British runner Marc Scott had a choice about where to race last weekend: the European Indoor Championships in Poland, or a low-key, low-pressure twilight meet in California. He , notching a personal best and Olympic qualifying time of 13:05 for 5,000 meters—just behind one of his training partners from the Portland-based Bowerman Track Club, and just ahead of another.
Scott’s run was reminiscent of last year’s in Portland, where teammates Shelby Houlihan and Karissa Schweizer both smashed the American 5,000-meter record. Or the previous year, before pandemic travel restrictions, when the Bowerman group’s unheralded Woody Kincaid , making him the fifth fastest American ever, in a race against his teammates, paced by another teammate, on the local track at Nike HQ where he trained regularly—a race, in other words, that sounds like .
At first blush, this seems odd. The alchemy of the big race is a key tenet of the runner’s faith: paces that seem impossibly hard in training will become manageable when the pressure is on, the crowds are watching, and your rivals are breathing down your neck. The Bowerman squad’s impressive results in what amount to time trials are a reminder that big races also take a toll: travel, unfamiliar environments, disrupted routines, stress, uncertainty about how the race will play out, and so on. By eliminating or controlling those factors, perhaps you conserve some hard-to-quantify element that frees you to run fast. But what is that element?
Ever since the publication of by Samuele Marcora, then at Bangor University, sports scientists have been debating the idea of “mental fatigue”—and more specifically, the notion that a tired brain impairs your physical performance. Marcora’s original protocol involved spending 90 minutes doing a cognitively demanding computer task, which is thought to trigger a build-up of that increases your perception of effort. The reason coaches and sports scientists are interested in the idea is that it offers a plausible physiological explanation for why, say, making a tight flight connection the day before a race or fretting about pacing during a race might hurt your performance.
That’s the theory. In practice, though, it’s not clear how well lab findings about mental fatigue translate into the real world. Marcora’s latest study, led by his former doctoral student at the University of Kent, Chiara Gattoni, explores this gap by testing the effects of mental fatigue on half-marathon performance. The researchers tapped into a neat initiative called , headed by a University of Verona researcher named Federico Schena, in which volunteers agree to run half marathons or marathons after being randomized into various interventions that scientists want to study. Gattoni’s results are (meaning they haven’t yet been peer-reviewed)—and they offer a reminder of just how hard it is to test these ideas outside the lab.
Over the course of three successive years of the Run4Science program, the research team managed to recruit 46 athletes to run a half marathon. Half were assigned to spend 50 minutes immediately before the race doing a mentally fatiguing computer task that involved pressing buttons as quickly as possible in response to cues on the screen. The other half were assigned to read magazines for 50 minutes. Ideally, you’d want every subject to run two half marathons, one with mental fatigue and one without, so that you could compare each runner to their own previous performance. But the generosity of volunteers has its limits.
The mentally fatigued runners averaged 106.2 minutes for the half marathon, compared to 102.4 for the control runners. Their heart rates were also about three percent lower throughout the run, and perceived effort was roughly the same between the groups. That’s exactly what you’d predict based on Marcora’s previous work: mental fatigue makes exercise seem harder, and perceived effort is how you judge the appropriate pace, so mentally fatigued runners should run at a lower physical effort (as reflected by heart rate) and finish in a slower time while feeling as though they’ve pushed just as hard.
Here’s what the data for running speed (top), heart rate (middle), and rating of perceived exertion (RPE, bottom) looked like. Measurements were taken every 7K during the race, and immediately after the finish. As you’d expect, speed drops over the course of the race, while heart rate and effort drift upward. But it’s the gaps between the mentally fatigued group (circles) and the control group (squares) that matter:
There’s a major catch, though: the differences between the two groups aren’t statistically significant. The effects of mental fatigue are expected to be subtle: based on the results of previous studies, the authors calculate that they would have needed 472 half marathon volunteers to discern a statistically significant difference between two groups who are only running one race each. Given that it took them three years to scrounge up 46 volunteers, this isn’t going to happen.
It’s tempting to gloss over the stats. After all, the results are more or less what you’d expect given the subtlety of the effect and the limitations of a one-shot measurement with a small sample size. But that’s a slippery slope. What if the results had been the opposite, suggesting that mental fatigue helps performance? In that case, you’d probably assume that the findings were a fluke, the kind of thing that happens now and then by chance when you only have a tenth as many subjects as you really need. As a result, you might not even bother publishing it. That’s how publication bias creeps into the literature, giving the illusion of real effects even when there’s nothing but random chance.
There have been several meta-analyses that aggregate the results of studies on mental fatigue and athletic performance, with varying results. One, by researchers at McMaster University, concluded that the effect is real and significant for endurance and strength tasks, but not for all-out sprints. Another, by a group led by Darías Holgado of the University of Granada, took a more skeptical line. They analyzed 21 studies focused on endurance performance, with 317 participants in total, and again found a significant effect—but concluded that it was probably the result of publication bias.
The moral here isn’t that Marcora and Gattoni shouldn’t have published their study. Quite the opposite: it’s that researchers need to commit, in advance, to publishing all their studies, regardless of whether the results conform to their expectations. That’s the best way to mitigate publication bias. And we, in turn, need to look at the results of these studies and conclude—well, nothing. The coach of the Bowerman Track Club, Jerry Schumacher, certainly isn’t scheduling his team’s low-key race plans on the basis of arcane mental fatigue research. Personally, I think that the body of research is intriguing, and it feels plausible, and it may well be one of the factors contributing to the spate of big performances in pandemic-limited small meets. But for now, that’s just a hunch.
For more Sweat Science, join me on and , sign up for the , and check out my book .