The purpose of this study was to evaluate demographic factors associated with foot function in a community-dwelling, population-based large sample of men and women. We found significant differences in CPEI in groups stratified by age, sex, and past high heel use, but not by BMI or physical activity. Continuous models showed similar results in both men and women. These results indicate that future studies of foot function should consider the effects of sex, age, and history of high heel use as these factors may affect an outcome of interest.
Similar to prior studies [20
], this study found foot function differs by sex. This work noted that women displayed a lower CPEI. Smaller CPEI values are associated with greater amounts of pronation [18
]. This in part may be due to women having a more planus foot structure then men. Recent literature also suggests that arch height is reduced post-partum [22
]. As over-pronation has been clinically observed to be associated with several pedal pathologies (hallux valgus, hallux limitus, hallux rigidus, posterior tibial dysfunction, etc.) [23
], this may mean women are at a greater risk of foot issues compared to men. The higher CPEI seen in men is consistent with prior work showing higher forces in the lateral metatarsals in men, which may correspond with a more supinated foot position [21
]. Thus, the current work extends the information available in the medical literature.
In this study, age older than 65 years was associated with lower CPEI in women, indicating more pronated foot function during gait. A similar magnitude of effect was noted in men, but was only significant in the continuous models. Clinicians have qualitatively observed that arches may become lower as persons age, and this observation is consistent with studies noting increased rates of flat feet [24
] and pronated feet [5
] with increasing age. If older individuals are becoming more planus they may be predisposed to a greater incidence of associated pathologies. As foot pathologies have been linked to functional limitations [25
] and fall risk [26
], this age related change might have significant consequences over time While there was no statistically significant difference in mean foot function between men ≥ 65 or < 65 years, both sexes had significantly lower CPEI among those 75 years or older, relative to those under 55, with a trend towards decrease over time in the other age groups. Future research should more thoroughly investigate biomechanical changes in the foot with age , as well as sex differences with age in foot function as perhaps it may help explain differences in rates of knee injury [27
] and joint degeneration [28
] between the sexes.
Individuals with higher BMI are suspected of having a higher prevalence of flat feet [24
], which may be associated with increased foot pronation during the stance phase of gait [23
]. However, it is unclear whether static measures are accurate predictors of foot function [29
], and few studies have directly assessed the relation between BMI and foot function. Previous research in small groups of adult volunteers has found that obese participants had larger plantar contact areas [32
] and higher pressure under the forefoot during stance [11
]. Messier et al. [33
] found obese participants had significantly greater rearfoot eversion relative to normal weight participants in a 2D motion capture analysis of female volunteers. In our current analysis, CPEI was unaffected by BMI, suggesting no relation between foot function and obesity. Several factors may account for the difference between this result and previous research. As CPEI measures foot function using the distribution of load under the foot over time [18
], it may not be directly comparable to previous studies of static plantar pressure [10
] and kinematics [33
]. Moreover, this study was population-based, while previous studies used small samples of volunteers [11
]. Further, our study population was older (with an age range of 36 to 98 years versus samples primarily in their twenties [11
] and forties [32
]) and had a lower mean BMI (mean BMI was 28.4 kg/m2
versus 41.1 kg/m2
for obese group in the Messier study [33
Given the link between foot deformities and muscle weakness in diabetic patients [34
] and fallers [35
], and between foot deformities and altered foot biomechanics [23
], physical activity may affect foot function. In the current study, however, CPEI was not affected by physical activity levels in either continuous or categorical analyses. While PASE does not measure physical function directly, it has been shown to be associated with a number of physiological measures of physical function [16
]. This result provides preliminary evidence that foot function as measured in the current study may not be significantly related to physical activity.
Both finite element modeling [36
] and a study of young volunteers in Taiwan [37
] found that high heels increased medial forefoot and toe loading in shod feet. However, the effects of habitually wearing high heels on barefoot gait are not well understood. In this study, women who always wore high heels over their adult lifespan had significantly lower CPEI than those who never wore them. Lower CPEI is consistent with the higher medial forefoot loading observed previously by these authors, and thus may indicate that changes from high heel use have a modifying effect on plantar loading. These results are in agreement with work in children showing that past shoe use can affect foot structure [9
]. Future research should look at the specific effects of past shoe-wear on plantar loading in both older women and men.
This study has several strengths and limitations worth noting. Because the study design was cross-sectional, causal relations between foot function and the factors under study cannot be inferred. Due to examination time constraints, only one plantar pressure scan per foot was obtained from each participant and thus, there is likely a larger degree of measurement error. This limitation is mitigated by the large sample size of the study, but if this error had an effect on the results, it would act to bias towards a null effect between variables rather than create a false positive [39
]. There were also several strengths to this population-based study. The study had a large sample size spanning a wide age range (36 to 98 years) and body size (BMI ranged 14.6 to 57 kg/m2
), in addition to including both men and women.