The present population-based cross-sectional study reported the prevalence of radiographically-assessed and self-recognized HV in approximately 2000 participants and the discrepancy between them. The prevalence of HV self-recognition was almost half that of radiographic HV and increased according to HV severity. The present study also demonstrated that valgus position of the proximal phalanx and female sex, but not interphalangeal HV, were independent positive factors for self-recognition of HV.
The present study demonstrated a 29.8% (1189 out of 3992 feet) prevalence of radiographically-assessed HV in two towns, including a mountainous region and a coastal region, which is similar to the investigation by Nishimura et al. reporting a 22.8% (184 out of 806 feet) prevalence in a single mountainous village from a different prefecture in the same country as ours [12]. A 2010 meta-analysis by Nix et al. including 78 articles written in several different languages reported a pooled prevalence estimate of HV of 23% in the population aged 18 to 65, and of 35.7% in the elderly over 65 years [4]. In Japan, where the present survey was conducted, traditional open-toed footwear like sandals and clogs had been preferably worn until the mid-20th century. However, the acceptance of westernized footwear and the declined usage of conventional footwear after World War II have resulted in a marked increase of HV surgeries [13]. Our survey demonstrated that Japan currently embraces a comparable population to other nations having a long-established culture of wearing shoes in having HV.
While the prevalence of HV varies by population group, ethnicity, geographic location, cultural differences in footwear, or socioeconomic status, methodological differences in investigation could also affect the reported value on prevalence. As for the sampling method, studies covering convenience samples such as people visiting foot clinics with foot problems are more likely to report higher prevalence estimates compared to those studies covering people randomly sampled from the general population[4]. One of the strengths of the present study is that we used population-based cohorts, even though our cohort did not include those who could not visit the survey venue or those who did not agree to the survey, and was thus not perfectly free from selection bias. As for the evaluation methods, the studies using clinical examinations showed a decisively higher prevalence than those using self-reporting measures such as interviews or questionnaires [4]. In a meta-analysis of HV prevalence, Nix et al. reported that only 16% of studies in their review used diagnostic criterion for HV based on radiographically- or clinically-measured angle, and that the larger the size of the studied population, the less likely it was to adopt radiographic evaluation for HV and more likely to adopt self-reporting or visual inspection [4]. To our knowledge, there has been no report on HV prevalence using radiographic angular criteria in community-based studies involving over 1,000 subjects.
Another problem to be faced when interpreting reports about HV prevalence is the use of the term “bunion” in questionnaires or interviews. The term "bunion" means the bursitis located at the medial of first metatarsal head caused by irritation with footwear and is sharply distinguished form the term "hallux valgus" which is a morphological deformation of the hallux [14]. However, these two terms are often used synonymously and cause discrepancies between surveys. For example, this kind of discrepancy is well presented as a large difference in HV prevalence between two surveys by Adams et al. in 1999 and Dunn et al. in 2004, respectively [5, 15]. Adams et al. reported a 0.9% prevalence of bunions among 63402 persons in the National Health Interview Survey of the USA using questionnaires about whether the questionees had "trouble with bunions" [15]; on the other hand, Dune et al. reported a 37.1% prevalence of bunion on clinical assessment among 784 randomly-sampled community-dwelling adults in the USA, although they seemed to use the term "bunion" to refer to "hallux valgus" [5]. In Japan, a term corresponding to "bunion" does not exist but a term corresponding to "hallux valgus" does. The condition of bunion is generally stated just as "pain or redness caused by HV"; therefore, our question about "hallux valgus" in the present survey can be regarded as asking simply about "deformity" but not about "symptoms".
The definition of the normal range of HVA also affects HV prevalence. An investigation in Korea by Cho et al. adopted a definition of HV as HVA >15°, which was broader than our definition of HVA > 20°, and reported a prevalence as high as 64.7% among community-dwelling subjects aged between 40 and 69 years [16]. If we adopted the same definition of HV as Cho et al., HV prevalence in the present study would be 45.2% in men and 62.7%, which is consistent with their study. The “normal range” of anatomical structures is a terminological conception and often arbitrary. It can be based on the distribution of relevant values among asymptomatic healthy subjects or the general population, prognostic prediction, treatment goals, and so on. HV is asymptomatic in a large part of the population; however, the deformity is clinically considered to be progressive, although there has been no consensus about the cut-off value predicting poor prognosis. Many studies using the HV definition of HVA > 15°in radiographic assessments are based on the study by Hardy and Clapham in 1951, reporting a mean 15.7° of HVA among convenient samples, including the staff and students at college [17]; however, it seems too strict to adopt the mean value as a cut-off value between normal and abnormal values. In the present study, there was a great leap in self-recognition between the two groups with HVA of 15-20° and 20-25° (4.3% vs. 12.8% in men, and 13.1% vs. 26.2% in women); therefore, we consider HVA > 20° to be valuable as one of the borders between normal and abnormal in terms of its influence on body image.
The prevalence of symptomatic HV was as low as approximately 10% in the present study, which might still be an overestimation as the presence of pain from HV was asked only to those participants who had self-recognition of HV. This low prevalence of pain would mean that the majority of HV cases are asymptomatic. Many patients do not visit clinics or seek surgical help until symptoms become troublesome years after they recognize the deformity. A survey of patients aged 20–66 years who were waiting for HV operation at a single hospital reported that 46% of the patients had noticed their deformity before they were 20 years old [17]. A cross-sectional investigation in our cohort consisting of community-dwelling adults demonstrated that participants with self-recognition of HV remembered having their deformity for an average of 17 years, but only 11.7% of them had a history of visiting the hospital for it. HV is generally a progressive deformity that can lead to functional disability and an elevated risk of falls in the elderly [2, 3, 18]. Mild deformity can be managed with conservative treatments such as advice on footwear, exercise, and orthosis in order to prevent progressive deformity and pain; however, symptomatic severe deformity regularly requires surgery. Increasing the general awareness of foot with mild HV deformity would help promote prevention of HV deformity.
The HV interphalangeus refers to a laterally deviated distal phalanx of the great toe, which is mainly attributed to its anatomical nature. The IPA was reported to be approximately 13° on average, among 346 British feet [19]. The HV interphalangeus could be presumed to contribute substantially to the total valgus deviation of the hallux, considering that normal HVA, a lateral deviation of the proximal phalanx against the 1st metatarsal, is 5° to 15° [20]. However, the effect of the HV interphalangeus on self-recognition of HV has not been elucidated. Some radiographic studies have described an inverse association between HVA and IPA, which was confirmed in the present study [10, 21, 22]. There are two proposed explanations for this inverse relationship [21, 23]. One explanation is based on the findings from a comparative study of radiographic parameters showing larger IPA in feet with hallux rigidus compared to normal feet or feet in hallux rigidus [23]. The increased stability in the horizontal plane at the metatarsophalangeal joint of hallux rigidus feet would let the laterally diverting force from shoe pressure or muscle activity toward the hallux concentrate at the interphalangeal joint; however, in the unstable metatarsophalangeal joint, these forces would likely affect the metatarsophalangeal joint, leading to HV instead of HV interphalangeus [21, 22]. Another explanation is based on findings from some studies demonstrating a postoperative increase in radiographic measurements of the HV interphalangeus following correction of HV, suggesting possible underestimation of HV interphalangeus due to pronation of the hallux in the foot with HV [24, 25]. A previous study by Strydom was the first to propose the concept of TVDH and evaluate the contribution of IPA to TVDH as 37.9%, which was slightly lower than the contribution of 48.9% found in the present study [10]. This lower contribution of IPA in the study by Strydom et al. might be attributed to the biased cohort of their study consisting of patients who visited clinics for foot problems. Their cohort is assumed to have more subjects with HV than the general population. A possible underestimation of the HV interphalangeus due to pronation of the hallux in the foot with HV has been proposed [25]. In the present study, multivariable analysis was conducted to detect the independent influence of HV interphalangeus on the self-recognition of HV, and IPA was detected as a negative factor for self-recognition, but its effect per unit increase was little compared to that of HVA.
Some limitations of our study must be considered in interpreting the results obtained. First, the present study adopted non-weightbearing foot radiographs instead of weightbearing ones. The difference in HVA measurement between weightbearing and non-weightbearing foot radiographs reported in previous studies are contradictory, with some researchers reporting no consistent change between two conditions [26], larger values under non-weightbearing conditions in normal or early HV, larger values under weightbearing conditions in advanced HV [27], or larger values under non-weightbearing conditions unrelated to the severity of HV [28]. Second, the present study was performed in limited areas, including two local towns, and might not reflect the situations in other areas. Although little has been clarified about the influencing factors on body image of HV in the general population, some factors besides those analyzed in the present study, such as educational, environmental, socioeconomic, psychological, and ethnocultural factors, might have the potential to affect self-recognition.
In conclusion, the present cross-sectional study is the first to show a discrepancy between self-recognition and radiographic diagnosis of HV in community-dwelling subjects. Valgus deformity at the metatarsophalangeal joint and female sex were independent positive factors for the self-recognition of HV, but not for HV interphalangeus.