Energy waveform (ew) ECG uses continuous wavelet transforms (CWT) to create time-frequency energy characterizations. Although ewECG changes reflect myocardial disease, optimal features for prediction of subclinical left ventricular dysfunction (Stage B heart failure [SBHF]) are unclear. Using machine learning (ML), we investigated whether ewECG could predict SBHF in at-risk community subjects to establish it's potential role in screening.