Despite the recent increase in market shares and the number of investigations on minimalist shoes, our study confirmed the need for a standardised definition of minimalist running shoes. Indeed, when experts initially suggested their own preferred definition of minimalist shoes, we observed many similarities but also several differences. In fact, commonly advocated characteristics of minimalist shoes, such as lightness, high flexibility as well as low stack height and heel to toe drop were observed in the majority of individual definitions and obtained high ratings for inclusion within the standardised definition. However, many experts emphasized that additional research is needed to determine if minimalist shoes really encourage lower limb kinematics, kinetics and muscle activations similar to barefoot running. Studies that have been published so far have reported divergent findings using shoe models that were all classified as minimalist according to their authors' definition; however, characteristics were sometimes very different. This research process confirmed that there was a definite need to develop a rating scale that would allow optimising study designs and comparability between studies to better determine the effects of minimalist shoes.
During Rounds 1 and 2, the panel strongly agreed on the fact that minimalist shoes should not restrict the natural movement of the foot, ideally by having a wide toe-box until the tip of the toes (anatomical fit) that contributes to natural expansion of the forefoot during gait. Since all these elements targeted the same construct, which did not require experimental validation, experts agreed to merge them together within the definition. Participants agreed that minimal interference with the natural movement of the foot could be achieved through high flexibility as well as low weight, stack height, heel to toe drop and the absence of motion control technologies, so that the definition's formulation was acceptable. Even though these characteristics might not consistently influence all runners' biomechanics, at least not according to the current evidence [39
], they are assumed to minimise the impact on barefoot biomechanics.
Two experts disagreed with the consensus definition; one (RWW) commented that the definition should only include shoe characteristics without mentioning the foot, while the other one (BMN) commented that the definition was too specific and included, in his opinion, characteristics that were not related to minimalist shoes (e.g. flexibility).
Similar to the definition, high ratings were received for weight, stack height, heel to toe drop, flexibility and motion control and stability devices to be included within the MI. As pointed out by experts during the process, it could be argued that a total score is not needed and that all characteristics could be reported to describe a shoe. From a research point of view, a summary of scores to all subscales of the MI is certainly pertinent to explain research findings in more details. However, a total score should be seen as pertinent and needed to inform the running community. A total score in percentage was chosen for its accessibility to the general public, where 100 % represents the highest degree of minimalism, and 0 % represents very maximalist shoes. Thus, the sum of all subscores is multiplied by 4 to obtain the total MI score (see Additional file 1
, Minimalist Index rating scale). For example, a runner seeking to buy new shoes can relate on a difference in MI score to guide his transition, and adjust training accordingly so that injury risk is minimised [32
]. Although more evidence is needed to establish guidelines, it could be reasonably hypothesised that transitioning from shoes rated 10 % to others rated 30 % within one month is more likely to be safer than switching to shoes rated 80 % within the same timeframe. The name Minimalist Index
was chosen to reflect higher degrees of minimalism in higher scored shoes, even though it covers the whole spectrum of running footwear. Since the panel determined that all subscores should equally influence the total score, a lighter shoe with higher stack height could be scored the same as a heavier shoe with a lower stack height, for example. Hence, no cut-off value for dichotomising between minimalist or not can be determined at this point. Indeed, comments received about similar weighing between all factors revealed that evidence was lacking to state that specific characteristics should account for a higher proportion of total MI score, therefore accounting for a higher degree of minimalism. To our knowledge, three studies have investigated the effects of different stack heights on running biomechanics [39
] and only one study was published regarding the influence of different heel to toe drops [41
]. Thus, we suggest that research be conducted to determine if specific characteristics should be perceived as more responsible than others for a decreased interference with the natural movement of the foot, and consequently, for a higher degree of minimalism.
Since the goal of an objective scale like the MI is to quantify the degree of minimalism of a given running shoe regardless of the person wearing it, it would have been suboptimal that items related to fitting (e.g. wide toe-box) influence the MI score given that every runner's foot shape is different. Even if comfort remains the key factor when selecting appropriate running shoes, it seems illogical that improper fitting would be a factor responsible for changing how minimalist is a shoe. Most of the aforementioned characteristics can be easily quantified using a standard man size (United States size 9; United Kingdom size 8; European size 42.5), especially when weight, stack height and heel to toe drop are accurately provided by the manufacturer.
Even though agreement thresholds were reached for all subscales during Round 3, it was decided by the research team to adopt modifications for stack height and flexibility ratings. As expected, agreement rates increased following adjustments, and the detailed assessment of flexibility containing descriptions and pictures of all possible scores was very well received by participants. As for the motion control and stability technologies subscale, it was not modified although it received the lowest agreement rate. In fact, despite a few suggestions for classifying technologies depending on their influence on biomechanics, no alternative was found that was supported by the literature; therefore, it was decided that all devices would account for the same influence within this specific subscore.
It is imperative that rating scales present with adequate psychometric properties before they can be utilised. Based on results from this study, the MI proved to be highly correlated with VAS, to have the ability to differentiate between different degrees of minimalism and to be highly reliable for both inter- and intra-rater assessments [43
]. Thus, the MI can be used by retailers and clinicians to provide insightful footwear recommendations to runners.
The rating guide on how to use the MI certainly contributed to high reliability indices obtained in this study. Indeed, detailed instructions provided to participants standardised the rating process so that inter- and intra-rater reliability were excellent. Obviously, weighting the shoe produced very consistent results, except in some shoes for which the weight was close to a cut-off between two possible scores. Similar variations occurred with stack height and heel to toe drop measurements, in addition to a possibility of error secondary to suboptimal use of the electronic caliper. Since heel to toe drop is calculated using two different caliper measurements, it was expected that its reliability be slightly inferior to that of stack height. Of note, means of measurements issued by participants sometimes differed from manufacturer's specifications, potentially because factory measures were calculated without shoe insole. Participants had to assess shoes with the insole, as determined by the experts' consensus on minimalist shoes.
Flexibility and technologies assessment were expected to result in slightly inferior indices of reliability. Despite possible variations in strength applied to determine shoe flexibility, we believe that standardised instructions on the use of only three fingers per hand along with specific descriptions of every rating optimised reliability. As for technologies, it must be noted that inter-rater reliability might have been influenced by variations in materials used by different manufacturers. Hence, different raters may not interpret equally the presence of devices like "rigid heel counter" and "supportive tensioned medial upper". Nonetheless, good reliability was found for this subscale.
This work should be seen as an effort from a group of international experts to improve knowledge in the field of running footwear. Still, this study contains some limitations. First, some continents were less represented in the panel (Asia, Africa, South America), which may have influenced the study results by adding a cultural bias regarding the perception of minimalist shoes. Second, the flexibility as well as motion control and stability devices assessments may contain an evaluator-related bias. The strength applied to the shoe, and the ability to identify technologies within the shoe may vary from one evaluator to the other. To address this issue, a detailed assessment guide was created to standardise evaluation procedures (see Additional file 2
, Minimalist Index instruction guide). Third, many more technologies currently exist and others will be developed in the future, but not all of them are included within the MI. However, this panel of experts helped determine the most important ones to include, which relate to the most widely used devices in the industry.
The authors acknowledge the panel of experts for their contribution, as well as Fradette Sports and La Vie Sportive for providing running shoes.
List of contributing experts:
Ian Adamson, M.Sc., M.Eng., Healthy Running, United States.
Shawn W. Allen, DC, Allen Chiropractic Orthopedics & The Gait Guys, Chicago, IL, United States.
Christian Barton, PT, PhD. Complete Sports Care, Melbourne, Australia.
Jason Bonacci, PT, Ph.D., School of Exercise & Nutrition Sciences, Deakin University, Australia.
Nicholas A. Campitelli, DPM, Northeast Ohio Medical Associates, Kent State University College of Podiatric Medicine, United States.
Roy T.H. Cheung, PT, Ph.D., Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong.
Mark Cucuzella, MD, West Virginia University School of Medicine, United States.
Irene S. Davis, PT, Ph.D., Spaulding National Running Center, Department of Physical Medicine and Rehabilitation, Harvard Medical School, United States.
Jay Dicharry, MPT, SCS, REP Biomechanics Lab, Bend, OR, United States.
Scott Douglas, Senior content editor, Runner's World, United States.
Marlène Giandolini, M.Sc., University of Saint-Étienne, France.
Allison Gruber, Ph.D., Department of Kinesiology, Indiana University in Bloomington, United States.
Bryan Heiderscheit, PT, Ph.D., University of Wisconsin, Madison, United States.
Luiz Carlos Hespanhol Junior, PT, M.Sc., VU University Medical Center Amsterdam, The Netherlands.
Alex Hutchinson, Ph.D., Senior editor, Canadian Running Magazine, Canada.
Casey D. Kerrigan, MD, OESH shoes, VA, United States.
Peter Larson, Ph.D., Performance Health Spine and Sport Therapy, Concord, NH, United States.
Greg Lehman, PT, M.Sc., The Urban Athlete, Toronto, Canada.
Daniel E. Lieberman, Ph.D., Department of Human Evolutionary Biology, Harvard University, United States.
Everett Lohman, III, D.Sc., PT, School of Allied Health Professions, Loma Linda University, United States.
Alexandre Dias Lopes, PT, Ph.D., Universidade Cidade de São Paulo (UNICID), São Paulo Running Injury Group (SPRunIG), Brazil.
Ray McClanahan, DPM, Northwest Foot and Ankle Clinic & Correct Toes, Portland, OR, United States.
Guillaume Y. Millet, Ph.D., Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Canada.
Benno M. Nigg, Dr.sc.nat., Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Canada.
Timothy Noakes, MD, DSc, Ph.D (hc), Department of Human Biology, University of Cape Town, South Africa.
Craig Payne, DPM, Australia.
Craig E. Richards, B.Med., School of Biomedical Sciences & Pharmacy, University of Newcastle, Australia.
Michael Ryan, Ph.D., Centre for Musculoskeletal Research, Griffith University, Australia.
Jacob Schelde, MD, Occupational Health Clinic, Odense University Hospital, Denmark.
Darren Stefanyshyn, Ph.D., P.Eng., Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Canada.
Jack Taunton, M.Sc., MD, Division of Sports Medicine, Faculty of Medicine, University of British Columbia, Canada.
Daniel Theisen, Ph.D., Sports Medicine Research Laboratory, Public Research Centre for Health of Luxembourg.
Ross Tucker, Ph.D., Research Unit for Exercise Science and Sports Medicine, University of Cape Town, South Africa.
Ivo F. Waerlop, DC, Summit Chiropractic & Rehabilitation & The Gait Guys, Dillon, CO, United States.
Joe Warne, B.Sc., School of Health and Human Performance, Dublin City University, Ireland.
John D. Willson, PT, Ph.D., East Carolina University, United States.
Richard W. Willy, PT, Ph.D., East Carolina University, United States.