Super Shoes Non-Responder or is the Test Flawed?

Are you familiar with the scientific proverb: "Garbage in, garbage out."?

Luv2Run wrote:

A 4% improvement in economy is likely about a 1% improvement in performance. Your heart rate is not going to be able to reflect that small of a change.

Based on what? A 4% improvement in economy should be more like a 3-4% improvement in performance.

Changes in heart rate are less sensitive but they're still there. One problem with using heart rate is that it can spike without an increase in effort, unlike VO2.

So here is the data for HR vs treadmill speed. I followed the same protocol: warmup - rest until HR goes to 70, start first step of test, run for 2 minutes, then run for another 3 minutes (data collected during these 3 minutes), change speed, run for 2 minutes, run for 3 minutes (data), change speed, run for 2 minutes, run for 3 minutes (data). Then rest until HR goes to 70 and repeat for 2 more steps, as follows:

1. 8.1 MPH - 8.0 MPH - 8.3 MPH

HR: 130 , 132, 135

2. 8.3 MPH - 8.0 MPH - 8.1 MPH

HR: 133, 133, 133

3. 8.0 MPH - 8.3 MPH - 8.1 MPH

HR: 129, 135, 133

As you can see, fisky was right, the noise in the data too large to see an effect for any single step. However, when one averages the HR for each speed over the three steps, one gets:

8.0 MPH - 131.3

8.1 MPH - 132.7

8.3 MPH - 133.7

Further complicating the result is that I didint' do every possible order of speeds (only 3 out of the 6 possibilities).

So, it seems that each 1.25% difference is about equivalent to a 1 bpm HR difference - pretty hard to detect, although a 4% difference in energy usage (efficiency) should lead to a difference in HR of about 3 bpm, something that probably should have shown up.

It would be nice to try all of this on asphalt, but this is probably too much for me.

Thanks to everyone for their input in helping me solve this

HobbyJogger2000 wrote:

So what would be a better test protocol?

Contact a local university with an exercise program and see if they will measure your oxygen consumption with a metabolic cart. The paper on the VaporFlys details the protocol so it would be easy to replicate it.

In fact, it would be interesting to have the study replicated.

Hardloper wrote:

Luv2Run wrote:

A 4% improvement in economy is likely about a 1% improvement in performance. Your heart rate is not going to be able to reflect that small of a change.

Based on what? A 4% improvement in economy should be more like a 3-4% improvement in performance.

Changes in heart rate are less sensitive but they're still there. One problem with using heart rate is that it can spike without an increase in effort, unlike VO2.

No. It is not a 1:1 relationship.

Find the Real Science of Sport podcast, Ross Tucker discusses this in at least one episode about the shoes.

This paper should help.

https://www.frontiersin.org/articles/10.3389/fphys.2019.00079/full

Shows at faster running speeds the improvement in RE is not 1:1.

Luv2Run wrote:

This paper should help.

https://www.frontiersin.org/articles/10.3389/fphys.2019.00079/full

Shows at faster running speeds the improvement in RE is not 1:1.

It's not exactly 1:1 (which is why I said 3-4%), but it's close, and it's definitely not 4:1. This study actually shows lots of cases where it's 1:1 or even more than 1:1.

"With an improvement in RE of 1% (due to training, footwear, nutrition, tailwind, etc.) a recreational athlete who could typically run at 2.60 m/s (4:30:00 marathon) would be predicted to run their race 1.17% faster, finishing in ∼4:26:53, a 3 min and 7 s improvement. Alternatively, with the same 1% improvement in RE, an elite marathoner running at 5.72 m/s (2:03:00 marathon), would be able to run only 0.65% faster, finishing in 2:02:13, only a 47 s improvement."

So I don't know where you would get 4:1. That's basically saying to run 1% faster you need to consume 4% more oxygen.

HR is not going to be as easy to detect real differences compared to VO2, as previously mentioned. If you are using HR, I might just use the final minute of the 5 minute trial. As far as testing sequence, I would suggest testing each shoe twice in a mirrored order (A, B, C, C, B, A). Take the average across the two trials and compare. I would just test at one speed for now to simplify things, something slower than threshold pace.