Atrial Fibrillation in Athletes: Different but the Same?

Andre La Gerche; Jonathan M. Kalman


Europace. 2019;21(12):1762-1763. 

Atrial fibrillation (AF) is associated with significant morbidity, health costs, and impact on quality of life that appears similar for athletes and non-athletes. Numerous case–control, cohort and large population studies have reasonably consistently reported an increased prevalence of AF amongst male endurance athletes[1] and recent evidence would suggest that long-term complications such as thromboembolic stroke occur with similar frequency.[2] For symptomatic athletes with paroxysmal AF, relatively limited available evidence suggests that pulmonary vein isolation (PVI) appears to have similar efficacy in preventing AF recurrence as in non-athletes with AF.[3–5] However, definitive evidence to guide management decisions remains elusive and significant questions remain.

Why should athletes be a particular focus for study? Apart from the aforementioned excess of AF, there is also evidence that exercise training results in a unique combination of structural, autonomic and inflammatory changes that promote triggers and create an adverse substrate predisposing to AF.[1] Thus, it is reasonable to question whether typical treatments for AF are equally efficacious in athletes and, more specifically, whether continued exercise training influences outcome. It stands to reason that if exercise training contributes to causing AF then removing this inciting stimulus may reduce recurrence. Discussions about modifications in training tend to dominate consults with athletes, and clinicians are left with frustratingly little information on which to lean.

Decroocq et al.[6] from Lille provide some useful evidence to start filling the vacuum. In a retrospective case–control study of 73 athletes and 73 inactive subjects treated with PVI, 5-year recurrence rates were similar. Also, success rates were similar amongst the athletic cohort regardless of whether they had ceased sports training, modified the amount or intensity or continued training without change. Three prior studies had also suggested similar efficacy between athletes and non-athletes[3–5] but none had specifically interrogated the effect of de-training on outcomes. Koopman et al.[3] had reported a trend towards greater 5-year AF recurrence amongst endurance athletes when compared with non-endurance athletes or non-athletes but that observation was not validated in the current study in which recurrence was almost identical between athletes and inactive subjects. The athletic cohort comprised approximately two-thirds endurance athletes and there was no association found between recurrence and duration, type or intensity of sport performed. Thus, the current study contributes significantly to the confidence that can be afforded to the electrophysiologist in counselling athletes that outcomes from PVI are as good as in other AF patients; no worse, but also no better.

There are several unique issues when studying athletic populations. Firstly, athletic status creates a binary classification for a continuum of exercise-related conditioning and cardiac remodelling. Ideally, athletic conditioning would be measured directly in the form of maximal oxygen consumption (VO2max) on a cardiopulmonary exercise test. VO2max has been demonstrated to be strongly associated with the extent of exercise-induced cardiac remodelling (often termed 'athlete's heart') as well as other athletic features such as sinus bradycardia that have been proposed as putative risks for AF in athletes.[1,7] Thus, just as it would be difficult to assess the role of hypertension without measuring individual blood pressures, we are left with an imperfect description of athletic status on which to interpret results. Perhaps this is an unfair critique given that most athlete studies rely on imprecise questionnaires subject to recall bias and issues associated with dividing the physical activity spectrum into intensity and time segments. It is notable that each of the studies assessing PVI efficacy has used a different definition of athlete. Importantly, Decrooq et al. have chosen a more stringent definition of performing more than 6 h of strenuous exercise per week and more than 2000 cumulative lifetime hours and this affords greater confidence that we may be able to generalize to a typical athletic population in whom exercise is more likely to have been a contributing causal factor. On the other hand, it is of some concern that there was no difference in the extent of cardiac remodelling between the athletes and inactive subjects in their study. The cohort size is adequate to expect a difference in cardiac mass and the prevalence of eccentric remodelling between the groups. Atrial size is confounded by AF itself but nonetheless, it might also have been expected that the atria may have been larger in athletes. Finally, the prevalence of risk factors, particularly the surprising 16% prevalence of cigarette smoking and high average body mass index, perhaps reflect a reduction in training and adoption of a more sedentary lifestyle in some athletes with increasing age. It would be interesting for future studies to assess PVI efficacy according to the degree of athletic conditioning and/or cardiac remodelling.

Although the precise mechanisms by which elite levels of exercise lead to an increased incidence of AF in athletes is uncertain, it is reassuring to have further confirmation that athletes with paroxysmal AF respond well to a PVI procedure. Pulmonary vein-left atrial stretch and autonomic remodelling, both potential factors in athletes, may play a significant role in initiation of pulmonary vein triggers. In contrast, it is noteworthy that athletes with persistent AF had very high AF recurrence rates (15 of 22, 68%) perhaps tempering any optimism that the normal markers for PVI success can be stretched for athletes. In this small patient subset it is not clear whether outcomes were different to the non-athletic population. Although not the subject of the current study, it seems probable that this poor outcome does indeed reflect more advanced atrial electrical remodelling as observed in any persistent AF population.

In this context, it may seem surprising that no relationship was observed between ongoing exercise and AF recurrence. Prior studies have demonstrated the possibility of reverse remodelling with treatment of risk factors such as obesity and sleep apnoea and conversely when risk factors are not managed, remodelling does progress even over similar time frames of follow-up.[8,9] It is possible that even the 5-year follow-up observed in the current study is an insufficient time span in which to see the impact of progression of remodelling, particularly when exercise intensity in this age group is likely to have declined. It would also be interesting to know whether the athlete with paroxysmal AF and more profound markers of cardiac remodelling including larger atrial volumes is indeed at increased risk of AF recurrence after simple PVI when matched to a non-athletic cohort. Ultimately, it is unknown whether atrial remodelling due to exercise is likely to be more or less extensive than the remodelling observed in the context of hypertension, obesity, sleep apnoea, or other AF risk factors. The current study would support the hypothesis that extreme physical training should simply be considered on the same spectrum as other more classical and emerging AF risk factors in terms of impact on atrial remodelling and of long-term AF outcomes. In this context, it has been previously argued that an athlete with AF should not be viewed as having lone AF.[10]

What is the next step on the pathway to identifying whether there is an important role for detraining in the management of AF in athletes? A prospective study design would be ideal as it necessitates a considered definition of athletic status and provides an opportunity for randomization to a training vs. de-training intervention. Whilst this might sound simple, there are substantial challenges faced when attempting a randomized lifestyle intervention. Not least, many athletes will not be willing to detrain and thus a very clear selection bias can be anticipated from the outset. The ideal of a traditional randomized control trial model may well prove too great an obstacle and it is possible that we will be relying on retrospective observational data for some time. These trials are helpful in their consistency. It can be stated with reasonable confidence that PVI is an effective treatment in (male) athletes with symptomatic paroxysmal AF. Furthermore, until larger studies of better trained athletes provide more definitive evidence, best contemporary evidence suggests that athletes can return to training without hastening recurrence.