Habitually Barefoot Children Inadvertently Resigned To Wear Ill-Fitting School Shoes: A Cross-Sectional Study


 Background

Research shows that ill-fitting shoes can negatively impact the development of the pediatric foot, in a very direct manner. The primary aim of the study was to determine if the dimensions of available prescribed school shoes fit the foot dimensions of habitually barefoot South African children and adolescents.
Methods

A cross-sectional observational study was conducted where static standing foot measurements of children and adolescents from urban and rural schools were obtained with a mobile caliper. The maximum heel-toe-length (HTL) and foot width (FW) with an added 10 mm (millimetres) toe and width fit allowance to each participant, were compared to the corresponding school shoe length and shoe width available in retail. A mixed model ANOVA was used to compare foot dimensions between gender, age and side.
Results

Six hundred and ninety-eight school children (N=698) (431 girls; 267 boys; average age 10.86 years, SD=2.55) were participants. A total of seventy-seven (N=77) black coloured prescribed school shoes currently available in retail ranging from different styles and brands were measured. Results show that, comparing the shoe length (SL) and HTL of participants, as well as taking a10 mm toe allowance into account, fifty-nine percent (59%) of children wore shoes that were not the correct length. With regards to the shoe width (SW) and the added 10 mm of width fit allowance, ninety-eight percent (98%) of the shoes worn by participants were too narrow for their feet.
Conclusions

Results confirmed that school shoes currently available in retail, are not suited for the habitually barefoot population studied. It is recommended that the shoe manufacturing industry should consider the shoe width of school shoes for children and adolescents in habitually barefoot populations to avoid the long-term negative effect of ill-fitting shoes on the pediatric foot.


Background
Foot morphology has been studied in various populations over the years. In recent years the function and development of the foot have sparked renewed interest (1). Areas on which studies have focused over the past ten years include foot movement patterns (2-6), the advantages and disadvantages of barefoot locomotion compared to shod walking (6-11), shoe development, shoe characteristics, shoe tting assessment methods (12)(13)(14)(15)(16)(17), and inter-continental differences in foot morphology and foot function (1,7,(17)(18)(19). A few recent studies have focused on the pediatric foot, and speci c footwear habits and foot development (2,7,14,20). It has been emphasized that the child's developing foot may have the same features as an adult foot but should not be treated as such. Ongoing development and speci c needs, through different stages of growth, make the pediatric foot vulnerable to internal and external in uences (15). The soft tissues and foot bones are still maturing during these stages, and the full development of these structures are only achieved late in adolescence (21). Internal factors such as age, gender, and body mass index (BMI) in uence the development of the pediatric foot. Rapid changes in foot shape and function throughout the rst fourteen years of life, coupled with high variability in static-and dynamic positions, have been reported (3,5,20,22). During standing and walking increased body height and weight lead to more foot loading and peak pressure in different areas of the foot (3). A wider midfoot section and atter foot arch were observed in boys throughout the rst nine years of development (3).
During the corresponding developmental stage, higher foot arches were observed in girls (23).
Most notable external factors in uencing the development of the pediatric foot are climate, socioeconomic status and shoe-wearing habits (18, 19,22,24). For example, growing up barefoot, compared to growing up shod, in uences pediatric foot arch morphology, foot pliability, the hallux valgus angle, rearfoot strike patterns and motor performance (6,7,25). Shoes have been identi ed as an external factor that could signi cantly in uence foot development, as well as gait in children (9,11).
Forefoot movement patterns are also signi cantly in uenced by shoes (9). Compared to barefoot walking, longer steps, increased knee and ankle range of motion, reduced foot movement, swing phase leg speed and shock absorption, as well as increased rearfoot strike patterns, were reported in children wearing shoes (11). Although shoes should primarily protect the foot from the external environment, they should still allow the foot to develop and function optimally (12,15,16,22,23). The impact of shoes on developing feet is illustrated by the fact that shoes are often used to treat foot deformities and certain musculoskeletal injuries (12). Therefore, it is important that shoes are developed to t properly, and not to interfere with the development of the pediatric foot.
Ill-tting shoes have been associated with injuries to the knees, hips, back, as well as other musculoskeletal conditions. This is due to changes in the biomechanical alignment of the body and impact on load transfer during walking (8, 12,13). It has been recommended that the tting of shoes, as well as shoe characteristics, should be considered by health professionals when treating patients with musculoskeletal injuries (12). Large portions of adults and children in Spain, Germany, the United Kingdom and South Africa are wearing ill-tting shoes (10,13,14,18,21,26), with associated pain and foot pathology (16, 27, 28). Previous research, mostly in adult populations, have shown that 46-81% of participants wore shoes that were too narrow for their feet (16). Most comprehensive studies have been done on adult feet and adult shoe design (29) and not the pediatric population.
There are several possible reasons for the prevalence of ill-tting shoes. Recent research has indicated toe-and width allowances are not applied effectively. Often too much or too little space is left for the toes inside the shoe to ensure a correct shoe t (10). The recommended toe allowance of 10mm should be considered in addition to the FL of the child to guarantee the correct t for SL (10). According to Barisch-Fritz et al. (10) toe allowance is the same for all shoe sizes and genders. A width allowance of 10mm should be considered when addressing the t for the SW (14). An important factor is that the development of children's shoes is often not predominantly in uenced by orthopaedic and biomechanical considerations, but on fashion trends (10,15,21). Shoes currently available in retail tend to focus on the correct t in SL but limited guidelines are given regarding the width of the shoe (10). South Africa's shoe manufacturers use a shoe design based on the British system, using HTL as the basic measurement. In this system, each increase in HTL will correspond with a standardized increase in foot girth, based on the Mondo Point System (28, 30). Thompson et al. (28) found that sixty-two percent (62%) of adult female participants in a study conducted in South Africa had a forefoot length-to-girth greater than the standardized length-to-girth ratio (10,15,21). Differences in inter-continental foot morphology should be considered when developing shoes for children. This should ensure the correct shoe t according to the unique foot dimensions of the child, based on previously mentioned external factors (18,19).
Unfortunately, a lack of available information on inter-continental foot morphology differences contributes to the current problems with child-friendly shoes (18). Researchers emphasize that the functional aspects of the shoe should cater for unhindered, age-appropriate development of the foot (13)(14)(15)31). Mauch et al.(18) have stressed the importance of obtaining comprehensive information on the differences in foot morphology in children across different continents, including habitually barefoot populations. South African children have an inherent culture of walking barefoot, with a signi cant majority of children between the ages of six and eighteen years reported to be habitually barefoot (7).
Warmer climates allow children to be barefoot more often, during most parts of the year, which could in uence foot development and foot morphology. For example, differences between the foot morphology of South African and German children have drawn attention to on a higher medial longitudinal arch, longer HTL, wider FW, and a difference in foot pliability in the South African study population (32). There has been limited research on school shoe t in habitually barefoot children and adolescents. In South Africa, school shoes typically form part of the prescribed school uniform. The aim of this study was to determine if the sizes of prescribed new schools typically available in retail are compatible with the foot dimensions (length and width) of habitually barefoot South African children and adolescents.

Methods
A cross-sectional observational study was conducted, with foot measurements taken of children and adolescents between the ages six and sixteen, from both urban (N=379) and rural (N=319) schools in a large region of South Africa. After receiving approval from the Department of Education, schools were randomly selected per stratum (representing a combination of regional and school models). These selected schools were contacted via email by the principal investigators.
All children and adolescents from each school, as well as their parents or legal guardians, were informed about the study through a Written Project Information Sheet. Consent forms were handed out two weeks prior to the commencement of testing. Inclusion criteria stipulated boys and girls between six and sixteen years of age, who attended one of the participating schools in the Western Cape, and who had submitted completed and signed Informed Consent and Assent forms. Participants were explicitly excluded if they had an acute foot injury at the time of testing or a severe foot deformity, in which case their feet could not be measured accurately. All participants were assessed for height and weight and self-reported their age and gender. Their shoe sizes at the time of data collection were self-reported, as well as documented by the researcher as seen inside the shoe. All measurements were taken once-off during school hours. Ethical approval was obtained from the Research Ethics Committee of the institution (REC-2018-7153) and the Health Research Ethics Committee (Project ID 14419, Ref number S20/01/008). The study was carried out in accordance with the Helsinki Declaration guidelines.

Foot measurements
Static foot measurements were done with a specially constructed foot caliper, used in previous research (7,24,26). The caliper consisted of heel cups for proper positioning and horizontal metal sliding indicators for accurate measurement, with a resolution of 1 mm (24) (Fig. 1). The reliability of this static foot measurement was shown to be in the range of "good to excellent" for children and demonstrated high reproducibility (33).
The foot HTL and FW were measured barefoot with the participant standing with the back of the heels touching the heel cups, both knees extended, and weight distributed evenly between both feet and looking straight ahead. In the standing position, the foot being measured is in an elongated position, with fty percent (50%) of the body weight distributed over each foot. For the purpose of this study the researchers wanted to obtain the largest measurement of HTL and FW to determine the static dimensions which should be accommodated by shoes (12,22). Both the right and left feet were measured. Measurements of the longest and widest feet were used for statistical analysis, with a coin ipped to determine the use of the measurements when both feet were of equal length or width (14). HTL was de ned as the distance, in millimetres, between the most posterior aspect of the foot and the most anterior part of the toes (22, 24). FW was measured, in millimetres, between the most medial part of the rst metatarsal head (MTH1) and the most lateral point of the fth metatarsal head (MTH5) (22). All measurements were recorded to the nearest 0.01 mm.

Shoe measurements
Measured sizes of new shoes ranged from United Kingdom (UK) child size number twelve (12) to UK adult size number ten (10) for boys. The girls' shoes ranged from UK child size number ten (10) to UK adult size number eight (8). Four different brands of shoes, readily available in the retail sector, were measured. No health branded shoes or shoes for special populations were included in these measurements. Measurements of SL and SW were performed on the right shoe of new school sizes. A exible plastic straw was used to measure the inside length of the shoe according to the guidelines recommend by Barton et al. (12) (Fig. 2). One end of the straw was placed in the toe area at the longest part of the shoe, with the other end, touching the heel cup. The straw was then bent at the heel and cut (12).
The straw was measured to the nearest millimetre using a steel ruler. An additional 5 to 20 mm toe allowance to the HTL measurement (12,14,15) has previously been suggested for the splaying and elongation during movement of the developing foot (12). This is to ensure a proper t of the foot in the length of the shoe. In the current study, the researcher performed standing static HTL measurements, and measured the foot during maximal extension. A toe allowance of 10 mm was added to the HTL to ensure SL t.
The width of the shoe was measured using a sliding caliper (34). The caliper was placed over the upper part of the shoe, and the measurement was taken on the most medial and lateral parts of the shoe. Speci c care was taken not to compress the shoe during measurement (Fig. 3). A SW allowance of 10 mm has been suggested to ensure a proper t (14).
The participant's shoe size and foot size were available to compare to the range of new school shoe sizes available. The HTL measurement with the added toe allowance of 10mm of each participant were compared to the ranges of SL of the corresponding new schools in retail of the same size the participant reported wearing the day of testing. The same method was used comparing the FW with the added width allowance of 10mm with the SW of the corresponding new school size.
Each participant was divided into a group depending on their HTL and added toe allowance compared to measurements of the new school shoe SL. The three groups for HTL compared to SL was named "too short", "proper t" or "too long". The FW and the added width allowance were compared to the SW and each participant was divided into one of the following groups "too narrow", "proper t" or "too wide".

Reliability of foot and shoe measurements
Two researchers completed a test-retest reliability assessment for both intra-and inter-tester reliability. The researchers did not have access to their previous measurements. Inter-rater reliability was calculated with an interval of one week. Inter-rater intraclass correlation coe cient (ICC) with con dence levels of ninety-ve percent (95%) (Cronbach's α) for foot measurements, was 0.99 for FL Left, 0.98 for FL Right, 0.99 for FW Left, and 0.93 for FW Right. Inter-rater ICC was 0.99 for SL, and 0.99 for SW.

Statistical analysis
Statistical analyses were done using Statistica 13.3.0 (TIBCO Software Inc, Palo Alto, CA, USA). Mixed model analysis of variance (ANOVA) was used to compare foot dimensions between various groups (including sex and age, as well as left and right foot). The participants were entered in the statistical model as a random effect. Gender, age, and side (left or right) were treated as xed effects. For post hoc testing, Fisher least Signi cant (LSD) difference testing was used. Statistical signi cance was accepted when p < 0.01.

Results
Foot measurements of six hundred and ninety-eight school children (N=698)(431 girls; 267 boys) between the ages of six and sixteen years were taken. The average age of the participants was 10.86 (SD=2.55). Table 1 shows the number of children per age group, as well as the number of boys and girls tested in each age group. In the total group of children, boys had signi cantly longer feet than the girls (p <0.01). Looking at agespeci c differences, no signi cant differences in HTL existed between boys and girls, until the age of twelve years. Between the ages of twelve and sixteen, signi cant differences (p<0.01) in HTL were found between boys and girls, with boys having longer feet than age-matched girls. For the whole group, the right feet were signi cantly longer than the left (p<0.01), and this applied equally so for boys and girls (non-signi cant interaction, p=0.86).
The boys in the group had signi cantly wider feet than the girls (p <0.01). Looking at age-speci c differences, there was a consistent trend for boys having wider feet until the age of 10.
Between ages eleven and fourteen, signi cant differences in FW were observed between boys and girls (p<0.01). Similar to HTL, the right feet of participants from the whole group were signi cantly wider (p<0.01) than the left feet, and this applied equally so for boys and girls (non-signi cant interaction, p=0.37). The difference in width between the left and right feet could be observed from the age of seven years in both genders.
Foot measurements of HTL and FW (with an added toe allowance of 10 mm) were compared to the range of SL and SW measurements of new shoes in the corresponding shoe size. The number of children that wore shoes that were either too short or too narrow, according to guidelines for shoe t in children, are shown in Table 2. N=number of participants Twenty-two percent (22%) (N=154) of the children tested wore shoes too short, forty-one percent (41%) (N=286) wore shoes of the correct length and thirty-seven percent (37%) (N=258) wore shoes that were too long. Concerning the SW, (with allowance of 10 mm), ninety-eight percent (98%) (N=682) of the shoes were too narrow for the feet, two percent (2%) (N=15) were of the correct width, and zero percent (0%) (N=1) were too wide for the feet.

Discussion
The current study aimed to determine if the dimensions of the currently available prescribed school shoes accommodate the dimensions (length and width) in a sample of habitually barefoot children and adolescents. The rst important nding was that for ninety-eight percent (98%) of the participants, school shoes were too narrow for their feet. This nding supports previous research, which indicated that sixtysix-point seven percent (66.7%) of participants from Spain wore shoes that were too narrow (14). The difference in results between the Spanish and the current South African study could be ascribed to the footwear habits (habitually barefoot) of South African children and adolescents (14). South Africa is regarded as a country with a warm climate and children are barefoot more often than in other countries with colder weather, which trend possibly also in uences the shape and development of the foot (18).
Researchers agree that habitually barefoot populations displayed an increased or above-average forefoot width, compared to habitually shod populations (7, 8, 23, 26, 28). In the current study, SW was measured with a sliding caliper to measure the widest part on the outside of the shoe, while previous studies measured the inner SW with specialized telescopic gauges containing protractors (14). Elena et al.(14) as well as the current study determined SW t with an added 10 mm for width t allowance. Measuring on the outside of the shoe has the limitation of the shoe material, adding to the width of the shoes. This would, however, mean that the added width of the shoe, due to the shoe material, especially supports the nding that the currently available shoes are not wide enough for the children's feet.
In the current study, fty-nine percent (59%) of children wore shoes that were not the correct length. Considering a recent study on shoe t on 505 school going children in Spain, also taking into account a toe allowance of 5-15 mm, seventy-two-point ve percent (72.5%) of school-going children were wearing ill-tting shoes (14). These ndings correlate with a previous study conducted in South Africa which reported sixty-seven percent (67%) of South African children (26) wearing ill-tting shoes. A possible reason for the differences between the current study and the previous studies is that previous studies measured the shoe after it had been worn by the participants. This poses the question whether, at the time the previous research was conducted, the shoes were in good condition or not, and if that factor could have contributed to the percentage of ill-ttings discovered. The current study investigated the shoe measurements using a newly manufactured shoe. It is, therefore, challenging to compare the results obtained from using a newly manufactured shoe to that of one that has been worn for a period. Some arguments could be made that the percentage of ill-tting shoes in the previous studies might have been either lower or higher had the shoe been measured before it was worn for some time.
Upon further inspection of the school shoes currently available in the South African retail market, it was found that a good range of school shoe lengths are available. The reason for ill-tting shoes, when considering HTL, might be the quality of information available on footwear t, which can, at times, be scarce and not scienti cally based (13). Other reasons for ill-tting shoes in HTL might be the rapid increase in HTL in children between the ages of six and fourteen years of age. Between these ages, HTL in children can increase up to 1.5 cm (centimetre) per year, after which it reaches a plateau (5,20,26).
Therefore, the shoe t needs to be checked regularly while keeping the appropriate toe allowance in mind. Accurate feedback on the shoe t, concerning toe allowance, cannot be obtained from the child, therefore, parents need to be educated appropriately to select the appropriate shoe size (15). Even though a toe allowance of 10 mm has been added the to the HTL results, the current study still reports thirty-seven percent (37%) of shoes being too long for the foot. This correlates with the ndings of Barisch-Fritz et al. (15) that toe allowance should be smaller than assumed. The recommended 90th percentile for toe allowance for females is 9.8 mm and for males 11.5 mm (15). Another possible reason for shoes not corresponding with the HTL might be the decision to change the SL in an attempt to accommodate the width of the foot for more comfort, as South African footwear do not offer a width adjustment option.
It is debatable if the children merely chose the wrong shoes size due to limited information available on footwear t. It needs to be pointed out that even if the correct school shoe size with regards to SL was selected, the SW in the ninety-eight percent (98%) of FW measurements gathered in this study would not have been catered for.
The current study supports the ndings of Mauch et al.(18) that intercontinental differences in foot dimensions are common. The reasons behind the differences might require more research. It is essential for the shoe industry to familiarize itself with these research-based differences and adapt footwear accordingly to ensure better-tting shoes. Upon investigation, it appears that shoe companies are not currently catering for inter-continental feet differences. Previous studies, which investigated foot shapes and morphological differences between habitually barefoot children and habitually shod children, reported longer feet in the barefoot population (7). This nding is supported by Shu et al. (35), who reported signi cantly larger feet in barefoot females. These results, however, do not align with the nding of a study on German habitually shod children, which reported signi cantly longer feet than their Australian, habitually barefoot, counterparts. This difference was, however, only present in younger children, and no signi cant difference was recorded for older subjects (18).
While the adult foot has been researched extensively, the pediatric foot is still a much-understudied eld (29). Even when looking at the external foot shape of a child, when compared to an adult, the structural and functional characteristics are different (22). Anthropometrical data on the pediatric foot is important when considering footwear design and shoe construction (21,29,36,37). The maximal forces and leverages of children's feet are vastly different from those of adults. Subsequently, a shoe which provides cushioning and stability to most adults, may feel hard and uncomfortable for children (21).
The market for children's shoes is not currently driven by orthopaedic and biomechanical considerations, but by consumer behaviour and trends. Children's shoes are mostly developed as downsizing of an adult shoe (21). According to previous studies, manufacturers are not taking the 3D shape of the foot into account and are unable (or unwilling) to produce a variety to cater for the differences in foot morphology (22,29).
Shoe designs for children in habitually barefoot populations should, therefore, produce a shoe to t the foot properly and mimic the shape and dimensions of the bare foot. Optimum foot development occurs when the natural shape and function of the foot is respected (8, 21,29,36,37). However, one of the most challenging tasks that remain in shoe manufacturing is the ability to access useful data to build a standard shape from available measurements (30). In most current designs, SL is used as a basic measurement, without considering the other dimensions (30).
Ill-tting shoes can have a lasting effect on the gait patterns and development of the foot, causing foot abnormalities such as a hammer-, clawed-and retracted toes (30,39). Associated pain and pathology, due to ill-tting footwear, are widely reported (11,13,16,28). Previously, eighty percent (80%) of South African adult females reported wearing shoes, which caused pain, blisters and callouses (28). Common foot deformities like hallux valgus are signi cantly induced and in uenced by shoe t being too narrow or too short (30,39). Shoes, which are too narrow for the foot, will restrict the slay of the forefoot, leading to biomechanical deviations and restriction in movement of the foot (10). Hallux valgus are believed to be avoided and corrected by selecting and wearing shoes that provide su cient space for the toes (30).
A limitation of the current study could be the assessment of only two static measurements, being the HTL and FW compared to the SL and SW. Foot height measurements should be considered when developing shoes. All school shoes in the South African retail market have the required shoelaces and/or adjustable straps over the upper part of the foot. The size over the upper part of the foot can therefore be adjusted.

Conclusions
The current study advances and elaborate on existing views on this highly relevant issue, as it places the onus on shoe manufacturing companies to enhance and improve on the basic dimensions of available shoes for habitually barefoot populations. Given the width dimensions of shoes available in the present study, the majority of children in South Africa will not have the option of a school shoe with a wider forefoot, as this is not available. Current shoe design practices in South Africa are also based on a Mondo Point System (30), which might not be the correct system for habitually barefoot populations.
Dynamic measurements will add another dimension to the development of the South African pediatric shoe. Future research to investigate the foot dimensions of habitually barefoot children and adolescents, focusing on shoe development, under dynamic conditions, is indicated. Future research should also investigate whether currently available shoes respect the foot development of the habitually barefoot population, with a speci c focus on foot pliability and arch height index.
Finding the correct shoes for habitually barefoot populations is not a new problem and, already at the start of the previous century, Hoffman (38) stated that "because of the tightness of the "modern" shoe, the habitually wide shape of the forefoot, with lots of space between the toes necessary for grasping functions, is lost." Results from the current study should create awareness of the current mismatch between the children's feet and available school shoes. Results could assist the shoe manufacturing industry, to provide welltting shoes for habitually barefoot populations. For the immediate future, better tting school shoes can assist with healthy foot development in children. In the long term, access to well-designed and footappropriate shoes could lessen foot deformities, pain and musculoskeletal injuries in adulthood. Availability of data and materials

List Of Abbreviations
The datasets used and/or analysed during the current study are not publicly available due to it currently being used in negotiations with local shoe manufacturing companies but are available from the corresponding author on reasonable request.

Competing interests
The authors declare that the research was conducted in the absence of any commercial or nancial relationships that could be construed as a potential con ict of interest.  Figure 1 Heel-to-toe-length and foot width measured with a specially constructed caliper.

Figure 2
The image shows a plastic straw cut to t the length of the shoe