Vibram FiveFingers and muscle strength … umm, err…

{POSTSCRIPT: The journal have removed this paper after they were informed of the conflict of interest by someone who had read what I wrote below. }
{POSTSCRIPT 2: The journal have published the paper with all the flaws mentioned still there. They should not have. It should be retracted}

I had heard about this study previously and was looking forward to it being published … boy, was I in for a surprise!

Effect of Vibram FiveFingers Minimalist Shoes on the Abductor Hallucis Muscle
Nicholas A. Campitelli, Kaitlyn N. Bernhard, Alan Kidon, Kristen Heard, Scott A. Spencer and Jill S. Kawalec
Kent State University College of Podiatric Medicine
Journal of the American Podiatric Medical Association; In-Press
Background: We investigated the effect of Vibram FiveFingers Bikila minimalist shoes on intrinsic foot musculature. We hypothesized that a gradual transition into minimalist shoes strengthens the intrinsic muscles of the foot, specifically the abductor hallucis muscle.
Methods: Forty-one individuals were enrolled into four groups: control (traditional shod) (n = 9), restricted walking in Vibram FiveFingers (n = 11), running in Vibram FiveFingers (n = 10), and unlimited walking in Vibram FiveFingers (n = 11). At baseline, 12 weeks, and 24 weeks, the thickness of the abductor hallucis muscle was determined using ultrasound. Statistical analysis was performed with either one-way repeated-measures analysis of variance or Friedman repeated-measures analysis of variance on ranks.
Results: The mean thickness of the abductor hallucis muscle at 24 weeks was significantly greater than that at baseline for the restricted walking (P = .005) and running (P < .001) groups. In the unlimited walking group, the mean thickness of the muscle at 12 weeks was significantly greater than that at baseline (P < .05) but not at 24 weeks. There were no significant differences in muscle thickness among the three time points for the control group (P = .432).
Conclusions: This study demonstrated that transitioning runners over a 6-month period using the 10% philosophy for increasing mileage significantly increases intrinsic muscle thickness of the abductor hallucis in all of the groups wearing the Vibram FiveFingers. The abductor hallucis muscle aids in support of the medial longitudinal arch, and an increase in this muscle thickness may help reduce running-related injuries thought to arise from arch weakness.

I just do not know where to start with this one. This meme comes to mind:
I have talked about the CONSORT statement in many posts. This is the document agreed to by the major medical journals for the reporting of clinical trials and covers the minimum and basic information that needs to be reported to judge the quality of the trial. A copy of the statement was previously published in the journal that the above study is in, so I double face palm that this paper made it to this stage with the woefully inadequate amount of information provided by the authors and the wrong analysis that they have done.

So here goes:

  • There appears to be no a priori sample size calculation done
  • There is no indication how the participants were allocated to the groups; where they properly randomized or not? What randomization method did they use?
  • The authors stated that “Demographic data such as existing activity levels and body mass index were not collected“. That is not acceptable. That information and more should have been collected. This is vital information to compare the characteristics of the groups to ensure that the randomization process was successful (assuming that is what they did). The lack of this is more than enough to reject the study from publication.
  • The authors stated that “no intra-rater reliability testing was performed because only one sonographer was performing the test.” That too is unacceptable as we have no idea if the measurements that were done in the study had any reliability and given this was a repeated measures study that information is vital. This should have been done.
  • The authors combined the left and right foot data into one data set. That is unacceptable; they should have only used one foot from each participant (for more see: Two feet, or one person? Problems associated with statistical analysis of paired data in foot and ankle medicine).
  • There is no information if there were any drop outs in the study or not. Typically, most studies do have drop outs. If they did, they should have used an intention-to-treat analysis. We do not know if they did or did not do this.
  • The authors did a within groups analysis when they should have done a between groups analysis. That too is very unacceptable and totally invalidates the study.
  • In the results, no standard deviations are given with the mean values (and oddly, the means are stated to 3 decimal points for the mm!)
  • Looking at the graph of results in figure one…a number of red flags are raised. The muscle in the right foot of the control group did appear to increase in size and did decrease in the left foot. The authors need to explain that, especially in the context of the failure to carry out intra-tester reliability of the measurements. It also appears that the baseline muscle size was bigger in the control group than the other groups. We have no idea if that is statistically significant or not and if it was, this needs to be controlled for in the appropriate between groups analysis. As the authors collected no data to compare the characteristics of each group, this can not be investigated further.

Finally, the authors state:

Conflict of Interest: None reported.

That is not the case at all. The lead author of the study has a clear conflict of interest as they are on the Vibram Five Finger Advisory Board. It is for an obvious reason that journals require these conflicts to be declared. The failure to declare this in the above case is very serious.

The irony of the timing of this publication has not escaped me. Just last week in the same journal was a review of the methodological quality of trials that were published in it which found the overall quality to only be moderate. The inclusion of the above study is certainly going to lower that even more. I urge JAPMA to have a serious look at its review processes and how such a publication can make it through the system. They should give serious consideration to withdrawing this paper (it is still in the pre-publication process) and get the authors to provide the required information and do the appropriate analysis (just like another journal did after I called out a paper of theirs in pre-publication). The University that these researchers represent need to ask questions about how something like this can happen.

The fan boys are going to love this study, just like they had ‘wet dreams‘ when the lead author previously posted faked photos on his blog. I guess the fan boys will resort to their usual trope of logical fallacies to defend the study, calling me biased and attack my analysis, but before they launch the ad hominem attack, perhaps they should answer these questions:
1. Is it ethical and honest of the author not to declare their conflict of interest?
2. Do you think it is acceptable to analyse a study with the above type of design by a within groups analysis rather than a between groups analysis?
3. Do you think its acceptable to not report that characteristics of each of the groups so they can be confirmed that they were the same at baseline?
4. Etc etc; I think you get my point.

As always: I go where the evidence takes me until convinced otherwise, and the results of the above study can not be trusted due to the woefully inadequate amount of information provided and the wrong analysis being done on the data.

Campitelli, N., Bernhard, K., Kidon, A., Heard, K., Spencer, S., & Kawalec, J. (2015). Effect of Vibram FiveFingers Minimalist Shoes on the Abductor Hallucis Muscle Journal of the American Podiatric Medical Association DOI: 10.7547/14-084.1

Last updated by .


5 Responses to Vibram FiveFingers and muscle strength … umm, err…

  1. Craig Payne August 8, 2015 at 12:34 am #

    The above study has now been withdrawn by the journal.

    • Craig Payne August 8, 2015 at 3:56 pm #

      …but its still up on PubMed:

      • Jane August 17, 2015 at 2:43 am #

        Hi Craig,

        I cannot access the original article anymore. It would be very useful for my research. Is there any other way to access it?


        • Craig Payne August 17, 2015 at 2:50 am #

          It has been taken down by the journal so is not available. I tried to email you via the email you used for the comment but it bounced. Can you get in touch via the contact form here.

  2. Mike August 10, 2015 at 7:30 am #

    Great job!

    More credible science is needed in this area, and these folks are not doing any favors to people like me looking for evidence of what makes the most sense for our feet.

Leave a Reply