This protocol is reported according to the SPIRIT statement for improved reporting of study protocols [24].
Study design
This is a binational multicenter, cross-sectional observational study looking at differences of foot mechanics and motor performance between habitually barefoot and habitually shod children and adolescents aged 6–18 years. Ethical approval has been obtained from the ethics committee of the medical association Hamburg (protocol number PV4971) and Stellenbosch University ethics committee (protocol number HS1153/2014). The regional separation of the recruitment is due to the obligation of footwear use in most German educational institution while in South Africa the habit of being barefoot prevails.
Participants
After pilot testing for reliability and validity of the measurement apparatus, recruitment of participants exclusively will take place in rural and urban primary and secondary schools with no restriction to school type. In South Africa, primary school attendees are aged 6–13/14 and secondary school children are between 13/14–18 years old, their German counterparts are 6–9/10 and 9/10–18. With approval from the German and South African supervisory school authorities, schools will be randomly selected per stratum (representing a combination of district and type of school) and contacted by the principal investigators. Schools (in blocks of five primary schools and five secondary schools) will be initially contacted via email and when interested visited by the study staff for further organisation. If the school wants to participate, consent forms for all pupils (and their parents) will be provided in the appropriate language (English, Afrikaans, Xhosa or German, Additional file 1). No limit will be set per school for maximum number of participating children per age. We will strive for an equal distribution of portion of participants per school to ensure equal representation. For participation, pupils will be requested to bring along the signed consent form on the testing day.
Inclusion criteria will consist of healthy children that are physically active for at least 120 accumulative minutes per week (parent reported). Children and young adolescents between the age of 6 and 18 will be recruited for this study. Exclusion criteria will be evaluated by parent proxy report and consist of current injuries of the lower extremity, abnormal gait or any neurological or neuromuscular abnormalities likely to influence the gait.
Testing procedure
Methodological planning stipulates all anthropometrical, foot and motor performance measurements to be performed during a physical education lesson. Prior to the testing, a physical activity questionnaire for children/adolescent (PAQ-C and PAQ-A [25, 26]) and a barefoot questionnaire will be distributed by the teachers and collected by the study staff on the testing day and briefly checked for completeness. All children bringing the signed consent form and voluntarily want to participate will be gathered and all relevant information for the testing will be given by the principal investigator. After a short warm up period (jogging for < 5 min), participants will be randomly allocated to the seven testing stations:
-
1)
Anthropemetry
-
2)
Foot caliper
-
3)
Pressure plate
-
4)
20 m sprinting
-
5)
Lateral jumping
-
6)
Standing long jump
-
7)
Backwards balancing
After their first testing station, a participant will be transferred to the next vacant station without a fixed order. Testing station 1–3 will be completed only barefoot, while testing stations 4–7 will be performed in barefoot and shod conditions. The order of the condition is randomized a priori on the registration sheet which is distributed to the participants on the testing day. A flowchart displaying the participant flow through the study is shown in Fig. 1.
Barefoot questionnaire
Participants will complete a six-item questionnaire used to decide if a child can be considered barefoot or shod. This questionnaire (Additional file 2) is developed specifically for this study and inquires if the child is barefoot on a three point Likert scale: most of the time, half of the time or barely/none of the time during 1) school 2) sports and 3) in and around the house. These questions will be asked twice: one for the warm weather and one for cold weather. For those in secondary school, questions are repeated for when they were in primary school. Due to the multi lingual culture of South Africa, the questions will be asked in Afrikaans, English and Xhosa. Children will be considered barefoot when they are always barefoot in and around the house and always barefoot in school or during sports -in warm weather- in primary school. In secondary school they have to have a similar level of barefootness during primary school and are always barefoot in and around the house currently.
Outcomes
Prior to the start of the investigations in the schools, the German and South African research team will perform a joint training session over several days in Germany to ensure the identical use of the equipment and collection of data. Furthermore, the principal investigator (KH) will attend the first weeks of testing in South Africa to ensure accuracy in methodology and identical data collection. Inter-rater reliability testing will be performed to improve data quality and interpretability.
Foot mechanics
Foot mechanical outcomes will consist of dynamic arch index (dAI), static arch height index (sAHI) whilst sitting and standing (double limb support), foot and arch height pliability ratios, hallux angle, and footstrike pattern while jogging and running.
The dAI describes the proportion of the middle third of a footprint compared to the whole footprint area (except for the toes) and was firstly described by Cavanagh and Rodgers [27] (Fig. 2). This method was shown to be valid and reliable in children [28, 29]. The dynamic footprints geometric will be acquired with a capacitance-based pressure platform system (Emed n50, Novel GmbH, Munich, Germany) using a two-step protocol [29, 30]. The platform has 6080 sensors in an area of 47,5 × 32 cm (4 sensors/cm2) and has been shown to be reliable in adult [31] and paediatric populations [29]. In order to level the platform to the ground, it will be embedded in a 300 cm wooden walkway.
Footprint data will be used to calculate the hallux angle according to R Donatelli and SL Wolf [32].
Measures of static foot anthropometrical data will be obtained with a specially constructed caliper (Fig. 3). This caliper consists of heel cups for the placement of both feet and sliding indicators for proper measurement of heel-to-toe length (HTL), foot width (FW) and dorsum height. HTL will be defined as the distance from the most posterior aspect of the foot to the most anterior part of the toes. Dorsum height will be measured at 50 % of HTL and the static arch height index (sAHI) will be defined as the ratio of dorsum height and HTL:
$$ Static\ arch\ height\ index = \frac{Dorsum\ height}{Heel-to- toe\ length} $$
Feet of the participants will be measured at sitting (10 % of body weight (BW)) and standing (50 % of BW) and the pliability ratio will be calculated according to Kadambande et al. [20]:
$$ Pliability\ ratio = \frac{HTL\ 50\%\ of\ BW \times FW\ 50\%\ of\ BW}{HTL\ 10\%\ of\ BW \times FW\ 10\%\ of\ BW} $$
Foot strike patterns will be captured during 20 m jogging and both sprinting trials in each condition at the 17.5 m mark using a wide-angle high speed camera (GoPro HD Hero 4, GoPro Inc., San Mateo, California, USA). The camera will be positioned 150 cm from the midline orthogonal to the marked running way and set to record 120 frames per second at a resolution of 1280 × 720 pixels. After testing, videos will be processed using a video editing software (Adobe Premiere Pro CS 6, Adobe Systems, San Jose, California, USA) and rated independently by two reviewers, with a third experienced reviewer for consensus. A rearfoot strike is defined as a first ground contact with the heel or the rear third of the foot, while a forefoot strike is present when the anterior part of the foot first contacts the ground. For a midfoot strike heel and anterior part of the foot contact the ground simultaneously. This method has been used successfully in other studies to determine the foot strike pattern [33].
Motor performance
Participants will complete multiple tests to assess motor performance; 20 m sprint [34], lateral jumping, standing long jump, and backwards balancing [35]. Each station will be performed in two conditions: barefoot and wearing sport shoes.
Preceding the lateral jumping (Additional file 3), participants will stand in one half of a square next to a line, indicated on the floor by masking tape. He/she will be instructed to jump sideways as fast as possible for 15 s. One minute rest will be given between the two trials per condition and the average score of each condition will be used for analysis. For the standing long jump (Additional file 4), a participant will be instructed to stand with their toes adjacent to a labelled start line, bend at the knees whilst swinging the arms backwards and subsequently, jump as far as possible on to a soft rebounding mat. One of the researchers will place a rod at the most posterior of where the participant landed and will read off the distance on the measuring tape attached parallel to the rebound mat on the floor. This will be repeated three times per condition and the distance of the best trial per condition is used for analysis. A sprinting lane of 20 m (Additional file 5) will be indicated on the floor by a start line labelled by tape and two pylons at 0, 10 and 20 m, as well as one meter after the 20 m mark as a dummy. A time sensor device will be placed at the start, 10 and 20 m. In Germany mobile magnetic timing gates from Humotion SmarTracks (Münster, Germany) will be used and Brower Timing Systems speed gates (Salt Lake City, UT, USA) in South Africa. The 10 m and 20 m time of the best trial per condition will be used for analysis. Lastly participants will be asked to walk backwards (Additional file 6) over a 6 cm, 4.5 cm and 3 cm wide balance beams with, respectively, 2 trials per beam [35]. The first step after the starting position will not be counted, every subsequent step will then be counted until one foot touches the ground or a maximum of 8 steps per beam is achieved. The children will be instructed to look up straight (with an X on eye level) and put their next step directly behind the other foot. The scores of two trials on each of the three beams will be added to a total score per condition for analysis (max 48 points).
Data collection and management
Data will primarily be collected on paper sheets and then transferred to electronic spreadsheets at each testing site (Germany and South Africa). The entered data will be checked by the principal investigators. The pedobarographic data will be collected within the provided software (Novel database pro m, Version 24.3.20, Novel GmbH, Munich, Germany) and then exported as ASCII files and included into the spreadsheet. Electronic data will not include any confidential participant information and will then be transferred via secure servers to Germany for further processing and statistical analysis. As specified in the ethically approved data protection declaration for the participants, the encoded data will be stored on external hard disks in a locked safety cabinet for 10 years. Participants have the right to obtain the personalised data. Data analysis and publication will only be done in an anonymised format.
Sample size
Although we did not opt to distinguish primary or secondary outcome measures, the dAI was used to enable a sample size calculation. Muller et al. [36] has shown that the average dynamic arch index (μ: 0,19) and the accompanying standard deviation (SD: 0065) are stable between the ages of 6 and 13. The minimal important difference was set at 20 % of the average (0.19; i.e., 0.038). With a two-sided significance level of 0.05, and assuming a power of 0.8, a minimum of 16 participants per age group, per country, had to be included for this study. Due to the diversity of the study, additional power calculations have been performed. Based on the arch height ratio derived from Waseda et al. [22] the sample size for arch height ratio (navicular height*100/foot length) should be n = 12 per age group (μ = 15,0; SD = 2,6). Furthermore, based on a preliminary study on children, the sample size based on the standing long jump should be n = 18 (μ = 152,9; SD = 31,5), for lateral jumping n = 22 (μ = 38; SD = 9,2) and for 20 m sprint n = 12 (μ = 3,81; SD = 0,64). Therefore, our n = 20 per group seems to be sufficient. In order to allow for these differences and other unknown variances within the other variables we choose to increase the number from 16 to 20 participants per age group, per country.
Statistical analysis
Descriptive data will be presented using descriptive statistics. The participant’s barefoot questionnaire and the PAQ outcomes will be compared to all attending children in 2 primary (1 rural, 1 urban) and 2 secondary schools to assess external validity. The outcome measures will be evaluated for normality using Shapiro-Wilkins and visually using P-P plots, if possible, non-normal distributed data will be adjusted. Mixed Models linear regression will be used to assess if the foot mechanical and motor performance outcomes differ between the barefoot and the shod participants (fixed factors). Furthermore, we will test differences due to being barefoot or shod change by age by adding an interaction of age*group to the linear regression. Differences in foot strike pattern during jogging and sprinting between barefoot and shod participants will be assessed using ordinal regression. The school of a participant will be added to the model as a random effect in order to adjust for possible differences between the schools and their geographical location. Furthermore, Gender, BMI, PAQ-Score, ethnicity and inside/outside testing will be tested for confounding in all models (regression coefficient changes >10 %). We hypothesize that older boys could have higher arches (lower dAI) [37] and perform better on motor performance tasks, while girls’ feet will show a higher pliability. BMI increases the dAI and possibly decreases motor performance [38, 39]. A higher level of physical activity (higher PAQ score) is probably related to better motor performance. And lastly, Caucasians could have higher arches (lower dAI) [40].
We anticipate that gender might modify the effect of being barefoot or shod on the outcome measures and therefore gender will additionally tested as an effect modifier (interaction between outcome and gender p < 0.1). In all cases, a significance level of 5 % is pre-stipulated.