Risk Prediction After Myocardial Infarction by Cyclic Variation of Heart Rate, a Surrogate of Sleep-Disordered Breathing Assessed From Holter ECGs.

Xu Cao,Petra Barthel,Alexander Müller,Daniel Sinnecker,Karl-Ludwig Laugwitz,Michael Dommasch,Georg Schmidt,Alexander Steger,Ralf J Dirschinger

doi:10.3389/fphys.2019.01570

Abstract

AimsSleep-disordered breathing (SDB) is common among cardiac patients, but its role as an independent risk predictor after myocardial infarction (MI) is unclear. SDB causes cyclic variation of heart rate (CVHR). The aim of this study was to score Holter ECGs of a large cohort of MI survivors for SDB-related CVHR to investigate its value for mortality prediction.MethodsA total of 1590 survivors of acute MI in sinus rhythm were prospectively enrolled and followed for 5-year all-cause mortality. Heart rate (HR) tachograms were generated from nocturnal (00:00–06.00 am) segments of Holter ECGs, and the minutes with CVHR were quantified by a previously developed algorithm. According to a pre-specified cutpoint, SDB was assumed if CVHR was present during ≥72 min.ResultsSeventy-seven patients (4.8%) had flat HR tachograms which prohibited analysis for SDB. Of the remaining 1513 patients, 584 (38.6%) were classified as having SDB. Mortality rates in groups stratified according to ECG-derived SDB did not differ significantly. Taken as a continuous variable, low CVHR duration was associated with increased mortality.The mortality of patients with flat HR tachograms was significantly increased, even after adjustment for age, sex, LVEF, GRACE score and diabetes mellitus. Mortality prediction by a flat HR tachogram was also independent of heart rate variability (HRV), heart rate turbulence (HRT), and deceleration capacity (DC).ConclusionIn Holter ECG recordings of survivors of acute MI, signs suggestive of SDB were frequently present, but not associated with mortality. A flat nocturnal HR tachogram was a strong, independent predictor of 5-year all-cause mortality.

Highlights

MATERIALS AND METHODSPatients who have survived the acute phase of an acute myocardial infarction (MI) are at increased risk of subsequent mortality within the years, which may be due to reinfraction, arrhythmias, or progressive heart failure, and related to co-morbidities
Several parameters related to respiration such as the respiratory rate (Barthel et al, 2013) [which can be measured from Holter ECG recordings as the nocturnal respiratory rate (Dommasch et al, 2014; Sinnecker et al, 2014)] or respiratory sinus arrhythmia (Sinnecker et al, 2016) have been demonstrated to be strong predictors of the mortality risk of MI survivors
It has been recognized that Sleep-disordered breathing (SDB) is accompanied by a typical pattern of heart rate decelerations followed by accelerations which has been termed cyclic variation of heart rate (CVHR), (Guilleminault et al, 1984) and various algorithms have been proposed to assess the likelihood of SDB from standard Holter ECG recordings (Mendonca et al, 2018)

Summary

Introduction

MATERIALS AND METHODSPatients who have survived the acute phase of an acute myocardial infarction (MI) are at increased risk of subsequent mortality within the years, which may be due to reinfraction, arrhythmias, or progressive heart failure, and related to co-morbidities. Several parameters related to respiration such as the respiratory rate (Barthel et al, 2013) [which can be measured from Holter ECG recordings as the nocturnal respiratory rate (Dommasch et al, 2014; Sinnecker et al, 2014)] or respiratory sinus arrhythmia (Sinnecker et al, 2016) have been demonstrated to be strong predictors of the mortality risk of MI survivors. It has been recognized that SDB is accompanied by a typical pattern of heart rate decelerations (during apnea episodes) followed by accelerations (during arousals) which has been termed cyclic variation of heart rate (CVHR), (Guilleminault et al, 1984) and various algorithms have been proposed to assess the likelihood of SDB from standard Holter ECG recordings (Mendonca et al, 2018). A simple method to manually score RR interval tachograms generated from ECG recordings obtained during polysomnography for the presence of CVHR (Stein et al, 2003) was able to predict the presence or absence of SDB (as assessed by standard polysomnography) with a positive and negative predictive accuracy of 86 and 100%, respectively (Stein et al, 2003)

Results

Discussion

Conclusion