Kinematic changes after 12 weeks of running in minimalist running shoes

One of the potential shortcomings of kinematic, kinetic and metabolic studies of barefoot or minimalist running versus traditional shod running is that acute interventions are often used, rather than a prolonged period of habituation to the different conditions under study. It is not known if what is found in these acute intervention studies will still have the same findings after neuromotor adaptation has taken place to different running forms over a period of time. A new study published last week in the International Journal of Sports Medicine set out to look at those kinematic changes after 12 weeks of a controlled transition program to minimalist running:

12 Weeks of Simulated Barefoot Running Changes Foot-Strike Patterns in Female Runners.
McCarthy C, Fleming N, Donne B, Blanksby B.
Int J Sports Med. 2013 Sep 18.
To investigate the effect of a transition program of simulated barefoot running (SBR) on running kinematics and foot-strike patterns, female recreational athletes (n=9, age 29±3 yrs) without SBR experience gradually increased running distance in Vibram FiveFingers SBR footwear over 12 weeks. Matched controls (n=10, age 30±4 yrs) continued running in standard footwear. A 3-D motion analysis of treadmill running at 12 km/h-1 was performed by both groups, barefoot and shod, pre- and post-intervention. Post-intervention data indicated a more-forefoot strike pattern in the SBR group compared to controls; both running barefoot (P>0.05), and shod (P<0.001). When assessed barefoot, there were significant kinematic differences across time in the SBR group for ankle flexion angle at toe-off (P<0.01). When assessed shod, significant kinematic changes occurred across time, for ankle flexion angles at foot-strike (P<0.001) and toe-off (P<0.01), and for range of motion (ROM) in the absorptive phase of stance (P<0.01). A knee effect was recorded in the SBR group for flexion ROM in the absorptive phase of stance (P<0.05). No significant changes occurred in controls. Therefore, a 12-week transition program in SBR could assist athletes seeking a more-forefoot strike pattern and “barefoot” kinematics, regardless of preferred footwear.

Basically they took 30 female runners, randomized them to a control group that kept running in their traditional shoes or to a group that was transitioned to Vibram FiveFingers over a 12 week period. They looked at number of kinematic variables in each condition while running on a treadmill. The main findings on barefoot kinematics at baseline were consistent with all the previous studies:

  • shorter ground contact time
  • more of a midfoot and forefoot strike pattern
  • more plantarflexed ankle at foot strike
  • lower peak knee flexion
  • lower knee range of motion

After the 12 week intervention the only further change in these differences was greater ankle plantarflexion at toe-off. This means that there are differences between the ‘acute intervention’ of barefoot/minimalist running vs a 12 week period of adaptation, but this was the only variable that they found a difference in. So they concluded that changes in the motor pattern can be made in the 12 week period used for the study:

The findings of this study indicate that changes in motor patterns in previously habitually shod runners are possible and can be accomplished within 12 weeks

There are a number of shortcomings of the study and contexts that the results need to be interpreted in:

  • the data was collected on a treadmill, so this may or may not be an issue, depending on how close you want to believe that treadmill running mimics overground running.
  • it was a study on kinematics and not kinetics. Kinematics is all about the motion. Kinetics is about the forces driving those motions and it is the loads associated with kinetics that are responsible for injury risk, so we do not know from this study what happened to the kinetics (ie loads in the tissues).
  • almost a third of the original participants dropped out of the study which is unusually high. Dropouts happen in studies, but not normally this many. However, the characteristics of the dropouts were similar in each group, so its not necessarily an issue.
  • Those fan boys who objected to the results of the bone stress study in those wearing Vibrams on the grounds that the 12 weeks was not long enough will of course be objecting to the results of this study on the same grounds (even though it was exactly what Vibram were recommending).
  • this was NOT a study on which of the two (minimalist vs traditional shod) was better than the other. It was just describing the differences after the 12 week transition.

While I have no problem with the conclusion of the study, I do have a problem with some of the unsubstantiated broad statements made by the authors, such as:

There is emerging evidence that a FFS [forefoot strike] pattern such as developed over time by SBR [simulated barefoot running] could have performance benefits [ 16 , 17 , 24 ] and perhaps lead to lower injury rates [ 8 , 13 ]

References 16 and 17 had nothing to do with performance or economy! Reference 24 was cherry picked to make a point, when the preponderance of studies are showing that this is NOT the case. References 8 and 13 on injury rates are really another poor example of cherry picking at its worse. The preponderance of evidence says the opposite, so the authors cherry picked two studies to support their point. They ignored all the other studies that contradicted their point. This shoddy poor use of references should not happen in an academic publication (my students would have been severely marked down for this if they did this in an essay). The editor and reviewers should not have let this through.

As always, I go where the evidence takes me until convinced otherwise, and this study tells us that neuromotor adaptation to minimalist running can occur over a 12 week period. Its just a shame that the paper gets ruined by their cherry picking, misinterpreting, misquoting and misuse of references.


McCarthy C, Fleming N, Donne B, & Blanksby B (2013). 12 Weeks of Simulated Barefoot Running Changes Foot-Strike Patterns in Female Runners. International journal of sports medicine PMID: 24048910

Last updated by .

, , ,

8 Responses to Kinematic changes after 12 weeks of running in minimalist running shoes

  1. Colm McCarthy September 29, 2013 at 3:42 am #

    Craig; thank you for reviewing our paper on your blog. Any website/blog which makes scientific literature digestible, and critiques new research in a field with so much new information is very valuable. I enjoy your sceptical outlook and think it is a good antidote to the over-enthusiasm and unproven claims which tend to flourish with new trends. I have a few disagreements with your review on our recent publication however.
    You have correctly identified limitations/contexts of the study, most of which we have discussed in the paper. The high dropout rate was due in part to our stringent criteria that injury, engaging in rehab or not training for >3 weeks might be confounders in any kinematic changes observed. As some kinematic changes (in barefoot in particular) approached, but did not achieve statistical significance, this high drop-out rate was unfortunate for us. A higher final “n” might have made things more interesting statistically. However being strict on adherence and injury allows us be confident that the changes we did observe were as a result of the intervention.
    The “broad unsubstantiated statements” line I cannot agree with. (“could have performance benefits” is rather tentative I think!!). To clarify, reference 16 (Hasegawa et al. 2007; “Foot Strike Patterns of Runners at the 15km Point during an Elite Half Marathon”) observed that the percentage of midfoot-strikers was highest in the first 50 runners, and decreased the further down the field they observed. I.e. the faster runners were more likely to have a non-rearfoot strike. Reference 17 (Hayes & Caplan 2012; “Foot Strike Patterns and Ground contact times during high calibre middle-distance races) observed that when all 181 runners in a seeded graded meet of 800m and 1500m races were analysed, forefoot and midfoot strikers had significant faster average race speeds than rearfoot strikers. Would you agree these papers have something to do with performance?
    I do agree that the findings on running economy are mixed, and dependent on the population studied and factors controlled for.
    Regarding the “cherry picking” of articles 8 (Daoud et al.) and 13 (Goss & Gross); in an attempt to locate the “preponderance” of fruit I apparently left on the tree regarding injury rates and foot-strike pattern (see your link), I failed to find Kleindienst’s 2003 study in PubMed, or in a google search. Walther 2005 was similarly elusive, although I did eventually find an article in German that may correspond? I would appreciate if you could post the links to these papers. Therefore it would appear that I cited 100% of the published articles available to me in your embedded link, imperfect though they may be for the reasons you give. Perhaps the editors and reviewers of my paper had the same difficulty finding these other studies/conference abstracts. I also would add that I cannot locate any studies that “say the opposite” i.e. that FFS running gives a greater risk of injury than RFS (with due respect to risk of metatarsal injuries, see my references to Ridge et al. and Slazler et al.). Maybe you have misinterpreted or misquoted? Again, if I have neglected an important study please provide the link.
    Finally on cherry-picking (or not picking), I note that you did not mention, at all, the findings of our study regarding shod gait following the SBR intervention. There are a number of statistically significant foot strike and kinematic changes to shod gait in regular shoes following the 12 week intervention (see abstract). We believe these may be clinically significant findings for those who might seek to alter gait or loading patterns in runners, for whatever reason that may be. To my knowledge this effect on shod gait has not been identified in the literature before (although I’m sure many older running coaches will find the concept familiar).
    We agree that healthy debate is good for science and for patients. We are also glad to have the opportunity to defend our findings, and interpretation thereof.
    Yours sincerely,
    Dr. Colm McCarthy.
    MRCPI, MICGP, FRACGP, MSc.(Sports and Exercise Medicine)

    • Craig Payne September 29, 2013 at 6:21 am #

      Thanks Colm. I have no problems with the conclusions of the study – its an important and timely finding. The only issue was the context in which the references were being cited to support the other statements being made when there are references showing the opposite.

      • Craig Payne October 1, 2013 at 9:02 pm #

        Sorry, the studies that you missed were linked to in my post above.

        eg on the injury rates. The conclusion was: “perhaps lead to lower injury rates [ 8 , 13 ]”
        Ref 8 was on 54 almost elite level track runners and ref 13 was a self selected web based study that no one is taking seriously. But both did show a higher rate of injury in heel strikers.
        Why did you leave out these two studies on big samples of typical runners that showed the opposite?:
        Kleindienst (2003) was a retrospective review of 471 typical runners that found no difference between rearfoot and forefoot strikers concerning the frequency of injury.
        Walther (2005) was a retrospective review of 1203 runners that also found no difference in incidence of injury between rearfoot and forefoot strikers in the rate of injury.

        Since then there are two more studies showing no differences (though they are only in abstract form): I discussed all this here (in the link I mentioned above): http://www.runresearchjunkie.com/foot-strike-pattern-and-injury-rates/

        re: “could have performance benefits [ 16 , 17 , 24 ] ”
        All the most recent references are showing heel striking is more economical or that there is no difference. You picked the only one that showed midfoot/forefoot was more economical: I reviewed most of the studies in the link above: http://www.runresearchjunkie.com/running-economy/

  2. Colm McCarthy October 2, 2013 at 6:17 am #

    Hi Craig, yes, I saw the link to your description of the studies, but I cannot access their full text, or abstract.

    Can you post a PubMed link, or journal link so I can try to read them?

    I have read 50 or more journal articles relating to barefoot/shod kinematics, injury rates, economy etc, and I have never seen either of these studies even referenced.

    I “left them out” because a standard literature review search could not find them. See a recent review in Sports Medicine by Murphy, Murray and Matzkin on barefoot running and injury, which does not cite them either. Is there a secret repository of articles the rest of us are not party to?

  3. Marc s May 9, 2014 at 12:34 am #

    Craig, why have you yet to respond to Colm’s apparently reasonable question?

  4. Neil Fleming September 17, 2014 at 3:56 pm #

    Hi Craig,
    i took a look at the reference list which you linked in your last response (dated May 9th).

    I was surprised to see the articles which you have quoted are either not fully peer-reviewed, conference abstracts, or not even available in English! You seem puzzled as to why “for some reason” these articles are never mentioned in the literature. Indeed, you asked my colleague Dr. McCarthy why we omitted 2 such articles (Kliendienst, 2003; Walthur, 2005). The reason would seem fairly obvious: they are not peer-reviewed and/or are not freely available to the greater scientific community.

    To maintain a high standard of scientific rigor, most authors will not reference work which has not yet undergone the full peer-review process. This is not “cherry-picking”; it is just sound scientific practice. The type of practice which i teach my students each semester (and certainly do not chastise them for). If we don’t have access to a full methodology which has been peer-reviewed and published, then we don’t consider the results of such work to be strong evidence of anything. To be clear, that policy applies to both sides of an argument. We ignore without bias all data which has not undergone peer-review, regardless of whether it supports or refutes our hypotheses.

    Your website’s motto “i go where the evidence takes me” suggests a level of objectivity and scientific rigor. Yet you repeatedly criticise peer-reviewed articles in favour of unpublished data, which most scientists would either readily ignore or at best view through a skeptical lens. As such, i feel it is a little disingenuous to accuse us of “cherry-picking” when you are the one quoting conference abstracts to support your arguments…

    Best regards,
    Neil Fleming, PhD.
    Asst. Prof. of Exercise Science,
    Indiana State University

  5. Luke January 6, 2015 at 10:34 am #

    Hi Craig,
    Could you please direct me to the post that summarises load patterns related to different pitches in particular reference to the hip. Thanks! Really enjoy your blog…

Leave a Reply