Keeping to the rhythm of cardiovascular health

This editorial refers to ‘Role of ideal cardiovascular health metrics in reducing risk of incident arrhythmias’, by Y.-J. Cheng et al., https://doi.org/10.1093/eurjpc/zwad357.

Cardiovascular disease is the leading cause of excess deaths worldwide. As a result, there has been recent emphasis on the optimization of cardiovascular health (CVH) metrics as a means to reduce the public health burden of atherosclerotic cardiovascular disease and heart failure. Comparably little attention has been paid to cardiac rhythm disorders, which are associated with substantial morbidity and mortality. Studies from the UK Biobank, a prospective cohort of over 500 000 participants, have shown that prevalent arrhythmias are found in ∼1% of the population under 55 years of age and in ∼5% of those aged 65 years or older. Atrial fibrillation (AF) is by far the most common arrhythmia and occurs with an incidence rate of 3.1 per 1000 person-years, while supraventricular arrhythmias (1.1 per 1000 person-years) and ventricular arrhythmias (0.5 per 1000 person-years) are relatively less common.¹

There is solid rationale to support the notion that the optimization of CVH metrics may have beneficial effects on arrhythmia risk. The best studied example is AF, where cardiometabolic factors such as hypertension and obesity appear to contribute to atrial fibrosis, left atrial enlargement, and changes to the conduction system that predispose to the initiation and maintenance of arrhythmia.² Furthermore, improvement in cardiometabolic parameters appears to have beneficial effects on AF incidence and burden. For example, people with AF who achieve weight loss of >10% have been shown to have decreases in left atrial volume and AF burden, while physical activity (PA) has been associated with incident AF in an almost dose-dependent manner.^3–5 Consequently, the recently published 2023 ACC/AHA/ACCP/HRS guidelines for the diagnosis and management of AF categorize the optimization of risk factors such as obesity and PA as pillars of AF treatment.⁶

But what about other important rhythm disorders such as ventricular arrhythmias or bradyarrhythmias? And rather than investigating individual risk factors, can we gain insight into the effects of CVH holistically using a composite guideline-supported framework, such as the AHA-endorsed CVH metric, sometimes called ‘Life’s Simple 7’?⁷

In this issue of the European Journal of Preventive Cardiology, Cheng et al.⁸ investigated the association of CVH with incident arrhythmias in the population-based Atherosclerosis Risk in Communities (ARIC) study, a community cohort recruiting over 15 000 people from four communities in the USA between 1987 and 1989. Individuals underwent standardized evaluation of cardiovascular risk factors, and returned for four follow-up visits. Cardiovascular health was defined systematically using Life’s Simple 7, which comprises seven domains of CVH. The seven domains comprise four health behaviours: dietary quality, PA, smoking, and weight, and three health measures: fasting blood glucose, total cholesterol, and blood pressure (BP). As recommended by the AHA, each domain was categorized as poor, intermediate, or ideal using previously defined thresholds.⁷ Arrhythmia outcomes were classified into three main groups: AF, ventricular arrhythmias, and bradyarrhythmias, and then ascertained using electrocardiograms performed at study visits, hospitalization records, and death certificates.

After excluding individuals with prevalent rhythm disorders, they analysed 13 078 participants for associations between CVH and incident AF, ventricular arrhythmias, and bradyarrhythmias, using separate Cox regression models adjusted for multiple baseline clinical factors. Over a median of 23 years of follow-up, there were a total of 2548 incident AF, 1363 incident ventricular arrhythmia, and 706 incident bradyarrhythmia events. Compared with people with poor health, those with ideal health were substantially less likely to develop AF, ventricular arrhythmias, and bradyarrhythmias, with absolute rate differences per 1000 person-years of −5.2 [95% confidence interval (CI): −6.3 to −4.0], −5.0 (−5.7 to −4.4), and −1.7 (−2.5 to −0.9), respectively. Multivariable-adjusted population attributable fractions (i.e. estimates of the proportion of disease attributable to non-ideal CVH) were estimated at 29.9% (95% CI 18.6–39.5%) for AF, 54.4% (39.4–50.4%) for ventricular arrhythmias, and 21.9% (41.8–42.0%) for bradyarrhythmias.

Overall, this epidemiologic study by Cheng et al. substantially advances our understanding of the potential broad-ranging effects of composite CVH on the risk of arrhythmias, with important implications for future efforts to reduce the public health burden associated with cardiac rhythm disorders. Although the key finding is clearly that the optimization of cardiometabolic risk factors and achievement of a healthy lifestyle appear critically important for lowering the risk of all incident arrhythmias, we would like to highlight several additional observations.

First, it seems that not all health metrics and risk factors are created equal with respect to arrhythmia risk. For instance, compared with having a given metric in the poor range, having an ideal measure for BMI [hazard ratio (HR) range 0.60–0.68], glucose (HR 0.55–0.66), and BP (HR 0.55–0.89) each seemed to result in particularly strong beneficial effects on incident arrhythmia risk. However, when compared with the intermediate state, effect sizes for ideal metrics were much more modest for risk factors other than BP, which retained a large effect (HR 0.68–0.75). Therefore, for many metrics, it may be that much of the benefit is realized by avoiding a ‘poor’ value, while for others such as BP, there remains substantial benefit in the achievement of an ideal state. The importance of achieving optimal BP control for arrhythmia risk is consistent with observations for general cardiovascular risk from randomized studies such as the SPRINT trial.⁹ Overall, given the challenges faced by public health interventions designed to achieve truly ideal CVH metrics, population-level strategies prioritizing a shift from the poor to at least the intermediate state across several metrics, rather than achievement of truly ideal values, may maximize utility for arrhythmia risk, although such approaches require prospective study.

Second, this study provides a reminder that the relationship between CVH and risk of different arrhythmias is likely driven by distinct underlying mechanisms (Figure 1). In the Kaplan–Meier curves provided, differences in the cumulative risk of arrhythmia among people with poor, intermediate, or ideal health seem to emerge after varying degrees of latency. In AF, the curves start to diverge at around 5 years, while for bradyarrhythmias, the divergence is not apparent until ∼10 years of follow-up. For ventricular arrhythmias, those with poor health separate from those with intermediate or ideal health almost immediately. Such findings may indicate different drivers, whether directly causal or not, that are associated with health status and risk of arrhythmia. We can assume a strong component of ischaemic heart disease in increasing the risk for ventricular arrhythmias, whereas the contribution of ischaemia to AF risk may be more limited.¹⁰ Likewise, autonomic function, inflammation, and cardiac remodelling may each associate with risk of incident AF, ventricular arrhythmias, and bradyarrhythmias to varying degrees (Figure 1).

Figure 1.

Potential relationships between cardiovascular health and incident arrhythmias. Overview of the association of American Heart Association cardiovascular health metrics with incident arrhythmias. The ‘Life’s Simple 7’ metric comprises seven domains, including four health behaviours: dietary quality, physical activity, smoking, and body weight; and three health measures: fasting blood glucose, total cholesterol, and blood pressure. The attainment of ideal health is associated with lower risk for atrial fibrillation, bradyarrhythmias, and ventricular arrhythmias. Possible mechanisms include ischaemia, cardiac structural remodelling, autonomic dysfunction, inflammation, and fibrosis. The relative contribution of these mechanisms to risk of each respective arrhythmia is not known but likely to vary across conditions.

Strengths of this study include analysis of a well-phenotyped cohort representative of the US populations from which they were sampled, with reasonable diversity in demographics and clinical risk factors, and with sufficient events and follow-up to facilitate a robust analysis. Nevertheless, this is a retrospective observational study from which no causal relation between risk factors and disease can be inferred. For example, although the findings suggest that the optimization of CVH metrics will lead to a lower burden of arrhythmias, this hypothesis cannot be confirmed without a prospective, randomized study. Furthermore, there is always potential for residual confounding, as well as reverse causation, in which individuals have lower CVH due to a subclinical arrhythmia that is only diagnosed after some latency. To this end, Chen et al. performed multivariable modelling with adjustment for an array of clinical risk factors, and undertook a blanking period analysis, which constitute important steps to evaluate for and mitigate these biases. Ascertainment of outcomes depended substantively on billing codes, which often possess limited sensitivity and specificity for arrhythmia.¹¹ The authors also used the 2010 version of the CVH metrics (Life’s Simple 7) which have since been updated to ‘Life’s Essential 8’ to include sleep health as well as updates to some other metrics.¹² Although this study highlights important relationships between risk factors and the development of arrhythmias, it does not advance our understanding of underlying mechanisms. Future work is warranted to elucidate specific processes by which suboptimal CVH may lead to the development of arrhythmias.

On balance, the current findings by Cheng et al. are a call to collective action: cardiac rhythm disorders have emerged as an important goal of cardiovascular prevention, and perhaps the best way to stay in rhythm is to keep to the beat of optimal CVH.

Funding

S.Ka. is supported by the Walter Benjamin Fellowship from the Deutsche Forschungsgemeinschaft (501100001659), award #521832260. S.Kh. is supported by the American Heart Association (23CDA150571) and the National Institutes of Health (K23HL169839).