A comprehensive new study zeroes in on pelvis motion and leg angle.
It’s hard to put your finger on exactly what makes a running stride ‘good’, as we’ve learned from an endless succession of inconclusive and often contradictory biomechanics studies over the years.
Still, if you stand on the sidelines of a big road race and watch the runners go by – first the effortless elites, then the dedicated locals, then the casual fitness types, and so on – you’re left with the impression that there are some clear differences in how well people run. So what factors really matter?
That’s what a new study at Loughborough University in the UK set out to test (abstract here, press release here). They put 97 runners with 10K season-best times ranging from 29:32 to 56:49 through a series of treadmill tests while using 3D motion capture to analyse 24 different variables related to the motion of various body parts. Then they analysed the results from a range of speeds between 5:00 and 6:00 per kilometre (which all the runners in the study could sustain comfortably during testing) to look for patterns.
The first thing that pops out in the results is the high level of variability: for most of the biomechanical parameters, there was a huge range in how people ran. Some runners had twice as much vertical oscillation (how much up-and-down bounce there is in your stride, basically) as others; some had almost three times more ‘braking’ when their foot hit the ground than others.
There was also a huge range in racing performances (as noted above) and in running economy, a measure of how much energy you burn to sustain a given pace. Again, some runners burned almost twice as much energy to cover a given distance compared to others, when resting metabolism was subtracted.
The key question, then, was whether the differences in biomechanics were related to the differences in speed and efficiency. And the answer was yes. In fact, 19 of the 24 biomechanical variables were significantly correlated with running economy, and 11 with seasonal best racing times.
In a sense, that’s not a very useful result. If everything matters, then you don’t get any guidance on what matters most. But many of the biomechanical variables were actually measuring similar things. For example, to measure vertical oscillation, should you look at where your center of mass is, or how much a core point on your body like the pelvis moves? And should you measure oscillation during the whole stride, or just when your foot is on the ground?
So the researchers then used statistical techniques to figure out which combination of variables offered the best independent predictions of running performance and economy.
For running economy, three variables stood out: vertical oscillation (measured by the up-and-down motion of the pelvis; less is better); how bent your knee is when your foot hits the ground (more bent is better); and braking (also measured by looking at the motion of your pelvis; less slowdown as your foot hits the ground is better).
Overall, these three variables explained 39.4 per cent of the individual differences in running economy – and the vast majority of that (27.7 per cent) came from vertical oscillation.
For running performance, four variables stood out: braking (as above); the angle of the shin when your foot hits the ground (closer to vertical is better); duty factor (basically a measure of how long your foot stays on the ground relative to your overall stride; quicker is better); and the forward lean of your trunk (more upright is better).
Overall, these four variables explained 30.5 per cent of individual variation in race times, with shin angle (10 per cent) and braking (9.9 per cent) as the biggest contributors.
So… clear as mud, right? The picture is actually a little simpler than it looks. Braking seems to matter. Vertical oscillation seems to matter. And the angle of your lower leg when your foot hits the ground is closely tied to how bent your knee is at that point – it’s essentially two ways of measuring very similar things, so that also seems relevant.
There are some notable factors that don’t show up in this analysis: what part of your foot hits the ground first, for example, or what your cadence is. That doesn’t necessarily mean these factors don’t matter, but they didn’t show up among the most significant predictors of running efficiency or performance.
The next question is what we should do with this information. In the press release, lead researcher Jonathan Folland says: “Runners and coaches are advised to be attentive to technique and the key aspects of good technique, as well as simply training hard and focusing on physical fitness.”
In the journal article, the authors are even more specific in their conclusions: “It is therefore recommended that runners and coaches be attentive to stride parameters (lower duty factor, shorter ground contact time and shorter stride length) and lower limb angles (more vertical shank and plantarflexed foot at touchdown, and a smaller range of motion of the knee and hip during stance) in part to optimise pelvis movement (minimal braking, vertical oscillation and transverse rotation), and ultimately enhance performance.”
It’s worth pausing to consider whether this advice is justified. One of the strengths of this study is that it considered such a wide variety of running abilities, allowing the researchers to see patterns that have eluded previous studies. If you look at group of people who are all relatively similar – all elite runners, say – then these patterns are much harder to see. That’s why, for example, there’s virtually no relationship (or perhaps even a negative relationship) between height and scoring in professional basketball. Height doesn’t matter only if you assume that pretty much everyone is already very tall.
But that same strength may also be a weakness when it comes to prescribing training advice, because the differences between 30-minute and 50-minute 10K runners go far beyond their shin angles. And even if the links they’ve observed really are causal (i.e. better shin angle causes more efficient running) rather than correlative (i.e. people with better shin angle tend to run more efficiently, perhaps for some reason related to both things), that still doesn’t mean that ‘being attentive’ to these factors will improve them.
For example, you could imagine a study that compared elite runners to ‘regular’ runners and found that the elite tend to have more highly defined calf muscles. It doesn’t necessarily follow that doing a whole bunch of hardcore calf exercises will make you faster. It’s more likely that a whole lot of training, combined with some genetics, has given elites more defined calves. Fixating on getting better calf muscles would be distraction that’s unlikely to help you, and takes away from things that really would make you faster, like running more.
Does running form fall into this ‘misleading correlation’ category? I don’t think it does entirely. There are likely some real, actionable insights about running form that can help people. Figuring out what insights are the useful ones is the first challenge; then figuring out the best way to alter your stride is the second, and even bigger, challenge.
This study seems like a great start on the first challenge, culling a list of three or four key parameters that seem to be linked to fast and efficient running. The next step is to try intervention studies that target these parameters, and see if they produce improvements in running. Is there a practical way of changing vertical oscillation? And if so, does it give you benefits without producing other unintended consequences that counteract those benefits?
These are hard questions, and of course such studies take a long time. In the meantime, technology has advanced to the point that relatively inexpensive consumer devices – hip pods like the Lumo Run, or the straps on some heart rate monitors – allow anyone to monitor some of these stride parameters and see how they respond to stride changes.
That sort of personal experimentation can be a lot of fun. My advice, though, is to start with a long period of baseline data so you have a good handle on what your stride parameters usually are. Then change one thing at a time, and give yourself plenty of time – weeks, if not months – to see how it affects your stride.
And remember the one stride-altering technique that has shaped the biomechanics of virtually all elite runners: years and years of daily running. It’s very hard to study, but it seems to work.