Effects of Shoe Cushioning and Body Mass on Injury Risk in Running

The design of this study is a randomized controlled trial with an intervention period of six months and a biomechanical analysis of running pattern the day of inclusion in the study. Running footwear is provided by a renowned sport equipment manufacturer working in close collaboration with the Sports Medicine Research Laboratory (SMRL) of the Luxembourg Institute of Health (LIH), which in turn is responsible for the management and execution of the study. Random allocation of the study shoes, recruitment and follow-up of the participants, biomechanical testing, as well as data management and analysis will be carried out by the SMRL. The study design is based on the comparison between 2 running shoe prototypes, which only differ with respect to the cushioning (i.e. stiffness). The cushioning properties of both shoe versions will be within the range of values of models available on the market.

Recruitment:

Participants will be recruited through advertisements in local newspapers and press releases within the Grand-Duchy of Luxembourg during the months of September 2017 to January 2018. The benefits for the participants are strongly highlighted during the recruitment phase, and are as follows: 1) the use of a sports diary developed by the SMRL and called TIPPS (Training and Injury Prevention Platform for Sports, www.tipps.lu) to record all their sporting activities, quantify progress, and keep track of their sport-related injuries, 2) a free pair of running shoes, and 3) a brief report with the results of the biomechanical testing as well as their anthropometric measurements. However participants will be made aware that they will randomly receive one of the 2 study shoe versions to be worn during the observation period for all their running activities.

Registration procedure:

The participants must understand and agree on the randomized design of the study and that they will be allocated to one of the two running shoe groups. Anyone wishing to take part and believing themselves to meet the inclusion criteria will be invited to follow the process described hereafter:

Volunteers first have to create a personal account on TIPPS, pre-register to the study via their personal account, and answer an online inclusion/exclusion questionnaire as well as a baseline questionnaire. To finalize the registration, the participant will have to set an appointment with the research team via the TIPPS platform. Answers to both questionnaires will be assessed by the investigators during the initial visit.

Data collection:

The data collection phase of this project will last from September 2017 until July 2018. It consists of 2 different parts: an epidemiological study with 6-month follow-up and a biomechanical analysis of the participants' running pattern.

The follow-up period will start as soon as the participants have received their study shoes (from September 2017 to January 2018) and will finish at the latest on the 31st of July 2018.

The biomechanical analysis of the participants' running pattern will occur during their initial visit, using the study shoes that they will have received.

The initial visit of the participants at the SMRL will be the occasion 1) to verify the inclusion criteria, 2) to validate their answers to the online questionnaire, 3) to determine the appropriate shoe size, 4) to allocate each participant to one of the study arms according to the randomization list (see stratified randomization), 4) to deliver the shoes, 5) to collect some anthropometric measures, and 6) to perform the biomechanical test on the treadmill.

Stratified randomization:

A stratified block randomization will be prepared by the Competence Center for Methodology and Statistics of the LIH before the beginning of the recruitment. Participants will be stratified according to their sex because body mass (one of the two variables of interest) as well as many anthropometric characteristics are largely dependent on sex. Each block will include 40 participants.

Shoe types:

The study shoes are prototypes and will be anonymized for the purpose of this trial. The sole of the shoes will be customized so that the two running shoe prototypes will be exactly the same (same midsole, same outsole, same upper), except for their cushioning properties which will differ by more than 35%, while remaining within the range of the shoes available on the market. The difference in cushioning properties between the shoe versions will be created by modifying the midsole material, i.e. chemistry, density and therefore the hardness of the Ethylene Vinyl Acetate (EVA) foam. In order to provide accurate data on the technical specifications (i.e. stiffness ) of each shoe model, a set of 40 shoes (10 pairs per condition) will be tested for cushioning properties by the manufacturer according to a standardized protocol (Impact test: ASTM1614, Procedure A).

Data on exposure:

Data on running practices is collected using the TIPPS system. Required information in the sport activity report includes the type of activity, context, duration, subjectively perceived intensity, distance, shoe pair used, running surface (hard or soft), and whether the participant had experienced any pain during the session forcing him/her to reduce practice volume or intensity, or to interrupt the practice. Session intensity is determined using the Borg's rating of perceived exertion scale, a subjective 10-point scale.

Data on outcome:

The primary outcome is the first running-related injury. A consensus definition of running-related injury in recreational runners has been recently published. The definition of running-related injury is a "running-related (training or competition) musculoskeletal pain in the lower limbs that causes a restriction on or stoppage of running (distance, speed, duration, or training) for at least 7 days or 3 consecutive scheduled training sessions, or that requires the runner to consult a physician or other health professional." Similar to uploading a training session or competition, the TIPPS provides a complete yet easy to fill in questionnaire when reporting an injury. Information regarding the following is required: injury date, context, sports discipline, injury mechanism (acute or progressive), anatomical location, type of injury, description and estimated return date. Running-related injury will be classified according to the Orchard Sports Injury Classification System version 10 (OSICS-10). Injury severity will be measured in days of modified or interrupted training.

Anthropometric measures:

The body composition (i.e. the proportion of muscle and fat tissues) will be assessed by bioelectrical impedance analysis (Tanita SC-240 MA).

Leg length will be defined as the measure between the anterior superior iliac spine and the medial malleolus, and is referred to as the "direct" clinical method. Additionally, the distance between the great trochanter and the ground will also be measured for the assessment of leg stiffness.

Biomechanical testing:

The biomechanical testing consists in a running test on an instrumented treadmill and will be performed in the study shoes, according to the random allocation. The test (10 minutes) consists of a 5-minute warm-up followed by a 5-minute run at the self-declared preferred (habitual) running speed. Two records of 45 seconds will be obtained over the last 2 minutes of the test.

Additionally, the participants who reported a preferred running speed equal to 10 km/h (+/- 1 km/h) will be invited to perform a second test at the end of the follow-up period. This second test will consist in 10 minutes of running in each shoe model. Records will be obtained during the last 2 minutes of each run. This will allow a within subject analysis of the shoe effect on running biomechanics at a standardized speed.

Sample size:

A sample size calculation for Cox regression was used to determine the number of participants needed for the primary hypothesis of the study. With an alpha of 0.05 and a power of 80%, an average injury rate of 30%, an expected HR=1.50 between groups, 50% of participants randomized to each shoe group and an expected drop-out rate of 20%, the total number of participants required is 802.

A within subject analysis will be performed on a subgroup of participants to investigate the effect of shoe condition on VGRF. A total sample of 39 participants will be required to detect a difference of 0.16 body weight (standard deviation: 0.25 body weight) between shoe conditions with 80% power and a significant level of 5%.

Statistical analysis:

Descriptive data for the personal, anthropometric, biomechanical and training-related characteristics will be presented as count and percentage for dichotomous variables, and as mean and standard deviation, or as median and range, respectively, for normally and non-normally distributed continuous variables. Average sport-related characteristics will be computed for each participant over their specific period of observation. Shock absorption properties of the two types of shoes will be compared using a Student's t test. A two-way analysis of variance (ANOVA) will be used to determine whether any difference in running biomechanics results from the shoe cushioning properties or body mass.

Cox proportional hazards regressions will be used to compute the hazard rates (HR) in the exposure groups, using first-time injury as the primary outcome. Date of inclusion (baseline evaluation date) and date of injury or of censoring will be basic data used to calculate the time at risk, which is expressed in hours spent running and defined as the time-scale. A participant will be right-censored if injury unrelated to running or severe disease caused a modification of the running plan, or at the end of follow-up. The assumption of proportional hazards will be evaluated by log-minus-log plots.

Unadjusted Cox regressions will be performed to present the crude estimates of HRs for shoe model, body mass and other potential risk factors such as running biomechanics variables and training-related characteristics. Body mass is an exposure that can change over time (time-dependent covariate). This means that each participant could move between exposure states continuously (every month in our study). A delayed entry will be used in the unadjusted Cox regression model for body mass.

Subsequently, the variables with a P value <0.200 will be included in the adjusted Cox regression analysis to determine whether shoe cushioning and/or body mass are associated with injury risk, controlling for potential confounders. The recommendation for using at least 10 injuries per predictor variable included in the Cox regression analysis will be strictly followed.

Finally, to investigate if the effect of shoe cushioning on injury risk is modified by body mass, a stratified analysis will be performed. HRs and their 95% confidence intervals (CI) will be determined within each stratum. All analyses will be performed using STATA/SE version 14.