Cardiovascular Disease Risk Prediction Research Articles

Risk Stratification for Cardiovascular Disease: A Comparative Analysis of Cluster Analysis and Traditional Prediction Models

Abstract Aim Primary prevention of cardiovascular disease (CVD) relies on effective risk stratification to guide interventions. Current models, primarily developed using regression analysis, can lead to inaccurate estimates when applied to external populations. This study evaluates the utility of cluster analysis as an alternative method for developing CVD risk stratification models, comparing its performance with established CVD risk prediction models. Methods Using data from 3,416 individuals (mean age of 66 years and no prior CVD) followed for an average of 5.2 years for incidence of CVD, we developed a risk stratification model using cluster analysis based on established CVD risk factors. We compared our model to the Systematic Coronary Risk Evaluation (SCORE2), the Pooled Cohort Equations (PCE) and the Predicting Risk of Cardiovascular Disease Events (PREVENT) models. We used Poisson and Cox regression to compare CVD risk between risk categories in each model. Predictive accuracy of the models was evaluated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and C-statistic. Results During the study, 161 CVD events were detected. The high-risk cluster had a sensitivity of 59.0%, a PPV of 7.5% a specificity of 64.2% and NPV of 96.9% to predict CVD. Compared to the high-risk groups of the SCORE2, PCE and PREVENT, the high-risk cluster had a high sensitivity and NPV, but a low specificity and PPV. No statistically significant differences were found in C-statistic between models. Conclusions Cluster analysis performed comparably to existing models and identified a larger high-risk group that included more individuals who developed CVD, though with more false positives. Further studies in larger, diverse cohorts are needed to validate the clinical utility of cluster analysis in CVD risk stratification.

Transforming Cardiovascular Risk Prediction: A Review of Machine Learning and Artificial Intelligence Innovations

Cardiovascular diseases (CVDs) remain a leading cause of global mortality and morbidity. Traditional risk prediction models, while foundational, often fail to capture the multifaceted nature of risk factors or leverage the expanding pool of healthcare data. Machine learning (ML) and artificial intelligence (AI) approaches represent a paradigm shift in risk prediction, offering dynamic, scalable solutions that integrate diverse data types. This review examines advancements in AI/ML for CVD risk prediction, analyzing their strengths, limitations, and the challenges associated with their clinical integration. Recommendations for standardization, validation, and future research directions are provided to unlock the potential of these technologies in transforming precision cardiovascular medicine.

Refining PREVENT prediction models for 10-year risk of cardiovascular disease using measures of anxiety and depression

Refining PREVENT prediction models for 10-year risk of cardiovascular disease using measures of anxiety and depression

Red cell distribution width as a cardiovascular risk predictor in adults with hypertension in sub-Saharan Africa.

Red cell distribution width (RDW) quantifies the degree of variation in erythrocyte size, is identified as a potential marker of adverse cardiovascular events, and may be a surrogate marker for assessing cardiovascular disease (CVD) risk in low-resource settings. We evaluated RDW as a predictor of CVD risk compared to the World Health Organization (WHO) CVD risk score among adults with hypertension attending primary healthcare centers (PHCs) in Ghana and Nigeria. Adults with hypertension attending selected PHCs in Ghana and Nigeria participated in a cross-sectional study. Each participant underwent blood pressure (BP) measurement and laboratory evaluation (RDW, total cholesterol, and fasting blood sugar) following standard methods. We recruited 319 adults aged 40-74 years from the study sites. The mean (standard deviation) RDW was 13.96 (1.1%). The median CVD risk score was 8.11% [interquartile range (IQR) 4.00 to 11.00]. For participants with hemoglobin (Hb) levels ≥ 12 g/dL, RDW showed positive correlations with age (r = 0.136; p = 0.042); systolic BP (r = 0.183; p = 0.006), diastolic BP (r = 0.206, p = 0.002) and WHO CVD risk scores (r = 0.166, p = 0.013). Multiple linear regression showed an independent association between RDW and WHO CVD risk scores with an upward gradient, and was most significant at 3rd quartiles. Using receiver operating characteristic curve, the C-statistic was 0.673 (95% confidence interval: 0.618 to 0.724), p = 0.031. With a cut-off of >14, the RDW demonstrated a sensitivity of 81.82% and specificity of 55.84%. This study shows that at Hb levels ≥ 12 g/dL, RDW modestly predicted CVD risk in adults with hypertension in sub-Saharan Africa.

Artificial intelligence-based cardiovascular/stroke risk stratification in women affected by autoimmune disorders: a narrative survey.

Women are disproportionately affected by chronic autoimmune diseases (AD) like systemic lupus erythematosus (SLE), scleroderma, rheumatoid arthritis (RA), and Sjögren's syndrome. Traditional evaluations often underestimate the associated cardiovascular disease (CVD) and stroke risk in women having AD. Vitamin D deficiency increases susceptibility to these conditions. CVD risk prediction in AD can benefit from surrogate biomarker for coronary artery disease (CAD), such as carotid ultrasound. Due to non-linearity in the CVD risk stratification, we use artificial intelligence-based system using AD biomarkers and carotid ultrasound. Investigate the relationship between AD and CVD/stroke markers including autoantibody-influenced plaque load. Second, to study the surrogate biomarkers for the CAD and gather radiomics-based features such as carotid intima-media thickness (cIMT), and plaque area (PA). Third and final, explore the automated CVD/stroke risk identification using advanced machine learning (ML) and deep learning (DL) paradigms. Analysed biomarker data from women with AD, including carotid ultrasonography imaging, clinical parameters, autoantibody profiles, and vitamin D levels. Proposed artificial intelligence (AI) models to predict CVD/stroke risk accurately in AD for women. There is a strong association between AD duration and elevated cIMT/PA, with increased CVD risk linked to higher rheumatoid factor (RF) and anti-citrullinated peptide antibodies (ACPAs) levels. AI models outperformed conventional methods by integrating imaging data and disorder-specific factors. Interdisciplinary collaboration is crucial for managing CVD/stroke in women with chronic autoimmune diseases. AI-based assisted risk stratification methods may improve treatment decision-making and cardiovascular outcomes.

Cardiovascular Risk Prediction Scores in Type 1 diabetes: A Systematic Review and Meta-Analysis

The extent of the performance and utility of scores for the risk of cardiovascular disease (CVD) in persons with type 1 diabetes (T1DM) largely remains unclear. The purpose of this study was to synthesize data on the performance of CVD risk scores in people living with T1DM. This study is a systematic review and meta-analysis. PubMed and EMBASE were searched through December 31, 2023. The included studies: 1) were retrospective, prospective, or cross-sectional in design; 2) included persons with T1DM; 3) assessed CVD outcomes; and 4) had data on at least on CVD risk score. Measures of calibration and discrimination qualitatively summarized. Measures of discrimination were combined using random-effects models stratified by type of risk model. In a meta-analysis of observational studies of CVD risk scores in T1DM individuals, including 11 studies and 73,664 participants (mean age of 34years, mainly White individuals and male [55%]), we evaluated 12 CVD risk prediction models (7 T1DM-specific, 1 type 2 diabetes-specific, and 4 general population models). Most risk scores had a moderate to excellent discrimination (C-statistic: 0.73-0.85) and predicted CVD risk well when compared to actual clinical events. CVD risk scores specifically developed in T1DM individuals exhibited a higher discriminative performance-pooled C-statistic of 0.81 vs 0.75 for risk scores developed in the general population or those with type 2 diabetes and also showed a better calibration. Among individuals with T1DM, CVD risk models had a moderate to excellent discrimination, with a better discrimination and accuracy for T1DM-specific scores.

Unlocking the diagnostic potential of electrocardiograms through information transfer from cardiac magnetic resonance imaging.

Cardiovascular diseases (CVD) can be diagnosed using various diagnostic modalities. The electrocardiogram (ECG) is a cost-effective and widely available diagnostic aid that provides functional information of the heart. However, its ability to classify and spatially localise CVD is limited. In contrast, cardiac magnetic resonance (CMR) imaging provides detailed structural information of the heart and thus enables evidence-based diagnosis of CVD, but long scan times and high costs limit its use in clinical routine. In this work, we present a deep learning strategy for cost-effective and comprehensive cardiac screening solely from ECG. Our approach combines multimodal contrastive learning with masked data modelling to transfer domain-specific information from CMR imaging to ECG representations. In extensive experiments using data from 40,044 UK Biobank subjects, we demonstrate the utility and generalisability of our method for subject-specific risk prediction of CVD and the prediction of cardiac phenotypes using only ECG data. Specifically, our novel multimodal pre-training paradigm improves performance by up to 12.19% for risk prediction and 27.59% for phenotype prediction. In a qualitative analysis, we demonstrate that our learned ECG representations incorporate information from CMR image regions of interest. Our entire pipeline is publicly available at https://github.com/oetu/MMCL-ECG-CMR.

Prognostic value of the fibrinogen-to-albumin ratio (FAR) in patients with chronic heart failure across the different ejection fraction spectrum

ABSTRACT The ratio of fibrinogen to albumin (FAR) is considered a new inflammatory biomarker and a predictor of cardiovascular disease risk. However, its prognostic value for patients with chronic heart failure (CHF) with different ejection fractions (EFs) remains unclear. A total of 916 hospitalized patients with CHF from January 2017 to October 2021 in the First Affiliated Hospital of Kunming Medical University were included in the study. Death occurred in 417 (45.5%) patients out of 916 patients during a median follow-up time of 750 days. Among these patients, 381 patients suffered from HFrEF (LVEF <40%) and 535 patients suffered from HFpEF or HFmrEF (HFpEF plus HFmrEF, LVEF ≥ 40%). Patients were categorized into high-level FAR (FAR-H) and low-level FAR (FAR-L) groups based on the optimal cut-off value of FAR (9.06) obtained from receiver operating characteristic (ROC) curve analysis. Upon analysing the Kaplan – Meier plots, the incidence of death was significantly higher in all patients with FAR-H and patients in both HF subgroups (p < 0.001). The multivariate Cox proportional hazard analyses indicated that the FAR was an independent predictor of all-cause mortality, regardless of heart failure subtype. (HR 1.115, 95% CI 1.089–1.142, p < 0.001; HFpEF plus HFmrEF, HR 1.109, 95% CI 1.074–1.146, p < 0.0001; HFrEF, HR 1.138, 95% CI 1.094–1.183, p < 0.0001) The optimal cut-off value of FAR in predicting all-cause mortality was 9.06 with an area under the curve value of 0.720 (95% CI: 0.687–0.753, p < 0.001), a sensitivity of 68.8% and a specificity of 65.6%. After adjusting for the traditional indicators (LVEF, Lg BNP, etc.), the new model with the FAR had better prediction ability in patients with CHF. Elevated FAR is an independent predictor of death in CHF and is not related to the HF subtype.

Ceramide-based risk score CERT-1 improves risk prediction for overall mortality and adverse cardiovascular outcomes in patients with and without cardiovascular disease: A prospective cohort study.

Whether the plasma-based ceramide-based risk score CERT1 improves risk prediction for cardiovascular disease (CVD) is uncertain. Baseline and follow-up data were combined from two cohorts, 334 patients with established/suspected CVD and 196 patients with type 2 diabetes followed for a median of 74 months (interquartile range 54-79 months). For the calculation of CERT1 risk score, we measured four specific plasma ceramides [Cer(d18:1/16:0), Cer(d18:1/18:0) and Cer(d18:1/24:1)] and their ratios to Cer(d18:1/24:0). Based on the CERT1 risk score, patients were split into four risk categories (low, moderate, increased or high risk). The primary outcome was a composite of overall mortality and incident nonfatal CVD outcomes (including myocardial infarction, ischaemic stroke or permanent atrial fibrillation). One hundred and thirty-nine patients developed the primary composite outcome (72 nonfatal CVD outcomes and 67 total deaths) during follow-up. Baseline CERT1 risk categories were significantly associated with the risk of developing the primary composite outcome (adjusted HR for high vs. low-risk category 2.43, 95% CI 1.39-4.22, p = 0.002, and adjusted HR for increased vs. low-risk category 2.16, 95% CI 1.28-3.63, p = 0.004). Receiver operator characteristic curve analysis showed that adding CERT1 risk score to traditional CVD risk factors and pre-existing CVD, improved the discriminatory capability of the regression model for predicting the primary composite outcome (AUROC 0.691 [95% CI 0.674-0.769] vs. 0.722 [95% CI 0.642-0.742], p = 0.0275). The ceramide-based risk score CERT1 risk score improves risk prediction for long-term risk of overall mortality and adverse cardiovascular outcomes in patients with and without CVD.

Acdim: A Cardiovascular Disease Risk Prediction Model Based on TabNet and AdaBoost for Smart Aging

Cardiovascular disease is a typical chronic disease. The incidence of cardiovascular disease in the elderly population has grown to exceed 50%. The further aging of the current population has brought additional pressure on public health services. Therefore, to reduce the service pressure of medical institutions, accurate prediction of cardiovascular disease risk has become an essential task in intelligent elderly care. In order to achieve accurate prediction of cardiovascular disease risk, we propose a cardiovascular disease risk prediction model Acdim (a cardiovascular disease risk prediction model) based on ResNet (Residual Network), zebra optimization algorithm (ZOA), TabNet (Attentive Interpretable Tabular Learning), and AdaBoost (Adaptive Boosting) algorithms. In training the Acdim model, we used the dat set of the World Health Organization (WHO) Behavioral Risk Factor Monitoring System records. In the experiment, the Acdim model achieved an accuracy of 96%, a precision of 94%, a recall of 93%, a specificity of 95%, an F1 score of 91%, and an AUC of 95%. The experimental results show that the proposed Acdim model can enable elderly care institutions to accurately predict the risk of cardiovascular disease in the absence of doctors.

Lipoprotein(a) Is Elevated and Inversely Related to Coronary Endothelial Function in People With HIV.

HIV-associated cardiovascular disease (CVD) is increasing in prevalence. The mechanisms underlying the heightened cardiovascular risk faced by people with HIV (PWH), however, remain poorly defined. Recent studies indicate an important role of lipoprotein(a) (Lp[a]) in predicting CVD risk in the general population, but little is known regarding its role in HIV-associated CVD. Thus, we sought to evaluate whether Lp(a) is elevated in PWH and if it is associated with impaired coronary endothelial function (CEF), a known mediator of CVD in PWH. In this cross-sectional study, cardiac magnetic resonance imaging with isometric handgrip exercise, an endothelial dependent stressor, was performed to assess CEF in 65 PWH and 52 controls without HIV. Percent changes in coronary cross-sectional area and coronary blood flow from rest to stress were used to quantify CEF. Lp(a) levels were assessed by immunoturbidimetric assay at the time of magnetic resonance imaging. Lp(a) levels were higher in PWH compared with controls (78 nmol/L [39-137 nmol/L] versus 45.5 nmol/L [18-102.5 nmol/L], P<0.01). Both percent change in coronary cross-sectional area (0.38% [-6.1% to 5.4%] versus 7.43% [2.4%-11.2%], P<0.0005) and coronary blood flow (9.1% [-1.3% to 23.1%] versus 24.1% [3.3%-39.8%], P<0.05) were lower in PWH compared with controls. In PWH, Lp(a) was inversely associated with percent change in coronary cross-sectional area (β=-6.18±1.01%/nmol/L, P<0.001) but not with percent change in coronary blood flow even after adjustment for confounding risk factors. No association between Lp(a) and measures of CEF was observed in individuals without HIV. Lp(a) concentrations are elevated in PWH and inversely related to CEF in PWH.

Cardiovascular wellness in low-resource settings: A mobile app-based risk prediction study among fuel filling station employees in Puducherry district.

India is witnessing a significant increase in the prevalence of non-communicable diseases (NCDs), and addressing this requires a comprehensive and multi-faceted approach. The burden of NCDs puts a strain on the healthcare system, requiring an increased focus on preventive measures, early detection, and management of chronic conditions. Adopting a risk-based approach to cardiovascular diseases (CVDs) in resource-poor settings offers several economic and social advantages. The aim of the study was to assess the prevalence of CVD risk factors among fuel filling station employees in the Puducherry district and the 10-year CVD risk prediction score among the study participants with the World Health Organisation package of essential non-communicable (WHO PEN) app and package for resource-poor settings. A community-based cross-sectional study was conducted among the fuel filling station employees in Puducherry. A universal sampling method was employed. The data were collected using a pilot-tested, predesigned, structured questionnaire and the WHO PEN app was used to estimate the CVD risk score. The data were collected from February 2021 to January 2022 and analysed using Statistical Package for Social Sciences (SPSS) version 20. Frequency distribution along with the Chi-square test was employed to test statistical significance. Out of 212 subjects, 170 (80.2%) were males, out of which 116 (54.7%) were between 40 and 50 years old. Nearly half the participants (48%) had CVD risk scores ranging from 5 to 20%, with an increased prevalence of CVD risk factors, namely, obesity/overweight (65.5%), physical inactivity (58.5%), hypertension (52%), alcohol consumption (51%) and tobacco consumption in any form (25.5%). This study sheds light on the sedentary nature of the occupation and the increased prevalence of CVD risk factors among the study participants. It is also evident that the participants had higher CVD risk scores for developing CVDs in the future. Recommendations: The use of mobile-based apps can be used as a feasible strategy to save scarce resources in delivering primary health care. We also propose that the nature of occupation be taken into account as one of the parameters for risk prediction. Risk prediction assessment should be made mandatory during the annual examination of employees.

Assessing the Accuracy of Cardiovascular Disease Prediction Using Female-Specific Risk Factors in Women Aged 45 to 69 Years in the UK Biobank Study.

Cardiovascular disease (CVD) is the leading cause of mortality in women. We aimed to assess whether adding female-specific risk factors to traditional factors could improve CVD risk prediction. We used a cohort of women from the UK Biobank Study aged 45 to 69 years, free of CVD at baseline (2006-2010) followed until the end of 2019. We developed Cox proportional hazards models using the risk factors included in 3 contemporary CVD risk calculators: Pooled Cohort Equation - Atherosclerotic Cardiovascular Disease, Qrisk2, and PREDICT. We added each of the following female-specific risk factors, individually and all together, to determine if these improved measures of discrimination and calibration for predicting CVD: early menarche (<11 years), endometriosis, excessive, frequent or irregular menstruation, miscarriage, number of miscarriages, number of stillbirths, infertility, preeclampsia or eclampsia, gestational diabetes (without subsequent type 2 diabetes), premature menopause (<40 years), early menopause (<45 years), and natural or surgical early menopause (menopause <45 years or timing of menopause reported as unknown and oophorectomy reported at age <45). In the model of 135 142 women (mean age, 57.5 years; SD, 6.8) using risk factors from Pooled Cohort Equation - Atherosclerotic Cardiovascular Disease, CVD incidence was 5.3 per 1000 person-years. The c-indices for the Pooled Cohort Equation - Atherosclerotic Cardiovascular Disease, Qrisk2, and PREDICT models were 0.710, 0.713, and 0.718, respectively. Adding each of the female-specific risk factors did not improve the c-index, the net reclassification index, the integrated discrimination index, the slope of the regression line for predicted versus observed events, and the Brier score or plots of calibration. Adding all female-specific risk factors simultaneously increased the c-index for the Pooled Cohort Equation - Atherosclerotic Cardiovascular Disease, Qrisk2, and PREDICT models to 0.712, 0.715, and 0.720, respectively. Although several female-specific factors have been shown to be early indicators of CVD risk, these factors should not be used to reclassify risk in women aged 45 to 69 years when considering whether to commence a blood pressure or lipid-lowering medication.

Associations of baseline and changes in the triglyceride glucose-weight adjusted waist index and cardiovascular disease risk: evidence from middle-aged and older individuals

BackgroundExisting researches have predominantly focused on the implications of dynamic alterations in the triglyceride-glucose (TyG) index and traditional obesity measures for cardiovascular disease (CVD) risk. However, the application of the weight-adjusted waist index (WWI), which incorporates the dynamically changing body composition factors of weight and waist circumference, alongside the TyG index for predicting CVD risk, remains unexplored. This study explores the relationships between baseline TyG-WWI index and its dynamic changes with CVD risk.MethodsSubjects were drawn from the China Health and Retirement Longitudinal Study. Logistic regression analyses were conducted to determine the relationships between baseline and longitudinal changes in the TyG-WWI index and CVD risk, quantified through odds ratios (ORs) and 95% confidence intervals (CIs). The robustness of results was confirmed via subgroup analyses and E-values. Additionally, restricted cubic spline and quartile-based methods evaluated the relationships between baseline and cumulative TyG-WWI indices and CVD risk.ResultsOver two survey waves, 613 CVD events were recorded. Analysis using adjusted multivariable models demonstrated a significant relationship between the cumulative TyG-WWI index and increased CVD risk, with an adjusted OR (95% CI) of 1.005 (1.000, 1.009). Class 2 of the TyG-WWI index change showed greater risk of CVD compared to Class 1, with ORs of 1.270 (1.008, 1.605). However, no significant connection was observed between the baseline TyG-WWI index and CVD risk (OR = 1.007, 95% CI: 0.996, 1.019). These findings were corroborated through extensive sensitivity analyses.Graphical

Zero End-Digit Preference in Blood Pressure and Implications for Cardiovascular Disease Risk Prediction-A Study in New Zealand.

Background/Objectives: Blood pressure (BP) readings are often rounded to the nearest zero end-digit. Guidelines permit rounding to the closest 2 mmHg. This paper investigated the effect of rounding systolic blood pressure (SBP) values on the prediction of cardiovascular disease (CVD) risk among the New Zealand population. A total of 427,299 individuals received opportunistic cardiovascular disease risk assessments at primary care facilities in New Zealand. Method: A total of 292,122 SBP readings possessed a non-zero terminal digit. These were rounded to the nearest zero end-digit. A survival model estimating a 5-year CVD risk was applied to both datasets, i.e., with and without rounding. Hazard ratios and misclassification rates were analysed to emphasise the notable differences. Financial impact was assessed by examining healthcare expenditures. Results: In total, 32% of SBP values exhibited a terminal digit of zero, and 2.85% and 4.24% of men were misclassified as moderate and high risk, respectively, while approximately 3.21% of women were misclassified into the same risk categories. Likewise, 1.19% and 0.47% of men, as well as 0.62% and 0.20% of women, were misclassified into the low and moderate risk categories, respectively. Conclusions: Precisely measuring SBP is crucial in accurately assessing CVD risk and managing healthcare resources effectively.

A Proteomics-Based Approach for Prediction of Different Cardiovascular Diseases and Dementia.

Many studies have explored whether individual plasma protein biomarkers improve cardiovascular disease risk prediction. We sought to investigate the use of a plasma proteomics-based approach in predicting different cardiovascular outcomes. Among 51 859 UK Biobank participants (mean age, 56.7 years; 45.5% male) without cardiovascular disease and with proteomics measurements, we examined the primary composite outcome of fatal and nonfatal coronary heart disease, stroke, or heart failure (major adverse cardiovascular events), as well as additional secondary cardiovascular outcomes. An exposome-wide association study was conducted using relative protein concentrations, adjusted for a range of classic, demographic, and lifestyle risk factors. A prediction model using only age, sex, and protein markers (protein model) was developed using a least absolute shrinkage and selection operator-regularized approach (derivation: 80% of cohort) and validated using split-sample testing (20% of cohort). Their performance was assessed by comparing calibration, net reclassification index, and c statistic with the PREVENT (Predicting Risk of CVD Events) risk score. Over a median 13.6 years of follow-up, 4857 participants experienced first major adverse cardiovascular events. After adjustment, the proteins most strongly associated with major adverse cardiovascular events included NT-proBNP (N-terminal pro B-type natriuretic peptide; hazard ratio [HR], 1.68 per SD increase), proADM (pro-adrenomedullin; HR, 1.60), GDF-15 (growth differentiation factor-15; HR, 1.47), WFDC2 (WAP four-disulfide core domain protein 2; HR, 1.46), and IGFBP4 (insulin-like growth factor-binding protein 4; HR, 1.41). In total, 222 separate proteins were predictors of all outcomes of interest in the protein model, and 86 were selected for the primary outcome specifically. In the validation cohort, compared with the PREVENT risk factor model, the protein model improved calibration, net reclassification (net reclassification index +0.09), and c statistic for major adverse cardiovascular events (+0.051). The protein model also improved the prediction of other outcomes, including ASCVD (c statistic +0.035), myocardial infarction (+0.023), stroke (+0.024), aortic stenosis (+0.015), heart failure (+0.060), abdominal aortic aneurysm (+0.024), and dementia (+0.068). Measurement of targeted protein biomarkers produced superior prediction of aggregated and disaggregated cardiovascular events. This study represents an important proof of concept for the application of targeted proteomics in predicting a range of cardiovascular outcomes.

Abstract 4121042: Novel genetic risk loci of arterial stiffness in Asian ancestry individuals

Background: Arterial stiffness serves as an independent predictor of cardiovascular disease risk. Pulse-wave velocity (PWV) is a non-invasive measurement of arterial stiffness. However, the genetic risk variants associated with arterial stiffness in the Asian ancestry population remain largely unknown. This lack of knowledge hinders the development of targeted preventive and treatment strategies. Hypothesis: Arterial stiffness is associated with genotyped SNP loci that have genetic relationships with metabolic syndrome phenotypes. Methods: For PWV, we performed genome-wide association study and gene-based genetic association study in 912 individuals from the Tongji-BGI research cardiovascular health (TORCH) cohort. We then used linkage disequilibrium score regression to estimate pairwise genetic correlations between arterial stiffness, blood pressure, lipids, and diabetes. Results: Eleven risk loci associated with arterial stiffness were discovered at genome-wide significance ( P &lt; 5 × 10 −8 ), including three in intronic (rs978453410 mapping to PPP2R5E; rs146270927 mapping to ECI1; rs189517402 mapping to CNN2 ). UBE2U and ECI1 were identified as the PWV-associated genes by gene-based analysis. Genetic correlation analyses revealed inter-related of PWV with systolic blood pressure, mean arterial pressure, total cholesterol and diabetes. Conclusions: This research firstly identified eleven novel genetic variants associated with arterial stiffness in Asian ancestry individuals. Further research aimed at reducing arterial stiffness may benefit from the identification of SNPs and genes like ECI1, which is related to the oxidation of unsaturated fatty acids in the mitochondria.

Abstract 4121454: Machine-extractable Markers in Chest Radiograph to Predict Cardiovascular Risk in Screening Population

Introduction: Recent research has shown that AI is able to assess biological aging and cardiovascular disease (CVD) risk using chest radiographs. However, the lack of explainability of such deep learning algorithms hinders clinical utility and adoption. This motivates the current study which searches for and tests the use of machine extractable quantitative features in chest radiographs to predict CVD risk in population screening. Method: Chest radiograph measurements characterizing cardiomediastinal geometry, aortic calcification and tortuosity were handpicked for development of a segmentation-based feature extraction algorithm. The algorithm was applied on the PLCO lung screening dataset for analysis. The association between measurement-based imaging features, clinical characteristics (age, sex, BMI, smoking status, hypertension, diabetes, liver disease) with CVD mortality and 10-year major adverse cardiovascular events (MACE) were analysed by using proportional hazard regression, with feature selection done by LASSO. Result: Of 29,453 eligible subjects, 5693 subjects from a single study centre were used for fitting of all models. The median follow-up time was 19 years. A total of 32 imaging features were extracted and analysed. For both 10-year MACE and CVD mortality, model using imaging features, age, and sex performed similarly to model using conventional risk factors, and a deep learning chest radiograph CVD risk model. Two imaging features, mediastinal width at valve-level [HR 1.36 (1.23-1.50)] and maximal lateral displacement of descending aorta [HR 1.29 (1.18-1.42)] were found to be prognostic. To the best of our knowledge, these features have not been reported previously. Conclusion: Quantitative imaging features can predict CVD risk in chest radiograph similar to deep learning models while providing feature interpretability and explainability. Two novel imaging features prognostic of CVD risk were found and shown to be complementary to conventional risk factors.

Abstract 4139194: Predicting Cholesterol Screening Behavior After Age 50 Using Machine Learning: Insights from the Health and Retirement Study

Background: In the U.S., about 8% of adults never received cholesterol screening. Although machine learning (ML) has been used to develop decision tools for Atherosclerotic Cardiovascular Disease (ASCVD) risk prediction, its application in behavioral forecasting has not yet been explored in the context of cholesterol screening behaviors. This study aimed to examine the performance and accuracy of ML algorithms in forecasting cholesterol screening behaviors in adults after age 50. Methods: This analysis used deidentified data from the Health and Retirement Study (HRS) 2004-2018. HRS is a longitudinal survey among 23,000 households in the U.S. Participants were excluded from the current analysis if they passed away by 2019, ever had ASCVD or stroke, were under age 50 at baseline, or had missing data in self-reported cholesterol screening. In total, 7176 participants (mean age [SD]=62 [8]) met the inclusion criteria; participants were randomly split into a training set (80%) and a testing set (20%). The synthetic minority oversampling technique was used to solve the imbalance distribution of the rare event. Five ML algorithms were used: random forest, gradient boosting machine (GBM), XGBoost, Support Vector Machine (SVM), and logistic regression. Accuracy, AUROC, and positive predictive value (PPV) were used to compare model performance. The average gain was evaluated for feature importance in the demographic and health domains. Results: In total, 232 (3.2%) respondents did not receive any cholesterol screening from 2008 to 2018. Experiments with five ML algorithms suggested that XGBoost with deeper trees and learning rate performed better in classifying those who did not screen for cholesterol levels over 10 years. Adding prior cholesterol screening history (2004-2006) into the model significantly improved model performance. Hypertension, self-rated health, and smoking were the major health features, while insurance, poverty, and work status were the major demographic features in the predictive model (accuracy=0.97; AUROC=0.88; PPV=0.42). Conclusion: Findings underscore the potential utility of ML models in predicting cholesterol screening behaviors after age 50. This could be the basis for developing decision tools for clinicians to identify those with a lower chance of cholesterol screening or make reminders accordingly. The low-cost predictive model might improve the uptake of preventive screening behaviors in middle-aged and older adults.

Abstract 4137836: Association of AHA PREVENT Risk Score with Cardiovascular Mortality Across Social Determinants of Health: A Nationwide Retrospective Cohort Study

Introduction/Background: The AHA PREVENT risk calculator was recently developed to accurately predict the risk of cardiovascular disease (CVD) risk, including incident heart failure, stroke, and coronary heart disease. However, the risk prediction of this tool for CVD mortality overall and among groups with increasing social determinants of health (SDOH) burden in a multi-ethnic nationally representative cohort of American adults is unclear. Methods: This retrospective study included adult NHANES participants (age &gt; 30 years) from 1998-2018 cycles without known CVD. Individuals without data for risk estimation were excluded. AHA PREVENT risk was scored from 0 to 100. Multivariable-adjusted Cox regression models were used to estimate the risk of CVD mortality, accounting for 7 SDOH measures (employment, family income, food security, health access, health insurance, housing instability, and being married or living with partner). Population was stratified based on cumulative SDOH burden (1, 2, 3, 4, 5, &gt; 5 unfavorable SDOH markers). Model discrimination was assessed using C-statistics in these groups. Results: Among 19,887 participants (median age 53 years, 49% Female, 54% White, 19% Black), a 1% increase in AHA PREVENT Risk score calculation was associated with 13% higher risk of CVD mortality (HR adj 1.13, 95% CI: 1.12-1.14) with C-statistic of 0.91. While AHA PREVENT Risk score was predictive of CVD mortality across the SDOH burden strata, there was a decline in C-statistic from 0.91 among those without any unfavorable SDOH to 0.81 among those with &gt; 5 unfavorable SDOH measures. Conclusion: The AHA PREVENT Risk score is predictive of CVD mortality even when accounting for SDOH measures. However, the risk discrimination decreases among individuals with a higher SDOH burden. Further investigation is needed to identify means to improve CVD risk prediction among disadvantaged people with high unfavorable SDOH burden.