In contemporary healthcare, effective risk stratification in the general population is vital amidst rising chronic disease rates and an ageing demographic. Deceleration Capacity of the heart rate (DC), derived from 24-hour Holter electrocardiograms, holds promise in risk stratification for cardiac patients. However, the potential of short-term electrocardiograms of five minutes duration for population screening has not been fully explored. Our study aims to investigate the utility of Deceleration Capacity derived from short-term electrocardiograms as a scalable, fully-automated screening tool for predicting long-term mortality risk in the general population. Within a cohort of a representative population-based survey in Germany (KORA-KMC-study), 823 participants with sinus rhythm aged 27 to 76 years at enrollment (females 47.4%) were followed for a median of 13.4 years (IQR 13.1-13.6). All-cause mortality was defined as the primary endpoint and observed in 159 participants. Deceleration Capacity was calculated from 5-minute 12-lead electrocardiograms by a fully automated approach. Participants were divided into three predefined risk categories: DCcategory0 - low-risk (> 4.5ms); DCcategory1 - intermediate-risk (2.5-4.5ms); and DCcategory2 - high-risk (≤ 2.5ms). More than two-thirds of the participants (n = 564, 68.5%) fell into DCcategory0, about one-fifth (n = 168, 20.4%) into DCcategory1, and about one-tenth (n = 91, 11.1%) into DCcategory2. Estimated 13-years mortality in the risk groups was 16.7%, 23.5%, and 49.1%, respectively (p < 0.001). Adjusting for age, life-style-related risk factors, and comorbidities, increased mortality risk was observed for DCcategory2 (HR 2.34, 95%-CI 1.56-3.50). Deceleration Capacity, derived automatically from brief 5-minute electrocardiogram recordings, emerges as a robust, feasible, and independent predictor of long-term mortality risk in the general population.
Keywords: Autonomic regulation; Deceleration capacity of the heart rate; Electrocardiogram; Fully automated biosignal analysis; General population screening; Non-invasive long-term risk stratification.
© 2024. The Author(s).