Article Text

Download PDFPDF

Odd couple: premature ventricular contractions and heart failure
  1. Marc W. Deyell,
  2. Nathaniel M Hawkins
  1. Heart Rhythm Services, Division of Cardiology, The University of British Columbia, Vancouver, British Columbia, Canada
  1. Correspondence to Dr Marc W. Deyell, Heart Rhythm Services, Division of Cardiology, The University of British Columbia, Vancouver, V6E 1M7, Canada; mdeyell{at}

Statistics from

The interplay between frequent premature ventricular complexes (PVCs) and the development of left ventricular systolic dysfunction (LVSD) and heart failure (HF) is not a simple one. The first described relationship between frequent PVCs and LV dysfunction was in 1998,1 but importantly this was a description of patients with frequent PVCs and established LVSD, that improved with suppression of PVCs using amiodarone. On the surface, the concept was simple: a high burden of PVCs leads to (or at least contributes to) LVSD and HF. However, the more we learn about PVCs and HF, the more complex the relationship becomes, further highlighted in the current study from Limpitikul and colleagues.2

The study is an analysis of patients with available raw Holter data nested within the prospective Cardiovascular Health Study (CHS). From the original randomly selected 5201 adults ≥65 years recruited in 1989–1990, 1429 were randomly selected to undergo 24 Holter monitor. After exclusions, 871 participants were available, of whom 316 had ≥10 PVCs on monitoring and form the basis of analysis. A further subgroup of 209 patients underwent both a baseline and 5-year echocardiographic evaluation. The association between PVC characteristics and cardiovascular outcomes was examined over a median follow-up of 11 years.

An uncertain epidemiology and pathophysiology

The current study, in combination with the group’s previous work,3 raises fundamental questions about the association between PVCs and HF. The first key observation is that even a relatively low PVC burden is associated with a higher incidence of LVSD, HF and death. The gradient of risk with PVC burden occurred at levels far below (<1% PVC burden) our notion of how PVCs might cause LV dysfunction and HF. The magnitude of risk is startling, with a population attributable risk of PVCs for incident HF of 8% (HR of 2.04 for every 0.1% increase PVC burden), similar to that of conventional risk factors such as hypertension and coronary artery disease.3 Yet, the outstanding animal work in this arena has shown that at least 20%–25% PVCs are required to induce LVSD over a period of roughly 3 months.4 While these animal models represent an ‘accelerated’ form of PVC-induced LVSD, there is a clear disconnect from the association between very low levels of PVCs in humans and the development of LV dysfunction and HF.

From characterisation to heterogeneity in PVC assessment

Numerous features of PVCs have been proposed as markers of risk for cardiomyopathy and other adverse outcomes, each representing one or more potential interacting pathophysiological mechanisms leading to adverse remodelling. These can broadly be considered as measures of intrinsic PVC characteristics, variability and exposure. They include PVC duration, morphology and location (causing intra-LV and interventricular dyssynchrony with associated haemodynamic and valvular disturbance), timing and irregularity (causing atrioventricular dyssynchrony, impairing ventricular filling, diastolic mitral and tricuspid valvular regurgitation, and altered calcium handling), and overall burden/duration.

One of the primary measures of PVC variability is the coupling interval. Limpitikul and colleagues focus on coupling interval heterogeneity, defined as the SD of the coupling interval divided by the mean.2 This effectively corrects for the underlying heart rate, conceptually similar to the coefficient of variation in statistics, a standardised measure of dispersion of a frequency distribution.

The pathophysiological rationale for this measure is important and is twofold. First, heterogeneity in coupling interval may have more profound impacts on myocardial dynamics and consequent neurohormonal changes than PVCs at a fixed coupling interval.5 Second, while coupling interval variations have minor influences such as variation of the preceding cycle length, the dominant determinant is thought to be the underlying arrhythmic mechanism. Fixed and low coupling interval variation is more indicative of re-entrant or triggered mechanisms for PVCs, that are associated with a higher risk of sudden death in structural heart disease.6 Conversely, high variability is more indicative of automaticity, and more specifically modulated parasystole, as the mechanism.7 High variability may reflect PVCs coming from subclinical myocardial disease, as myocyte uncoupling is necessary for the development of parasystole. Consequently, high variability in coupling interval is more predictive of HF rather than ventricular tachyarrhythmias.

In the large, well-characterised, community-based cohort of the CHS, the current study found a strong and independent (adjusting for PVC burden) association between variation in the PVC coupling interval and decline in LV function and incident HF (adjusted HR of incident HF 2.98 (95% CI 1.13 to 7.84) for high coupling interval heterogeneity).2 The study adjusted for conventional risk factors for HF (age, sex, race, body mass index, diabetes, hypertension, smoking and myocardial infarction) and multiple other PVC measurements including PVC frequency, duration, coupling interval and coupling interval heterogeneity. Only coupling interval heterogeneity predicted decline in left ventricular ejection fraction, while both coupling interval heterogeneity and PVC burden predicted incident HF. Crucially again, a low PVC burden had a strong association with clinical outcomes (0.1% increase in baseline PVC frequency was associated with a more than twofold increased risk of incident HF).

A word of caution

The current study2 and the previous work by Marcus and colleagues3 have leveraged the data from the CHS, with its population-based design, rigorous data collection and a median follow-up period of over 11 years. Yet, despite the rich nature of the data from the CHS, it was not designed specifically to look at the impact of PVCs on cardiovascular health outcomes. Notably, ambulatory ECG was only performed at baseline, without repeated measures during follow-up. Our group has previously demonstrated a high rate of spontaneous resolution of idiopathic frequent PVCs over a 5-year time period,8 highlighting the dynamic nature of PVCs over time. Thus, an association between PVC characteristics at one point in time and subsequent health outcomes over the next 10 years should be treated with caution. As acknowledged by the authors, the current study could not account for all potential confounding factors that may have influenced the association between PVC characteristics and LV dysfunction/HF. The sample size for the echocardiographic subgroup was limited, reflected in the wide CI of the adjusted OR (aOR) for LVEF reduction (aOR 15.1 (1.9 to 118.7)).

The road ahead

Trying to synthesise existing knowledge of the relationship between PVCs and the development of HF is daunting. Quite possibly, PVCs are an epiphenomenon in certain patients and causative of LV dysfunction in others (figure 1). In broader populations with lower levels of PVCs, PVCs may be a marker of subclinical myocardial disease. However, in more select populations with very frequent PVCs, the link between PVC characteristics and adverse outcomes is more direct.

Figure 1

Conceptualisation of the relationship between PVCs and outcomes. CAD, coronary artery disease; HTN, hypertension; LV, left ventricle/ventricular; PVC, premature ventricular complex.

As a cardiology and electrophysiology community, we have done the easy work. We have demonstrated, in highly selected populations with very frequent PVCs and LV dysfunction, that contemporary treatment of PVCs can improve outcomes. However, we have only just begun to tackle the more important fundamental work of how to manage patients who have frequent PVCs with normal cardiac function. Before we embark on randomised trials of therapy, we first need to add to our fundamental epidemiological knowledge with well-designed longitudinal studies of frequent PVCs. Critically, we need to progress from highly selected populations (those who filter to tertiary academic centres) to broader population cohorts. We need to have rigorous, structured follow-up to ascertain which patients are at risk of adverse outcomes with PVCs. Although cardiologists (the authors included) are quick to be distracted by advances in therapy, I have no doubt that, as a community, we can roll up our sleeves and put on our epidemiological hats to shed more light on the enigma of frequent PVCs.

Ethics statements

Patient consent for publication



  • Twitter @MarcDeyell

  • Contributors MWD was responsible for the initial draft of the editorial. MWD and NMH each provided critical revisions of the manuscript and each approved the final manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles