Coronary Disease. Journal of the American College of Cardiology. 2017 Aug;70(7):813–26. Liu S, Yang S, Xing A, Zheng L, Shen L, Tu B, et al. Machine learning-based long-term outcome prediction in patients undergoing percutaneous coronary intervention. Cardiovasc Diagn Ther. 2021 Jun;11(3):736–43. Lin G, Liu Q, Chen Y, Zong X, Xi Y, Li T, et al. Machine Learning to Predict Long-Term Cardiac-Relative Prognosis in Patients With Extra-Cardiac Vascular Disease. Front Cardiovasc Med. 2021 Nov 25;8:771504. Kigka VI, Georga E, Tsakanikas V, Kyriakidis S, Tsompou P, Siogkas P, et al. Machine Learning Coronary Artery Disease Prediction Based on Imaging and Non-Imaging Data. Diagnostics. 2022 Jun 14;12(6):1466. Motwani M, Dey D, Berman DS, Germano G, Achenbach S, Al-Mallah MH, et al. Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis. Eur Heart J. 2016 Jun 1;ehw188. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med. 2001 Sep;16(9):606–13. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Arch Intern Med. 2006 May 22;166(10):1092. Ganz P, Heidecker B, Hveem K, Jonasson C, Kato S, Segal MR, et al. Development and Validation of a Protein-Based Risk Score for Cardiovascular Outcomes Among Patients With Stable Coronary Heart Disease. JAMA. 2016 Jun 21;315(23):2532. Xing F, Luo R, Liu M, Zhou Z, Xiang Z, Duan X. A New Random Forest Algorithm-Based Prediction Model of Post-operative Mortality in Geriatric Patients With Hip Fractures. Front Med. 2022 May 11;9:829977. Hong W, Lu Y, Zhou X, Jin S, Pan J, Lin Q, et al. Usefulness of Random Forest Algorithm in Predicting Severe Acute Pancreatitis. Front Cell Infect Microbiol. 2022 Jun 10;12:893294. Breiman L. Random Forests. Machine Learning. 2001 Oct 1;45(1):5–32. Liu Y, Song J, Wang W, Zhang K, Qi Y, Yang J, et al. Association of ejection fraction with mortality and cardiovascular events in patients with coronary artery disease. ESC Heart Failure. 2022 Oct;9(5):3461–8. Chang HC, Tseng CH, Huang WM, Lee CW, Yu WC, Cheng HM, et al. Supranormal Left Ventricular Ejection Fraction, Concentric Remodeling, and Long-Term Survival. JACC: Asia. 2024 Dec;4(12):928–37. Chan NI, Atherton JJ, Krishnan A, Hammett C, Stewart P, Mallouhi M, et al. Diastolic Dysfunction and Survival in Patients With Preserved or Mildly Reduced Left Ventricular Ejection Fraction Following Myocardial Infarction. Journal of the American Society of Echocardiography. 2025 May;38(5):380–91. Wu P, Zhang X, Wu Z, Chen H, Guo X, Jin C, et al. Impaired coronary flow reserve in patients with supra-normal left ventricular ejection fraction at rest. Eur J Nucl Med Mol Imaging. 2022 Jun;49(7):2189–98. Krittayaphong R, Songsangjinda T, Jirataiporn K, Yindeengam A. Outcomes and Left Ventricular Ejection Fraction in Cardiac Magnetic Resonance: Challenging the “Higher Is Better.” JAHA. 2025 Apr 15;14(8):e039889. Huffman JC, Celano CM, Beach SR, Motiwala SR, Januzzi JL. Depression and Cardiac Disease: Epidemiology, Mechanisms, and Diagnosis. Cardiovascular Psychiatry and Neurology. 2013 Apr 7;2013:1–14. Xu L, Zhai X, Shi D, Zhang Y. Depression and coronary heart disease: mechanisms, interventions, and treatments. Front Psychiatry. 2024 Feb 9;15:1328048. Löwe B, Toussaint A, Rosmalen JGM, Huang WL, Burton C, Weigel A, et al. Persistent physical symptoms: definition, genesis, and management. The Lancet. 2024 Jun;403(10444):2649–62. Sohier L, Dallaire-Habel S, Turcotte S, Foldes-Busque G. Prevalence of Mood and Anxiety Disorders in Canadians with Cardiovascular Disease: A Cross-Sectional Study. Heart and Mind. 2024 Jan;8(1):40–6. DiCaro MV, Ogurek I, Tak N, Dawn B, Tak T. Optimizing Cardiovascular Health: A Narrative Review of Lifestyle, Psychobehavioral, and Alternative Strategies for Management and Prevention. Heart and Mind. 2025 Jan;9(1):29–39. Jering KS, Claggett BL, Pfeffer MA, Granger CB, Køber L, Lewis EF, et al. Prognostic Importance of NT-proBNP (N-Terminal Pro-B-Type Natriuretic Peptide) Following High-Risk Myocardial Infarction in the PARADISE-MI Trial. Circ: Heart Failure [Internet]. 2023 May [cited 2025 Aug 3];16(5). Available from: https://www.ahajournals.org/doi/10.1161/CIRCHEARTFAILURE.122.010259 Ozbaltan OC, Cakmak S, Sogut O, Az A, Ogur H. Predictive value of NT-proBNP for major adverse cardiovascular events within a 6-month period in patients with acute coronary syndrome. Ir J Med Sci. 2025 Feb;194(1):71–7. Vergaro G, Gentile F, Meems LMG, Aimo A, Januzzi JL, Richards AM, et al. NT-proBNP for Risk Prediction in Heart Failure. JACC: Heart Failure. 2021 Sep;9(9):653–63. Holl MJ, Van Den Bos EJ, Van Domburg RT, Fouraux MA, Kofflard MJ. NT-proBNP is associated with mortality and adverse cardiac events in patients with atrial fibrillation presenting to the emergency department. Clinical Cardiology. 2018 Mar;41(3):400–5. Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 25 Feb, 2026 Reviews received at journal 22 Feb, 2026 Reviews received at journal 20 Feb, 2026 Reviews received at journal 17 Feb, 2026 Reviews received at journal 15 Feb, 2026 Reviewers agreed at journal 10 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers agreed at journal 09 Feb, 2026 Reviewers invited by journal 09 Feb, 2026 Editor assigned by journal 09 Feb, 2026 Editor invited by journal 30 Jan, 2026 Submission checks completed at journal 29 Jan, 2026 First submitted to journal 29 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8616117","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":588980876,"identity":"db7c97c6-5ddf-4caf-a0f6-1dfc90090c4e","order_by":0,"name":"Zhongxing Jiang","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences, Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zhongxing","middleName":"","lastName":"Jiang","suffix":""},{"id":588980877,"identity":"4a7b2d9c-e534-47ad-a23c-dc551643da36","order_by":1,"name":"Haofeng Zhou","email":"","orcid":"","institution":"Beijing Hospital","correspondingAuthor":false,"prefix":"","firstName":"Haofeng","middleName":"","lastName":"Zhou","suffix":""},{"id":588980878,"identity":"5bbbb707-4142-4d66-a21a-1c0f818f6a44","order_by":2,"name":"Yindu Liu","email":"","orcid":"","institution":"The Third People's Hospital of Foshan","correspondingAuthor":false,"prefix":"","firstName":"Yindu","middleName":"","lastName":"Liu","suffix":""},{"id":588980880,"identity":"869a3f5f-5c1e-465c-867d-dbc47eb4c00a","order_by":3,"name":"Han Yin","email":"","orcid":"","institution":"Shenzhen People's Hospital, Jinan University, Southern University of Science and Technology)","correspondingAuthor":false,"prefix":"","firstName":"Han","middleName":"","lastName":"Yin","suffix":""},{"id":588980882,"identity":"9c2e5ef6-0d5f-44fd-bed2-f9b58e645c99","order_by":4,"name":"Junshuo Zhu","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Junshuo","middleName":"","lastName":"Zhu","suffix":""},{"id":588980883,"identity":"b218596d-5bd9-40b9-a9b0-4b27e29309aa","order_by":5,"name":"Xiaoya Xiong","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xiaoya","middleName":"","lastName":"Xiong","suffix":""},{"id":588980885,"identity":"de558a1c-91b6-4b4c-ab33-04351ca2f1a1","order_by":6,"name":"Jinna Chang","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jinna","middleName":"","lastName":"Chang","suffix":""},{"id":588980887,"identity":"0357232c-f742-44b9-bf20-3982688886d5","order_by":7,"name":"Rou Wang","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences, Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Rou","middleName":"","lastName":"Wang","suffix":""},{"id":588980889,"identity":"31a3cb64-1978-4e9e-874e-19eb56e95526","order_by":8,"name":"Huan Ma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYFACHhDBDGF/MLCxI0ELGwMD44yCtGTStDDzfDjE2EBIg2772WMSP3dYy/HPbz722cbgADMD++GjG/BpMTuTlybZeybdWOIYW/LsHIM7fAw8aWk38Go5kGMmwdt2OHEDG48xc47BM2YGCR4z/FrOvzGT/AvTYmFwmLGBoJYbOWbScFsYiNPyxthatg3kl7Rkxh6DtGQ2gn45n2N4820bMMSaDx9m+PHHxo6f/fAxvFowARtpykfBKBgFo2AUYAMAtpVEaEyeD7QAAAAASUVORK5CYII=","orcid":"","institution":"Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences, Southern Medical University","correspondingAuthor":true,"prefix":"","firstName":"Huan","middleName":"","lastName":"Ma","suffix":""}],"badges":[],"createdAt":"2026-01-16 08:29:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8616117/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8616117/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102442012,"identity":"d30aef12-765b-40c3-a3c9-f5074d436c17","added_by":"auto","created_at":"2026-02-11 17:02:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":36435,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves of six prediction models with 11 variables for predicting 5-year MACE in patients with coronary artery disease in the training cohort (A) and validation cohort (B). The x-axis represents specificity [probability of correctly identifying patients who did not experience MACE], and the y-axis represents sensitivity [probability of correctly identifying patients who did experience MACE].\u003c/p\u003e\n\u003cp\u003eAbbreviations: KNN, k-nearest neighbors; SVM, support vector machine; GBM, gradient boosting machine.\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/e4a0b3a828e7b11f7a586546.png"},{"id":102745848,"identity":"6a20339c-ad69-4078-ad68-5474d469bf21","added_by":"auto","created_at":"2026-02-16 08:54:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":36031,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis of six prediction models with 11 variables for predicting 5-year major adverse cardiovascular events (MACE) in patients with coronary artery disease in the training cohort (A) and validation cohort (B). The x-axis represents the threshold probability of 5-year MACE. The y-axis represents net benefit.\u003c/p\u003e\n\u003cp\u003eAbbreviations: LR, logistic regression; KNN, k-nearest neighbors; SVM, support vector machine; RF, random forest; GBM, gradient boosting machine; LightGBM, Light Gradient Boosting Machine.\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/5d1e918a897ce62a53d8fe69.png"},{"id":102442016,"identity":"cb0b6588-ec9a-4ccd-9da3-22cb7e71f0e7","added_by":"auto","created_at":"2026-02-11 17:02:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":457482,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP summary plot of the random forest prediction model with 11 variables for predicting 5-year major adverse cardiovascular events (MACE) in patients with coronary artery disease.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/561729950817d38d9c10b32f.png"},{"id":102746309,"identity":"4081f941-b755-441b-9f74-63060c186f93","added_by":"auto","created_at":"2026-02-16 08:56:40","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":117377,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP dependence plots for LVEF (A), NT-proBNP (B), nitrates (C), previous PCI (D), CAD duration (E), BUN (F), age (G), smoking status (H), number of diseased vessels(I), depressive symptoms (J), and NYHA (K) in the random forest prediction model for predicting 5-year MACE in patients with coronary artery disease.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/4dbd694539d05565505c5e64.png"},{"id":103503839,"identity":"ca1277a7-b15a-4a1d-9dc1-855ce646ca9c","added_by":"auto","created_at":"2026-02-26 13:03:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1912558,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/832f658e-9e72-406e-88e7-5115061ff96c.pdf"},{"id":102442014,"identity":"b4cb6a07-2756-4109-8ed5-05d242bbf07d","added_by":"auto","created_at":"2026-02-11 17:02:32","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1048834,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-8616117/v1/a2b4e517da660a0872448822.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Development and Validation of an Interpretable Machine Learning Model for Predicting 5-year Major Adverse Cardiovascular Events in Patients with Coronary Artery Disease","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eCoronary artery disease (CAD) remains a leading global cause of morbidity and mortality, posing a significant public health burden(1,2). According to the Global Burden of Disease study, ischemic heart disease, the most common manifestation of CAD, consistently ranks among the top causes of death and disability-adjusted life years globally(3). Despite advances in medical therapy, revascularization techniques, and secondary prevention strategies, patients with CAD continue to face substantial long-term risks(4). Critically, the post-discharge period represents a high risk phase rather than an endpoint of vulnerability(5,6). These events can significantly impact long-term survival, quality of life, and healthcare utilization. Consequently, effective long-term risk stratification is essential\u0026mdash;not only to identify high-risk individuals but also to personalize post-discharge secondary prevention strategies.\u003c/p\u003e \u003cp\u003eTraditional prognostic models for CAD patients, such as simple integer risk score(7) and the biomarker-based ABC-CHD model(8), rely on a relatively limited set of clinical or biomarker variables and were developed primarily in stable outpatient populations, which may limit their applicability to broader hospitalized CAD cohorts.\u003c/p\u003e \u003cp\u003eTheir performance in real-world clinical settings, particularly for long-term outcomes, remains suboptimal(9). Machine learning (ML) approaches offer the ability to model complex, nonlinear relationships and integrate a broader range of routinely collected features, including advanced imaging, detailed laboratory values, and granular clinical parameters(10,11). This expanded feature space enables improved risk stratification and predictive performance. While ML has demonstrated improved predictive capabilities in CAD, many prior studies have either concentrated on short-term outcomes or been restricted to specific clinical contexts\u0026mdash;such as percutaneous coronary intervention (PCI)-treated patients or those with suspected CAD\u0026mdash;limiting their generalizability to broader cohorts and long-term secondary prevention(9,12).\u003c/p\u003e \u003cp\u003eIn this study, we aimed to develop and validate ML-based models for predicting 5-year MACE in hospitalized CAD patients. By leveraging routinely collected clinical data, we sought to improve long-term risk stratification and support individualized care in secondary prevention.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Design and participants\u003c/h2\u003e \u003cp\u003eThis prospective cohort study included 705 consecutive inpatients diagnosed with CAD at Guangdong Provincial People's Hospital between October 2017 and January 2018. Baseline clinical data were extracted from electronic medical records, including routine examinations, coronary angiography (CAG) results, and discharge diagnoses. Participants were randomly assigned to a training cohort (n\u0026thinsp;=\u0026thinsp;564) and a validation cohort (n\u0026thinsp;=\u0026thinsp;141) in an 8:2 ratio. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Medical Ethics Committee of Guangdong Provincial People's Hospital. Written informed consent was obtained from all participants.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Study variables\u003c/h2\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.2.1. Outcome variables\u003c/h2\u003e \u003cp\u003eThe primary outcome was 5-year MACE defined as a composite of cardiac death, unplanned revascularization, cardiac rehospitalization, non-fatal myocardial infarction, and stroke. Outcome data were extracted from electronic medical records and independently verified through structured annual telephone interviews with patients or their families. Participants were prospectively followed on an annual basis for up to five years. Follow-up was censored at the earliest occurrence of a MACE event, all-cause death, loss to follow-up, or study termination on January 31, 2023.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.2.2. Predictors\u003c/h2\u003e \u003cp\u003eA total of 54 candidate predictors were selected for model development based on clinical relevance, routine availability, and evidence from prior literature. These variables included five categories: (1) demographic and lifestyle factors, such as age, sex, and body mass index (BMI); (2) psychological factors, such as depressive symptoms and anxiety symptoms, assessed using two standardized self-report questionnaires\u0026mdash;the Patient Health Questionnaire-9 (PHQ-9)(13) and the Generalized Anxiety Disorder-7 (GAD-7)(14)\u0026mdash;administered by the same researcher to all patients on the night before surgery during hospitalization. Depressive symptoms and anxiety symptoms were defined as PHQ-9 and GAD-7 scores\u0026thinsp;\u0026ge;\u0026thinsp;9, respectively; (3) clinical history and comorbidities, such as CAD duration, hypertension, and diabetes mellitus; (4) medication and intervention history, such as aspirin use, statin therapy, and previous PCI; and (5) imaging and laboratory findings, such as left ventricular ejection fraction (LVEF), fasting glucose, and N-terminal pro-B-type natriuretic peptide (NT-proBNP). Detailed definitions, measurement methods, and coding schemes for all predictors are provided in Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Statistical analysis\u003c/h2\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1. Data preparation\u003c/h2\u003e \u003cp\u003eMissing data were handled via multiple imputation by chained equations among candidate predictors under the assumption of missing at random. Baseline characteristics were compared between patients with and without MACE, as well as between the training and validation cohorts. Student\u0026rsquo;s t-test was used for continuous variables, and the chi-square test was used for categorical variables. A two-sided p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant. All analyses were conducted using R software (version 4.4.3).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2. Variable selection\u003c/h2\u003e \u003cp\u003eTo reduce model complexity and identify the most informative predictors of 5-year MACE, the least absolute shrinkage and selection operator (LASSO) regression was used. This penalized regression technique applies an L1 penalty to shrink less relevant coefficients toward zero. All predictors were standardized before modeling. The optimal regularization parameter was determined via 10-fold cross-validation in the training cohort, using the area under the receiver operating characteristic curve (AUC) as the selection criterion. The final set of variables corresponding to the lambda value that achieved the highest mean AUC was selected for subsequent model construction.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e2.3.3. Model development\u003c/h2\u003e \u003cp\u003eSix classification algorithms were used to develop prediction models for 5-year MACE: logistic regression (LR), k-nearest neighbors (KNN), support vector machine (SVM), random forest (RF), gradient boosting machine (GBM), and LightGBM. Hyperparameters for each model were tuned using grid search with 10-fold cross-validation in the training cohort. Final model parameters were selected based on a comprehensive assessment of the highest mean AUC, lowest Brier score, and optimal F1 score within the training data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e2.3.4. Performance evaluation\u003c/h2\u003e \u003cp\u003eModel performance was assessed in terms of discrimination, calibration, and clinical utility. Discrimination was evaluated using AUC derived from receiver operating characteristic (ROC) curves, and AUC comparisons among models were conducted using the DeLong test. Calibration was assessed via Brier scores and visualized with calibration plots. Clinical utility was evaluated using decision curve analysis (DCA). To enhance model interpretability, Shapley Additive Explanations (SHAP) were computed to quantify the marginal contribution of each predictor to model output, both at the global and individual levels.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Participant characteristics\u003c/h2\u003e \u003cp\u003eAmong 705 hospitalized CAD patients, 221 (31.3%) developed a MACE within 5 years. The baseline characteristics of patients with and without MACE are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Compared with non-MACE patients, those who experienced MACE were more likely to be older, have a longer CAD duration, lower LVEF, higher NT-proBNP levels, a history of stroke, advanced New York Heart Association (NYHA) functional class, and elevated blood urea nitrogen (BUN) and creatinine levels. They were also more likely to have received nitrate therapy, undergone multi-stent PCI, and present with extensive coronary artery disease. Baseline characteristics of the training and validation cohorts are shown in Supplementary Table S2, and the two cohorts were generally comparable in demographic, clinical, and biochemical profiles.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristics of subjects with or without MACE\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePredictors\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNon-Mace\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;484)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMace\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;221)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003ex\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e/t\u003c/em\u003e \u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e- value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eAge, Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e62.32\u0026thinsp;\u0026plusmn;\u0026thinsp;10.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.98\u0026thinsp;\u0026plusmn;\u0026thinsp;10.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-3.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.002*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSex, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e129(26.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e65(29.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.503\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e355(73.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e156(70.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eBody mass index(kg/m2), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.55\u0026thinsp;\u0026plusmn;\u0026thinsp;3.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24.17\u0026thinsp;\u0026plusmn;\u0026thinsp;3.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.184\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eMarriage, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDivorced or Widowed or Single\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32(6.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9(4.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.245\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMarried\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e452(93.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e212(95.9)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eEducation, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLess than 6 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e131(27.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e75(33.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e4.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e0.204\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7\u0026ndash;9 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e149(30.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54(24.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u0026ndash;12 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e101(20.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e45(20.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMore than 12 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e103(21.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e47(21.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eSleep quality, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGood\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e104(21.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e50(22.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e5.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e0.122\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFair\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e202(41.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e78(35.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePoor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e120(24.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e53(24%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVery poor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58(12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e40(18.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoking status, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePrevious/current\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e210(43.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e104(47.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.408\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDepressive symptoms\u003c/p\u003e \u003cp\u003e, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56 (12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e38 (17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.055\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnxiety symptoms, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32 (7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20 (9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.320\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eCAD duration(mouth), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31.06\u0026thinsp;\u0026plusmn;\u0026thinsp;48.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e40.29\u0026thinsp;\u0026plusmn;\u0026thinsp;51.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-2.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.025*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eNYHA, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClass I\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e186 (38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e68 (31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e17.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClass II\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e248 (51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e107 (48)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClass III\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42 (9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e33 (15)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClass III-IV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8 (2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13 (6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003eMedical history\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHypertension, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e283(58.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e133(60.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.730\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiabetes mellitus, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e149(30.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74(33.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.530\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStroke\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22 (5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e21 (10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.017*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003eMedication use\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAspirin,n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e397 (82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e188 (85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.374\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClopidogrel or Ticagrelor,n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e360 (74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e179 (81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnticoagulants,n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23 (5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14 (6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.489\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatin, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e436(90.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e208(94.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.104\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eβBlocker, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e389(80.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e187(84.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.212\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eACEI or ARB, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e325(67.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e158(71.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.287\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNitrates, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e50(10.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36(16.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.034*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCCB, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e108(22.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e53(24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.695\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003ePrevious PCI, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNever\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e359(74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e146(66)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e14.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e0.039*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOne stent\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60(12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27(12)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTwo stents\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35(7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23(10)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThree or more stents\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30(6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25(11)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eNumber of diseased vessels, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo vessel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e87 (18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32 (14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e13.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e0.004*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSingle vessel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e99 (20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25 (11)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTwo vessels\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90 (19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41 (19)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThree or more vessels\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e208 (43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e123 (56)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eLeft main CAD, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e396 (82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e179 (81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.876\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e88 (18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e42 (19)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eHbA1c(%), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.52\u0026thinsp;\u0026plusmn;\u0026thinsp;1.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.76\u0026thinsp;\u0026plusmn;\u0026thinsp;1.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.054\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eFast glucose(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.28\u0026thinsp;\u0026plusmn;\u0026thinsp;2.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.66\u0026thinsp;\u0026plusmn;\u0026thinsp;2.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.099\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eCREA(\u0026micro;mol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e94.87\u0026thinsp;\u0026plusmn;\u0026thinsp;31.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e105.8\u0026thinsp;\u0026plusmn;\u0026thinsp;64.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-2.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.018*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eBUN(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.17\u0026thinsp;\u0026plusmn;\u0026thinsp;2.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7.17\u0026thinsp;\u0026plusmn;\u0026thinsp;3.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-3.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eTC(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.46\u0026thinsp;\u0026plusmn;\u0026thinsp;1.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.34\u0026thinsp;\u0026plusmn;\u0026thinsp;1.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.212\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eHDL-C(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.96\u0026thinsp;\u0026plusmn;\u0026thinsp;0.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.030*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eLDL-C(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.83\u0026thinsp;\u0026plusmn;\u0026thinsp;0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.403\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eTG(mmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.74\u0026thinsp;\u0026plusmn;\u0026thinsp;1.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.77\u0026thinsp;\u0026plusmn;\u0026thinsp;1.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.804\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eApoA(g/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.14\u0026thinsp;\u0026plusmn;\u0026thinsp;0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.13\u0026thinsp;\u0026plusmn;\u0026thinsp;0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.277\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eApoB(g/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.91\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.9\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.479\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eLpa(nmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e303.63\u0026thinsp;\u0026plusmn;\u0026thinsp;359.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e315.45\u0026thinsp;\u0026plusmn;\u0026thinsp;367.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.690\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eFT3(pmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.86\u0026thinsp;\u0026plusmn;\u0026thinsp;0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.77\u0026thinsp;\u0026plusmn;\u0026thinsp;0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.189\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eFT4(pmol/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12.5\u0026thinsp;\u0026plusmn;\u0026thinsp;9.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13.37\u0026thinsp;\u0026plusmn;\u0026thinsp;12.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.360\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eTSH(mIU/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.92\u0026thinsp;\u0026plusmn;\u0026thinsp;3.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.82\u0026thinsp;\u0026plusmn;\u0026thinsp;1.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.578\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eCK(U/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e115.39\u0026thinsp;\u0026plusmn;\u0026thinsp;98.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e128.77\u0026thinsp;\u0026plusmn;\u0026thinsp;146.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.218\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eCKMB(U/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.86\u0026thinsp;\u0026plusmn;\u0026thinsp;7.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12.72\u0026thinsp;\u0026plusmn;\u0026thinsp;8.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.208\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eNT-proBNP(ng/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e743.61\u0026thinsp;\u0026plusmn;\u0026thinsp;2067.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1537.15\u0026thinsp;\u0026plusmn;\u0026thinsp;3514.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-3.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.002*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003ePT(s), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.65\u0026thinsp;\u0026plusmn;\u0026thinsp;1.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13.86\u0026thinsp;\u0026plusmn;\u0026thinsp;1.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-2.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.031*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eAPTT(s), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37.99\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e38.04\u0026thinsp;\u0026plusmn;\u0026thinsp;4.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.883\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eFIB(g/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.94\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.08\u0026thinsp;\u0026plusmn;\u0026thinsp;1.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.143\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eTT(s), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17.14\u0026thinsp;\u0026plusmn;\u0026thinsp;8.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17.57\u0026thinsp;\u0026plusmn;\u0026thinsp;8.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.527\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eINR, Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.06\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-1.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eD-dimer(\u0026micro;g/mL), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e674.61\u0026thinsp;\u0026plusmn;\u0026thinsp;940.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e676.23\u0026thinsp;\u0026plusmn;\u0026thinsp;612.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.978\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003ehs-CRP(mg/L), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.83\u0026thinsp;\u0026plusmn;\u0026thinsp;17.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.53\u0026thinsp;\u0026plusmn;\u0026thinsp;13.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.806\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eLVEF(%), Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e59.74\u0026thinsp;\u0026plusmn;\u0026thinsp;9.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54.81\u0026thinsp;\u0026plusmn;\u0026thinsp;13.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eAbbreviations: CAD, coronary artery disease; NYHA, New York Heart Association functional classification; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; PCI, percutaneous coronary intervention; HbA1c, glycated hemoglobin A1c; Fast glucose, fasting blood glucose; CREA, creatinine; BUN, blood urea nitrogen; TC, total cholesterol; HDL-C, high density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; TG, triglyceride; ApoA, apolipoprotein A1; ApoB, apolipoprotein B; Lpa, lipoprotein(a); FT3, free triiodothyronine; FT4, free thyroxine; TSH, thyroid-stimulating hormone; CK, creatine kinase; CKMB, creatine kinase MB isoenzyme; NT-proBNP, N-terminal pro-brain natriuretic peptide; PT, prothrombin time; APTT, activated partial thromboplastin time; FIB, fibrinogen; TT, thrombin time; INR, international normalized ratio; hs-CRP, high sensitivity C-reactive protein; LVEF, left ventricular ejection fraction.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003csup\u003ea\u003c/sup\u003eThe comparative analyses for continuous variables and categorical variables were conducted using Student\u0026rsquo;s t test and Chi-square test, respectively.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003csup\u003e*\u003c/sup\u003e \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Variable selection results\u003c/h2\u003e \u003cp\u003eTo identify the most relevant predictors of 5-year MACE, we employed LASSO regression using the glmnet package (version 4.4.3) in R. All 54 candidate variables were z-score normalized before modeling. A 10-fold cross-validation procedure was used to determine the optimal regularization parameter, and the value corresponding to the minimum mean cross-validated deviance (lambda.min) was selected for final variable selection (Supplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAt this lambda value, 11 variables with non-zero coefficients were retained (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), including age, stroke, NYHA functional class, previous PCI, number of diseased vessels, CAD duration, depressive symptoms, nitrates use, NT-proBNP, BUN, and LVEF. Most of these predictors have well-established associations with adverse cardiovascular outcomes. The selected predictors were subsequently used for model development across six classification algorithms.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Prediction performance for 5-year MACE\u003c/h2\u003e \u003cp\u003eThe discriminatory performance of all six models is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, and detailed performance metrics are provided in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. In the training cohort, the RF model achieved the highest AUC of 0.887 (95% CI: 0.859\u0026ndash;0.915), followed by GBM (AUC: 0.805) and LightGBM (AUC: 0.777). The SVM, KNN, and LR models showed moderate discrimination, with AUCs of 0.731, 0.679, and 0.660, respectively. Pairwise comparisons confirmed that RF significantly outperformed all other models in terms of AUC.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerformance comparison of 6 MACE prediction models with 11 variables\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAUC(95%CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThreshold*\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ePPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eNPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eF1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eBrier score\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.660(0.611\u0026ndash;0.708)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.470\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.775\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.497\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.756\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.483\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.201\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.679(0.632\u0026ndash;0.725)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.260\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.606\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.441\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.789\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.201\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.731(0.685\u0026ndash;0.777)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.296\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.751\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.634\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.493\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.844\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.595\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.201\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.887(0.859\u0026ndash;0.915)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.185\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.603\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.915\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.708\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.152\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGBM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.805(0.768\u0026ndash;0.843)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.326\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.707\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.773\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.595\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.848\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.646\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.181\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLightGBM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.777(0.737\u0026ndash;0.816)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.705\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.533\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.839\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.610\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.196\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.697(0.604\u0026ndash;0.790)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.802\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.524\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.537\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.185\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.649(0.538\u0026ndash;0.760)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.475\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.812\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.796\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.487\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.190\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.644(0.533\u0026ndash;0.755)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.303\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.525\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.782\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.488\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.806\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.506\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.199\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.753(0.656\u0026ndash;0.849)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.675\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.782\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.551\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.859\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.607\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.177\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGBM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.746(0.647\u0026ndash;0.845)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.355\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.575\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.881\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.840\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.613\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.175\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLightGBM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.735(0.631\u0026ndash;0.840)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.575\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.891\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.676\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.841\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.184\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"9\"\u003eAbbreviations: AUC, area under the receiver operating characteristic curve; PPV, positive predictive value; NPV, negative predictive value; F1, F1 score; LR, logistic regression; KNN, k-nearest neighbors; SVM, support vector machine; RF, random forest; GBM, gradient boosting machine; LightGBM, Light Gradient Boosting Machine.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"9\"\u003e*The optimal threshold for binary classification was determined using the maximum Youden index from the ROC curve.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn addition to AUC, RF also demonstrated the best overall predictive performance in the training cohort, with the highest sensitivity (0.856), and F1 score (0.708), as well as the lowest Brier score (0.152). GBM and LightGBM also performed well, with F1 score of 0.646 and 0.610, respectively.\u003c/p\u003e \u003cp\u003eModel calibration was evaluated using calibration plots in Supplementary Fig.\u0026nbsp;2\u0026thinsp;\u0026minus;\u0026thinsp;1 (training cohort) and Supplementary Fig.\u0026nbsp;2\u0026ndash;2 (validation cohort). In the training cohort, the RF model showed the best calibration, with predicted risks closely aligning with actual outcomes. In the validation cohort, GBM and RF exhibited superior calibration performance compared with other models, with calibration curves closer to the ideal diagonal line.\u003c/p\u003e \u003cp\u003eIn the validation cohort, RF and GBM remained the top performers, with AUCs of 0.753 (95% CI: 0.656\u0026ndash;0.849) and 0.746 (95% CI: 0.647\u0026ndash;0.845), outperforming the remaining models (AUC range: 0.644\u0026ndash;0.735). GBM and LightGBM achieved the highest specificity (0.881 and 0.891, respectively), while RF offered the best balance between sensitivity (0.675) and negative predictive value (NPV\u0026thinsp;=\u0026thinsp;0.859). Among the traditional models, LR showed moderate performance (AUC\u0026thinsp;=\u0026thinsp;0.697), whereas SVM and KNN performed relatively poorly.\u003c/p\u003e \u003cp\u003eDecision curve analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) further confirmed the clinical utility of the RF, GBM, and LightGBM models. These approaches consistently provided greater net benefit across a wide range of decision thresholds compared with the \u0026ldquo;treat-all\u0026rdquo; or \u0026ldquo;treat-none\u0026rdquo; strategies, supporting their potential role in individualized risk assessment and clinical decision-making.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Importance of predictive variables for 5-year MACE\u003c/h2\u003e \u003cp\u003eVariable importance was evaluated based on the random forest model trained with the 11 LASSO-selected predictors. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, LVEF, NT-proBNP, and nitrates use were identified as the top three predictors contributing most to model performance. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e also depicts several clinically relevant and directionally consistent feature\u0026ndash;outcome relationships through distinct SHAP value distributions. NT-proBNP shows a strong positive association, with higher values\u0026mdash;represented by red data points\u0026mdash;mainly concentrated in the positive SHAP region, highlighting its adverse impact on 5-year MACE. Furthermore, the SHAP dependence plot illustrates the effect of each feature on individual risk prediction (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs illustrated in the SHAP dependence plots, LVEF and depressive symptoms showed distinct associations with the predicted 5-year MACE risk. For LVEF, a U-shaped pattern was observed: SHAP values below 50% or above 70% were associated with increased SHAP values, suggesting elevated risk at both ends of left ventricular function. Patients identified as having depressive symptoms exhibited higher SHAP values, suggesting a possible association between depressive symptoms and elevated MACE risk. Other variables associated with higher SHAP values included age above 70 years, elevated NT-proBNP (\u0026gt;\u0026thinsp;1100 pg/ml), nitrate use, higher NYHA functional class, history of PCI, elevated BUN, longer CAD duration, and greater number of diseased coronary vessels. These features were all linked to an increased likelihood of adverse cardiovascular events over the 5-year follow-up period.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eCAD remains a major global contributor to long-term cardiovascular morbidity and mortality. Accurate post-discharge risk prediction is essential for optimizing secondary prevention. In this prospective study, we developed six ML models to predict 5-year MACE in hospitalized CAD patients, identifying RF as the optimal model.\u003c/p\u003e \u003cp\u003eML offers distinct advantages over traditional statistical models in prognostic prediction, particularly for diseases with multiple and complex risk factors such as coronary heart disease. Unlike conventional linear models that rely on predefined assumptions, ML algorithms can automatically learn nonlinear relationships and interactions from large-scale, high-dimensional clinical data, thereby improving prediction accuracy and enabling personalized risk stratification. In cardiovascular medicine, ML has been increasingly applied to diagnosis, classification, and outcome prediction, showing potential to refine prognostic assessments and guide more targeted interventions for patients undergoing PCI(9).\u003c/p\u003e \u003cp\u003ePrevious models for predicting cardiovascular risk in coronary artery disease patients have shown varying approaches and performance. Ganz et al. developed a 9-protein risk score with moderate discrimination (C-statistics 0.64\u0026ndash;0.75) outperforming the Framingham model for 4-year outcomes(15). Lindholm et al. created the \u0026ldquo;ABC-CHD\u0026rdquo; model combining biomarkers and clinical factors, achieving high discrimination for cardiovascular death (c-index 0.78\u0026ndash;0.81)(8). Liu et al. applied machine learning to PCI patients, with a random forest model predicting 5-year all-cause mortality with an AUC of 0.71(9). Our findings similarly support the superior predictive value of ML methods like random forest for long-term prognosis in CAD, underscoring the potential of integrating complex clinical and biomarker data to enhance risk stratification.\u003c/p\u003e \u003cp\u003eRF\u0026mdash;an ensemble method aggregating multiple decision trees\u0026mdash;outperforms traditional logistic regression by capturing complex nonlinear relationships without strict parametric assumptions(16\u0026ndash;18). In our cohort, RF achieved superior discrimination (training AUC: 0.887, 95% CI: 0.859\u0026ndash;0.915; validation AUC: 0.753), sensitivity (0.856), F1-score (0.708), and calibration (Brier score: 0.152).\u003c/p\u003e \u003cp\u003eImportantly, we improved the interpretability of the RF model by applying SHAP. Notably, our model revealed a U-shaped relationship between LVEF and 5-year MACE risk, with increased risk observed both below 50% and above 70%. This aligns with prior studies: Liu et al. reported that LVEF below 55% was associated with increased all-cause mortality and major adverse cardiac and cerebrovascular events in PCI patients(19); Chang et al. demonstrated that LVEF exceeding 70% conferred higher mortality risk, potentially linked to left ventricular concentric remodeling, especially in women(20). These findings underscore the importance of monitoring both impaired and supranormal LVEF values in CAD prognostication. While the adverse prognostic impact of reduced LVEF is well established, the mechanisms underlying the risk associated with supranormal LVEF are less widely appreciated. One explanation is that supranormal LVEF often reflects concentric left ventricular remodeling, which can impair diastolic filling and elevate filling pressures, particularly in women(21). In addition, excessive contractility may be accompanied by coronary microvascular dysfunction, where increased myocardial oxygen demand induces microvascular ischemia, myocardial injury, and interstitial fibrosis, ultimately compromising cardiac performance(22). Furthermore, patients with supranormal LVEF frequently exhibit diffuse interstitial myocardial fibrosis\u0026mdash;manifested by increased extracellular volume\u0026mdash;rather than focal scarring, a pathophysiological substrate closely linked to heart failure with preserved ejection fraction and adverse cardiovascular outcomes(23).\u003c/p\u003e \u003cp\u003eDepressive symptoms were also significantly associated with higher MACE risk, highlighting the crucial role of psychosocial factors in CAD prognosis. The interplay between depressive symptoms and cardiovascular outcomes may involve behavioral and biological pathways, including systemic inflammation, autonomic dysfunction, endothelial impairment, hypothalamic\u0026ndash;pituitary\u0026ndash;adrenal axis dysregulation, platelet hyperactivation, and genetic predisposition(24,25). Clinically, patients with emotional distress often present with concomitant somatic complaints such as chest tightness, chest pain, palpitations, and dyspnea(26,27). These symptoms are frequently misattributed by patients to recurrent cardiac events, prompting more aggressive medical consultation and intervention(28). This heightened vigilance may, in turn, increase the likelihood of subsequent healthcare utilization and rehospitalization. Future research should further explore the contribution of this symptom-perception\u0026ndash;driven pathway to long-term outcomes in CAD patients with emotional comorbidities.\u003c/p\u003e \u003cp\u003eNT-proBNP was identified as the second most important predictor, reflecting myocardial wall stress and volume overload often resulting from ischemia or myocardial injury. Elevated NT-proBNP (\u0026gt;\u0026thinsp;1100 pg/ml in our cohort) predicted increased long-term MACE risk, corroborating previous research demonstrating its prognostic value in chronic cardiac conditions(29,30). Vergaro et al. analyzed 12,763 patients with stable heart failure (mean LVEF 33%) in the BIOS consortium and reported optimal NT-proBNP cutoffs across BMI categories for 5-year all-cause mortality ranging from 3,785 pg/ml in underweight to 755-1,554 pg/ml in overweight or obese patients(31). In a prospective study of 355 patients presenting to the emergency department with atrial fibrillation, Holl et al. reported that NT-proBNP\u0026thinsp;\u0026gt;\u0026thinsp;500 pmol/L was associated with increased risk of death and MACE over 2 years(32). The cutoff identified in our study (\u0026gt;\u0026thinsp;1,100 pg/ml) is lower than those in acute AF cohorts but close to the range observed in overweight/obese HF patients in Vergaro et al. Possible explanations include differences in underlying disease type and severity, BMI distribution, outcome definitions (MACE vs. mortality), follow-up duration, and measurement conditions. These factors should be considered when applying NT-proBNP thresholds across different clinical settings.\u003c/p\u003e \u003cp\u003eThe strengths of this study include its prospective design with complete 5-year follow-up and clinically verified endpoints, comprehensive variable selection encompassing psychosocial factors, and rigorous validation through multiple performance metrics coupled with SHAP-enhanced interpretability. However, several limitations warrant consideration. First, the single-center data may limit generalizability to diverse healthcare settings. Second, external validation in multicenter cohorts is needed to confirm robustness despite internal validation. Finally, while SHAP improves transparency, it does not fully resolve the inherent 'black box' nature of ensemble ML models.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn this prospective cohort study of hospitalized CAD patients, we developed and validated ML-based models for predicting 5-year MACE, with RF demonstrating the best performance across discrimination, calibration, and clinical utility. By incorporating routinely available clinical, laboratory, and psychosocial variables, our models achieved accurate long-term risk stratification and enhanced interpretability through SHAP analysis. These findings support the integration of ML-based risk prediction into secondary prevention strategies to improve individualized patient management after hospital discharge. External validation in diverse populations is warranted to confirm generalizability.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCAD = Coronary artery disease\u003c/p\u003e\n\u003cp\u003eMACE = major adverse cardiovascular events\u003c/p\u003e\n\u003cp\u003eML = Machine learning\u003c/p\u003e\n\u003cp\u003ePCI = percutaneous coronary intervention\u003c/p\u003e\n\u003cp\u003eCAG = coronary angiography\u003c/p\u003e\n\u003cp\u003ePHQ-9 = Patient Health Questionnaire-9\u003c/p\u003e\n\u003cp\u003eGAD-7 = Generalized Anxiety Disorder-7\u003c/p\u003e\n\u003cp\u003eLASSO = least absolute shrinkage and selection operator\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAUC = area under the receiver operating characteristic curve\u003c/p\u003e\n\u003cp\u003eLR = logistic regression\u003c/p\u003e\n\u003cp\u003eKNN = k-nearest neighbors\u003c/p\u003e\n\u003cp\u003eSVM = support vector machine\u003c/p\u003e\n\u003cp\u003eRF = random forest\u003c/p\u003e\n\u003cp\u003eGBM = gradient boosting machine\u003c/p\u003e\n\u003cp\u003eROC = receiver operating characteristic\u003c/p\u003e\n\u003cp\u003eDCA = decision curve analysis\u003c/p\u003e\n\u003cp\u003eSHAP = Shapley Additive Explanations\u003c/p\u003e\n\u003cp\u003eNT-proBNP = N-terminal pro-B-type natriuretic peptide\u003c/p\u003e\n\u003cp\u003eNYHA = New York Heart Association\u003c/p\u003e\n\u003cp\u003eBUN = blood urea nitrogen\u003c/p\u003e"},{"header":"Declarations","content":" \u003cp\u003e \u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e \u003cp\u003e Ethical approval was given by the medical ethics committee of Guangdong General Hospital with the following reference number: No.GDREC2017203H. All participants gave written informed consent.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no competing interests\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis project is supported by grant from the Guangzhou Municipal Science and Technology Program key projects (2023B03J1249), China Heart House-Chinese Cardiovascular Association TCM fund (CCA-TCM-032; 202342) and Su Ke'an Pharmaceutical Research and Development Project (202460).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eZ.X.J., H.F.Z., Y.D.L., H.Y. and H.M. contributed to the conception and design of the work. Y.D.L., H.Y., J.S.Z., X.Y.X., J.N.C. and R.W. were responsible for the acquisition of data. Z.X.J. and H.F.Z. performed all data analyses. Z.X.J., H.F.Z. and Y.D.L. drafted the manuscript. All the authors gave comments and revised the manuscript. All the authors approved the final version to be submitted.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated and/or analyzed during the current study are not publicly available due to patient privacy and ethical restrictions but are available from the corresponding author upon reasonable request. Requests for access to the data should be directed to Dr. Ma (email: [[email protected]]).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWang Z, Du A, Liu H, Wang Z, Hu J. Systematic Analysis of the Global, Regional and National Burden of Cardiovascular Diseases from 1990 to 2017. J Epidemiol Glob Health. 2022 Mar;12(1):92\u0026ndash;103.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeto MG, Saquetto MB, Roever L, Carvalho VO. The Effect of Yoga Intervention on Psychological Symptoms, Health-Related Quality of Life, and Cardiovascular Risk Factors in People with Coronary Artery Disease: A Systematic Review and Meta-Analysis. Heart and Mind. 2024 Oct;8(4):300\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVos T, Lim SS, Abbafati C, Abbas KM, Abbasi M, Abbasifard M, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990\u0026ndash;2019: a systematic analysis for the Global Burden of Disease Study 2019. The Lancet. 2020 Oct;396(10258):1204\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreites A, Hernando L, Salinas P, C\u0026aacute;novas E, De La Rosa A, Alonso J, et al. Incidence and prognosis of late readmission after percutaneous coronary intervention. Cardiol J. 2023 Oct 27;30(5):696\u0026ndash;704.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVallejo-Vaz AJ, Dharmayat KI, Nzeakor N, Carrasco CP, Fatoba ST, Fonseca MJ, et al. Recurrent cardiovascular and limb events in 294,428 patients with coronary or peripheral artery disease or ischemic stroke on antiplatelet monotherapy: The RESRISK cohort study. Atherosclerosis. 2024 Nov;398:118589.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown TM, Bittner V, Colantonio LD, Deng L, Farkouh ME, Limdi N, et al. Residual risk for coronary heart disease events and mortality despite intensive medical management after myocardial infarction. Journal of Clinical Lipidology. 2020 Mar;14(2):260\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBavry AA, Kumbhani DJ, Gong Y, Handberg EM, Cooper-DeHoff RM, Pepine CJ. Simple Integer Risk Score to Determine Prognosis of Patients With Hypertension and Chronic Stable Coronary Artery Disease. JAHA. 2013 Aug 22;2(4):e000205.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLindholm D, Lindb\u0026auml;ck J, Armstrong PW, Budaj A, Cannon CP, Granger CB, et al. Biomarker-Based Risk Model to Predict Cardiovascular Mortality in Patients With Stable Coronary Disease. Journal of the American College of Cardiology. 2017 Aug;70(7):813\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu S, Yang S, Xing A, Zheng L, Shen L, Tu B, et al. Machine learning-based long-term outcome prediction in patients undergoing percutaneous coronary intervention. Cardiovasc Diagn Ther. 2021 Jun;11(3):736\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin G, Liu Q, Chen Y, Zong X, Xi Y, Li T, et al. Machine Learning to Predict Long-Term Cardiac-Relative Prognosis in Patients With Extra-Cardiac Vascular Disease. Front Cardiovasc Med. 2021 Nov 25;8:771504.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKigka VI, Georga E, Tsakanikas V, Kyriakidis S, Tsompou P, Siogkas P, et al. Machine Learning Coronary Artery Disease Prediction Based on Imaging and Non-Imaging Data. Diagnostics. 2022 Jun 14;12(6):1466.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMotwani M, Dey D, Berman DS, Germano G, Achenbach S, Al-Mallah MH, et al. Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis. Eur Heart J. 2016 Jun 1;ehw188.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKroenke K, Spitzer RL, Williams JBW. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med. 2001 Sep;16(9):606\u0026ndash;13.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpitzer RL, Kroenke K, Williams JBW, L\u0026ouml;we B. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Arch Intern Med. 2006 May 22;166(10):1092.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGanz P, Heidecker B, Hveem K, Jonasson C, Kato S, Segal MR, et al. Development and Validation of a Protein-Based Risk Score for Cardiovascular Outcomes Among Patients With Stable Coronary Heart Disease. JAMA. 2016 Jun 21;315(23):2532.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXing F, Luo R, Liu M, Zhou Z, Xiang Z, Duan X. A New Random Forest Algorithm-Based Prediction Model of Post-operative Mortality in Geriatric Patients With Hip Fractures. Front Med. 2022 May 11;9:829977.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong W, Lu Y, Zhou X, Jin S, Pan J, Lin Q, et al. Usefulness of Random Forest Algorithm in Predicting Severe Acute Pancreatitis. Front Cell Infect Microbiol. 2022 Jun 10;12:893294.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreiman L. Random Forests. Machine Learning. 2001 Oct 1;45(1):5\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y, Song J, Wang W, Zhang K, Qi Y, Yang J, et al. Association of ejection fraction with mortality and cardiovascular events in patients with coronary artery disease. ESC Heart Failure. 2022 Oct;9(5):3461\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang HC, Tseng CH, Huang WM, Lee CW, Yu WC, Cheng HM, et al. Supranormal Left Ventricular Ejection Fraction, Concentric Remodeling, and Long-Term Survival. JACC: Asia. 2024 Dec;4(12):928\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan NI, Atherton JJ, Krishnan A, Hammett C, Stewart P, Mallouhi M, et al. Diastolic Dysfunction and Survival in Patients With Preserved or Mildly Reduced Left Ventricular Ejection Fraction Following Myocardial Infarction. Journal of the American Society of Echocardiography. 2025 May;38(5):380\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu P, Zhang X, Wu Z, Chen H, Guo X, Jin C, et al. Impaired coronary flow reserve in patients with supra-normal left ventricular ejection fraction at rest. Eur J Nucl Med Mol Imaging. 2022 Jun;49(7):2189\u0026ndash;98.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKrittayaphong R, Songsangjinda T, Jirataiporn K, Yindeengam A. Outcomes and Left Ventricular Ejection Fraction in Cardiac Magnetic Resonance: Challenging the \u0026ldquo;Higher Is Better.\u0026rdquo; JAHA. 2025 Apr 15;14(8):e039889.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuffman JC, Celano CM, Beach SR, Motiwala SR, Januzzi JL. Depression and Cardiac Disease: Epidemiology, Mechanisms, and Diagnosis. Cardiovascular Psychiatry and Neurology. 2013 Apr 7;2013:1\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu L, Zhai X, Shi D, Zhang Y. Depression and coronary heart disease: mechanisms, interventions, and treatments. Front Psychiatry. 2024 Feb 9;15:1328048.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026ouml;we B, Toussaint A, Rosmalen JGM, Huang WL, Burton C, Weigel A, et al. Persistent physical symptoms: definition, genesis, and management. The Lancet. 2024 Jun;403(10444):2649\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSohier L, Dallaire-Habel S, Turcotte S, Foldes-Busque G. Prevalence of Mood and Anxiety Disorders in Canadians with Cardiovascular Disease: A Cross-Sectional Study. Heart and Mind. 2024 Jan;8(1):40\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDiCaro MV, Ogurek I, Tak N, Dawn B, Tak T. Optimizing Cardiovascular Health: A Narrative Review of Lifestyle, Psychobehavioral, and Alternative Strategies for Management and Prevention. Heart and Mind. 2025 Jan;9(1):29\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJering KS, Claggett BL, Pfeffer MA, Granger CB, K\u0026oslash;ber L, Lewis EF, et al. Prognostic Importance of NT-proBNP (N-Terminal Pro-B-Type Natriuretic Peptide) Following High-Risk Myocardial Infarction in the PARADISE-MI Trial. Circ: Heart Failure [Internet]. 2023 May [cited 2025 Aug 3];16(5). Available from: https://www.ahajournals.org/doi/10.1161/CIRCHEARTFAILURE.122.010259\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOzbaltan OC, Cakmak S, Sogut O, Az A, Ogur H. Predictive value of NT-proBNP for major adverse cardiovascular events within a 6-month period in patients with acute coronary syndrome. Ir J Med Sci. 2025 Feb;194(1):71\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVergaro G, Gentile F, Meems LMG, Aimo A, Januzzi JL, Richards AM, et al. NT-proBNP for Risk Prediction in Heart Failure. JACC: Heart Failure. 2021 Sep;9(9):653\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoll MJ, Van Den Bos EJ, Van Domburg RT, Fouraux MA, Kofflard MJ. NT-proBNP is associated with mortality and adverse cardiac events in patients with atrial fibrillation presenting to the emergency department. Clinical Cardiology. 2018 Mar;41(3):400\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-cardiovascular-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcar","sideBox":"Learn more about [BMC Cardiovascular Disorders](http://bmccardiovascdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcar/default.aspx","title":"BMC Cardiovascular Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Coronary artery disease, Major adverse cardiovascular events, Machine learning, SHAP value","lastPublishedDoi":"10.21203/rs.3.rs-8616117/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8616117/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eCoronary artery disease (CAD) remains a major contributor to global cardiovascular mortality. The accurate prediction of prognosis is critical for guide clinical decision-making. This study aimed to develop and validate interpretable machine learning (ML) models for predicting 5-year major adverse cardiovascular events (MACE) in hospitalized CAD patients.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA prospective cohort of 705 CAD patients was included and randomly divided into training (n\u0026thinsp;=\u0026thinsp;564) and validation (n\u0026thinsp;=\u0026thinsp;141) sets. Eleven key predictors were selected using Least Absolute Shrinkage and Selection Operator (LASSO) regression. Six ML algorithms were developed, and model performance was assessed using discrimination, calibration, and decision curve analysis. Shapley Additive Explanations (SHAP) were applied to enhance model interpretability.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eKey predictors identified by LASSO regression included left ventricular ejection fraction (LVEF), N-terminal pro-B-type natriuretic peptide (NT-proBNP), nitrate use, CAD duration, depressive symptoms, and age. The random forest (RF) model demonstrated superior performance, achieving the highest Area Under the Curve (AUC) in both training (0.887, 95% CI: 0.859\u0026ndash;0.915) and validation (0.753, 95% CI: 0.656\u0026ndash;0.849) cohorts, along with optimal balance of sensitivity (0.856), F1 score (0.708), and Brier score (0.152). The LASSO method revealed that LVEF, NT-proBNP, and nitrate use were the top 3 predictors of 5-year mace. Depressive symptoms were also associated with increased MACE risk.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThis interpretable RF-based model provides accurate and interpretable 5-year MACE prediction in CAD patients. By integrating clinical and psychosocial features, it supports personalized secondary prevention. External validation is warranted to assess real-world applicability.\u003c/p\u003e","manuscriptTitle":"Development and Validation of an Interpretable Machine Learning Model for Predicting 5-year Major Adverse Cardiovascular Events in Patients with Coronary Artery Disease","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-11 17:02:27","doi":"10.21203/rs.3.rs-8616117/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-25T11:07:56+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-23T00:13:34+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-20T05:33:10+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-17T11:58:27+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-15T12:03:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"228279057702544053390166253592415834955","date":"2026-02-10T12:49:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"160402955222009558955745576622952992167","date":"2026-02-10T04:49:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"59873530625122799165364249117798581777","date":"2026-02-09T20:01:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"201326186480567893102925258808013815339","date":"2026-02-09T14:45:36+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"6353996771738383354782653605055642434","date":"2026-02-09T13:28:30+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-09T12:47:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-09T12:37:11+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-01-30T07:06:28+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-29T15:57:14+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Cardiovascular Disorders","date":"2026-01-29T15:36:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-cardiovascular-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcar","sideBox":"Learn more about [BMC Cardiovascular Disorders](http://bmccardiovascdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcar/default.aspx","title":"BMC Cardiovascular Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e9082f27-bc97-4796-bc2d-a897e4510554","owner":[],"postedDate":"February 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-11T13:28:44+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-11 17:02:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8616117","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8616117","identity":"rs-8616117","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}