Abstract
Background It is unclear at what age the medial longitudinal arch (MLA) of the foot becomes stable in children. The influence of footwear on MLA development also is unknown.
Objective The purpose of this study was to examine the MLA development of children using a longitudinal approach. The relationship between wearing different types of footwear and MLA development also was explored longitudinally.
Design This was a longitudinal cohort observational study.
Methods The MLA of 111 healthy children (mean age=6.9 years, SD=0.3) was evaluated using 3 parameters (arch index [AI], midfoot peak pressure [PP], and maximum force [MF]) extracted from dynamic foot loading measurements at baseline (t0), 10-month follow-up (t1), and 22-month follow-up (t2). Information on footwear usage was surveyed. Linear mixed modeling was used to test for differences in MLA over time.
Results The MLA of the children remained stable over time (AI: t0/t1/t2=0.25 [95% confidence interval (CI)=0.24, 0.26]/0.25 [95% CI=0.24, 0.26]/0.25 [95% CI=0.24, 0.26]; P=.95). When the children's sex was considered, the AI of boys decreased (higher arch) with age (0.26 [95% CI=0.24, 0.27]/0.25 [95% CI=0.24, 0.27]/0.25 [95% CI=0.23, 0.27]; P=.02). Boys also displayed a flatter MLA than girls at age 6.9 years (AI: mean difference=0.02 [95% CI=0.01, 0.04]; P=.02). At baseline, children who wore closed-toe shoes displayed the lowest MLA overall (AI: closed-toe shoes/sandals/slippers=0.26 [95% CI=0.24, 0.28]/0.24 [95% CI=0.23, 0.25]/0.25 [95% CI=0.24, 0.26]; P<.01). Children who used slippers at toddlers' age experienced a higher PP (flatter arch) in later childhood than those who wore sandals (mean difference=31.60 kPa [95% CI=1.44, 61.75]; post hoc P=.04).
Limitations Information on the type of footwear worn was self-reported and, therefore, may be subjected to recall bias.
Conclusions The MLA of children remained stable from 7 to 9 years of age. The child's sex and the type of footwear worn during childhood may influence MLA development.
Flexible flatfoot may affect musculoskeletal performance and the risk of injury in children.1,2 Consequently, some clinicians recommend treatment before the age of 6 years,3,4 whereas others discourage any intervention for asymptomatic flatfoot, as this condition may resolve on its own as the children grow.5–11 However, a recent review suggested that the chances of this condition resolving naturally after the age of 10 years remains slim.12 In general, a decreasing trend of flatfoot that corresponded with increasing age has been reported.13–16 This observation might explain that the medial longitudinal arch (MLA) of the foot is actively developing alongside the child's growth.14
The MLA is thought to develop most rapidly until the age of 6 years,8,17,18 after which changes are less apparent.7,19 There is, however, much debate on MLA development during the critical age range of 7 to 9 years. Within this age range, some studies demonstrated that the MLA remained fairly stable,19–22 whereas other studies showed that the foot arch was becoming higher.23 The foot arch also was reported to become flatter from 7 to 8 years of age and to show a reversal to a higher arch from 8 to 9 years of age.15,16 Therefore, it is still uncertain at which point the MLA stops developing in children. In addition, it is currently unknown whether MLA development differs between boys and girls, although cross-sectional studies suggested that boys tend to have a flatter foot type than girls.11,13–16,24,25 There is a need to clarify the conflicting reports on MLA development in boys and girls during the critical age range from 7 to 9 years using a longitudinal approach.26 Subsequently, health care practitioners can make informed decisions when managing flexible flatfoot in children.
The use of shoes in early childhood also may have an influence on MLA development.27–30 Optimum foot development seems to occur in the barefoot environment,27,28,31 whereas flatfoot has shown to be closely associated with wearing shoes at an earlier age.13,29,32,33 Sachithanandam and Joseph33 proposed that shoe wearing before the age of 6 years would predispose a child to a flatter foot later in life compared with another child who wears shoes only after the age of 6 years. Hence, some researchers have discouraged wearing of shoes in children because they believed that barefoot walking would enhance ligamentous strength, which eventually may help in arch development.2,24 The confirmation of the relationship between footwear and foot arch development from a longitudinal data set will be useful for clinicians to advise parents on the type of footwear most suitable for healthy foot arch development in children.
In short, there is (1) disagreement on whether the MLA is still developing higher or lower or remains stable in children aged 7 to 9 years and (2) a lack of longitudinal data displaying the relationship between footwear usage and MLA development in children. Therefore, the primary aim of this study was to clarify whether the MLA was still developing higher or lower or remained stable in children from age 7 to 9 years using a longitudinal study approach. A secondary aim was to confirm the association between footwear usage and MLA development. Our first hypothesis was that there would be no significant changes in the MLA of children from 7 to 9 years of age. This hypothesis was in line with the only other longitudinal study that showed a stable foot arch in children within the same age range.19 Second, we hypothesized that there would be no significant differences in MLA development among children who wore different types of footwear.
Materials and Method
Study Design
A longitudinal study was conducted to examine the MLA of a cohort of primary 1 children (mean age=6.9 years, SD=0.3) at baseline (t0) and followed up at 10 months (t1) and 22 months (t2) using dynamic footprint measurements. We decided to follow this cohort over 22 months because of the discrepancy in the literature on foot arch stabilization from 7 to 9 years of age.15,16,19–23
Participants
Two hundred sixty-nine primary 1 children (equivalent to grade 1 elementary school children in North America), approximately 7 years of age, from a local neighborhood school were invited to enroll in this study through convenience sampling. Parents or legal guardians of 122 children gave written informed consent for their children or wards to participate. Children were excluded if they had a history of foot surgery; conditions that affected the nerves, muscles, foot, and leg postures; or pain or injury that required non-weight bearing greater than 1 month at the time of the study.20,24 Nine children met the exclusion criteria, and 2 children were unable to follow the MLA evaluation protocol. These 11 children were subsequently excluded from the study. In total, 111 participants were enrolled at baseline and followed up at 10 and 22 months (Tab. 1).
Demographics of the Children at Each Time Point Over the Study Perioda
Background Survey
At baseline, survey questions were developed to collect information on the exclusion criteria (eAppendix). Information on footwear usage, such as type and frequency of footwear worn by the participants, also was collected. Details of the footwear factors are presented below.
Footwear Factors
Four questions that pertained to the age of the children when footwear was first used, type of footwear worn, and frequency of using the footwear were developed:
When did your child or ward first put on footwear?
At the onset of footwear use, what kind of footwear did your child or ward put on most of the time?
At the onset of footwear use, how often did your child or ward put on footwear?
Excluding school hours, what kind of footwear does your child or ward put on most of the time?
Questions 1 through 3 provided information on footwear usage at toddlers' age during which walking was developing. Question 4 referred to the type of footwear worn by the children at baseline (age 7 years). These 4 footwear factors were chosen because previous studies reported that the type29 and frequency33 of footwear worn and age when footwear was first used33 had influences on the foot arch structure. The parents or legal guardians were provided with pictorial examples on the types of footwear worn by most children in Singapore (Fig. 1). These footwear included sandals, slippers, and closed-toe shoes with and without fastenings. Each parent or legal guardian was allowed to select only one choice of footwear each for questions 2 and 4 for their child.
Pictorial examples of footwear typically worn by Singaporean children. (A) Sandals: open-toe footwear with adjustable fastenings (eg, straps, self-adhesive fasteners, or buckles) at the forefoot, midfoot, or heel to hold the foot firmly. (B) Slippers: open-toe footwear without adjustable fastenings, commonly known as thongs, flip-flops, and slip-ons; excludes bedroom slippers meant for nighttime use. (C) Shoes without fastenings: closed-toe footwear without adjustable fastenings but with a heel counter. (D) Shoes with fastenings: closed-toe footwear with adjustable fastenings (eg, laces, straps, self-adhesive fasteners, buckles) at the midfoot to hold the foot firmly and with a heel counter.
From the responses to the survey, footwear factors were categorized into various levels (Tab. 2). In order to ensure an adequate number of children in each footwear subgroup during analysis, those who wore closed-toe shoes with fastenings (n=11) and without fastenings (n=10) at baseline were combined into a single closed-toe shoes subgroup. Children who first put on footwear later than 2 years of age (n=8) were not analyzed for this particular footwear factor due to the small sample size.
Categorization Levels of Footwear Factors
MLA Arch Assessment
The MLA was assessed from dynamic footprints obtained from a 2-step gait approach using the emed M pressure platform (Novel GmbH, Munich, Germany) at all 3 time points. The pressure platform consisted of 2,792 sensors (4 sensors/cm2) arranged within an area of 395 mm × 240 mm, preset in the factory to record footsteps at a sampling frequency of 50 Hz. It was placed flat on the ground, embedded midway along a raised (2.25-cm) jigsaw-fitted, high-density foam walkway (7.06 m × 0.86 m).34 Data were recorded utilizing proprietary emed/E software (version 23.1.14e). Each child was positioned 0.7 m away from the edge of the platform and, upon instruction by the tester, commenced walking with self-selected speed along the foam walkway, placing the second footstep onto the platform and halting at the fifth step. A total of 3 footprints from each foot were collected, alternating between the left and right feet. This method has shown to be reliable for evaluating footprint geometry and foot loading in children by Tong and Kong.35
Gait speed during the 2-step protocol also was measured using timing gates (TC-System, Brower Timing Systems, Draper, Utah), as this protocol has been shown to affect dynamic footprint loading characteristics.36 The infrared light timing gates were mounted on tripods adjusted to each child's neck height (Fig. 2). The first set of timing gates (start time) was placed across the near edge of the emed platform, and the second set of timing gates (stop time) was placed 1.0 m ahead.
Footprint and midfoot loading measurement protocol with placement of the first set of timing gates at (a) and placement of the second set of timing gates 1 m ahead at (b). Gait speed is computed as 1 m divided by the time taken for the child to walk from the first set of timing gates to the second set of timing gates.
Data Reduction
Subsequently, 3 geometric and loading parameters were extracted from each footprint using emed “Multiprojects-E” software (version 23.1.14e). These parameters included the arch index (AI), midfoot peak pressure (PP), and maximum force normalized to body weight (MF [percentage of body weight]), which characterized the MLA.35 The AI is calculated as the midfoot contact area divided by the total foot excluding the toes, and the PP and MF are defined as the maximum pressure and force (expressed as a percentage of body weight), respectively, experienced on the midfoot. A preliminary scan of the AI acquired from the children showed variations among all 3 footprints on the left and right feet. Step-to-step variability can be expected even when measurements were taken under the same conditions.37 As the footprint data may potentially be skewed when the mean value of the 3 footprints is analyzed, the median AI was used in the analysis instead.38 Similarly, the footprint that corresponded to the median AI was used for the extraction of midfoot PP and MF for each participant.
Data Analysis
Linear mixed modeling was used to examine for changes in the MLA over time with post hoc Bonferroni corrections. This modeling approach is able to correct for missing data that arise from any loss of follow-up at a given time point.39 A first model was built to test for differences between the left and right feet in all 3 parameters over time. The second model included participant's sex as a factor during analysis to observe any differences in MLA development between boys and girls. Thereafter, each footwear factor (Tab. 2) was entered separately into the third model to observe any differences in the 3 outcome parameters among footwear groups. Gait speed was entered as a covariate to adjust for its influence on the 3 MLA parameters in all models. Subgroup analyses were conducted for any parameter that displayed an interaction with time. All analyses were conducted with IBM SPSS software, version 19 (IBM Corp, Armonk, New York), with significance set at P<.05.
Role of the Funding Source
This study was funded by the National Institute of Education Academic Research Fund. The funding source did not play a role in conducting or reporting any aspect of this work.
Results
No significant differences were found in AI (P=.15), midfoot PP (P=.84), or MF (P=.91) between the left and right feet. There is currently no consensus among investigators whether to analyze both feet separately or only one represented foot.40 As symmetrical foot loading has shown to improve with increasing age,41 only the trial that corresponded to the median AI out of the 3 right footprints measured was used for further analysis.
Changes in MLA Over Time
In the total sample, there were no significant changes in AI (P=.95), midfoot PP (P=.61), or MF (P=.79) (eTab. 1). These results suggested that the MLA remained unchanged over time during the study. The results of each MLA parameter also are presented as the 3rd, 50th, and 97th percentiles over the 3 time points separately for boys and girls (Fig. 3).
Third, 50th, and 97th percentiles of the arch index and midfoot loading over the 3 time points between boys and girls: (A) arch index for girl, (B) arch index for boys, (C) midfoot peak pressure for girls, (D) midfoot peak pressure for boys, (E) midfoot maximum force for girls, (F) midfoot maximum force for boys. * Significant difference between boys and girls at 6.9 years of age (P<.02). †Main effect of time in boys (P=.02).
Influence of Sex on MLA Development
No overall differences between boys and girls in any MLA parameter were observed. A significant interaction between sex and time was found for AI (P<.01) but not for other MLA parameters. In the subgroup analysis, the AI of boys decreased (higher arch) significantly over time (P=.02), but post hoc comparisons were insignificant (Fig. 3B). This finding implied that the MLA was still developing for boys at age 6.9 years. No significant changes in the AI of girls were noted. Moreover, the AI of boys at t0 (age=6.9 years) was significantly higher (flatter arch) than that of girls (mean difference=0.02; 95% confidence interval [CI]=0.01, 0.04; P=.02; Figs. 3A and 3B). No differences in the AI between male and female participants were found at t1 and t2.
Influence of Footwear Factors on MLA Development
Age of children when they first put on footwear.
There were no significant differences in the MLA parameters (AI: P=.56; PP: P=.73; MF: P=.59) between children who first put on footwear before 1 year of age and those who commenced footwear use at 1 to 2 years of age (AI: mean difference=0.00, 95% CI=−0.02, 0.02; PP: mean difference=−6.83 kPa, 95% CI=−22.08, 8.41; MF: mean difference=1.4%, 95% CI=−3.7, 6.6). No significant interaction between the age of children at the onset of footwear use and time was observed in any MLA parameter.
Type of footwear worn by the children at the onset of using footwear.
Overall, there was a significant difference in midfoot PP among children who wore different footwear at the onset of footwear use (slippers: mean=118.38 kPa, 95% CI=100.00, 136.76; closed-toe shoes with fastenings: mean=99.39 kPa, 95% CI=88.55, 110.23; closed-toe shoes without fastenings: mean=90.17 kPa, 95% CI=74.53, 105.81; sandals: mean=86.78 kPa, 95% CI=74.19, 99.37; P=.02). Post hoc results showed that children who wore slippers frequently at the onset of footwear use experienced higher midfoot loading (more pronated) than those who wore sandals (mean difference=31.60 kPa; 95% CI=1.44, 61.75; post hoc P=.04). No overall differences between footwear subgroups were noted in other MLA parameters. A significant interaction between the type of footwear worn at onset and time was found for AI (P=.02) but not for other MLA parameters. Subgroup analysis showed that the AI of children who used slippers increased (flatter arch) over time (P<.01; eFig. 1), but no post hoc differences between any time point were noted. There were no significant changes in AI for the other footwear subgroups.
Frequency of wearing footwear at the onset of footwear use.
Regardless of the wearing frequency at the onset of footwear use, no significant differences in any MLA parameters were noted overall (eTab. 2). A significant interaction between footwear usage frequency at onset and time was observed in midfoot MF (P=.01) but not in other MLA parameters. Subgroup analysis showed that only the MF of children who used footwear (onset) less than 3 days per week increased (more pronated) significantly over time (P=.01; eFig. 2), but no post hoc differences between time points were noted. No significant changes in the MF were observed in the other footwear (onset) usage frequency subgroups.
Type of footwear worn by the children at baseline.
Participants who frequently wore closed-toe shoes at baseline (age=6.9 years) displayed the highest AI (flattest arch; P<.01) and midfoot loading (PP and MF [more pronated]: P=.04) among all footwear groups (Fig. 4). These results suggested that children who wore closed-toe shoes at 6.9 years of age tended to exhibit a lowered arch than those who wore sandals and slippers. However, post hoc comparisons showed no significant differences in any parameter among the types of footwear. There was no significant interaction between the type of footwear worn at baseline and time in any MLA parameter.
Overall differences in arch index and midfoot loading among children grouped by the footwear worn at baseline. Data are presented as mean difference (95% confidence interval). AI=arch index, MF=maximum force, PP=peak pressure.
Discussion
Our study showed that the MLA in children remained unchanged from 7 to 9 years of age, clarifying the contradictory debates in the literature regarding the age at which the foot arch stabilizes.13–16,19,29 The current study also demonstrated some relationships between footwear usage and foot arch development from a longitudinal data set.
MLA Development
In our study, Singaporean children exhibited a consistent AI at 0.25 and experienced stable midfoot loading from 7 to 9 years of age. These results provide strong evidence that the foot arch structure has already reached a stable state of development during this period. When sex was considered, boys displayed a flatter foot type (higher AI) than girls at 7 years of age, and their arches became slightly less flat (AI decreased) from 7 to 9 years of age. These findings suggest a delay in reaching a stable MLA in boys compared with girls. The later development in boys may explain why many previous cross-sectional studies showed a flatter MLA in boys than girls of the same age.11,13–16,24,25 Taking flexible flatfoot as a developmental process in the first decade of life, one study showed that more than 90% of the treatments on children with flexible flatfoot were unnecessary.24 Findings from our study confirmed that flexible flatfoot is unlikely to develop much further in children from age 7 to 9 years. Such information may guide clinicians to make informed decisions when managing flexible flatfoot in children. For example, when a child is seen with painful flexible flatfoot, the podiatrist or physical therapist may choose to intervene with orthotic therapy knowing that the child's foot arch structure might not develop much further.
Our findings that foot arch structure remained stable from 7 to 9 years of age are in line with other studies reporting a relatively stable trend in the AI of Australian20,21 and German19,22 children (eFig. 3). It must be highlighted, however, that Singaporean children displayed a flatter foot type (higher AI=0.25) than children of European (lower AI=0.19–0.20)19,22 and Australasian (lower AI=0.20–0.21)20,21 descent. These conflicting results in the literature may be attributed to the ethnic influence in MLA development. In previous studies, Brazilian children demonstrated higher dorsal arches than children in the United States and Japan,42 whereas Australian children showed a lower prevalence of flatfoot compared with German children.30 Mauch et al30 postulated that the observed ethnic difference may be associated with footwear, as the Australians preferred barefoot walking or wearing thong-styled and open sandals as a result of a warmer climate. Although Singapore is hot and humid, the children in our study used both open-styled footwear and closed-toe shoes (Tab. 2). These findings suggest that the disparity in AI between Singaporean and European children is likely related to the ethnic difference and not footwear usage between the populations.
Although all MLA parameters remained stable over time in the current study, we must caution that our results are unable to show the MLA development before 7 years of age or after 9 years of age. The only other longitudinal study showed that the foot arch of German children remained fairly stable from 7 to 9 years of age (AI≈0.20).19 Given that the MLA parameters of Singaporean children remained very stable for 22 months, it is unlikely to expect substantial changes in foot arch structure after 9 years of age.
Footwear Factors Associated With MLA Development
Currently, the literature suggests that children do not require shoes until they begin to walk43 or when the environment necessitates shoe wearing (eg, to protect the foot from frostbite in winter).31,44 When shoes are worn, they should be well-fitting, soft, and lightweight and have soft cushions, essentially protecting the feet.28,31,43 Optimum foot development appears to occur in the barefoot environment.27,28,31 However, our findings provide reasoning to re-examine these recommendations in greater detail.
Footwear use for school-aged children.
Our study showed that children who often wore closed-toe shoes at age 7 years displayed a lower MLA than those who wore sandals and slippers. This finding is in line with the notion that children who wore closed-toe shoes exhibited a flatter foot type than those who were habitually barefooted13 or used sandals or slippers29,33 and other forms of footwear.45 Children who walked with closed-toe shoes when they were aged 7 to 10 years13,29,45 consistently displayed a higher proportion of flat feet than those who walked barefoot or with sandals or slippers. Our study agrees with these findings that wearing closed-toe shoes frequently in school-aged children is associated with a flatter foot type. Generally, it is believed that slipper users exercised their intrinsic foot muscles readily by gripping the slippers to prevent them from falling off.29 As a result, the MLA could be strengthened to achieve the truss-like structure. Although such a theory was recently supported by an increase in foot arch height before and after intrinsic foot muscle training on a cohort of adult participants,46 a randomized controlled trial showed no differences in navicular height after similar training.47 Other researchers observed an increased overpronation (flatter foot) and pressure at the medial midfoot after intrinsic muscular fatigue.48,49 Collectively, these studies suggest that intrinsic muscular strength may play a role in influencing foot arch structure.
Footwear use during walking development.
We also examined the association between footwear usage at toddlers' age and MLA development in later childhood. We found that children who often wore slippers at a young age developed a flatter foot type (higher midfoot pressure) than those who used sandals when they reached 7 to 9 years of age. This finding suggests that wearing loosely bound footwear at toddlers' age may lead to a flatter arch in later childhood. In our study, the majority of the children commenced footwear use at 2 years of age or earlier (n=103; Tab. 2), which corresponds well to the age of walking onset for healthy children.31,50 Wearing slippers at a very young age (≤2 years) may have caused intrinsic foot muscle fatigue instead of strengthening the muscles, leading to lower arch development in later childhood. Although both slippers and sandals are open-toe footwear, slippers do not have adjustable fastenings that could secure the foot firmly to the footwear. On the other hand, closed-toe shoes may hold the toddlers' feet too firmly in place. We postulated that toddlers' foot arches require some fastenings to support walking development but not to the extent of the support offered by closed-toe shoes, which may be too restrictive. Therefore, toddlers can benefit from wearing sandals instead of slippers for healthy foot arch development. Closed-toe shoes may be an option, although their influence on foot arch development is unclear.
The finding that wearing loosely bound footwear at toddlers' age may lead to a flatter arch in later childhood appears to contrast with findings of previous studies showing a lower prevalence of flatfoot among children who were habitually unshod13 and those who wore sandals or slippers.29,33 It should be noted, however, that the referenced age at which footwear type data were collected varied substantially among studies. Earlier studies reported the footwear type at the time of study where children were aged 6 to 12 years,45 4 to 13 years,29 and 3 to 12 years.13 Only Sachithanandam and Joseph33 reported the age at which footwear was first used, but the range was wider (1–5 years/6–15 years/16 years and older). Taken together, these findings indicate it is likely that the footwear worn at different ages during childhood may influence MLA development. Further studies may be warranted to investigate the influence of footwear usage on foot arch development at different ages during childhood.
Clinical Recommendations
Clinicians should be informed that footwear might influence MLA development. They may advise parents to discourage toddlers from wearing slippers during the stage of walking development. More securing footwear, such as sandals, can be considered instead. Once the children reach school age, parents can consider loosely bound footwear, which is closer to the barefoot environment,29,31,51,52 rather than closed-toe shoes to promote healthy foot arch development. It is inconclusive whether less frequent use of footwear at toddlers' age may be beneficial to foot arch development.
Limitations and Future Directions
First, the follow-up age range of the children was too narrow. To provide more insight, future studies can consider following 3 separate cohorts of children with overlapping ages simultaneously (eg, 5–8 years, 7–10 years, and 9–12 years). Second, the approach to measure footwear factors was limited to self-reporting methods. Recall bias could have skewed the results, especially for the age at which the children first used footwear and the type of footwear worn at that time. Third, we surveyed footwear usage at onset only by age (<1 year, 1–2 years, >2 years), without considering the footwear type worn at the onset of walking, which may play an important role in influencing foot arch development. More information on the age and footwear usage when a child commences walking will be useful. Fourth, the survey questions were not validated. Future studies may need to assess footwear by visually observing the type of footwear worn by children in situ. Finally, we did not measure intrinsic foot strength, which may influence foot arch height, as suggested by the literature. Therefore, it will be beneficial to quantify intrinsic foot strength among children who use various types of footwear and between children with neutral and flatter foot types.
In conclusion, this study showed that foot arch structure of children, defined by footprint geometry and foot loading, remained stable from 7 to 9 years of age. Boys showed a slight delay in the arch development compared with girls. Children who often wore closed-toe shoes at age 7 years had a flatter foot arch than those who used sandals and slippers. Those who used slippers at toddlers' age developed a flatter foot arch when they reached later childhood compared with those who wore sandals. Clinicians may choose to commence therapy when a child is seen with painful flexible flatfoot at age 7 years and may discourage younger children from wearing slippers when they commence using footwear. However, it must be cautioned that our study was unable to show the MLA development of children before 7 years of age and after 9 years of age.
Footnotes
Both authors provided concept/idea/research design, writing, data collection, data analysis, project management, and fund procurement. The authors thank Ms Crystal Wee and Mr Kelvin Chua, who coordinated the data collection; Dr Chan Yiong Huak for rendering his biostatistics expertise; Dr Masato Kawabata for his helpful discussion on data analysis; and all students and teachers who assisted in the project.
This study was approved by the Nanyang Technological University Institutional Review Board.
This study was funded by the National Institute of Education Academic Research Fund.
- Received March 30, 2015.
- Accepted January 25, 2016.
- © 2016 American Physical Therapy Association
References
Issue highlights
