SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

Objective

To determine the population prevalence and examine factors associated with hallux valgus in a primary care population.

Methods

A questionnaire was mailed to all adults age >30 years registered with 2 general practices. Validated instruments assessed self-reported hallux valgus, nodal osteoarthritis, and knee pain. The questionnaire also asked about big toe pain, joint replacement, and history of osteoarthritis and rheumatoid arthritis. Hallux valgus prevalence was calculated and standardized by the source population in terms of age, sex, knee pain, osteoarthritis, and rheumatoid arthritis. A nested case–control study was undertaken and age-sex adjusted odds ratios (ORs) were calculated between hallux valgus and age, sex, body mass index, nodal osteoarthritis, knee pain, big toe pain, joint replacement, self-reported osteoarthritis, and self-reported rheumatoid arthritis, using a binary logistic regression model.

Results

A total of 13,684 questionnaires were mailed and 4,249 (32%) responses were received. The standardized prevalence of hallux valgus was 28.4%. Hallux valgus was associated with age (adjusted OR 1.61 per decade; 95% confidence interval [95% CI] 1.52–1.69), female sex (adjusted OR 2.64; 95% CI 2.26–3.08), nodal osteoarthritis (adjusted OR 1.66; 95% CI 1.26–2.17), knee pain (adjusted OR 1.96; 95% CI 1.65–2.32), big toe pain (adjusted OR 3.28; 95% CI 2.48–4.33), self-reported osteoarthritis (adjusted OR 1.41; 95% CI 1.15–1.72), and self-reported rheumatoid arthritis (adjusted OR 2.04; 95% CI 1.43–2.91).

Conclusion

Hallux valgus is prevalent in the community and is associated with age, female sex, and components of generalized osteoarthritis such as nodal osteoarthritis, knee pain, big toe pain, and self-reported osteoarthritis.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

Hallux valgus is a common deformity characterized by abnormal angulation, rotation, and lateral deviation of the great toe at the first metatarsophalangeal joint. Estimates of the community prevalence of hallux valgus vary widely, ranging from 21–70% in epidemiologic studies (1–5). This variation is attributable, in part, to different study populations and definitions of hallux valgus used, in particular confusion of the terms hallux valgus and bunion. Hallux valgus poses a significant health problem, and is associated with foot pain (1), poor balance (6), immobility (6), and risk of falling (6, 7).

The first metatarsophalangeal joint is a target joint for osteoarthritis and it was included in Kellgren and Moore's seminal description of generalized osteoarthritis (8). More recently, radiographic evidence of osteoarthritis at the first metatarsophalangeal joint has been shown to be associated with radiographic evidence of osteoarthritis at the knee and hand (9). It is postulated that hallux valgus occurs as a consequence of osteoarthritis at the first metatarsophalangeal joint, and therefore is expected to be associated with osteoarthritis at other sites. Pain and osteoarthritis at a number of sites have been shown to be associated with foot pain (5, 10, 11), yet an association with osteoarthritis has not been examined to date. We used a cross-sectional study in a primary health care setting in order to determine the population prevalence of self-reported hallux valgus, and to examine the factors associated with hallux valgus.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

A cross-sectional study was conducted with patients from 2 general practices in Nottingham, UK to determine the population prevalence of self-reported hallux valgus, and examine associations with hallux valgus. Ethical approval was obtained from the Nottingham Local Research Ethics Committee 2.

Each practice compiled a list of all registered adult patients age >30 years, excluding those with a history of major psychiatric disease, dementia, or recently diagnosed malignancy. A questionnaire and a prepaid reply envelope were then mailed to all listed individuals. The questionnaire requested standard demographic information (age, sex, height, and weight). Self-reported hallux valgus was assessed using a validated line drawing instrument and was dichotomized as present or absent at each foot (12). Cases were defined as subjects with hallux valgus affecting either big toe. The questionnaire also included a validated line drawing instrument for self-reported nodal osteoarthritis (13), and asked about the occurrence of knee pain (14) and big toe pain on most days for at least 1 month within the last year, history of knee or hip replacement, and previous diagnosis of osteoarthritis, rheumatoid arthritis, or ankylosing spondylitis. An abbreviated questionnaire was mailed out to nonrespondents in 1 practice.

Cross-sectional prevalence study.

The crude prevalence of hallux valgus was calculated. Age was stratified into decades and sex- and age-specific prevalences were calculated. Subjects were then further classified as having bilateral hallux valgus (i.e., present in both feet) or unilateral hallux valgus (i.e., affecting only the right or left big toe). The overall prevalence and sex- and age-specific prevalence were again calculated.

Hallux valgus prevalence was then standardized using the Nottingham knee pain population survey as the reference population (14). The Nottingham knee pain survey was a survey of people ages 40–80 years in the same geographic area of Nottingham, and had an 82% response rate. Direct standardization was undertaken for age, sex, knee pain (14), self-reported osteoarthritis, and self-reported rheumatoid arthritis.

Comparison of cases and controls.

Cases with hallux valgus identified from the cross-sectional survey were compared with subjects without hallux valgus (controls) for associated factors. The crude odds ratios (OR) and the 95% confidence intervals (95% CIs) were calculated for hallux valgus (present/absent), age (decades), sex (female/male), body mass index >30 kg/m2 (present/absent), nodal osteoarthritis (defined as at least 2 rays affected by Heberden's or Bouchard's nodes on each hand including the thumb but excluding the first carpometacarpal joint) (present/absent) (15), knee pain in the last year (yes/no), big toe pain in the last year (yes/no), and a history of knee replacement (yes/no), hip replacement (yes/no), osteoarthritis (self-reported or a history of knee replacement in the absence of rheumatoid arthritis or ankylosing spondylitis) (yes/no), and self-reported rheumatoid arthritis (yes/no). Each variable was then entered into a comprehensive binary logistic regression model adjusting for age and sex, with hallux valgus as the dependent variable and the adjusted ORs (95% CIs) were calculated. The analysis was repeated separately for women and men.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

Questionnaires were mailed to 13,684 patients, and completed responses were received from 4,249 patients (age range 30–95 years). Notifications were received that 294 questionnaires were sent to the wrong addresses; therefore, the adjusted response rate was 32%. The hallux valgus self-report instrument was not completed by 381 subjects.

Prevalence of hallux valgus.

Hallux valgus affecting either big toe was reported by 1,194 subjects (crude prevalence 31%). The remaining 2,674 subjects reporting the absence of hallux valgus in both feet served as the control group. The prevalence of hallux valgus overall was greater in women (38%) compared with men (21%) and also greater in women across all age groups. The prevalence increased as patient age increased (Figure 1). Bilateral hallux valgus was reported by 16.9% of subjects compared with subjects with unilateral hallux valgus (6.0% in the left foot only, 7.0% in the right foot only; P < 0.001). In women, 21.2% reported bilateral hallux valgus, 7.6% in the left foot only, and 8.2% in the right foot only. In men, 11.0% reported bilateral hallux valgus, 3.9% in the left foot only, and 5.4% in the right foot only. The frequency of bilateral hallux valgus increased as patient age increased (Figure 2). The standardized prevalence of hallux valgus was 28.4% for patients 40–80 years old.

thumbnail image

Figure 1. Prevalence of hallux valgus by age and sex. Diamonds = men; squares = women; triangles = both.

Download figure to PowerPoint

thumbnail image

Figure 2. Prevalence of unilateral and bilateral hallux valgus by age. Diamonds = left foot only; squares = right foot only; triangles = bilateral.

Download figure to PowerPoint

Associations with hallux valgus.

The sample characteristics are shown in Table 1. On univariate analysis, hallux valgus was associated with age, female sex, nodal osteoarthritis, knee pain, big toe pain, knee replacement, hip replacement, osteoarthritis, and self-reported rheumatoid arthritis (Table 2). On multivariate analysis, hallux valgus remained associated with age, female sex, nodal osteoarthritis, knee pain, big toe pain, osteoarthritis, and self-reported rheumatoid arthritis.

Table 1. Characteristics of the study population*
 Total population (n = 4,249)Hallux valgus (n = 1,194)Control (n = 2,674)
  • *

    Values are the number (percentage) unless indicated otherwise. BMI = body mass index; OA = osteoarthritis; RA = rheumatoid arthritis.

  • Self-reported OA or joint replacement with no history of RA or ankylosing spondylitis.

Age, mean ± SD years57.6 ± 14.563.2 ± 13.454.5 ± 14.0
Female2,445 (57)853 (71)1,404 (53)
BMI, mean ± SD kg/m226.3 ± 4.626.0 ± 4.726.3 ± 4.6
BMI >30 kg/m2713 (17)191 (16)452 (17)
Nodal OA275 (7)142 (12)116 (4)
Knee pain905 (22)365 (31)449 (17)
Big toe pain266 (7)148 (12)102 (4)
Knee replacement59 (1)29 (2)23 (1)
Hip replacement87 (2)31 (3)43 (2)
OA607 (15)257 (22)299 (11)
Self-reported RA172 (4)83 (7)67 (3)
Table 2. Associations between hallux valgus and age, sex, BMI, and self-reported OA in the total study population*
 UnadjustedAdjustedP
  • *

    Values are the odds ratio (95% confidence interval). BMI = body mass index; OA = osteoarthritis; RA = rheumatoid arthritis.

  • Age adjusted for sex, sex adjusted for age. Other variables adjusted for age and sex.

  • Per decade.

  • §

    Self-reported OA or joint replacement with no history of RA or ankylosing spondylitis.

Age, years 1.61 (1.52–1.69)< 0.001
 30–391  
 40–491.61 (1.16–2.24)  
 50–593.18 (2.34–4.32)  
 60–694.62 (3.41–6.25)  
 70–795.92 (4.33–8.08)  
 ≥809.79 (6.75–14.18)  
Sex2.26 (1.95–2.62)2.64 (2.26–3.08)< 0.001
BMI >30 kg/m20.95 (0.77–1.14)0.98 (0.80–1.19)0.809
Nodal OA2.98 (2.31–3.84)1.66 (1.26–2.17)< 0.001
Knee pain2.25 (1.92–2.65)1.96 (1.65–2.32)< 0.001
Big toe pain3.74 (2.87–4.86)3.28 (2.48–4.33)< 0.001
Knee replacement2.91 (1.67–5.04)1.63 (0.91–2.93)0.099
Hip replacement1.63 (1.02–2.61)0.88 (0.53–1.44)0.602
OA§2.22 (1.85–2.67)1.41 (1.15–1.72)0.001
Self-reported RA2.91 (2.09–4.04)2.04 (1.43–2.91)< 0.001

The overall pattern of associations in the sex-specific analysis was similar, although hallux valgus was not associated with self-reported osteoarthritis in men (Tables 3 and 4). Hallux valgus appeared to be more strongly associated with increasing age and big toe pain in women compared with men, and more strongly associated with nodal osteoarthritis and rheumatoid arthritis in men compared with women.

Table 3. Association between hallux valgus and age, BMI, and self-reported OA in women*
 UnadjustedAge-adjustedP
  • *

    Values are the odds ratio (95% confidence interval). BMI = body mass index; OA = osteoarthritis; RA = rheumatoid arthritis.

  • Self-reported OA or joint replacement with no history of RA or ankylosing spondylitis.

Age, years   
 30–391  
 40–491.76 (1.21–2.58)  
 50–593.50 (2.45–4.99)  
 60–695.95 (4.16–8.52)  
 70–797.55 (5.18–10.99)  
 ≥8012.56 (7.89–20.00)  
BMI >30 kg/m20.89 (0.72–1.12)0.95 (0.75–1.20)0.948
Nodal OA2.26 (1.70–3.00)1.42 (1.05–1.91)0.023
Knee pain2.21 (1.80–2.71)1.79 (1.44–2.22)< 0.001
Big toe pain4.20 (2.99–5.91)3.63 (2.55–5.18)< 0.001
Knee replacement3.73 (1.76–7.91)1.94 (0.89–4.26)0.097
Hip replacement1.45 (0.80–2.62)0.68 (0.37–1.27)0.226
OA2.36 (1.87–2.97)1.43 (1.12–1.83)0.004
Self-reported RA2.80 (1.84–4.25)1.71 (1.10–2.66)0.018
Table 4. Association between hallux valgus and age, BMI, and self-reported OA in men*
 UnadjustedAge-adjustedP
  • *

    Values are the odds ratio (95% confidence interval). BMI = body mass index; OA = osteoarthritis; RA = rheumatoid arthritis.

  • Self-reported OA or joint replacement with no history of RA or ankylosing spondylitis.

Age, years   
 30–391  
 40–491.51 (0.77–2.94)  
 50–593.09 (1.66–5.76)  
 60–694.05 (2.20–7.46)  
 70–795.68 (3.12–10.70)  
 ≥808.58 (4.28–17.21)  
BMI >30 mg/kg20.91 (0.65–1.29)1.03 (0.72–1.47)0.866
Nodal OA4.10 (2.23–7.55)3.09 (1.66–5.78)< 0.001
Knee pain2.44 (1.87–3.20)2.26 (1.71–2.98)< 0.001
Big toe pain2.84 (1.81–4.46)2.72 (1.71–4.32)< 0.001
Knee replacement2.08 (0.82–5.25)1.28 (0.50–3.28)0.607
Hip replacement2.00 (0.92–4.35)1.33 (0.60–2.94)0.483
OA1.83 (1.32–2.55)1.35 (0.96–1.90)0.090
Self-reported RA2.99 (1.71–5.22)2.68 (1.51–4.77)0.001

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

This study is one of the largest epidemiologic surveys of hallux valgus to date. Hallux valgus is a common condition that appears to be strongly associated with age and female sex. It is also associated with nodal osteoarthritis, knee pain, big toe pain, self-reported osteoarthritis, and self-reported rheumatoid arthritis after adjustment for age and sex, suggesting that hallux valgus is a component of generalized osteoarthritis. Several studies have found that pain in the foot, although not specifically the first metatarsophalangeal joint, is associated with pain and self-reported osteoarthritis at other body sites (5, 10, 11). In 1952, Kellgren and Moore included the first metatarsophalangeal joint in their seminal description of primary generalized osteoarthritis (8). More recently, radiographically-confirmed osteoarthritis of the first metatarsophalangeal joint has been shown to be associated with radiographic osteoarthritis at the distal and proximal interphalangeal joints, first carpometacarpal joint, and the knee (9). Hallux valgus, therefore, appears to be a component of generalized osteoarthritis and is likely to be a marker of osteoarthritis of the first metatarsophalangeal joint, although the cross-sectional design of our study does not definitively confirm this relationship ahead of the possibilities that hallux valgus and first metatarsophalangeal joint osteoarthritis are coexisting conditions or that hallux valgus is a precursor of first metatarsophalangeal joint osteoarthritis.

It is therefore interesting that hallux valgus was more frequently found to be a bilateral phenomenon with little difference between the right or left foot. Other components of generalized osteoarthritis (e.g., radiographic knee and interphalangeal joint osteoarthritis) have been found to be asymmetrical with a predominance of disease on the right side (16, 17). This finding has been interpreted in the hands as evidence of the role of biomechanic factors in the development of osteoarthritis. Hence, 2 possible explanations for the symmetry of hallux valgus exist. First, the development of osteoarthritis may relate to constitutional or genetic factors or alternatively, biomechanic factors may apply equally to both feet, in contrast with the hand where use of the dominant side is favored.

Hallux valgus was found to be associated with big toe pain. Results of other studies have disagreed with this finding, although definitions of big toe problems vary, in particular the distinction between hallux valgus and bunions. The Cheshire Foot Pain and Disability survey found an association between disabling foot pain and self-reported bunions (3,044 subjects, ages 18–80 years) but not with podiatrist-assessed hallux valgus (281 subjects) (10). An Italian survey of 459 persons age >65 years reported a significant association between physician-assessed hallux deformity and foot pain when standing (1). A study of 172 retirement village residents found an association between hallux valgus severity (as assessed by the Manchester scale [18]) and disabling foot pain (11). In contrast, a study of 1,002 women with physical disability found no difference in the frequency of chronic, severe foot pain between subjects with and without nurse-assessed bunions (5). Similarly, the Feet First study assessed 784 community-derived subjects age ≥65 years and found no association between foot pain and hallux valgus/bunion confirmed by physical examination (19). Although there is some discordance between the findings of our present study and the published literature, the Nottingham study is one of the largest studies to date and has the advantages of using a validated instrument to assess hallux valgus and studying a wide age range (adults age >30 years).

Several important caveats apply to the results of this study, the most important being the low response rate to the postal questionnaire, raising the possibility of a differential response rate between those with and without foot pain and deformity and risking an overestimate of the prevalence and associations of hallux valgus. However, the prevalence of hallux valgus was standardized against a reference population from a previous community survey of knee pain in Nottingham (14). The response rate in the Nottingham Knee Pain survey was much higher than in the current study (81.9%). The reduced standardized prevalence (31–28%) suggests a potential bias due to the lower response rate and we would suggest use of the standardized prevalence if a population estimate is required. However, care must be taken as this standardized estimate is only for people ages 40–80 years. It may be lower if younger people (age <40 years) are included.

A further limitation is that the associations of hallux valgus identified by the study are based on self-reported data and are worthy of further verification by clinical assessment or radiographs. The hallux valgus self-report instrument depicts 5 grades of hallux valgus, and a 5-grade instrument has been validated in addition to the dichotomized version used in this study (12). Use of the 5-grade instrument may have provided insights into factors influencing hallux valgus severity. Although important associations of hallux valgus with age, sex, and pain and osteoarthritis at other sites have been identified, lifestyle factors (in particular footwear) may be an important consideration. Hallux valgus is uncommon in unshod populations (20–22), yet few studies have examined a link between footwear characteristics and hallux valgus. A study of 176 people age >65 years found that hallux valgus was associated with narrow shoes and heel elevation (23). In contrast, a hospital-based study of 89 subjects found a negative correlation between hallux valgus severity and a history of wearing high-heeled shoes (24). It is possible that the stronger association of hallux valgus with big toe pain in women compared with men could be due to footwear differences between men and women. The risk of hallux valgus associated with different styles of footwear warrants further study.

In summary, hallux valgus is strongly associated with age and female sex and also associated with components of generalized osteoarthritis such as nodal osteoarthritis, knee pain, big toe pain, and self-reported osteoarthritis.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

Dr. Roddy had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Roddy, Zhang, Doherty.

Acquisition of data. Roddy.

Analysis and interpretation of data. Roddy, Zhang, Doherty.

Manuscript preparation. Roddy, Zhang, Doherty.

Statistical analysis. Roddy, Zhang.

Guarantor. Doherty.

ROLE OF THE STUDY SPONSOR

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

The financial support from AstraZeneca UK, GlaxoSmithKline USA and Ipsen, France was unrestricted. The study design, data collection, data analysis, writing of the manuscript, and submission of the manuscript were undertaken by the authors independently of the sponsors.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES

We would like to thank the staff and patients of Arnold Health Centre and The Calverton Practice in Nottingham, UK.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. Acknowledgements
  10. REFERENCES