Machine Learning for Sudden Cardiac Death Prediction in the Atherosclerosis Risk in Communities Study

Introduction Sudden cardiac death (SCD) is a devastating consequence often without antecedent expectation. Current risk stratification methods derived from baseline independently modeled risk factors are insufficient. Novel random forest machine learning (ML) approach incorporating time-dependent variables and complex interactions may improve SCD risk prediction. Methods Atherosclerosis Risk in Communities (ARIC) study participants were followed for adjudicated SCD. ML models were compared to standard Poisson regression models for interval data, an approximation to Cox regression, with stepwise variable selection. Eighty-two time-varying variables (demographics, lifestyle factors, clinical characteristics, biomarkers, etc.) collected at four visits over 12 years (1987-98) were used as candidate predictors. Predictive accuracy was assessed by area under the receiver operating characteristic curve (AUC) through out-of-bag prediction for ML models and 5-fold cross validation for the Poisson regression models. Results Over a median follow-up time of 23.5 years, 583 SCD events occurred among 15,661 ARIC participants (mean age 54 years and 55% women). Compared to different Poisson regression models (AUC at 6-year ranges from 0.77-0.83), the ML model improved prediction (AUC at 6-year 0.89). Top predictors identified by ML model included prior coronary heart disease (CHD), which explained 47.9% of the total phenotypic variance, diabetes mellitus, hypertension, and T wave abnormality in any of leads I, aVL, or V6. Using the top ML predictors to select variables, the Poisson regression model AUC at 6-year was 0.77 suggesting that the non-linear dependencies and interactions captured by ML, are the main reasons for its improved prediction performance. Conclusions Applying novel ML approach with time-varying predictors improves the prediction of SCD. Interactions of dynamic clinical characteristics are important for risk-stratifying SCD in the general population.

Download Full-text

Vital exhaustion and sudden cardiac death in the Atherosclerosis Risk in Communities Study

Heart ◽

10.1136/heartjnl-2017-311825 ◽

2017 ◽

Vol 104 (5) ◽

pp. 423-429 ◽

Cited By ~ 2

Author(s):

Brittany M Bogle ◽

Nona Sotoodehnia ◽

Anna M Kucharska-Newton ◽

Wayne D Rosamond

Keyword(s):

Risk Factors ◽

Sudden Cardiac Death ◽

Cardiac Death ◽

Disease Risk ◽

Ventricular Tachyarrhythmia ◽

Atherosclerosis Risk In Communities ◽

Vital Exhaustion ◽

Increased Risk ◽

Atherosclerosis Risk ◽

Aric Study

ObjectiveVital exhaustion (VE), a construct defined as lack of energy, increased fatigue and irritability, and feelings of demoralisation, has been associated with cardiovascular events. We sought to examine the relation between VE and sudden cardiac death (SCD) in the Atherosclerosis Risk in Communities (ARIC) Study.MethodsThe ARIC Study is a predominately biracial cohort of men and women, aged 45–64 at baseline, initiated in 1987 through random sampling in four US communities. VE was measured using the Maastricht questionnaire between 1990 and 1992 among 13 923 individuals. Cox proportional hazards models were used to examine the hazard of out-of-hospital SCD across tertiles of VE scores.ResultsThrough 2012, 457 SCD cases, defined as a sudden pulseless condition presumed due to a ventricular tachyarrhythmia in a previously stable individual, were identified in ARIC by physician record review. Adjusting for age, sex and race/centre, participants in the highest VE tertile had an increased risk of SCD (HR 1.48, 95% CI 1.17 to 1.87), but these findings did not remain significant after adjustment for established cardiovascular disease risk factors (HR 0.94, 95% CI 0.73 to 1.20).ConclusionsAmong participants of the ARIC study, VE was not associated with an increased risk for SCD after adjustment for cardiovascular risk factors.

Download Full-text

Abstract MP083: Autonomic Imbalance at the Level of Atrioventricular Node, but Not at the Level of Sinus Node, is Associated With Sudden Cardiac Death: The Atherosclerosis Risk in Community Study

Circulation ◽

10.1161/circ.135.suppl_1.mp083 ◽

2017 ◽

Vol 135 (suppl_1) ◽

Author(s):

Srini V Mukundan ◽

Muammar M Kabir ◽

Jason Thomas ◽

Golriz Sedaghat ◽

Jonathan W Waks ◽

...

Keyword(s):

Sudden Cardiac Death ◽

Cardiac Death ◽

Cox Regression ◽

Sinus Node ◽

Atrioventricular Node ◽

Av Node ◽

Atherosclerosis Risk In Communities ◽

Autonomic Imbalance ◽

Summary Effect ◽

Atherosclerosis Risk

Introduction: Autonomic imbalance, quantified by decreased heart rate variability (HRV), is associated with increased cardiovascular mortality. It is unknown if autonomic influences on sinus and atrioventricular (AV) nodes are equally important for the risk of sudden cardiac death (SCD). Hypothesis: Autonomic influences on sinus and AV node are equally strongly associated with increased SCD, non-sudden cardiac death (non-SCD), and non-cardiac death. Methods: Baseline visit 10-second ECGs (n=14,250) of the Atherosclerosis Risk in Communities (ARIC) cohort were analyzed. Normalized variance of P-onset to P-onset intervals (PPVN) and QRS-onset to QRS-onset intervals (QQVN) was calculated to assess autonomic influence on sinus and AV node respectively. Normalized variance of Rpeak - Rpeak intervals was determined as HRV measure. Values were log-transformed to normalize distribution. SCD served as primary outcome. Secondary outcomes were non-SCD and non-cardiac death. Three Cox regression models were constructed for dichotomized at 20 th percentile predictor variables. Results: Over median follow-up of 24.4 years, there were 497 SCDs (incidence 1.66 [95%CI 1.52-1.82], 742 non-SCDs (incidence 2.48 [95%CI 2.31-2.67], and 3,753 non-cardiac deaths (incidence 12.6 [95%CI 12.1-13.0]) per 1,000 person-years. In paired analysis, LogPPVN was significantly larger than LogQQVN (-7.28±1.06 vs. -7.72±1.24; P<0.0001). There was no difference between LogQQVN and Log RRVN (-7.72±1.24 vs -7.72±1.23; P=0.364). After full adjustment, LogRRVN and LogQQVN were significantly associated with non-SCD and SCD. Association with non-SCD was stronger. LogPPVN was independently associated with non-SCD but not SCD. No value was associated with non-cardiac death. Conclusion: Autonomic imbalance at the AV node, with likely summary effect at the bundle of His, is associated with SCD and non-SCD. Autonomic imbalance at the SA node is associated with non-SCD only. Autonomic input to SA and AV node should be further studied.

Download Full-text

The Metabolic Syndrome and Risk of Sudden Cardiac Death: The Atherosclerosis Risk in Communities Study

Journal of the American Heart Association ◽

10.1161/jaha.117.006103 ◽

2017 ◽

Vol 6 (8) ◽

Cited By ~ 14

Author(s):

Paul L. Hess ◽

Hussein R. Al‐Khalidi ◽

Daniel J. Friedman ◽

Hillary Mulder ◽

Anna Kucharska‐Newton ◽

...

Keyword(s):

Metabolic Syndrome ◽

Sudden Cardiac Death ◽

Cardiac Death ◽

Atherosclerosis Risk In Communities ◽

The Metabolic Syndrome ◽

Atherosclerosis Risk

Download Full-text

P5645Lifetime sex-specific sudden cardiac death prediction using ECG global electrical heterogeneity: the atherosclerosis risk in communities (ARIC) study

European Heart Journal ◽

10.1093/eurheartj/ehz746.0588 ◽

2019 ◽

Vol 40 (Supplement_1) ◽

Author(s):

S Howell ◽

E Perez-Alday ◽

D German ◽

A Bender ◽

N Rogovoy ◽

...

Keyword(s):

Sudden Cardiac Death ◽

Prediction Model ◽

Cardiac Death ◽

Prediction Models ◽

Cox Regression ◽

Density Lipoprotein ◽

Screening Methods ◽

Atherosclerosis Risk In Communities ◽

Clinical Variables ◽

Atherosclerosis Risk

Abstract Background Sex-based differences in sudden cardiac death (SCD) exist and screening methods for SCD are inadequate. Purpose To develop sex-specific lifetime risk prediction models using electrocardiographic (ECG) global electrical heterogeneity (GEH) and clinical characteristics. Methods Participants from the Atherosclerosis Risk in Communities study with analyzable ECGs (n=14,725; age, 54.2±5.8 yrs; 55% female, 74% white) were followed up for 24.4 years (median). Traditional ECG and GEH variables were measured on 12-lead ECGs. A Cox regression model was used to develop a prediction model. In women, the final model included race, age, coronary heart disease (CHD), stroke, hypertension, diabetes, smoking, high-density lipoprotein, albumin, uric acid, education level, heart rate, QTc, sum absolute QRST integral, spatial peak QRS-T angle. In men, the final prediction model included age, race, CHD, stroke, hypertension, diabetes, total cholesterol, physical activity, smoking, serum phosphorus, albumin, chronic kidney disease, spatial area QRS-T angle, area spatial ventricular gradient (SVG) elevation and magnitude, and peak SVG magnitude. Results There were a total of 530 SCDs. Our prediction models showed robust prediction of SCD in both sexes [(Harrell's C-statistic women 0.863 (95% CI 0.845–0.882), men 0.786 (95% CI 0.786–0.803)]. In women when ECG and GEH variables were added to clinical variables, the net reclassification improved by 9% (P=0.001) (Table). In men there was no significant reclassification improvement. Net reclassification Lifetime SCD Risk: Clinical + ECG + GEH Variables Women Men <5% 5–15% >15% Total <5% 5–15% >15% Total SCD Cases <5% 82 14 0 96 103 16 0 119 5–15% 7 59 10 76 12 116 12 140 >15% 0 0 20 20 0 5 74 79 Lifetime SCD Risk: Total 89 73 30 192 115 137 86 338 Clinical Variables Only Non-Cases <5% 6,956 131 2 7,089 4,411 264 0 4,675 5–15% 180 509 42 731 210 1,059 48 1,317 >15% 0 28 84 112 0 56 214 270 Total 7,136 668 128 7,932 4,621 1,379 262 6,262 Conclusions We were the first to develop sex-specific lifetime SCD prediction models. The addition of ECG GEH to clinical variables improved SCD risk reclassification in women, but not in men. Prediction of SCD was more accurate in women as compared to men.

Download Full-text