1

Barefoot versus shod: physiological effort on an outdoor athletics track (Read 877 times)

Dear All,

I have attempted to use heart beats per km (which I have termed physiological effort) to assess whether running barefoot is more efficient than shod on an outdoor athletics track. The results are very preliminary, and need repeating - but, as a barefoot novice 3 km is about all I can do on my first session!

http://christofschwiening.blogspot.co.uk/2012/06/1-athletic-track-test-of-barefoot.html

Anyway, as usual, suggestions and comments are very welcome.

Greetings,

Christof

"Tentatively, it would appear that barefoot running is associated with lower physiological effort (perhaps 2%) when compared to lightweight running shoes (400g per pair), but another test with better controls is required."

I disagree with the statement that this data suggests lower physiological effort for barefoot running relative to shod running (I don't know that the reverse is true either but I don't think that even tentative conclusions can yet be drawn from the data) and I agree that another test with better controls is required.

I think that you need to generate the 10x set-beats/km curve for barefoot and for shod for both you and Max as a baseline. Without a baseline shape of the b/km versus successive sets looks like independent of shoe selection, you really cannot attribute changes to the shape of that curve to the presence or absence of a shoe. Without any sort of baseline one could argue from the data presented that because 6 of 7 shod beats/km data points (both for you and for Max) are at or below the minimum beat/km from the barefoot test, that shod running produces a lower physiological effort (this conclusion is obviously also not valid without a baseline).

Ultimately, in addition to the baseline, you will need to understand the noise in your measurement. If hypothetically, the bpm heart rate noise turns out to be +/-0.5% bpm from test to test when holding all else the same, you are looking at close to 1bpm variation independent of physiological effort which would be quite significant relative to the total range of about +/-1.5-2bpm from average that you see both for your samples and for Max's.

If possible, you may want to find a colleague who can record and compile the data for you and list the shod and barefoot tests as condition "1" and condition "2" without telling you which is which so that when you do your analysis you can prevent your knowledge of previous studies from influencing how you interpret the data.

Best of luck with your continued study!

James

For full disclosure: I think that the five finger shoes look silly and will never wear them but I have nothing against barefoot running in general

Interval Junkie --Nobby

If possible, you may want to find a colleague who can record and compile the data for you and list the shod and barefoot tests as condition "1" and condition "2" without telling you which is which so that when you do your analysis you can prevent your knowledge of previous studies from influencing how you interpret the data.

While the post deserves to be quoted in full, because it's really good advice, this last bit should be a bare minimum on how you should continue.

Confirmation bias is a b!#ch.

2016 Goals: Lose the 10lbs I gained for not having goals

Bushrat Runner

Christoph,

It appears that you may have primarily measured the importance of running a proper warm-up. If you truly want to measure the barefoot versus shod effects, the warmup should be done prior to data collection, and the barefoot segment should be randomly assigned in terms of sequence, meaning that some people would be barefoot for segments one and three while others were barefoot for segment two. And more than two subject, but you already knew that. Interesting approach to the question, though. Good luck.

Troy Hamon

Also on Kindle and Nook

I think that you need to generate the 10x set-beats/km curve for barefoot and for shod for both you and Max as a baseline. Without a baseline shape of the b/km versus successive sets looks like independent of shoe selection, you really cannot attribute changes to the shape of that curve to the presence or absence of a shoe. Without any sort of baseline one could argue from the data presented that because 6 of 7 shod beats/km data points (both for you and for Max) are at or below the minimum beat/km from the barefoot test, that shod running produces a lower physiological effort (this conclusion is obviously also not valid without a baseline).

[snip]

If possible, you may want to find a colleague who can record and compile the data for you and list the shod and barefoot tests as condition "1" and condition "2" without telling you which is which so that when you do your analysis you can prevent your knowledge of previous studies from influencing how you interpret the data.

Agreed, another better test is required! Given that this was our first barefoot outing, there was a limit to what we could do - 3km was the limit. In hindsight another two warm-up runs would have provided a better baseline. I much prefer to get a stable baseline and then change the shoes rather than do tests on different days. The inter-day variability is much higher than the set to set variability on one day. I have done lots of repeat sets in the past and the general trend is a rapid followed by a slower rise in heart beats per km. I have not seen a dip before in heart beats per km whilst using the same shoes. I agree that I need to show that - easy to do!

Hopefully, a nice stable baseline will deal with your main concern (and a decent set of post-test controls). As to the data analysis - I can post a link to the original data so that you can judge any transcription bias. Analysis bias can easily be judged by the reader from the plots of the original data. As you will have noticed there may be a 'carry' forward effect of shoe type. But, more tests are needed - no doubt about that. I will see if I can do better. And, be even more tentative with; "possible, maybe, perhaps statements!

Christoph,

It appears that you may have primarily measured the importance of running a proper warm-up. If you truly want to measure the barefoot versus shod effects, the warmup should be done prior to data collection, and the barefoot segment should be randomly assigned in terms of sequence, meaning that some people would be barefoot for segments one and three while others were barefoot for segment two. And more than two subject, but you already knew that. Interesting approach to the question, though. Good luck.

Yes, the warm-up was too short. I believe in recording and plotting the warm-up since it tells us whether we are in steady-state. The randomization would be good, but only if there are no residual effects of the previous run. I think longer breaks and longer runs would be required to fully separate the effects of each set on the subsequent one. Random isn't that important if the baseline is stable and there is a pre- and post- control. But, I don't mind either way. The problem is that there is a limited distance that can be run barefoot in one session. I will try again. I suspect it will take a good few repeats until everyone is happy that for Max and me there is or isn't a difference between the two conditions. The next question is whether  you get the same result in a race on real roads for other people. It seems like a simple question, but it takes a lot of effort to prove!

slowEET

Based on my own experience with barefoot-ish running (in my Merrell Road Gloves), it seems that an important factor that's tough to control for is technique. Before running in minimalist shoes, if I measured my effort shod 'normally' vs shod 'minimally', I would not be surprised to see equal or greater effort for the same pace in minimal shoes.

Fast forward a few months during which I ran exclusively in minimalist shoes. The other day I decided to go for a run in my Nike Free's and after 50m I almost turned around and came back. The perceived effort felt VASTLY greater than a normal run.

Seems that the only fair trial would be to observe runners who consistently alternate between shod and unshod running and have developed good muscle memory to automatically adjust technique for their shoe-ed-ness (or lack thereof). Otherwise, any difference you observe may be just the difference between well-honed technique vs not.

I agree that technique is going to be an important variable - quite how much I don't know. It would be interesting to compare how 'heatbeats per km' changes with the same shoes, same speed but different technique. One simple test might be a heel-strike versus a more midfoot/forefoot strike. But, I suspect that the differences will also be speed-dependent.

Much is often made of technique, relaxing the shoulders and arms etc but I don't have a good feel for quite how much they actually matter.