The Physiology of the Finishing Kick
If even pacing is so great, why do the best runners in the world always seem to have another gear at the end?
I used to see my finishing kick as a sign of toughness. Nobody passed me in the closing stages of a race, I’d tell myself, because nobody wanted it more than me.
But as time went on, I began to see it from a different perspective. No matter how a race played out, whether it was fast or slow and whether I was way ahead or way behind, I would always manage to sprint the last quarter-mile or so. Why did I always have energy left for a sprint, even if I’d been dropped by the leaders? Shouldn’t I have used that energy to avoid being dropped in the first place? Eventually, my kick became a source of frustration. I tried to race hard enough that I’d have nothing left for a kick, but I almost never managed it.
As a result, I’ve always been fascinated by attempts to unravel the physiology and psychology of the finishing kick. The most recent addition: a study in Medicine & Science in Sports & Exercise, led by graduate student Rafael Azevedo at the University of Calgary under the direction of Juan Murias and Guillaume Millet, using an interesting new methodology to tease out levels of fatigue in the brain and body before and after the finishing kick.
Some important background: I always come back to a 2006 study by University Cape Town researchers Ross Tucker, Michael Lambert, and Tim Noakes that analyzed pacing patterns in a century’s worth of men’s world record performances over distances between 800 and 10,000 meters. As I discussed in more detail here, they observed a remarkably consistent U-shaped pacing template for races longer than 800 meters, featuring a fast start, even-paced middle, and fast finish, as shown in this graph:
The presence of a finishing kick even in elite athletes running at world-record pace, they argued, reflected a hardwired tendency to maintain a physiological reserve during intense exercise. In other words, it was evolution rather than cowardice that made me hold back energy for a sprint.
This big-picture explanation makes intuitive sense, but actually unraveling what’s going in your body at different stages in a race has turned out to be more complicated than expected. To that end, Azevedo’s new study involved 12 male volunteers performing a series of 4K cycling time trials. The trials lasted a little over six minutes on average, and as expected they followed a U-shaped pacing profile with a fast start, even-paced middle, and finishing sprint. On average, according to a mathematical analysis, the fast start lasted 827 meters, and the finishing kick started with 410 meters left.
After a couple of familiarization trials, the subjects completed three experimental trials in randomized order: one in which they were stopped after the fast start, a second in which they were stopped after the even-paced section, and a third in which they continued to the finish. As soon as they stopped, they underwent a battery of tests to assess fatigue in various ways. The measurements used force sensors mounted in the pedals of the bike—a crucial detail, since fatigue starts dissipating within a few seconds. Previous experiments have involved getting subjects off the bike and then strapping them into a separate apparatus to measure fatigue, so this is a key technical innovation.
The simplest way of measuring muscle fatigue is with a maximal voluntary contraction: you ask the subject to contract the relevant muscle (in this case the quads) as hard as possible. Using more sophisticated techniques, you can also break it down into two subcomponents. “Central fatigue” is how much the signal from the brain to the muscles has decreased; “peripheral fatigue” is how much weaker the muscle fibers themselves are when you stimulate them with electricity. The researchers performed all three of these measurements.
The results showed a rapid increase in fatigue during the initial fast start: the max voluntary contraction dropped by 23 percent, central fatigue was 8 percent, and peripheral fatigue was 40 percent. Then things stabilized: by the end of the even-paced phase, which accounts for about 70 percent of the overall race, all three of the fatigue markers were essentially unchanged compared to just after the fast start. But after the finishing sprint, fatigue ramped up again, for example to 34 percent for max voluntary contraction.
In other words, muscle fatigue doesn’t accumulate in a nice straight line. After the initial excitement of the start, we settle into a sustainable pace that seems to have very little impact on muscle function. The sensation that your jellied legs couldn’t take another step after a race is produced almost entirely by the finishing sprint, not by the miles that preceded it. One way to understand this is in terms of critical speed (or, equivalently, critical power), a concept I’ve written about in detail a few times recently. Your critical speed is essentially the threshold of what’s metabolically sustainable. You can run above critical speed for a while, but you’re using up your finite reserves of anaerobic capacity—and once they’re done, you’re cooked.
If you were to run a race at a perfectly even pace, you’d use up your anaerobic capacity gradually, hitting zero as you cross the finish line if you judge it right. In contrast, what most of us tend to do is use up a chunk of anaerobic capacity at the start. (There may be good physiological reasons for that, since a fast start ramps up your oxygen-processing capacities more quickly.) Then we settle into a pace relatively close to critical power, where we’re only nibbling away very slowly at anaerobic capacity. Then, as we approach the finish, we use it all up with a glorious sprint.
Sure enough, in Azevedo’s data, the cyclists settled into a pace barely above critical power for the middle portion of the race, meaning that they used most of their anaerobic capacity at the beginning and end. The big question is whether this approach is suboptimal. That’s certainly my intuition. When Joshua Cheptegei broke the 5,000-meter world record last summer, I argued that the Wavelight pacing lights flashing around the perimeter of the track at a perfectly even pace helped Cheptegei by enabling him to run the most evenly paced world record ever. It must be more efficient, right?
But it’s perhaps not as obvious as I thought. Back in 2013, a study from Andy Jones’s lab at the University of Exeter compared different pacing strategies in three-minute cycling trials: the typical self-paced U-shape, an all-out-from-the-start effort, and an even-paced trial. Here’s what those pacing patterns looked like, with the amount of work done above critical power (i.e. the anaerobic capacity) shaded in grey. Panel A is an incremental test to exhaustion, B is all-out from the start, C is even pacing, and D is self-paced.
The numbers indicate the total anaerobic capacity shown by the shaded areas, and there are no significant differences between them. Even pacing produced an anaerobic capacity of 12.9 kJ; self-pacing with a finishing kick produced 12.8 kJ. There’s a big difference in how these different strategies feel, though. The closer you are to emptying your anaerobic capacity, the worse you feel. “My interpretation/hunch,” Jones told me by email, “is that athletes have learnt, or know intuitively, that a pacing strategy involving an end spurt results in the same performance outcome as other strategies, BUT that this same performance can be achieved with less pain for most of the race! The athletes will be just as knackered at the end but that middle section won’t be quite so excruciatingly intolerable if they implement an end spurt strategy.”
It’s an interesting idea. And it would explain why U-shaped pacing patterns are so ubiquitous even among the greatest runners in the world. It has always puzzled me that a seemingly suboptimal pacing strategy could produce so many world records. Even if we’re wired to pace ourselves cautiously, you’d still expect that world records would happen when athletes accidentally started too fast if having a finishing kick was really so bad for performance.
On the other hand, as Ross Tucker has noted, the pacing in world records does seem to be getting more and more even. The gains from smoothing out your pacing may be marginal, but at that level you have to look for every possible edge. Personally, though, I find Andy Jones’s argument very tempting—because if U-shaped pacing doesn’t cost you anything, then I can start thinking of my finishing kick as a badge of pride again, rather than a mark of shame.
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.