In the beginning of 2011, I participated in a study for the “On Cloudsurfer” shoe, where their shoe was compared to the old shoes of the participants, from a physiological and a biomechanical point of view. I wrote about it here, about the first (maximum effort test) here, then a step test with my old shoe here and finally with the Cloudsurfer here.

I got the results a long time ago and wanted to wait until the overall results of the study were there as well, which they have been for some time now; I just never got around to write about it, so here goes.

In the first test, I had to run at an increasing pace until I couldn’t any more. This was the result:

(Top down: maximal velocity, maximal heart rate, maximal blood lactate concentration, maximal relative oxygen consumption)

Looks all very plausible except the VO₂ max, which seems rather high (like, much too high). The anaerobic threshold was determined as well:

(Velocity, heart rate)

But puh-lease, this pace seems extremely low. The protocol was not designed to determine the anaerobic threshold, so this statement could be expected to be not so accurate. That’s a real bummer because I hoped to gain some insight I could apply to my own training, but no.

More interesting was the comparison of the shoes. This graph compares my heart rate when running at 60% and 70% of the max velocity from above, for 15 minutes each, with my old shoes (blue curve) and the Cloudsurfer (red). My heart rate was lower when wearing the cloud surfer, but as with all statistics, a sample size of 1 is hardly lots of evidence, but still, interesting to know.

The graph below shows the run cadence, but for the higher pace with the old shoes, the sensor had some problem; my own sensor shows values around 97 strides per minute.

Finally, blood lactate concentration was measured, and again the Cloudsurfer looks better. I didn’t even seem to reach steady state with my old shoes!

The graph for running economy shows how much energy per kilogram body mass and per kilometre I required; a fairly constant value for my old shoes, increasing with the pace for the Cloudsurfer.

So my personal results suggest that my heart rate and blood lactate concentration are lower when using the Cloudsurfer. How does it look for the complete sample of about 40 runners?

The next graph shows the relative difference in heart rates (y-axis) at various velocities (x-axis); specifically, how much higher the heart rate was with the old shoes.

The dot at 11 km/h and 6% (top left) for example means that this guy, who was running at 11 km/h for the test, had a heart rate that was 6% lower when using the Cloudsurfer than when using his old shoes. Everything within the red stripe is considered to be due to the form of the day, everything below or above is considered to be “more different than what could be just due to form of the day” (my words).

Simply put, there are more outliers in favour of the Cloudsurfer than the other way around. The next graph shows the distribution of average blood lactate concentration at 60% and 70% of the maximum running speed.

The Cloudsurfer values are again lower, significant on a 5% level (i.e., the probability that this difference is just coincidence is 5%).

Finally, they checked the correlation between blood lactate concentration and heart rate, or more precisely, the correlation of the difference of each between old shoe and Cloudsurfer, for 70% of the maximum pace.

The graph has to be read like this: the dot to the far right for example was somebody who had a heart rate 15 bpm lower (x-axis) and a lactate concentration about 0.75 mmol/l lower (y-axis). People in the top right quadrant had both lower heart rates and lactate concentrations, in the top left quadrant lower concentrations but higher heart rates and so on.

The black line is the straight line that, of all the straight lines you could draw, is the one with the least distance to all the data points (“least squares fit”). The R² value indicates the correlation of the concentration and heart rate differences; a value of 1 would be “perfectly correlated”, 0 would be “no correlation”, -1 would be “perfectly anti-correlated”. In this case, “somewhat correlated”.

The biomechanical test didn’t show any significant differences between the Cloudsurfer and other shoes.

What to make of this? My personal, very short analysis: seems that people could have a heart rate and lactate concentration a bit lower than with their old shoes. The people who conducted the study said “Because we didn’t find a difference in running economy, it is hard to estimate how much the reduction of heart rate and blood lactate concentration would affect race performance.”

Now, the producers of the shoe thought “let’s say something fancy about the study” and they produced this nice video with the message “Go ahead. Save 2 beats per minute.”

Sounds fair, right? If only there wasn’t the stupid facebook ad. It says “Are you training for a personal best at Greifenseelauf [large half marathon in Switzerland]? With On, you’ll run up to 8 minutes faster (ETH study). Test now.”

Well, On, that’s exactly not what the study said. The study said “it’s hard to say what the effect would be”, so where the hell did your eight minutes come from? Pfffff…

I have put close to 600 kilometres on my Cloudsurfers so far and I use them for triathlons as well since barefoot running is quite comfortable in them. One of the cloud elements has suffered a little:

But they’re still okay to use. I’m not sure I’d be willing to pay the steep price (I got mine for free for participating in the study), but maybe when I’ll be in the US, I’ll get another pair. Maybe.

“8 minutes”… come on!