health as it happens

There's New Evidence on Heart Health in Endurance Athletes - Outside Magazine

About a decade ago, a series of studies emerged with the counterintuitive message that modest amounts of exercise might actually be bad rather than good for your heart. The newspaper headlines—“One Running Shoe in the Grave,”—were almost gleeful. The evidence, on the other hand, was weak.

That debate has mostly faded out of the headlines, but the questions haven’t been completely dismissed. Some of the initial fearmongering about the dangers of, say, training for a marathon, or even running more than a few times a week, were clearly nonsense. But there are plenty of ultrarunners and Ironman triathletes and cycling junkies racking up huge training numbers, and for these people the evidence about possible risk is sparser and more ambiguous.

The good thing about all the debate is that it launched a bunch of studies, whose results have begun slowly trickling in over the last few years. A new review in the European Heart Journal , led by Gemma Parry-Williams and Sanjay Sharma of St. George’s University of London, sums up the current evidence about “the heart of the aging endurance athlete.” While there are still plenty of unanswered questions, there have been some advances since my last comprehensive look at this debate back in 2016.

Calcium in the Arteries

The most solid evidence for potentially negative heart changes associated with long-term serious endurance training relates to elevated coronary artery calcium (CAC) scores. The gradual build-up of calcium-rich plaques causes narrowing and stiffening of the coronary arteries, which supply blood to your heart muscles. These plaques can also rupture, blocking the artery completely and causing a heart attack. The CAC score measures how much calcium has accumulated in your arteries, so anything that increases it seems like bad news.

What’s new is how we interpret those scores in runners. When Sharma’s group studied 152 masters endurance athletes with an average age of 54, 11 percent of them had a CAC score of greater than 300 (which is considered very high), compared to none of the 92 people in the age-matched control group. That’s concerning, and other studies have reached similar conclusions. Not all plaques are equal, though. Some are smooth, hard, and calcified, and these are considered stable and less likely to rupture. Others are a softer mix of cholesterol, fats, calcium, and other substances, and these mixed plaques are more dangerous and likely to rupture. The athletes, it turns out, had 72 percent stable calcified plaques, while the controls had just 31 percent.

It’s not clear why exercise seems to produce more plaques in a small minority of endurance athletes, nor why those plaques seem to be more stable. There are various theories involving mechanical stress on the arteries, high blood pressure during exercise, inflammation, hormone levels, and so on. But there’s an interesting parallel that has emerged in recent research: statins, a type of drug widely prescribed to people with high cholesterol, also seem to increase arterial calcium levels and produce denser, more stable plaques, which may be one of the reasons they’re so effective at lowering the risk of heart attacks. That doesn’t necessarily mean that calcified arteries are nothing to worry about, but it does suggest that CAC scores (which don’t differentiate between calcium in stable versus unstable plaques) don’t carry the same meaning in endurance athletes as they do in non-athletes.

Atrial Fibrillation

Another issue that seems more common in endurance athletes is atrial fibrillation (a-fib), an irregular heartbeat originating in the atria, the upper chambers of your heart, that comes and goes. It’s relatively common, affecting 2.7 million Americans, and way less serious than other arrhythmias like ventricular fibrillation. It is, however, linked to an elevated risk of strokes, heart failure, and other issues—at least in non-athletes.

By some estimates, middle-aged endurance athletes are five times more likely than non-athletes to develop a-fib. The evidence isn’t perfect: one alternate theory is that athletes are simply more likely to notice that their heart is fluttering, and thus are diagnosed more frequently. And athletes are also less likely to have other risk factors like high blood pressure that combine with a-fib to raise the risk of serious outcomes like strokes.

The most interesting question is why athletes would be more susceptible to a-fib. Again, there are a bunch of different mechanisms proposed, including changes in the electrical pulse of the heartbeat, and stretching, inflammation, or scarring of the heart muscles. There are also some more benign possibilities: one is simply that bigger hearts (a common result of endurance training) have bigger atria, so there’s more tissue where a slight irregularity could trigger a bout of a-fib. In support of that idea: tall people are more likely to have a-fib, and large animals like elephants and horses get more a-fib than smaller animals.

Scarring of the Heart

The idea that prolonged endurance exercise causes fibrosis, or scarring, of the heart is probably the most worrying possibility. The evidence here is pretty convoluted, as the review explains. When you fire up the cardiac MRI machine, there are several types of scarring that you might find.

One is small focused points of scarring at the “hinge points” where two chambers of the heart connect. By some estimates, 40 percent of male masters athletes and 30 percent of female masters athletes have this type of scarring. It’s thought to be the result of years of high-pressure pumping of large volumes of blood, but studies haven’t found any evidence that it leads to health problems.

You might also find more diffuse patches of scar tissue in the heart muscle, which can either be the result of a temporary shortage of oxygen or the after-effects of a mild bout of myocarditis, an inflammation of the heart muscle. It’s not entirely clear whether endurance athletes have more of this type of scarring, in part because it’s hard to figure out what control group to compare them to. Sedentary people tend to have a bunch of other cardiac risk factors like high cholesterol and blood pressure, which biases the comparison in one direction. But if you select only sedentary people without those risk factors, you have an unnaturally healthy group that biases the comparison in the other direction.

In Sharma’s group of 152 masters athletes, 11 percent had diffuse patches of scar tissue. Of those, two thirds had a pattern consistent with myocarditis, which likely has nothing to do with exercise. The other third had a pattern consistent with temporary oxygen shortage—or, in other words, a subclinical mini-heart attack. But only half of those had signs of blockage in the relevant coronary artery, and in each case the blockage was less than 50 percent, making it unlikely to have caused a mini-heart attack. Note that, at this point, we’re talking about 3 of the 152 athletes.

Sharma’s control group was healthy, sedentary non-athletes matched to have the same cardiac risk factors as the athletes. A similar study published last year by University of Toronto researchers instead compared endurance athletes training an average of 7.6 hours a week to active controls meeting public health guidelines by exercising up to three hours a week. In this case, both groups had equivalent patterns of scarring, suggesting that it’s simply wear-and-tear that accumulates with age and activity.


As I write this, I can’t help noticing a whiff of motivated reasoning. “Sure, there’s more calcium in the arteries—but that’s good, not bad! And atrial fibrillation is just a minor nuisance for healthy athletes. And we’ve all got scarring! It’s normal! Really!” That’s the problem with proxy outcomes: they leave room for bargaining, and don’t tell us what we really want to know, which in this case is whether doing lots of endurance exercise will shorten or lengthen our lives.

The only death data we have so far is observational: ask a large group of people how much exercise they do, then wait and see how long it takes them to die. But this kind of data is deeply flawed, because there may be many differences between people who run 100 miles a week and people who run 0 miles a week. Maybe runners live longer because they cook a lot of healthy recipes from running magazines, not because they run a lot.

I dug deep into this problem in that 2016 article, and I don’t want to repeat all that here, but I do want to give one key example. The biggest epidemiological study that claimed to find negative effects of too much running (in this case, that was more than 20 miles a week) was from an analysis of 55,000 patients at the Cooper Clinic in Dallas. This finding made headlines around the world, and continues to be cited as evidence of the dangers of too much endurance exercise.

What got far less attention was another analysis of the same group of Cooper Clinic patients, by the same researchers, in 2018. This time they looked for links between strength training and longevity—and they found basically the same pattern. Strength training up to twice a week produced a modest reduction in the risk of dying (either of heart disease or of any cause) during the study follow-up period. Lifting three times a week erased most of the benefit, and lifting four or more times a week was worse than not lifting at all.

Here’s the data from that paper, showing relative risk of death as a function of days of strength training per week:

Resistance mortality data chart Photo: Medicine Science in Sports Exercise

As I wrote when the study came out, I don’t think strength training four times a week is actually dangerous. I suspect this pattern is a spurious artifact of the statistical adjustments they had to do in order to compare healthy strength trainers (with generally lower weight, blood pressure, cholesterol, and so on) with non-exercisers. That’s also the explanation the lead author gave me when I asked him. No one put out any press releases suggesting that pumping too much iron might kill you. Newspapers and blogs around the world didn’t amplify the message. And yet it’s exactly analogous to what they found for running.

One other point worth noting: if you skip the flawed studies where you try to guesstimate how much people exercise based on a brief questionnaire, and instead put them on a treadmill and measure their VO2 max to get an unambiguous assessment of aerobic fitness, the outcome is clear. The fitter you are, the longer you’re expected to live, and there’s no evidence whatsoever that the pattern reverses once you get really, really fit. You do get diminishing returns as you get fitter and fitter, but it’s always better to be a little more fit rather than a little less fit.

There’s still lots to learn about exactly what’s happening with calcium in the arteries, arrhythmias, and heart scarring. But in terms of the outcome that matters—death—the data presented in the new review makes me think the envelope of possible outcomes is getting narrower and narrower. It’s highly unlikely that we’re going to discover tomorrow that running ultramarathons takes ten years off your life expectancy (or, to be fair, adds ten years) compared to simply hitting the exercise guidelines. Any effects, if we can ever separate them from the noise, are likely to be pretty marginal. I take comfort from that.

Overall averages don’t tell the whole story, of course. Maybe running lots of ultramarathons adds a few months of life expectancy for 99 percent of us, but shortens it by a decade for an unlucky fraction of a percent who have some sort of underlying issue or genetic predisposition. That’s why this research continues to be important, in the hope that we can eventually figure out what those red flags might be. In the meantime, if you’re playing the odds, I’d suggest you keep running, swimming, and cycling to your heart’s content.

For more Sweat Science, join me on Twitter and Facebook, sign up for the email newsletter, and check out my book Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance .