On a Formula One track in Italy this month, three men tried to break the seemingly impregnable two-hour marathon barrier in a Nike-sponsored exhibition race.
Even in the early miles, one runner – Eliud Kipchoge, the 2016 Olympic champion from Kenya – stood out for the conspicuous smoothness of his running form. While his rivals bounced and swayed and leaned with each stride, Kipchoge looked as though he was riding on a moving sidewalk.
Do such impressions tell us anything useful about who the fastest or most efficient runner is? For years, that question has been debated by researchers and by the burgeoning subfield of running-form coaches, with little consensus. Looks, after all, can be deceiving: Sometimes the race goes to the most awkward-looking runner, with smooth-striding rivals left far behind.
A new study from researchers at Loughborough University in Britain offers perhaps the most comprehensive attempt yet to search for links between running form and performance, using 3-D motion analysis to assess and quantify subtle movement patterns that might be invisible to the naked eye.
The results suggest that the fastest runners do indeed have better form – but whether this observation will help the rest of us run better remains unclear.
The study involved 97 runners whose best recent 10-kilometre race times ranged from 29:32 to 56:49. They each completed a series of treadmill tests at paces between five and six minutes per kilometre while the researchers measured 24 different stride variables.
The findings, which were published in Medicine & Science in Sports & Exercise, confirm there are vast differences in how people run. Factors such as the up-and-down bounce in each stride or the amount of braking when your foot hits the ground differed by a factor of two or more from one runner to the next.
The researchers then compared this data to the participants' race times and lab-measured running economy (an efficiency statistic comparable to the fuel economy of a car), looking for patterns. Statistical analysis picked out the strongest independent predictors, which included up-and-down oscillation (less is better), braking (less is better) and the angle of the lower leg when it hits the ground (closer to vertical is better). Together, these factors explained about a third of the variation in running performance and efficiency.
Notably, some of the most talked-about aspects of running form, such as cadence (how many steps per minute you take) and the angle of the foot upon contact with the ground (which determines whether you land on your heel or forefoot), didn't show up among the most significant predictors.
"Did this surprise me? Not really," said Jonathan Folland, the study's lead author. Foot angle, he explained, is likely best in an intermediate range rather than at either extreme of toe or heel landing, which makes it hard to pinpoint the best value.
Similarly, the role of cadence is blurred by the effect of height, because if all else is equal, taller runners will take longer strides, he said. And if you get bounce, braking and lower-leg angle right, cadence will likely take care of itself.
So now that we have this information, what can we do with it? Having a coach assess your form can be useful, Folland said, along with using new forms of wearable technology that track hard-to-estimate parameters such as vertical bounce.
Still, applying these average results to any individual runner should be done with caution, noted Chris Napier, a Vancouver physiotherapist and biomechanics researcher at the University of British Columbia.
"For every Eliud Kipchoge, there is a Priscah Jeptoo or a Paula Radcliffe," he said, referring to two world-beating marathoners with famously ungainly strides.
And even if you identify a flaw in your form, you have to be sure that the medicine isn't worse than the ailment.
For example, a study published this month by researchers at East Carolina University offered real-time visual feedback to help runners adopt a forefoot strike, reduce the force of their landings or increase their cadence. All three interventions worked, but the first two also increased the load on the ankle joint, which could trigger an injury.
So if you decide to tinker with your form, start by taking several weeks of baseline data – with video or wearable technology – over a wide range of paces. Figure out what's "normal" for you, so you can watch for both intended and unintended consequences when you start making changes.
Caveats aside, Folland's results bolster the case that form really does matter. Kipchoge, in the end, lived up to his impeccable form in Italy by winning the race in 2:00:25, the fastest marathon time in history by more than two minutes (though it won't count as a world record due to the use of a rotating crew of pacemakers).
Fortunately, if you want to run more like Kipchoge, the first step is simple. A British study in 2012 found that as little as 10 weeks of regular running significantly improved running efficiency thanks to a series of subtle and unconscious form changes.
"Clearly," Napier said, "some 'economical' benefits come with just running more."
Alex Hutchinson's latest book is Which Comes First, Cardio or Weights? Follow him on Twitter @sweatscience