Running: The Risk of Coronary Events: Prevalence and Prognostic Relevance of Coronary Atherosclerosis in Marathon Runners

Stefan Möhlenkamp; Nils Lehmann; Frank Breuckmann; Martina Bröcker-Preuss; Kai Nassenstein; Martin Halle; Thomas Budde; Klaus Mann; Jörg Barkhausen; Gerd Heusch; Karl-Heinz Jöckel; Raimund Erbel


Eur Heart J. 2008;29(15):1903-1910. 

In This Article


The present study was designed to examine the prevalence of subclinical atherosclerosis in relation to cardiovascular risk factors and their role for myocardial damage and outcome in accomplished recreational marathon runners. To our surprise, given the substantial evidence that physical activity reduces CAD event rates,[1,2,21] a CAC score ≥100 was present in 36% of runners, which was not different from age-matched controls, even though the FRS was much lower. Further, the CAC score exceeded that in controls matched for age and FRS. The CAC score was predictive of myocardial damage, which was evidenced by LGE in 12% of runners. Our outcome data suggest that higher CAC scores may be associated with higher rates of events. There are several possible explanations for these findings.

With respect to the mismatch between FRS and the extent of CAC in marathon runners, many of the runners have commenced marathon running in middle-age. Consequently, their cardiovascular risk factors could have been reduced by exercise training and may not reflect their life-long risk exposure. In fact, more than half of our runners were previous smokers and 5% of runners reported active smoking. This would also explain the comparatively low CAC scores in controls matched for age and FRS, which may have had life-long protection from the atherogenic effect of cardiovascular risk factor exposure. The clinical implication of this possibility is that standard risk factor estimates may lead runners and their physicians to underestimate the athletes' true risk. Risk stratification in marathon runners is further rendered difficult by improved microvascular function in marathon runners, which can compensate for severe epicardial plaque burden and thereby concealing the true extent of coronary atherosclerosis.[11] This may in part explain why all runners in our study were asymptomatic at rest and during running despite considerable atherosclerosis in so many.

In our study, CAC scores were not related to any measure of physical activity in any of the groups. Given the expected population-wide annual increase in CAC of 15-20%,[18] regular marathon running seems not to protect runners from CAC progression once CAC is present. In fact, we even cannot exclude the possibility that exercise to this degree has deleterious effects on coronary arteries. This seems unlikely given the substantial experimental[22] and clinical[1,2,23,24] evidence for the benefits of regular physical activity, but no epidemiological studies have so far examined individuals engaged in such prodigious amounts of exercise as our marathon runners. Several mechanisms may be involved: regular exhaustive exercise during marathon and its required training may induce a rise in vascular oxidative stress because of a high-flow, high-pressure condition, to a point at which it challenges anti-oxidative capacity.[22] Bursts of inflammatory cytokines, which almost invariably occur during marathon running,[25] may also accelerate the atherosclerotic disease process and impair intramyocardial microvascular integrity, whereas no ischaemia is detected during a short-term diagnostic exercise protocol.

The possibility that marathon running and the required training aggravates pre-existing non-calcified atherosclerosis and has a role in LGE development is clearly speculative based on our cross-sectional data. However, it is interesting given recent reports of myocardial injury during marathon running and the observation of prevalent myocardial damage in our runners. Others have demonstrated increases in myocardial troponin levels in recreational marathon runners and other endurance athletes.[26,27,28,29] The cause of such possible myocardial damage is unclear, but unlikely owing to epicardial coronary artery obstruction. Evidence of myocardial damage was found in 12% of our 108 runners. A pattern highly suggestive of myocardial ischaemia was seen in 42% of these, whereas the others demonstrated more patchy defects. We have previously shown that infusion of small particles of 10-100 µm in diameter into porcine coronary arteries leads to haemorrhagic and patchy patterns of myocardial damage, depending on particle size.[30] Embolization of microthrombi or atherosclerotic plaque material into the microvasculature is also conceivable during marathon running, because excessive mechanical forces may put strain on plaques and thereby cause plaque erosion or fissuring with subsequent epicardial thrombus formation[8] and microembolization.[31] An increased exercise-induced thrombogenicity from increased catecholamine-induced platelet aggregation, or an imbalance in fibrinolytic/prothrombotic factors[32,33] may also have a role in such thrombus formation. It is therefore possible that both the myocardial injury reported after marathon running[26,27,28,29] and the myocardial damage in our marathon runners are in part due to small thrombotic or even atherosclerotic emboli.

The presence of myocardial LGE has recently been shown to predict cardiac events in patients with CAD.[14] In marathon runners, such damaged myocardium may be a substrate for an increased susceptibility to arrhythmias in response to increased exercise-related catecholamine levels,[8] and hence may contribute to cardiac events. In our study, CAC scores were higher in those runners with LGE, supporting a pathophysiological link between epicardial subclinical plaque burden and intramyocardial microvascular damage, as indicated above. Even though CAC is not a measure of plaque vulnerability at that site,[34] the increasing rate of coronary events in parallel with increasing CAC scores may indicate increased plaque vulnerability or susceptibility for plaque rupture or fissuring somewhere else in the coronary tree. The precise mechanism by which elevated epicardial plaque burden may increase the likelihood of myocardial damage and vulnerability, remains to be shown.


Our data do not apply to women and may not be representative for all marathon runners, as we cannot exclude recruitment bias. Participants may have had previous risk factors, a recent reduction in exercise capacity or recently discovered CVD in a relative. However, we have excluded all subjects with known CVD, diabetes, or any symptoms of CVD, and participants have been running regularly for nine years. The risk factor profile in our cohort is therefore typical for many marathon runners and is expected to be worse in many others. This pertains in particular to diabetic athletes, who have been excluded from this study.

Because of different recruitment strategies and inclusion criteria between the two studies, our findings are subject to selection bias: marathon runners were self-referred because they could not be randomly selected as the participants in the HNRS. They also had to be fit and healthy beyond age 50 to be included, while this was not the case in the HNRS.

Our cohort of marathon runners is heterogeneous with regard to duration of regular physical exercise and marathon running, as evidenced by the interquartile ranges of years of active running, number of marathons completed, and the weekly training mileage. Participants may also differ in their engagement in sports other than marathon running. We believe, though, that our findings reflect the typical spectrum of risk factor and atherosclerosis burden in males >50 years participating in marathons.

It is conceivable that the ratio of calcified and non-calcified atherosclerotic plaque differs among persons who are regularly exposed to exhaustive exercise and those who are not, because repetitive increases in shear stress and mechanical forces may predominantly impact on the calcified plaque component. Currently, there are no data to support this hypothesis, which can only be assessed non-invasively by additional administration of contrast agent and high-resolution computed tomographic techniques.

We have discussed oxidative stress, microembolization, bursts of inflammation or increased thrombogenicity as potential mechanisms for the pathogenesis of LGE. Yet, other mechanisms, such as subclinical myocarditis, vasculitis, or cardiomyopathy[19] as well as coronary vasospasm or anomalies including myocardial bridging[5] may also have been involved in its development in some athletes, even though marathon runners with any known current or previous CVD were excluded from this study. Further, the HNRS cohort did not undergo CMR scanning at baseline investigation, which precludes a comparison of the prevalence of LGE among these cohorts.

Two events were revascularizations and were in part subject to surveillance bias. The significant stenoses in runners free of symptoms may have been missed outside this prospective study. Even though statistically significant, our event data should therefore be interpreted with caution and longer follow-up in larger cohorts is required. Ideally, event rates in marathon runners should be compared with those in the control groups. However, outcome data from the HNRS will only be available in 2009. Our findings are in a similar magnitude, though, as previously reported from other asymptomatic low-risk cohorts and are in line with existing evidence on the prognostic value of CAC.[13]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: