Author + information
- Published online May 29, 2017.
- Philip F. Binkley, MD, MPH∗ ()
- ↵∗Address for correspondence:
Dr. Philip F. Binkley, Department of Medicine and Cardiology, Heart and Lung Institute, The Ohio State University, 473 West 12th Avenue, Columbus, Ohio 43210-1252.
In this issue of JACC: Heart Failure, Patel et al. (1) report findings that have the potential to markedly advance the clinical utility of measures of heart rate variability (HRV) in addition to providing important insight into the pathogenetic mechanisms governing the early stages of heart failure.
Measurement of HRV has been 1 of the fundamental techniques used to assess autonomic function in patients with heart failure, in addition to methods such as microneurography and measurement of norepinephrine spillover. Time and frequency domain HRV analyses have defined the aberrations in autonomic control that are operative in the progression of heart failure and increased risk for sudden cardiac death. A few studies with small patient cohorts have demonstrated HRV measures that can serve as markers for arrhythmias and adverse clinical outcomes (Online Appendix). However, for the most part, these measures have remained important tools for the mechanistic investigation of heart failure with limited clinical application.
Patel et al. (1) have demonstrated in a retrospective analysis of patients enrolled in the CHS (Cardiovascular Health Study) that abnormal HRV is associated with the onset of heart failure years in advance of the first clinical event. A range of measures of HRV previously shown to reflect autonomic imbalance in patients with existing heart failure were found to significantly differ between asymptomatic patients who ultimately did and did not develop heart failure symptoms when evaluated individually in univariable models. These data support that changes in autonomic function predate overt heart failure, and are consistent with studies that show that autonomic imbalance is present in early stages of left ventricular (LV) dysfunction (2). As such, it is perhaps not surprising that these abnormalities may have a clinical role as important predictors of incident heart failure.
Only a subset of measures of HRV remained significantly associated with incident heart failure in the context of clinical variables that constitute the Health ABC predictive model (3). These HRV measures may contain unique information regarding autonomic function and control that are not determined by the clinical condition of the patient before the development of ventricular dysfunction. As such, they potentially constitute a new echelon in the hierarchy of risks for heart failure and may reflect more primary mechanistic determinants of disease.
Some of the measures independently associated with incident heart failure have been referred to as novel additions to the scope of HRV analytic methods (4). Heart rate turbulence is a measure of changes in heart rate following a premature ventricular contraction. The normal multiphasic response is reminiscent of that following the Valsalva maneuver, which indeed was used in the past as a bedside indicator of heart failure (Online Appendix). Heart rate turbulence appears in part to be governed baroreflex activity, and abnormal responses may be secondary to blunting of this regulatory system, which is known to have a major effect on survival in patients with acute coronary events (Online Appendix). The association of heart rate turbulence with future incident heart failure suggests that impaired baroreflex responsiveness may set the stage for the evolution of autonomic imbalance in patients with heart failure.
A second novel measure is the detrended frequency analysis, which is a short-term fractal scaling exponent calculated over a small number of interbeat intervals (5). It has been shown to be associated with mortality and adverse outcomes in patients with heart failure (Online Appendix). As a nonlinear dynamic of HRV, it likely reflects abnormalities in heart rate control that are not captured in time or frequency domain measures. It is speculated that it reflects the fundamental organizational architecture governing heart rate variability, and similar to heart rate turbulence, may represent a higher echelon of both cardiovascular control and long-term antecedent risk for clinical heart failure.
As Patel et al. (1) note, very low-frequency HRV was in fact increased in patients who ultimately developed heart failure. It is important to recognize that increased HRV does not always reflect a healthy condition. This has been well demonstrated in the realm of intrauterine fetal monitoring, in which increases in HRV may in fact indicate what was formerly termed “fetal distress” (Online Appendix). Rapid variations in heart rate that are atypical for patients with heart failure have been shown to accompany the fluctuations in blood pressure encountered in pulsus alternans (6). In this case, increased HRV does not reflect normal autonomic balance. It is speculated that this atypical increase in high-frequency HRV is due to a “low gain” feedback control system that fails to contain fluctuations in physiological signals. Therefore, the increase in very low-frequency HRV in those who ultimately developed heart failure further suggests that defective regulatory mechanisms predate and promote the evolution of the autonomic imbalance that is reflected by changes in HRV.
It is not certain whether the patients with abnormal HRV reported in this study in fact had asymptomatic LV dysfunction. Abnormalities in HRV were not seen in the subset of patients with normal measures of brain natriuretic peptide. It is known that asymptomatic LV dysfunction may predate clinical heart failure for a matter of years, and in some cases, symptomatic heart failure may never occur (Online Appendix). This fact does not mitigate the potential clinical importance of the findings of the study, because identifying patients who are asymptomatic despite impairment of ventricular performance signals the need for disease monitoring and possible implementation of preventive strategies.
Ejection fractions at the time of onset of heart failure symptoms were not consistently available in the CHS study from which the current analysis was derived (3). In those for whom measures were available, the mean ejection fraction was 43.4 ± 16.1%. Therefore, we do not know what proportion of participants who developed heart failure and who underwent Holter monitoring had heart failure with reduced ejection fraction versus preserved ejection fraction. In future studies, it would be of interest to determine whether different measures of HRV that antedated overt heart failure symptoms segregate with heart failure with either reduced or preserved ejection fraction or are, in fact, common to both. Such an analysis may provide further mechanistic information regarding the differential evolution of these 2 forms of heart failure. Practically, it is important that measures of HRV may indicate risk for the development of symptomatic heart failure of any form.
As the authors note, this is a retrospective analysis. Accordingly, it can only be said that the HRV measures they identify are associated with future incident heart failure, but are not truly predictive. Statistical models are dependent on the population from which they are derived, which demands evaluation of their generalizability. For instance, the current model was based on a sample of older participants, and it is both mechanistically and clinically important to establish whether the HRV measures associated with incident heart failure are predictive in younger populations. Future prospective analyses will be required to determine whether these measures truly predict future clinical heart failure events and in which populations. Alternatively, an existing large dataset, such as the one used for this analysis, can be divided into a “training set” for model development and a “validation set” in which the subsequent model can then be “prospectively” evaluated in the remaining subjects. The promising findings of this investigation should motivate such analyses to allow HRV measures to be applied as a useful predictive clinical tool.
As noted, the work by Patel et al. (1) provides both new mechanistic and clinical insights into the evolution of heart failure. It suggests a model of autonomic pathogenesis in which higher-order control systems and the organizational architecture of heart rate control are themselves disturbed in advance of overt changes in autonomic function as reflected by heart rate variability. Accordingly, these systems are not capable of an adaptive response to an insult to the cardiovascular system, which then elicits the well-known abnormalities of autonomic function characteristic of patients with heart failure. This raises the question as to what promotes the early disturbance of these primary control mechanisms.
The potential clinical application of these HRV measures raises intriguing questions. If indeed they are found to be predictive of the risk for development of heart failure, what is the clinical response? Other than treating underlying disease processes, there are not clear evidence-based preventive strategies for those known to be at risk for heart failure. Would frequent monitoring by either biomarkers or imaging be warranted? Would pharmacological interventions, which have been shown to help restore or maintain autonomic balance, be of benefit in preventing the occurrence of heart failure in these patients (Online Appendix)? Would these patients now be considered stage B (patients with existing structural heart disease at risk for heart failure) rather than stage A?
Significant research findings, such as those reported by Patel et al. (1), motivate new generations of investigative questions. Their work should drive future research that will further explore the use of measures of HRV and establish their role as a clinical as well as investigative tool.
For supplemental references, please see the online version of this article.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
Dr. Binkley has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Patel V.N.,
- Pierce B.R.,
- Bodapati R.K.,
- Brown D.L.,
- Ives D.G.,
- Stein P.K.
- Eaton G.M.,
- Cody R.J.,
- Nunziata E.,
- Binkley P.F.
- Kalogeropoulos A.,
- Psaty B.M.,
- Vasan R.S.,
- et al.
- Stein P.K.,
- Barzilay J.I.,
- Chaves P.H.,
- et al.