Coronary Heart Disease Risk Prediction Research Articles

Rs420137, rs386360 and rs7763726 polymorphisms in fibronectin type III domain containing 1 are associated with susceptibility to coronary heart disease: Analysis in the Han population.

BackgroundNumerous genetic studies have shown that genes are related to the pathogenesis of coronary heart disease (CHD). The main aim of this study was to confirm whether fibronectin type III domain containing 1 (FNDC1) polymorphisms correlate with the risk of CHD.MethodsIn this study, in order to assess the association between three FNDC1 single nucleotide polymorphisms (SNPs) and the risk of CHD, we conducted a case-control study involving 630 patients with CHD and 568 healthy controls using Agena MassARRAY (Agena Bioscience, San Diego, CA, USA). Genotype distribution in case and control groups was analyzed by Chi square test. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated by logistic regression models adjusted for age, sex, smoking, and alcohol consumption to assess the correlation between SNPs and CHD risk.ResultsOur results indicated that FNDC1-rs420137, -rs386360, and -rs7763726 played important roles in enhancing the risk of CHD. Subgroup analysis revealed that rs420137 increased the susceptibility to CHD in males, smokers, and patients aged ≤62 years. Rs360 had an increased risk of CHD in males, patients at aged ≤62 years, smokers, and non-drinkers. Furthermore, the association of rs7763726 with increased CHD risk was also observed in males, patients aged ≤62 years, smokers, and drinkers. Last but not least, these three SNPs we selected were protective factors against hypertension in CHD individuals.ConclusionOur research suggest that FNDC1-rs420137, -rs386360, and -rs7763726 variants may be regarded as novel biomarkers for predicting CHD risk and other specific mechanisms of action of CHD need to be further studied.

Development and Validation of a Coronary Heart Disease Risk Prediction Model in Snorers with Hypertension: A Retrospective Observed Study.

To develop and validate a risk prediction model for coronary heart disease (CHD) in snorers with hypertension, including traditional and new risk factors. Twenty factors were evaluated in the records of 2810 snorers with hypertension. Training (70%) and validation (30%) sets were created by random allocation of data, and a new nomogram model was developed. The model's discrimination and calibration were measured by calculating the area under the receiver operating curve (AUC) and creating calibration charts. The performance of the nomogram model was compared with that of the Prediction for ASCVD Risk in China (China-PAR) and Framingham models by decision curve analysis. An optimal cutoff point for the risk score in the training set was computed to stratify patients. In the nomogram model, the AUCs for predicting CHD at 5, 7 and 9 years in the training set were 0.706 (95% confidence interval [CI] 0.649-0.763), 0.703 (95% CI 0.655-0.751) and 0.669 (95% CI 0.593-0.744), respectively. The respective AUCs were 0.682 (95% CI 0.607-0.758), 0.689 (95% CI 0.618-0.760) and 0.664 (95% CI 0.539-0.789) in the validation set. The calibration chart showed that the predicted events from the nomogram score were close to the observed events. Decision curve analysis indicated that the nomogram score was slightly better than the Prediction for ASCVD Risk in China (China-PAR) and Framingham models for predicting the risk of CHD in snorers with hypertension. A cutoff point was identified for being CHD-free (a nomogram score of ≤121), which could be helpful for the early identification of individuals at high-risk of CHD. The nomogram score predicts the risk probability of CHD in snorers with hypertension at 5, 7 and 9 years, and shows good capability in terms of discrimination and calibration. It may be a useful tool for identifying individuals at high risk of CHD.

Marital status and genetic liability independently predict coronary heart disease incidence.

Married individuals have a lower coronary heart disease (CHD) risk than non-married, but the mechanisms behind this are not fully understood. We analyzed whether genetic liability to CHD may affect these associations. Marital status, a polygenic score of CHD (PGS-CHD), and other risk factors for CHD were measured from 35,444 participants (53% female) in Finnish population-based surveys conducted between 1992 and 2012. During the register-based follow-up until 2020, there were 2439 fatal and non-fatal incident CHD cases. The data were analyzed using linear and Cox regression models. Divorced and cohabiting men and women had a higher genetic risk of CHD than married individuals, but the difference was very small (0.023-0.058 standard deviation of PGS-CHD, p-values 0.011-0.429). Both marital status and PGS-CHD were associated with CHD incidence, but the associations were largely independent. Adjusting for behavioral and metabolic risk factors for CHD explained part of these associations (11-20%). No interaction was found between marital status and PGS-CHD for CHD incidence. We showed minor differences between the marital status categories in PGS-CHD and demonstrated that marital status and genetic liability predicted CHD incidence largely independently. This emphasizes the need to measure multiple risk factors when predicting CHD risk.

All-Cause Death Prediction Method for CHD Based on Graph Convolutional Networks

Coronary heart disease (CHD) has become one of the most serious public health issues due to its high morbidity and mortality rates. Most of the existing coronary heart disease risk prediction models manually extract features based on shallow machine learning methods. It only focuses on the differences between local patient features and ignores the interaction modeling between global patients. Its accuracy is still insufficient for individualized patient management strategies. In this paper, we propose CHD prediction as a graph node classification task for the first time, where nodes can represent individuals in potentially diseased populations and graphs intuitively represent associations between populations. We used an adaptive multi-channel graph convolutional neural network (AM-GCN) model to extract graph embeddings from topology, node features, and their combinations through graph convolution. Then, the adaptive importance weights of the extracted embeddings are learned by using an attention mechanism. For different situations, we model the relationship of the CHD population with the population graph and the K-nearest neighbor graph method. Our experimental evaluation explored the impact of the independent components of the model on the CHD disease prediction performance and compared it to different baselines. The experimental results show that our new model exhibits the best experimental results on the CHD dataset, with a 1.3% improvement in accuracy, a 5.1% improvement in AUC, and a 4.6% improvement in F1-score compared to the nongraph model.

Coronary Heart Disease Risk Prediction Using Binary Logistic Regression Based on Principal Component Analysis

Based on data from the World Health Organization (WHO), one type of heart disease namely coronary heart disease is the deadliest disease in the world. In 2016 at least 9,4 million people died caused by coronary heart disease. In Indonesia, deaths caused by heart disease, blood vessel (CVD), and respiratory disorders are the fourth highest in ASEAN (23,1%). Because of the danger of coronary heart disease, we need a system or model that can predict heart disease early, so that it can be treated early and can reduce the death rate caused by heart disease. This study uses principal component analysis (PCA) to make a linear combination of variables that have a high correlation so that the assumption of multicollinearity in the data can be resolved. For the prediction, this study uses binary logistic regression to predict heart disease based on existing factors. The result of the PCA there is 7 component variables with a total variance that can be explained as much as 72,9%. From the Bartlett test of the PCA data, the obtained p-value is 1 which means that there is no multicollinearity in the data. Predictive analysis using binary logistic regression based on PCA’s data was proven to increase the accuracy to 85%.

Development and Validation of a Prognostic Model to Predict High-Risk Patients for Coronary Heart Disease in Snorers With Uncontrolled Hypertension.

PurposeSnoring or obstructive sleep apnea, with or without uncontrolled hypertension, is common and significantly increases the risk of coronary heart disease (CHD). The aim of this study was to develop and validate a prognostic model to predict and identify high-risk patients for CHD among snorers with uncontrolled hypertension.MethodsRecords from 1,822 snorers with uncontrolled hypertension were randomly divided into a training set (n = 1,275, 70%) and validation set (n = 547, 30%). Predictors for CHD were extracted to construct a nomogram model based on multivariate Cox regression analysis. We performed a single-split verification and 1,000 bootstraps resampling internal validation to assess the discrimination and consistency of the prediction model using area under the receiver operating characteristic curve (AUC) and calibration plots. Based on the linear predictors, a risk classifier for CHD could be set.ResultsAge, waist circumference (WC), and high- and low-density lipoprotein cholesterol (HDL-C and LDL-C) were extracted as the predictors to generate this nomogram model. The C-index was 0.720 (95% confidence interval 0.663–0.777) in the derivation cohort and 0.703 (0.630–0.776) in the validation cohort. The AUC was 0.757 (0.626–0.887), 0.739 (0.647–0.831), and 0.732 (0.665–0.799) in the training set and 0.689 (0.542–0.837), 0.701 (0.606–0.796), and 0.712 (0.615–0.808) in the validation set at 3, 5, and 8 years, respectively. The calibration plots showed acceptable consistency between the probability of CHD-free survival and the observed CHD-free survival in the training and validation sets. A total of more than 134 points in the nomogram can be used in the identification of high-risk patients for CHD among snorers with uncontrolled hypertension.ConclusionWe developed a CHD risk prediction model in snorers with uncontrolled hypertension, which includes age, WC, HDL-C, and LDL-C, and can help clinicians with early and quick identification of patients with a high risk for CHD.

Small Dense LDL: Scientific Background, Clinical Relevance, and Recent Evidence Still a Risk Even with 'Normal' LDL-C Levels.

Residual cardiovascular disease event risk, following statin use and low-density lipoprotein cholesterol (LDL-C) reduction, remains an important and common medical conundrum. Identifying patients with significant residual risk, despite statin drug use, is an unmet clinical need. One pathophysiologic disorder that contributes to residual risk is abnormal distribution in lipoprotein size and density, which is referred to as lipoprotein heterogeneity. Differences in low density lipoprotein (LDL) composition and size have been linked to coronary heart disease (CHD) risk and arteriographic disease progression. The clinical relevance has been investigated in numerous trials since the 1950s. Despite this long history, controversy remains regarding the clinical utility of LDL heterogeneity measurement. Recent clinical trial evidence reinforces the relevance of LDL heterogeneity measurement and the impact on CHD risk prediction and outcomes. The determination of LDL subclass distribution improves CHD risk prediction and guides appropriate treatment.

Association between the Suita Score and Stroke Recurrence in Patients with First-ever Ischemic Stroke: A Prospective Cohort Study.

Objective The Suita score is used to predict the 10-year prognosis of developing coronary heart disease (CHD). This study examined the association between the Suita score and stroke recurrence within one year in Japanese patients who experienced first-ever ischemic stroke. Methods This prospective cohort study at a stroke center in Japan included patients who experienced first-ever acute ischemic stroke (AIS) or a transient ischemic attack (TIA). During hospitalization, the Suita score was measured as the main exposure. Patients with a ≥5% predicted CHD risk were classified into the high-risk group. The primary outcome was stroke recurrence within one year of the stroke onset. A multivariate Cox regression analysis was conducted and adjusted for confounding and prognostic factors. Results Among the 1,204 patients evaluated, 937 (78%) were classified as having a high risk of developing CHD. Stroke recurrence was observed in 66 patients during the follow-up period. In the multivariate analysis, after adjusting for confounding and prognostic factors, such as non-small vessel occlusion and prescription of lipid-lowering agents at the time of discharge, a ≥5% predicted CHD risk was associated with the 1-year stroke recurrence after the initial onset [adjusted hazard ratio (HR)=2.20, 95% confidence interval (CI)=1.00-4.91, p=0.049; adjusted HR=2.00, 95% CI=1.01-4.14, p=0.048; adjusted HR=0.42, 95% CI=0.24-0.73, p=0.002]. Conclusion The Suita score, adapted for use in ischemic stroke with the same mechanism, correlated with the short-term recurrence within one year. Our findings suggest that the Suita score may be useful for predicting the long-term prognosis of developing CHD as well as the short-term recurrence for patients with first-ever AIS and TIA.

Prediction and risk stratification from hospital discharge records based on Hierarchical sLDA

BackgroundThe greatly accelerated development of information technology has conveniently provided adoption for risk stratification, which means more beneficial for both patients and clinicians. Risk stratification offers accurate individualized prevention and therapeutic decision making etc. Hospital discharge records (HDRs) routinely include accurate conclusions of diagnoses of the patients. For this reason, in this paper, we propose an improved model for risk stratification in a supervised fashion by exploring HDRs about coronary heart disease (CHD).MethodsWe introduced an improved four-layer supervised latent Dirichlet allocation (sLDA) approach called Hierarchical sLDA model, which categorized patient features in HDRs as patient feature-value pairs in one-hot way according to clinical guidelines for lab test of CHD. To address the data missing and imbalance problem, RFs and SMOTE methods are used respectively. After TF-IDF processing of datasets, variational Bayes expectation-maximization method and generalized linear model were used to recognize the latent clinical state of a patient, i.e., risk stratification, as well as to predict CHD. Accuracy, macro-F1, training and testing time performance were used to evaluate the performance of our model.ResultsAccording to the characteristics of our datasets, i.e., patient feature-value pairs, we construct a supervised topic model by adding one more Dirichlet distribution hyperparameter to sLDA. Compared with established supervised algorithm Multi-class sLDA model, we demonstrate that our proposed approach enhances training time by 59.74% and testing time by 25.58% but almost no loss of average prediction accuracy on our datasets.ConclusionsA model for risk stratification and prediction of CHD based on sLDA model was proposed. Experimental results show that Hierarchical sLDA model we proposed is competitive in time performance and accuracy. Hierarchical processing of patient features can significantly improve the disadvantages of low efficiency and time-consuming Gibbs sampling of sLDA model.

Comparative Performance Analysis of Random Forests against AutoPrognosis for predicting Coronary Heart Disease Risk and Metabolic Syndrome: A Retrospective Cohort Study

Cardiovascular Disease (CVD) is the leading cause of mortality worldwide. Amongst them, Coronary Heart Disease (CHD) is the most common type of CVD. The consequences of the presence of CVD risk factors often manifest as Metabolic Syndrome (MetS). In this study, data from the Framingham Heart Study (FHS), consisting of 4240 records and 17 variables, was used to build two types of 10-year CHD risk prediction models based on Random Forests (RF) and AutoPrognosis. The Framingham Risk Score model (AUC-ROC: 0.633) was used as a baseline model for performance evaluation. Results showed that the RF model with optimized hyperparameters had the best performance (AUC-ROC: 0.728). Furthermore, a dataset of 7821 records and 77 variables from the National Health and Nutrition Examination Survey (NHANES) was used to assess the predictive performance of RF against AutoPrognosis for determining the presence of MetS. The RF model with optimized hyperparameters had the best performance (AUC-ROC: 0.851). The performance of RF against AutoPrognosis on different sample sizes of data, ranging from 100 to 4900, was tested. The RF model with optimized hyperparameters had the best overall performance, followed by AutoPrognosis with an ensemble pipeline, then AutoPrognosis with a single pipeline and finally the RF model with default hyperparameter values.

Use of Artificial Intelligence in Coronary Artery Calcium Scoring.

Coronary artery calcium (CAC) scoring improves traditional risk factor-based coronary heart disease (CHD) risk stratification. Here, the contribution of CAC scoring to a traditional 10-year CHD risk prediction scores and new artificial intelligence methods used to automate CAC scoring were reviewed. Research shows that traditional risk factors tend to overestimate or underestimate the actual risk of CHD, meaning that including CAC score in the risk stratification has potential to reduce over- and undertreatment. The automated CAC scoring methods are shown to be accurate and significantly more time-effective than the commonly used semi-automated method.

Improvement in the Prediction of Coronary Heart Disease Risk by Using Artificial Neural Networks.

Cardiovascular diseases, such as coronary heart disease (CHD), are the main cause of mortality and morbidity worldwide. Although CHD cannot be entirely predicted by classic risk factors, it is preventable. Therefore, predicting CHD risk is crucial to clinical cardiology research, and the development of innovative methods for predicting CHD risk is of great practical interest. The Framingham risk score (FRS) is one of the most frequently implemented risk models. However, recent advances in the field of analytics may enhance the prediction of CHD risk beyond the FRS. Here, we propose a model based on an artificial neural network (ANN) for predicting CHD risk with respect to the Framingham Heart Study (FHS) dataset. The performance of this model was compared to that of the FRS. A sample of 3066 subjects from the FHS offspring cohort was subjected to an ANN. A multilayer perceptron ANN architecture was used and the lift, gains, receiver operating characteristic (ROC), and precision-recall predicted by the ANN were compared with those of the FRS. The lift and gain curves of the ANN model outperformed those of the FRS model in terms of top percentiles. The ROC curve showed that, for higher risk scores, the ANN model had higher sensitivity and higher specificity than those of the FRS model, although its area under the curve (AUC) was lower. For the precision-recall measures, the ANN generated significantly better results than the FRS with a higher AUC. The findings suggest that the ANN model is a promising approach for predicting CHD risk and a good screening procedure to identify high-risk subjects.

Risk of coronary heart disease among cancer survivors with different prediagnosis body mass index

Association between body mass index (BMI) and coronary heart disease (CHD) in cancer survivors is not clearly established. This study analyzed the prediagnosis BMI-CHD association by examining 13,500 cancer survivors identified from the National Health Insurance Service-Health Screening Cohort from January 1, 2004 to December 31, 2009 including the patients who were free of cardiovascular disease at enrollment. The Cox proportional hazards model (adjusted for socioeconomic, health behavior, health status, and medical characteristics) was used for calculating hazard ratios (HR) and 95% confidence intervals (95% CI) for CHD in each prediagnosis BMI category among cancer survivors. Compared to cancer survivors with a prediagnosis BMI between 18.5 and 22.9 kg/m2, those with a prediagnosis BMI of 23.0–24.9 kg/m2 and ≥ 25.0 kg/m2 had significantly higher CHD risk (HR = 1.51; 95% CI: 1.13–2.01 and HR = 1.38; 95% CI: 1.04–1.84, respectively). Cancer survivors with a low prediagnosis BMI (< 18.5 kg/m2) also had significantly higher CHD risk (HR = 1.97; 95% CI: 1.20–3.24) compared to those with a BMI of 18.5–22.9 kg/m2. Similar associations were found after stratifying analyses based on first cancer site and sociodemographic and medical characteristic subgroups. Our study suggests that prediagnosis underweight among patients with cancer is a predictor of CHD risk.

Predictive Value of Small Dense Low-Density Lipoprotein in Coronary Heart Disease in the Chinese Population.

Coronary heart disease (CHD) is a global public health concern, and CHD risk assessment remains a major challenge. Therefore, in this study, the ability of small dense low-density lipoprotein (sdLDL) alone to sufficiently predict CHD risk in the Chinese population was evaluated. Patients with CHD (139) and healthy controls (58) were included in this study. Serum sdLDL was measured using the peroxidase method. Other lipid parameters were also determined. The sdLDL level in the CHD group was significantly higher than that in the control group (p < 0.001). Logistic regression analysis revealed that sdLDL was an independent risk factor for CHD. Based on the receiver operating characteristic curves, the area under the curve (AUC) of sdLDL alone for CHD was 0.722. The AUC of triglycerides (TG), high-density lipoprotein (HDL), and sdLDL combined was 0.763, which was larger than that of the independent ones or combinations of any two; however, the value was not significant. sdLDL alone can predict CHD risk efficiently similar to the combination of TG, HDL, and sdLDL. This finding suggests that sdLDL can be considered as an ideal parameter for the preliminary diagnosis of CHD in Chinese people.

Real-time ECG signal pre-processing and neuro fuzzy-based CHD risk prediction

Coronary heart disease (CHD) is a major chronic disease which is directly responsible for myocardial infarction. Heart rate variability (HRV) has been used for the prediction of CHD risk in human beings. In this work, neuro fuzzy-based CHD risk prediction is performed after performing pre-processing and HRV feature extraction. The pre-processing is used to remove high frequency noise which is modelled as white Gaussian noise. The real-time ECG signal acquisition, pre-processing and HRV feature extraction are performed using NI LabVIEW and DAQ board. A 30 seconds recording of ECG signal was selected in both smokers and non-smokers. Various statistical parameters are extracted from HRV to predict coronary heart disease (CHD) risk among the subjects. The HRV extracted signals are classified into normal and CHD risky subjects using neuro fuzzy classifier. The classification performance of the neuro fuzzy classifier is compared with the ANN, KNN, and decision tree classifiers.

Abstract 15273: Estimating Coronary Heart Disease Risk and Treatment Benefit in Patients With Familial Hypercholesterolemia to Enable Shared Decision Making

Introduction: Current risk equations for coronary heart disease (CHD) do not take into account family history or cumulative exposure to LDL-C. We modified CHD risk equations for use in familial hypercholesterolemia (FH) patients and developed a decision aid (DA) for FH to facilitate shared decision making (SDM). Methods: Literature review, evidence synthesis and expert input were used to adapt the 30-year Framingham Risk Score (FRS) and 10-year ACC/AHA pooled cohort equations to calculate the CHD risk in individuals with FH aged 20-39 years and ≥40 years, respectively. Based on published data, estimated risk was multiplied by 2 for family history of early CHD or 3 in the presence of a pathogenic variant. Treatment benefit of lipid lowering was estimated based on the ‘cumulative exposure to LDL-C’ model ( JACC 2012; 60 ) using the following equations: Post-treatment 30-yr risk in those aged 20-39 years = Baseline Risk (FRS) x 0.54 delta LDL-C (mmol/L) Post-treatment 10-yr risk in those aged ≥40 years = Baseline Risk (ACC/AHA Pooled Cohort) x 0.72 delta LDL-C (mmol/L) To convey CHD risk and treatment benefit, we developed a DA using an iterative, patient-centered design process for use in FH encounters. Results: Using the DA, providers can discuss baseline CHD risk estimates, lipid lowering drug (LLD) options and the expected post-treatment risk reduction with FH patients. The estimated % LDL-C reduction for each LLD and the predicted risk reduction per mmol/L of LDL-C lowered were obtained from literature review. Lowering LDL-C by 1 mmol/L (39 mg/dL) was associated with a 28% CHD risk reduction over 10 years and a 46% reduction over 30 years ( Eur Heart J 2017; 38 ). We describe 2 examples of estimating CHD risk and treatment benefit in FH cases in Table 1 . Conclusion: We adapted CHD risk prediction equations to account for family history of CHD and presence of a pathogenic FH variant and designed a DA to convey personalized CHD risk estimates and treatment benefit to facilitate SDM with FH patients.

South Asian ethnicity as a risk factor for coronary heart disease

Background and aimsSouth Asian ethnicity has been associated with increased cardiovascular risk, in the context of a greater burden of traditional cardiovascular disease risk factors. We aimed to identify the 10-year incidence of coronary heart disease (CHD) in South Asians, as compared to other racial-ethnic groups, in a retrospective cohort study, using the Kaiser Permanente Northern California integrated health care system, and to understand whether traditional risk factors could explain differences in outcomes. Methods and resultsA study cohort established in 2006, with 10-year follow-up, identified a total n = 341,309 patients, including 5149 South Asians, aged 30–70 years who underwent a screening lipid panel and had no prior history of CHD. There were a total of 460 (8.93%) patients with at least one cardiovascular event, including cardiovascular death, coronary artery revascularization, and myocardial infarction, in South Asians, as compared to 5.66% patients in the total population (p < 0.0001). In a multivariable logistic regression analysis, South Asian ethnicity, compared to White race, was associated with an adjusted odds ratio (aOR) = 2.04 (95%CI: 1.83, 2.28) in predicting the cardiovascular outcomes. ConclusionsSouth Asian ethnicity, even after adjustment for traditional risk factors, is associated with an increased risk of coronary heart disease outcomes. This risk was greater than other studied racial-ethnic groups and second only to diabetes in CHD risk prediction.

Accurate Prediction of Coronary Heart Disease for Patients With Hypertension From Electronic Health Records With Big Data and Machine-Learning Methods: Model Development and Performance Evaluation.

BackgroundPredictions of cardiovascular disease risks based on health records have long attracted broad research interests. Despite extensive efforts, the prediction accuracy has remained unsatisfactory. This raises the question as to whether the data insufficiency, statistical and machine-learning methods, or intrinsic noise have hindered the performance of previous approaches, and how these issues can be alleviated.ObjectiveBased on a large population of patients with hypertension in Shenzhen, China, we aimed to establish a high-precision coronary heart disease (CHD) prediction model through big data and machine-learningMethodsData from a large cohort of 42,676 patients with hypertension, including 20,156 patients with CHD onset, were investigated from electronic health records (EHRs) 1-3 years prior to CHD onset (for CHD-positive cases) or during a disease-free follow-up period of more than 3 years (for CHD-negative cases). The population was divided evenly into independent training and test datasets. Various machine-learning methods were adopted on the training set to achieve high-accuracy prediction models and the results were compared with traditional statistical methods and well-known risk scales. Comparison analyses were performed to investigate the effects of training sample size, factor sets, and modeling approaches on the prediction performance.ResultsAn ensemble method, XGBoost, achieved high accuracy in predicting 3-year CHD onset for the independent test dataset with an area under the receiver operating characteristic curve (AUC) value of 0.943. Comparison analysis showed that nonlinear models (K-nearest neighbor AUC 0.908, random forest AUC 0.938) outperform linear models (logistic regression AUC 0.865) on the same datasets, and machine-learning methods significantly surpassed traditional risk scales or fixed models (eg, Framingham cardiovascular disease risk models). Further analyses revealed that using time-dependent features obtained from multiple records, including both statistical variables and changing-trend variables, helped to improve the performance compared to using only static features. Subpopulation analysis showed that the impact of feature design had a more significant effect on model accuracy than the population size. Marginal effect analysis showed that both traditional and EHR factors exhibited highly nonlinear characteristics with respect to the risk scores.ConclusionsWe demonstrated that accurate risk prediction of CHD from EHRs is possible given a sufficiently large population of training data. Sophisticated machine-learning methods played an important role in tackling the heterogeneity and nonlinear nature of disease prediction. Moreover, accumulated EHR data over multiple time points provided additional features that were valuable for risk prediction. Our study highlights the importance of accumulating big data from EHRs for accurate disease predictions.

Family history of cardiovascular disease and risk of premature coronary heart disease: A matched case-control study.

Background: Self-reported family history of cardiovascular disease (CVD) is an independent risk factor for future coronary heart disease (CHD) events. However, inclusion of family history of CVD in the traditional risk scores failed to improve risk prediction of CHD. It is proposed that family history of CVD may substantially increase the risk of CHD among younger individuals. Methods: We conducted a matched case-control study with 170 hospital-based premature CHD patients (<55 years in men and <65 years in women) from a tertiary care centre in Thiruvananthapuram, Kerala and age and sex matched community-based controls in 1:1 ratio. Conditional logistic regression analysis was conducted to assess the independent association of family history of cardiovascular disease (CVD) and premature CHD. We estimated McNemar's odds ratios and their 95 percent confidence intervals. Results: The prevalence of any family history of CVD and CHD in the control population was 24% and 21%, respectively. The family history of CVD was independently associated with premature CHD (odds ratio (OR) = 9.0; 95% confidence interval (CI) 4.7-17.3). There was a dose-response relationship between family history and premature CHD as the risk increased linearly with increase in number of affected family members. Conclusions: Family history of CVD is an independent risk factor for premature CHD. The risk of premature CHD increases linearly with increase in number of affected family members. Collecting family history beyond parental history of CVD is important for risk stratification. Targeting young individuals with family history of CVD for intensive risk reduction interventions may help to prevent future events.

Family history of cardiovascular disease and risk of premature coronary heart disease: A matched case-control study

Background: Self-reported family history of cardiovascular disease (CVD) is an independent risk factor for future coronary heart disease (CHD) events. However, inclusion of family history of CVD in the traditional risk scores failed to improve risk prediction of CHD. It is proposed that family history of CVD may substantially increase the risk of CHD among younger individuals. Methods: We conducted a matched case-control study with 170 hospital-based premature CHD patients (<55 years in men and <65 years in women) from a tertiary care centre in Thiruvananthapuram, Kerala and age and sex matched community-based controls in 1:1 ratio. Conditional logistic regression analysis was conducted to assess the independent association of family history of cardiovascular disease (CVD) and premature CHD. We estimated McNemar's odds ratios and their 95 percent confidence intervals. Results: The prevalence of any family history of CVD and CHD in the control population was 24% and 21%, respectively. The family history of CVD was independently associated with premature CHD (odds ratio (OR) = 9.0; 95% confidence interval (CI) 4.7–17.3). There was a dose-response relationship between family history and premature CHD as the risk increased linearly with increase in number of affected family members. Conclusions: Family history of CVD is an independent risk factor for premature CHD. The risk of premature CHD increases linearly with increase in number of affected family members. Collecting family history beyond parental history of CVD is important for risk stratification. Targeting young individuals with family history of CVD for intensive risk reduction interventions may help to prevent future events.