Introduction

In an attempt to better understand some of the musculoskeletal complications associated with childhood obesity, several studies have investigated the effects of obesity on foot structure and function (1, 2, 3, 4). Although these investigations have repeatedly documented that obese primary school children typically display flatter feet relative to those of their leaner counterparts, the cause of this increased area of contact between the feet of obese children and the ground is unknown. It has been postulated that the flatter feet of obese children may be caused by the existence of a plantar fat pad underneath the midfoot region. It is known that a fat pad is present underneath the medial longitudinal arch of the infant foot while the arch develops, although this fat pad is thought to resolve between the ages of 2 and 5 years as the arch of the foot is formed (5, 6). Riddiford-Harland et al. (2) speculated that this midfoot plantar fat pad might remain in the feet of obese children as a protective adaptation to cushion the loads associated with their excess mass, in turn causing their feet to be characteristically flatter than those of their leaner counterparts.

Alternatively, it has been suggested that the flatter feet of obese children may be caused by a collapse of the medial longitudinal arch due to excessive loading of the feet as a result of continually bearing additional body mass. Such a structural collapse can develop into a potentially disabling problem in later life, as proper mechanics of the longitudinal arch are critical to normal foot function (7). Although this notion of a longitudinal arch collapse is purely speculative, it highlights the need to understand the cause of flat feet in obese children. However, to date, no studies have investigated this issue. Therefore, the purpose of this study was to determine whether the flat feet displayed by young obese and overweight children are due to the presence of a thicker midfoot plantar fat pad or a lowering of the longitudinal arch relative to that in non-overweight children.

Research Methods and Procedures

Subjects

Five male and 14 female preschool children (mean age, 4.3 0.9 years; mean height, 1.07 0.1 m; mean BMI, 18.6 1.2 kg/m²), without other health problems, who were classified as overweight or obese according to international standard BMI cut-off points on the basis of age and sex (8), were selected as experimental subjects. An additional 19 non-overweight children, matched for age, height, and sex to the overweight/obese children, were selected as non-overweight controls (mean age, 4.3 0.7 years; mean height, 1.05 0.1 m; mean BMI, 15.7 0.7 kg/m²). The subjects were recruited from a larger sample of all consenting children (n = 95) from 10 randomly selected preschools in the Illawarra region of New South Wales, Australia. All recruiting and testing procedures were approved by the University of Wollongong Human Research Ethics Committee, and all parents gave written informed consent for their children to participate in the study.

BMI

Each child's height was measured to the nearest 0.1 cm using a PE87 portable stadiometer (Mentone Educational, Victoria, Australia), and the mass of each child was measured to the nearest 0.05 kg using calibrated electronic scales (Model HD646; Tanita Corp., Tokyo, Japan). Both mass and height were measured, according to standard procedures (1, 2), with the children's shoes and socks removed. These data were then input to calculate each child's BMI, an internationally accepted field measure for defining childhood obesity, using the standard Quetelet Index protocol: body mass divided by height squared (kg/m²). The classification system proposed by Cole et al. (8), which is based on the age and sex of the child, was used to classify each child's overweight and obesity status.

External Foot Anthropometry

To characterize the external structure and arch height of each child's feet, 10 anthropometric dimensions were directly measured twice (three times if the values were not within 3 mm of each other) for the right and left foot of each child, while the children stood erect, eyes looking forward. Foot, instep, and ball of foot lengths; instep and ball of foot circumferences; heel and ball of foot widths; and dorsal arch, plantar arch, and ball of foot heights were measured to the nearest 0.1 cm, according to the protocols described by Parham et al. (9), using an anthropometer (Lafayette Instrument Co., Lafayette, IN), a metal tape measure (KDS Corp., Kyoto, Japan), a combination square (Empire Level, Mukwonago, WI), and a custom-made footboard. To maximize reliability of the data, the chief investigator (K.J.M.) measured all anthropometric data from each of the 38 subjects using the same apparatus and procedures.

Arch Index

Footprints were recorded for each child's right and left foot using a pedograph (Suavepie, Capital Federal, Argentina), replicating the procedures of Riddiford-Harland et al. (2). Each child slowly lowered one foot (of which the underside was inked) onto the membrane of the pedograph and then stood motionless in an equal weight-bearing anatomical position for 2 seconds, while looking straight ahead, before quickly removing the foot. Two footprints of both the right and left feet were taken to obtain a permanent image of the plantar surface of the foot in contact with the ground during weight bearing.

To characterize the degree of flat-footedness, each child's arch height was estimated from the footprints using the arch index (10). To calculate arch index, the footprints were initially scanned using ArcSoft Photo Studio 5 software for Canon (ArcSoft, Freemont, CA) and digitized using SigmaScan 3.0 (Jandel Scientific, San Rafael, CA). The length of the footprint (excluding the toes) was then divided into three equal parts (rearfoot, midfoot, and forefoot), whereby the arch index was calculated by dividing the area of the midfoot by the area of the whole foot (excluding the toes). The arch index was selected to estimate the flatness of each child's feet because it has been found to be reliable and has a moderate to high relationship with the vertical height of the medial longitudinal arch in children, quantified by directly measuring the height of the navicular tuberosity (3).

Midfoot Plantar Fat Pad Thickness

A portable SonoSite 180PLUS ultrasound system (SonoSite, Bothell, WA) with a 38-mm broadband linear array transducer (10-5 MHz, maximum depth 7 cm) was used to measure the thickness of the plantar fat pad (cm) in the midfoot region of each child's right and left foot. Each non-weight bearing ultrasound image was collected with the child sitting on a chair with the leg extended and the foot resting in a relaxed position on the lap of the chief investigator (K. J. M.), who took all measurements. To standardize the site at which the midfoot plantar fat pad thickness was assessed, the ultrasound probe was placed longitudinally, vertically below the dorsal navicular landmark on the plantar surface of the foot and aligned with the second phalange (Figure 1). When a clear image was obtained, the system was paused and the thickness of the fat pad (accuracy, 0.01 cm) was quantified using the measurement function of the ultrasound system. Three trials were collected on both the right and left foot of each child, with the mean value per foot taken as representative of left and right midfoot plantar fat pad thickness.

Figure 1.

Placement of the ultrasound probe on the foot, directly below the dorsal navicular landmark and aligned to the second phalange.

Full figure and legend (51K)

Reliability of the Anthropometric and Plantar Fat Pad Measurements

Before testing, all anthropometric and plantar fat pad measurements were assessed for six children (age range, 3.8 to 4.4 years) unrelated to the main study, on three different days and three times per foot. Of the 20 anthropometric measurements (10 per foot), intraclass correlation coefficient values ranged from R = 0.70 to 0.99, with 18 values above R = 0.80. The intraclass correlation coefficient values for the midfoot plantar fat pad thicknesses were R = 0.85 and R = 0.96 for the right and left foot, respectively. Therefore, the external foot anthropometry and plantar fat pad thickness measurements were considered highly reproducible (11).

Statistical Analysis

Preliminary analysis of the data indicated very few significant differences in foot structure between the right and left feet of each child. Therefore, data for one foot (either right or left) were selected, based on a random number generation (Microsoft Excel for Microsoft Office 2000; Microsoft Corp., Redmond, WA), to represent each child's foot structure, and data pertaining to that foot were included in the statistical analysis. A Kolmogorov-Smirnov test (with Lilliefors' correction) was used to test all data for normality, and non-parametric tests were used if the data violated the assumption of normality. Independent t tests were then conducted to compare external foot anthropometry, arch index, and midfoot fat pad thickness data obtained for the overweight/obese children relative to those for the children with normal BMI to determine whether there were any significant differences (p 0.05) between the two subject groups. All statistical analyses were conducted using SPSS software (SPSS 11.5 for Windows; SPSS Corp., Chicago, IL).

Results

Descriptive data pertaining to the seven foot dimensions characterizing foot lengths, breadths, and circumferences of the two subject groups are illustrated in Figure 2. The overweight/obese preschool children displayed significantly larger foot dimensions than their leaner counterparts for all seven foot dimensions. Furthermore, the overweight/obese children had a significantly higher arch index (0.26 0.05; p = 0.03) than their non-overweight counterparts (0.20 0.9) (Table 1), whereby a higher arch index is associated with a flatter foot structure. However, no significant difference was found when comparing the thickness of the midfoot plantar fat pad recorded for the overweight/obese children (4.3 0.6 mm; p = 0.39) with the values recorded for the non-overweight children (4.1 0.6 mm) (Table 1).

Figure 2.

Means ( standard error of the mean) for each of the seven general foot dimensions displayed by the overweight/obese (n = 19) and non-overweight (n = 19) children. * Significant difference between the two subject groups. BOF, ball of foot; circumf, circumference.

Full figure and legend (71K)

Table 1. - Summary of differences in the key measures (mean standard deviation) between the overweight/obese children and their age-matched non-overweight counterparts.

Full table

When the overweight/obese children were compared with their non-overweight counterparts with respect to arch height, no significant between-subject group differences were found for either dorsal arch height (p = 0.59) or ball of foot height (p = 0.54) (Figure 3). In contrast, overweight/obese children displayed a significantly lower plantar arch height compared with that in the non-overweight children (p = 0.04) (Table 1 and Figure 3).

Figure 3.

Means ( standard error of the mean) for each of the arch height foot dimensions displayed by the overweight/obese (n = 19) and non-overweight (n = 19) children. * Significant difference between the two subject groups.

Full figure and legend (52K)

Discussion

The foot anthropometry results reported for the preschool children (3 to 5 years of age) in the present study are consistent with the previous research of Dowling (12), who reported larger foot dimensions for obese primary school-aged children (7 to 12 years of age) relative to those of their non-obese peers. These findings suggest that overweight/obese children as young as 3 years of age display foot dimensions that are significantly greater than those of non-overweight children.

The increased arch index displayed by the overweight/obese children in the present study supports previous investigations in which obese children have been shown to have flatter feet relative to those of their non-obese counterparts (1, 2, 3). For example, Gilmour and Burns (3) reported a significant difference between the arch index values of 15 obese children (0.25 0.07) and 257 non-obese children (0.21 0.07) who were 5.5 to 10.9 years of age. This flatter foot characteristic has also been reported in obese adults; in a study by Gravante et al., 38 obese men and women displayed significantly increased midfoot-to-forefoot width ratios than 34 non-obese men and women, indicating greater midfoot contact with the ground (13). Furthermore, Wearing et al. (14) found that the dynamic arch index had a significant positive relationship with percentage of body fat mass in a pilot study of 24 overweight and obese adults. Although the authors suggested that body adiposity was likely to be a confounding factor in interpreting footprint estimates of arch height, they failed to directly measure arch height or the amount of fat present underneath the foot.

The results of the present study not only support the notion that increased adiposity is associated with flatter feet, it highlights the fact that foot structure can be affected by overweight and obesity as early as the preschool years. Furthermore, this is the first investigation to show that overweight children, not only obese children, display greater midfoot contact with the ground. It must be acknowledged, however, that a higher arch index only indicates the presence of an increased midfoot surface contact area and not the direct mechanism causing this increased contact. It is also acknowledged that, as the sample of obese and overweight children were recruited via preschools, further research is warranted to confirm whether the findings of the current investigation are applicable to young Australian children outside the preschool sector. However, despite the small sample from which the two subject groups were drawn (n = 95), the percentage of those children who were overweight/obese (20% ) is consistent with the overall incidence of overweight/obesity in Australian children (15).

The present study is novel, as it is the first, to our knowledge, to report midfoot plantar fat pad thickness results in either children or adults. Interestingly, all children, irrespective of their BMI, displayed a midfoot plantar fat pad, with the thickness of this fat pad ranging from 2.8 to 5.4 mm. However, as the mean thickness of the midfoot plantar fat pad recorded for both subject groups differed by only 0.2 mm, with no significant between-subject group difference, it is postulated that differences in midfoot plantar fat pad thickness is not the cause of the flatter feet displayed by the young overweight/obese preschool children assessed in the present study.

The overweight/obese children in the present study displayed larger foot lengths, breadths, and circumferences; however, there were no between-subject group differences in either the dorsal arch height or the ball of foot arch height (Figure 3). On the basis of these results, it is speculated that the flatter foot structure displayed by the young overweight/obese preschool children assessed in the present study was attributable to between-group differences in the structure of the medial longitudinal arch. That is, as all other foot dimensions recorded for the overweight/obese were greater when compared with their leaner counterparts, it would be expected that the dorsal arch height would also be significantly higher in the overweight/obese children. However, if the navicular and cuneiform bones in the overweight/obese children were depressed, this could cause the dorsal arch heights of the two subject groups to be similar. Furthermore, depression of the navicular and cuneiform bones would cause the plantar arch height, which is measured on the medial plantar curvature of the foot, to be lower in the overweight/obese children than in the non-obese children, as the results of this study have found (Figure 3).

Therefore, it was concluded that the lack of difference in fat pad thickness between the two subject groups in the present study implies that the flatter foot structure displayed by young overweight/obese children relative to that of their non-overweight peers is not likely to be caused by a thicker midfoot plantar fat pad. Instead, the significantly lower plantar arch height found in the overweight/obese children suggests that their flatter feet may be caused by a lowering of the longitudinal arch, most probably caused by their feet continually bearing excess mass. It is postulated that these structural changes, which may adversely affect the functional capacity of the medial longitudinal arch, might be exacerbated if excess weight bearing continues throughout childhood and into adulthood. Therefore, urgent interventions, which take into consideration the unique structural characteristics of overweight and obese young children, are required to prevent further weight gain and structural and functional complications to their developing feet.

References

1. Dowling, A. M., Steele, J. R., Baur, LA. (2001) Does obesity influence foot structure and plantar pressure patterns in prepubescent children? Int J Obes Relat Metab Disord. 25: 845–852. | Article | PubMed | ChemPort |
2. Riddiford-Harland, D. L., Steele, J. R., Storlein, LH. (2000) Does obesity influence foot structure in prepubescent children? Int J Obes Relat Metab Disord. 24: 541–544. | Article | PubMed | ChemPort |
3. Gilmour, J. C., Burns, Y. (2001) The measurement of the medial longitudinal arch in children. Foot Ankle Int. 22: 493–498. | PubMed | ChemPort |
4. Dowling, A. M., Steele, J. R., Baur, LA. (2004) What are the effects of obesity on plantar pressure distributions? Int J Obes Relat Metab Disord. 289: 1514–1519. | Article |
5. Fixsen, JA. (1982) The foot in childhood. Klenerman, L eds. The Foot and Its Disorders 55–82. 2nd ed , Blackwell Scientific Publications Oxford, England.
6. Hefti, F., Brunner, R. (1999) Flexible arch of the foot. Orthopade. 28: 159–172. | PubMed | ChemPort |
7. Shiang, T. Y., Lee, S. H., Lee, S. J., Chu, WC. (1998) Evaluating different footprint parameters as a predictor of arch height. IEEE Eng Med Biol Mag. 17: 62–66. | Article | PubMed | ChemPort |
8. Cole, T. J., Bellizzi, M. C., Flegal, K. M., Dietz, WH. (2000) Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 320: 240–243. | Article | PubMed |
9. Parham, K., Gordon, C., Bensel, C. (1992) Anthropometry of the Foot and Lower Leg of U.S. Army Soldiers: Fort Jackson, S.C. Army Natick Research, Development and Engineering Center Natick, MA.
10. Cavanagh, P. R., Rodgers, MM. (1987) The arch index: a useful measure from footprints. J Biomech. 20: 547–551. | Article | PubMed | ChemPort |
11. Vincent, WJ. (1999) Statistics in Kinesiology 182–185. Human Kinetics Champaign, IL.
12. Dowling, AM. (2001) Does Obesity Affect Foot Structure and Function, Foot Sensation and Plantar Pressure Distribution in Children? University of Wollongong Wollongong, NSW, Australia Thesis.
13. Gravante, G., Russo, G., Pomara, C., Ridola, C. (2003) Comparison of ground reaction forces between obese and control young adults during quiet standing on a baropodometric platform. Clin Biomech. 18: 780–782. | Article | ChemPort |
14. Wearing, S. A., Hills, A. P., Byrne, N. M., Hennig, E. M., McDonald, M. (1994) The Arch Index: a measure of flat or fat feet? Foot Ankle Int. 25: 575–581.
15. Baur, LA. (2002) Child and adolescent obesity in the 21st century: an Australian perspective. Asia Pac J Clin Nutr. 11: S524–S528. | Article | PubMed |

Acknowledgments

This study was funded by a University of Wollongong Research Council grant. We acknowledge the support of the Child Obesity Research (CORe) center staff, particularly Dylan Cliff, Alison Crowshaw, and Lief Smith, during subject recruitment and data collection.