A radiographic study of the relationship between metatarsus adductus and hallux valgus
Article Outline
Abstract
The relationship between metatarsus adductus and hallux valgus was evaluated on 100 dorsoplantar weight-bearing radiographs. The metatarsus adductus angle varied positively with the degree of hallux valgus. By using measures of correlation, a significant linear association was found for women (r = .53, P < .001) and men (r = .48, P < .001). The relationship was strongest in women when all cases of abnormal metatarsus adductus (>24°) were associated with abnormal degrees of hallux valgus (>15°). This relationship was different in men in that abnormal metatarsus adductus angles were not always associated with abnormal hallux valgus angles. With male and female data combined, the prevalence of metatarsus adductus was 55% in subjects with hallux valgus deformity compared with 19% in subjects without hallux valgus. A X2 test showed this to be a significant difference in the distribution of the data (P = .002). The data of this study suggests that there may be a clinical association between metatarsus adductus and hallux valgus. The need to further evaluate the role of metatarsus adductus angle in hallux valgus surgery is emphasized. (The Journal of Foot & Ankle Surgery 42(1):9-14, 2003)
Keywords: metatarsus adductus, hallux valgus, gender
Metatarsus adductus (MA) has been cited as a cause of hallux valgus (HV), particularly in the juvenile foot 1, 2, 3 and it has been suggested that unrecognized MA deformity is a cause of recurrence of HV deformity after surgery (4). MA is described as a structural deformity occurring at the Lisfranc joint (tarsometatarsal joints) (5), with the metatarsals being deviated medially with reference to the lesser tarsus. It would not be unexpected to find HV associated with MA given that HV is related to a medially deviated first metatarsal 6, 7, 8 as seen by an increased intermetatarsal angle on radiograph. One study found that not all cases of HV were associated with an increased intermetatarsal angle (9). In those instances, MA would account for the deformity because the intermetatarsal angle remains small.
HV deformity has been linked with MA in shoe-wearing populations (5). When wearing shoes, the hallux cannot align with the adducted first metatarsal but is forced laterally at the metatarsophalangeal joint. It also has been suggested that compensatory pronation occurs to abduct the forefoot in MA. The pronation and subsequent forefoot instability is involved in the formation of HV 1, 10, 11.
Radiologic evaluation of the MA angle is obtained by taking the angular measurement formed between the line bisecting the second metatarsal and the longitudinal line bisection of the lesser tarsus on standard weight-bearing, dorsoplantar radiographs. The normal adult value is between 5° to 17° (5) and is considered pathologic above 20° 12, 13, 14. A simplified method that measures the angle between the second metatarsal bisection and the second cuneiform bisection has been shown to be valid (14) and has been used in several trials to examine the relationship between MA and HV. The simplified method produces values that are 3° greater than the values of the traditional method so that with this approach MA angles greater than 24° are considered abnormal in an adult.
Four studies have evaluated previously the relationship between MA and HV 4, 10, 13, 15. When considered in 2 groups, one group with HV deformity and a control group without HV, a 35% prevalence of MA was found in the HV group compared with 13% in the control group (P < .001) (13). This was lower than the prevalence found in adolescent patients presenting for HV surgery (−75.4%) (4). The investigators stated that it was their contention “that unrecognised metatarsus adductus is one factor that heavily contributes to the high recurrence in adolescent and adolescent-onset hallux abductovalgus” (4). A threshold of 15° (MA angle) has been suggested as the level of MA that will cause HV deformity. No gender difference was found for MA angles or HV angles 10, 14.
None of these studies used radiographs of the foot in a subtalar joint neutral position with the midtarsal joint pronated fully. The presence of subtalar joint pronation would be expected to decrease the MA angle whereas inversion of the forefoot would increase the MA angle. Failure to align the foot in a neutral position for the radiograph may therefore obscure the true prevalence of the condition. However, it would be particularly difficult to maintain such a position to the same degree in all patients for weight-bearing radiographs and so the resultant error perhaps has to be accepted.
However, all of the previous studies preselected the subjects for either HV or MA. The purpose of this study was to compare the HV angle and MA angle in healthy patients who were not preselected for these malalignments. We also wanted to determine if there was any difference in the relationship between them according to sex.
Materials and methods
Sample
One hundred radiographs (50 male and 50 female) were selected randomly from the radiographic files of our institution. Only radiographs showing a dorsoplantar, weight-bearing view were included in the study. At our hospital, the standard radiographic procedure for the dorsoplantar projection includes the central radiograph beam directed 15° proximally, centered on the navicular at a distance of 100 cm from the foot.
Radiographs of patients only were included when there was a clear view of the metatarsophalangeal joint articular surfaces. Only radiographs of patients who were younger than 40 years of age were used to reduce the likelihood of osteoarthritic changes obscuring the joint margins. Patients were in good general health and did not have a medical history of joint disease such as rheumatoid arthritis, osteoarthritis, or diabetes (Charcot arthropathy). No previous surgery on the first ray had been undertaken. Patients with and without HV were included.
Each radiograph was digitized using a digital camera. The camera was mounted on a tripod at a fixed distance from the radiograph viewer. A spirit level was used to ensure that the camera was not tilted between radiographs and a small torch was fixed to the camera so that the light beam was centered on the metatarsal head for each picture. The camera was adjusted to the minimum zoom position and the picture was taken with a remote control to prevent movement of the camera.
The OSIRIS software package was used to analyze the radiograph (available at http://www.expasy.ch/www/UIN/html1/projects/osiris/osiris.html .). An initial pilot study confirmed the reliability of the measurement technique. A ruler was included in each image in the initial pilot studies to provide a simple check against parallax with five 1-cm distances being measured down the ruler to ensure that the distances stayed evenly spaced. Each image was enlarged to 200% for maximum visibility.
The MA angle was measured using the simplified method described by Engel et al. and the HV angle was measured using the method described by LaPorta et al. 14, 16 (Fig. 1).
Fig. 1.
Dorsoplantar radiograph showing the determination of (A) MA angle measurement (180° angle shown) and (B) HV measurement.
Results
In 5 of the male and 1 of the female radiographs, the tarsal area was obscured so that the MA angle was not measured. The HV angle was measured in all of the 100 radiographs. Retrospective analysis of sample size comparing the MA angle with the HV angle (women, n = 49; men, n = 45) showed that the study had a power of 93% to detect a significant difference between the groups.
Frequency histograms were constructed to show the distribution of the MA angle and HV angle (Fig. 2).
Fig. 2.
Histograms of MA and HV data for men and women.
There appeared to be little difference in the distribution of the metatarsus angle between men and women but a more obvious difference in the HV angles. This was assessed by using the Student t test (Table 1).
Table 1. Data table for student t test for MA angle
| Women | Men | |
|---|---|---|
| Mean | 23.30 | 22.91 |
| Variance | 30.27 | 29.08 |
| Observations | 49 | 45 |
| Df | 92 | |
| t Stat | 0.35 | |
| P (T ≤ t) 2-tail | 0.73 |
A scatter graph (Fig. 3) of the MA and HV measurements suggested that a positive trend existed between the measurements for the female data and to a lesser extent for the male data.
Fig. 3.
Scatter graph of MA and HV measurements.
Table 2. Linear regression analysis
| Regression Statistics | ||
|---|---|---|
| Females (n = 49) | Males (n = 45) | |
| Multiple R | 0.53 | 0.48 |
| R2 | 0.28 | 0.23 |
| Adjusted R2 | 0.26 | 0.21 |
| Standard error | 4.73 | 4.80 |
| P values | .0001 | .0009 |
A contingency table (Table 3) was created that identified the number of cases of normal (<24°) and abnormal MA (>24°) with the number of cases of normal and abnormal HV angles.
Table 3. Contingency table for abnormal MA angles (>24°) and abnormal HV angles (>16°)
| Men | Women | |||||||
|---|---|---|---|---|---|---|---|---|
| MA angle | ≤24° | >24° | ≤24° | >24° | ||||
| HV angle | ≤15° | >15° | ≤15° | >15° | ≤15° | >15° | ≤15° | >15° |
| Count | 21 | 8 | 7 | 11 | 16 | 16 | 0 | 17 |
When considering men and women with and without HV together, the prevalence of MA was 55% in the HV group compared with 19% in the group without HV (Table 4).
Table 4. Subjects with MA (>24°) in subjects with and without HV (≤15°)
| HV Group | No HV | ||
|---|---|---|---|
| MA | No MA | MA | No MA |
| 23 | 28 | 8 | 36 |
Discussion
An association has been identified between the degree of MA and the degree of HV in male and female subjects. The relationship concurs with the findings of earlier studies 10, 15. The present study found no difference in the MA angles between men and women (P = .73) but did find a significant difference in the HV angle (P < .001). Banks et al. did not test for differences between male and female subjects and did not report the percentage of male and female patients in their study (10). Griffiths and Palladino tested for differences between the sexes and found no significant differences in the measurements for MA or HV angles (15). This latter observation is surprising given that most studies show the high female prevalence of HV. Only 1 study to date has shown a greater number of men with HV (17). The lack of a significant difference in HV measurements may be caused by the initial selection criteria that excluded any patient with clinical signs and symptoms of HV from their study.
La Reaux and Lee (13) found that 13% of their control group (no HV) had a MA deformity compared with 35% of their HV group (women:men ratio, 4:1). The present study had similar findings when considering men and women together (ratio, 1:1), with an 18% prevalence of MA in subjects without HV and a 55% prevalence in subjects with HV deformity. Pontious et al. (4) found that 75.4% of their HV patients (37 women, 17 men) had an MA deformity. Such a high prevalence was not explained, but because the study group consisted of juvenile patients awaiting surgery, it may be a result of the increased angle seen in children plus the correlation established here between HV and MA in women.
Our study has shown that the distribution of HV deformity, when considered in subgroups with and without MA, is significantly different between men and women (P = .0004). In men with a normal MA angle, the majority of subjects also had a normal HV angle, whereas the women had equal numbers of subjects with and without HV deformity. In both men and women with abnormal MA angles, the frequency of abnormal HV angles increased but to such an extent in women that all cases of abnormal MA angles were accompanied by abnormal HV angles. This was in agreement with the respective correlation coefficients. HV deformity can exist in women without MA, showing that the deformity has other etiologies, but this study has found that when an MA deformity is present, HV always accompanies it. An earlier study (4) had suggested that the existence of MA may be a cause for recurrence of deformity in adolescent-type HV deformity after surgery. The findings of this study suggest that recurrence of deformity in adult women warrants further consideration of the impact of MA.
Footwear may be responsible for the correlation between MA and HV in women. Rothbart suggested that HV deformity occurs with MA in shoe-wearing populations (5). With an adducted first metatarsal, the medial side of the toe-box prevents the hallux from aligning with the metatarsal and relatively abducts the hallux at the metatarsophalangeal joint. The more pointed the toe-box, the greater the abduction at the joint. The footwear worn by the subjects in this study was not known. The subjects were less than 40 years old and, given the present trends in footwear today, it is very likely that the women's shoes were more pointed than the men's shoes.
Other factors may be involved. An earlier study has found that the first metatarsal head in women had a more curved articular surface than men and this curvature was related significantly to the degree of HV deformity (18). Therefore, even if the abductory force were equal between men and women, the female hallux may deform more easily than in men because the joint would be more stable owing to the flatter-shaped metatarsal head. In addition, women's joints are known to be more flexible than men's 19, 20, which may lead to greater deformity when an abductory force is applied.
Conclusion
An association has been shown between MA and HV deformity. Abnormal MA angles are associated with abnormal HV angles. The relationship is significant for both men and women although the correlation is strongest in women. It is suggested that in women presenting for HV surgery, the coexistence of MA should be assessed.
References
- . Juvenile hallux valgus with compensated metatarsus adductus. J Am Podiatr Med Assoc. 1991;81:525–530
- . The pathophysiology of the juvenile bunion. Instructional course lectures. Chicago: American Academy of Orthopaedic Surgeons,; 1987;
- . Podopaediatrics. In: Baltimore: Wilkins and Wilkins,; 1980;p. 356–372
- . Characteristics of adolescent hallux abducto valgus. A retrospective review. J Am Podiatr Med Assoc. 1994;84:208–218
- . Metatarsus adductus and its clinical significance. J Am Podiatry Assoc. 1972;62:187–191
- . The First metatarsocuneiform articulation and its relationship to metatarsus primus varus. J Am Podiatry Assoc. 1979;76:700–706
- . Hallux valgus in the younger patient. The structural abnormality. J Bone Joint Surg. 1979;61B:176–177
- . Observations on hallux valgus. J. Bone Joint Surg. 1951;33B:376–391
- . The natural history of hallux valgus in adolescents and early adult life. J Bone Joint Surg. 1960;42B:749
- . Juvenile hallux abducto valgus association with metatarsus adductus. J Am Podiatr Med Assoc. 1994;84:219–224
- . Hallux valgus and metatarsus adductus: the surgical dilemma. Clin Podiatr Med Surg. 1989;6:103
- . Review of adult foot radiology. In: Hetherington VJ editors. Hallux Valgus and Forefoot Surgery. New York: Churchill Livingstone,; 1994; edited by
- . Metatarsus adductus and hallux abducto valgus: their correlation. J Foot Surg. 1987;26:304–308
- . A simplified metatarsus adductus angle. J Am Podiatry Assoc. 1983;73:620–628
- . Metatarsus adductus and selected radiographic measurements in the first ray of normal feet. J Am Podiatr Med Assoc. 1992;82:616–622
- . X-ray evaluation of hallux abducto valgus deformity. J Am Podiatr Med Assoc. 1974;64:544–566
- . Survey of foot deformities among 166 geriatric inpatients. Foot Ankle. 1995;5:156–164
- . The shape of the metatarsal head as a cause of hallux abductovalgus. Foot Ankle Int. 2002;23:236–242
- . Clinical assessment of hypermobility of joints: assembling criteria. J. Rheumatol. 1992;19:115–122
- . Hypermobility: features and differential incidence between the sexes. Arthritis Rheum. 1987;30:1426–1430
PII: S1067-2516(03)70046-5
doi:10.1053/jfas.2003.50002
© 2003 American College of Foot and Ankle Surgeons. Published by Elsevier Inc. All rights reserved.
