Why Ultrarunning Success Is Hard to Predict
Scientists can forecast marathon performance fairly accurately. But for longer distances, things get complicated.
Back in January, ultrarunning star Jim Walmsley eked out a qualifier for next February’s Olympic Marathon Trials at the Houston half marathon, hitting the 1:04:00 entry standard right on the nose. The tantalizing prospect of mountain man Walmsley challenging for an Olympic berth on the roads stirred up yet another round of the long-running marathoner vs. ultramarathoner debate.
Of course, such comparisons depend on the details of the challenge. Even a truly world-class road specialist with a marathon best of 2:05, Walmsley told the Citius podcast, wouldn’t be able to keep up with him on the long, mountainous trails he specializes in: “The way I attack the downhills… I will break your quads and you won’t be able to jog the flats anymore,” he said. “Find your 2:05 guy, give me a couple hours in the canyon, I’ll be the first one out.”
Bold words. But does he have a point? Marathons are generally thought to be a reasonably straightforward test of physiology: VO2max (how quickly you deliver oxygen for your muscles to use), running economy (how efficiently they use it), and lactate threshold (a proxy for how long you can sustain a hard effort). Ultramarathons, according to conventional wisdom, are different. Durability is more important than efficiency, which is why ultrarunners wear heavier shoes and take shorter strides. An iron stomach, to handle lots of eating and running, is more useful than a heart of steel. And the mind is the most important factor of all.
Few studies have actually tested these claims, though, and there’s a countervailing school of thought that if a few second-tier Kenyan and Ethiopian marathoners dove into the trail ultra scene, they’d immediately mop up the competition. One of the most hilarious passages in Adharanand Finn’s recent book The Rise of the Ultra Runners follows his efforts to crowdfund Kenyan and Ethiopian runners to big ultra races in Britain and the United States. A series of mishaps, ranging from a sore toe to the raging wildfires that canceled the North Face 50-mile race in California last year, scuppers every attempt, so the question remains unanswered.
But with the growth of ultrarunning, researchers are becoming increasingly interested in the topic. At a recent conference on sport innovation, University of Guelph researcher and former elite triathlete Alexandra Coates presented some preliminary data from a study of runners at the Sulphur Springs Trail Race, an event in Ontario that includes 50K, 80K, and 160K distances around a course with 20K loops that gain and lose 620 meters each time. Coates and her colleagues put 42 runners (25 men, 17 women) through a series of baseline tests, including the usual physiology tests in the lab, and also assessed their training history and monitored their hydration status during the race. Then they asked a simple question: which variables best predict race performance?
When all the racers were grouped together, performance (quantified by percent of winning time) was best predicted by the highest speed the subjects reached in their treadmill VO2max test. By looking at the speed during the treadmill test, rather than just VO2max itself, you get a measure that reflects both your VO2max and your running economy. In other words, as Coates put it during her talk, those with the best running fitness notched the highest finishes—not a huge surprise, and the same thing you’d expect to see in a regular road marathon.
The second-best predictor is perhaps a little more surprising: those who lost the most weight during the race tended to finish faster. That’s also consistent with previous findings from road marathons and other endurance races. What this means is up for debate. One possibility is that drinking too much slows you down by making you heavier, compared to simply drinking when you’re thirsty and allowing yourself to get a bit dehydrated. Alternately, the causal arrow may be the other way around: faster and more experienced runners may be better at judging how much (or how little) fluid they need, while slower newbies tend to be overly cautious and drink more. That’s a debate for another time, but it’s interesting that this pattern once again shows up even in longer races.
Notably missing from the picture is anything related to training history. How much the subjects reported running each week, or over the past year, how much strength training they did, how many years they’d been running, how many previous marathons or ultramarathons they’d completed—none of this had any significant impact on finishing performance. Now, don’t misread this to imply that training doesn’t matter for ultrarunning! But in this particular cohort of recreational runners training, on average, about eight hours per week, the differences between those doing a little more and those doing a little less weren’t what determined finishing place.
Another interesting detail is what happens when you break down the results by finishing distance. There were 21 runners in the 50K, 13 in the 80K, and just 8 in the 160K, so you have to take the sub-analyses with a big grain of salt. But for what it’s worth, here’s what you find.
The 50K looks a lot like the marathon, with plenty of straightforward physiological predictors. Speed at VO2max is the best one for both men and women; when you combine the genders, percent change in body mass is the best. Other basic health measures like body mass index, age, resting heart rate, and blood pressure also have some predictive power.
But as you go to longer distances, the picture changes. At 80K, the only significant predictor is speed at VO2max. And at 160K, none of the variables measured had any relationship with the eventual performance of the runners. In a sense, this corroborates what ultrarunners like Jim Walmsley have been saying for years: the traits that make you a great marathoner won’t necessarily make you a great ultramarathoner, and the differences get wider and wider as the distance gets longer.
I asked Coates what sorts of additional measurements she would add to her study if she were doing it again and wanted to get a better prediction of ultra success. She mentioned strength testing and neuromuscular fatigue resistance, to get a sense of how well your leg muscles can stand up to the prolonged pounding. In-race nutrition strategies is another big one, though it’s tricky to study in the field: some of the subjects were hallucinating from fatigue by the time they got to her research tent, which makes it hard to get reliable information on what they had eaten. And psychological characteristics are undoubtedly important: how well do you handle adversity and pain?
Of course, all of those “ultra-specific” traits are important to varying degrees even in shorter road races. They’re not two different sports, just two ends of a spectrum. And that’s what makes Walmsley’s upcoming appearance at the Olympic Marathon Trials so much fun to contemplate. Even in the marathon, the standard physiological prediction equations don’t always capture everything that matters.
My book, Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance, with a foreword by Malcolm Gladwell, is now available. For more, join me on Twitter and Facebook, and sign up for the Sweat Science email newsletter.