Abstract
BACKGROUND: To date, prevalence and clinical significance of noninvasive arrhythmia risk predictors in apparently healthy, middle-aged persons are largely unknown. METHODS: A total of 110 apparently healthy persons 20-75 years old were enrolled in this prospective observational monocenter study and followed up for 32 +/- 15 months. Baseline investigations included symptom-limited bicycle ergometry, echocardiography, time-domain analysis, and spectral turbulence analysis of the signal-averaged electrocardiogram (ECG), ventricular arrhythmias, and heart rate variability on 24-hour Holter ECG, baroreflex sensitivity, and t-wave alternans in all persons. RESULTS: The prevalence of an abnormal signal-averaged ECG was 1% for spectral turbulence analysis and varied between 1% and 37% for time-domain analysis depending upon the definition used for an abnormal time-domain analysis. A reduced heart rate variability defined as a standard deviation of normal-to-normal intervals < or =105 ms, <100 ms and <70 ms was found in 12%, 9%, and 1% of persons. A baroreflex sensitivity <6 ms/mmHg and <3 ms/mmHg was present in 15% and 2% of persons. Microvolt t-wave alternans was found to be positive in 5%, negative in 88%, and indeterminate in 7% of persons, respectively. During the 32 +/- 15 months follow-up, no arrhythmic events and no cardiovascular mortality were observed in this population. CONCLUSIONS: Abnormal findings of noninvasive arrhythmia risk stratification can be found in 1-37% of healthy, middle-aged persons when previously reported cut-off values are used.