Background
First-degree heart block, referred to as prolongation of PR interval beyond >200 ms on the electrocardiogram (ECG), has long been considered a benign functional variation [
1]. Nevertheless, recent major studies including the ARIC [
2] and Framingham Heart Study [
3], re-ignited controversies over the pathological nature of PR prolongation by consistently demonstrating that PR prolongation was linked to risk of atrial fibrillation in the population. In the Framingham study, PR prolongation was further associated with increased risk of pacemaker implantation and all-cause death [
3]. Nevertheless, the roles of PR prolongation in atherosclerotic cardiovascular (CV) disease, as well as outcome prediction in CV patients under clinical care, remained largely unknown. The Heart and Soul Study showed that PR prolongation predicted heart failure hospitalization and CV mortality in patients with stable coronary artery disease (CAD) [
4]. We further showed that PR prolongation when incorporated into the CHADS2 and CHA2DS2-VASc scores, augments their predictive power for new-onset CV events [
5]. Nevertheless, the individual role of PR prolongation, a precursor to atrial fibrillation, for new-onset ischemic events of stroke and myocardial infarction was not clearly established. Importantly, underlying pathophysiological mechanistic basis of how PR prolongation may lead to adverse CV outcomes remained also poorly understood.
In a prior study of 88 healthy subjects, PR prolongation was found independently correlated with endothelial dysfunction and increased pulse-wave velocity even in the absence of clinically manifest atherosclerotic disease [
6], suggesting that adverse vascular function could be an intermediate pathological phenotype in subjects with PR prolongation progressing towards clinical events. As endothelial dysfunction indicates widespread functional adversity of the CV system as a collective, it will be of interest to dissect the role of PR prolongation in closely related CV pathological entities commonly characterized by abnormal vascular function, including myocardial infarction (MI) [
7], ischemic stroke [
8], and congestive heart failure (CHF) [
9].
Therefore, we investigated in this prospective cohort study the role of PR prolongation as an independent predictor in a continuum of pathophysiologically-linked clinical events including ischemic stroke, MI, CHF, and CV death in high-risk CV patients, and the relationships between PR prolongation and indicator of abnormal vascular function.
Discussion
This is our first recognized study to show that PR prolongation alone is a strong and independent predictor for CV death, new-onset ischemic stroke and MI, and combined CV endpoints including CHF among patients with CAD or risk equivalent. We also assessed the vascular phenotype of these high-risk CV patients and found that PR prolongation was independently related to increased carotid IMT. Interestingly, despite the stated observations, adjustment for carotid IMT did not materially alter the prediction of PR prolongation for incident CV events, which suggests that atherosclerosis alone does not fully explain the pathophysiological mechanism along the pathway of PR prolongation leading to clinical events.
The clinical implications of this study are multi-fold. Firstly, as opposed to the traditional school of thought that considered PR prolongation as a benign ECG feature [
1], this study robustly showed that PR prolongation is strongly predictive of a collective plethora of pathophysiologically related CV outcomes, including MI, ischemic stroke, CHF and CV death. Recent epidemiological studies (ARIC (2) and Framingham [
3]) found higher risk of atrial fibrillation associated with PR prolongation. Our findings suggest that the pathophysiological implications of PR prolongation is far more than atrial fibrillation, and entails MI, ischemic stroke and CHF, all of which could be commonly characterized by impaired systemic vascular function [
7‐
9]. PR prolongation alone, regardless of other pertinent CV risk factors [
12] such as those delineated in the CHADS2 scores [
5], is a simple and independent CV risk predictor. Secondly, the magnitude of excess risk of PR prolongation was substantial for various adverse outcomes consistently (new-onset ischemic stroke: 4-fold excess risk; new-onset MI, 7-fold excess risk; CV death: 15-fold excess risk), thus reflecting PR prolongation has strong utility for CV prediction among high-risk patients. Thirdly, while the recent Heart and Soul Study [
4] showed that PR prolongation predicted heart failure hospitalization and CV death amongst patients with stable CAD, our study further showed that PR prolongation has strong predictive applicability for CV events in a much wider spectrum of high-risk patients with prior CAD or other risk equivalents, including diabetes and stroke. Fourthly, the escalated risk of new-onset MI emerged at a lower PR interval cut-off of >162 ms, consistent with our prior exploratory study in terms of subclinical changes in endothelial dysfunction [
6], and calls into question that early pathological changes may well precede the conventional cut-off of PR prolongation at >200 ms.
Mechanisms underlying our clinical observation are likely multi-dimensional. Firstly, PR prolongation representing delayed atrioventricular conduction could be an ECG manifestation of CV ageing [
13]. Atherosclerosis is increasingly recognized as a multi-factorial degenerative disease that is closely related to the aging process. Our analysis indicated that PR prolongation could be an excellent clinical marker of CV ageing since adjustment for chronological age did not obliterate the prediction of carotid IMT and CV events by PR prolongation. Secondly, myocardial fibrosis may involve the cardiac conduction system and is closely linked to the presence of CV risk factors including CAD [
14], hypertension [
15], diabetes [
16], hyperlipidemia [
17] and inflammation [
18]. Thus PR prolongation may be a risk marker that reflects the clustering of these important risk factors. Nevertheless, PR prolongation may also reflect higher underlying vagal tone [
19], especially in the young [
1], other than degenerative damage of the cardiac conduction system which predominantly affects the elderly [
13]. In our study, comprehensive adjustment for important CV risk factors, even when including widened QRS complex signifying overt ventricular electrical dyssynchrony or widespread conduction system disease, did not abolish PR prolongation as an independent CV predictor. Therefore a third potential explanation arises: that PR prolongation may be causal to adverse CV events. Indeed, PR prolongation, be it vagal- or degenerative-predominant, may mediate its adverse vascular effects via increased intra-atrial pressure consequential to slowed atrioventricular conduction, resulting in neurohormonal activation [
20]. Adverse neurohormonal changes including raised aldosterone may be reversible on sinus rhythm reversion [
21]. Aldosterone is associated with a constellation of pro-atherosclerotic changes including raised inflammatory response [
22], reduced circulating endothelial progenitor cell quantity and function [
23] resulting in increased arterial stiffness [
24]. Importantly, mineralocorticoid receptor blockade has been shown clinically to improve atherosclerotic changes [
25]. Indeed, if neurohormonal activation is the explanatory mechanism that underlies PR prolongation-related vascular dysfunction, such findings may translate into potentially important preventive implications through pharmacological modulation of the neurohormonal system.
Vascular function as indicated by carotid IMT is a strong, independent and consistent predictor of both incident ischemic stroke and CV events overall among healthy subjects from diverse ethnicities [
26]. Increased carotid IMT also predicts elevated risks of CV complications in patients with diabetes [
27], hypertension [
28] and prior atherosclerotic diseases [
29]. In this study, the magnitude of increased carotid IMT by +0.07 mm accounted for by PR prolongation is clinically significant, since each 0.1 mm increase of carotid IMT could translate into 18% excess risk of recurrent CV events [
30]. On top of their greater susceptibility to CV events in general, patients with prior CAD also have increased risk of incident ischemic stroke, with diabetes being an additional independent predictor [
31,
32]. Our study further suggests that the adverse prognostic value of PR prolongation is independent and over that of carotid IMT for CV events in high-risk patients. Whether impaired vascular function is indeed along the pathophysiological pathway or simply a bystander will require further mechanistic studies. Future studies could be strengthened by incorporating multiple vascular assessment modalities to ensure consistent findings and scrutinise further mechanisitc insights.
It should be noted that from our study, the magnitude of risk associations of PR interval prolongation with new-onset CV events was striking, and that IMT did not fully explain such associations, suggesting that alternative mediating pathophysiological mechanisms could be present. Also, the HR estimates are relatively large. Despite our meticulous efforts in the study design and data analysis to carefully minimize biases or residual confounding, potential risk of bias and confounding cannot be excluded totally. Findings from the exploratory endpoint analyses should also best be verified in another independent study. Furthermore, the study of incident atrial fibrillation was not included in the current study, which should be explored in the future.