Few studies have documented the outcome of conservative treatment of hallux valgus deformities on pain and muscle strength. We sought to determine the effects of foot mobilization and exercise, combined with a toe separator, on symptomatic moderate hallux valgus in female patients.
As part of the randomized clinical trial, 56 women with moderate hallux valgus were randomly assigned to receive 36 sessions for 3 months or no intervention (waiting list). All patients in the treatment group had been treated with foot joint mobilization, strengthening exercises for hallux plantarflexion and abduction, toe grip strength, stretching for ankle dorsiflexion, plus use of a toe separator. Outcome measures were pain and American Orthopedic Foot and Ankle Society (AOFAS) scores. Objective measurements included ankle range of motion, plantarflexion and abduction strength, toe grip strength, and radiographic angular measurements. Outcome measures were assessed by comparing pretreatment, posttreatment, and 1-year follow-up after the intervention. Mixed-model analyses of variance were used for statistical assessment.
Patients who were treated with 3 months of foot mobilization and exercise combined with a toe separator experienced greater improvement in pain, AOFAS scores, ankle range of motion, hallux plantarflexion and abduction strength, toe grip strength, and radiographic angular measurements than those who did not receive an intervention 3 months and 1 year postintervention (P < .001 for all comparisons).
These results support the use of a multifaceted conservative intervention to treat moderate hallux valgus, although more research is needed to study which aspects of the intervention were most effective.
Hallux valgus is a common structural foot deformity in which the angular deviation of the hallux is greater than 15° toward the lesser toes in relation to the first metatarsal bone; it appears as a medial bony prominence of the first metatarsal head.1 The hallux valgus angle is the angle between the bisection line of the first metatarsal bone and the proximal phalanx.2 A normal angle is defined as 15° or less, mild as less than 20°, moderate as 20° to 40°, and severe as greater than 40°.3 Hallux valgus is a deformity with many etiologies, such as familial history, female sex, occupational foot stress, shoe shape and style, a long first metatarsal, and an oval or curved metatarsophalangeal joint articulation surface.4 Hallux valgus has been stated to have a prevalence of 21% to 70% based on epidemiologic studies.5 Women are significantly more likely to develop hallux valgus than men.6
Hallux valgus is a major reason for orthopedic foot and ankle surgery each year,7 and it is accompanied by functional disability, foot pain, impaired balance, and high fall risk in older adults.8 Available treatments for hallux valgus are classified into operative and nonoperative treatments. Although surgery is considered the most common approach, it is expensive and may cause significant complications.9 Nonoperative treatments aim to decrease the angle of the first metatarsal big toe by stretching the contracted soft tissue around the joints by using night splints, improving muscle strength through foot exercises, or resolving abnormal function with orthoses.10
The dynamics of hallux valgus deformity are explained by examining the muscles of the first ray. The tendons and muscles that move the big toe are arranged around the metatarsophalangeal joint in four groups: the long and short extensors, the short flexors, and the two tendons of the abductor and adductor halluces.6
The intrinsic muscles of the foot when subjected to unsuitable footwear or because of inherent fragility may chronically malfunction and become an important factor in the pathomechanics of hallux valgus and lesser toe deformity.10
A toe separator can correct the alignment of the toes, and this functional realignment helps the foot to propel better in gait cycles.11
The purpose of this randomized clinical trial was to determine the effect of a foot mobilization and exercise program combined with the use of a toe separator to treat symptomatic moderate hallux valgus in female patients at the end of treatment and at 1-year follow-up.
In this study, we hypothesized that patients who participated in a foot mobilization and exercise program and also used a toe separator would have greater improvement of symptomatic moderate hallux valgus in terms of pain, functional disability, muscle strength, and radiographic measurements of the big toe than would patients who did not receive any intervention (natural course of the disease).
This was a randomized controlled clinical trial. Ethics committee approval for the study was obtained from the institutional review board of Cairo University, Cairo, Egypt.
Participants and Inclusion and Exclusion Criteria
Fifty-six women with a diagnosis of symptomatic moderate hallux valgus were included in this study. All of the patients had been referred by a physician to the outpatient physical therapy clinic between September 9, 2012, and August 30, 2015. Only 21% of eligible patients participated. All of the participants had pain and functional disability associated with symptomatic moderate hallux valgus deformity. In addition, all patients reported that they were not undergoing physical therapy or using orthotic devices or dynamic splints as treatment for hallux valgus. The study population comprised adult female patients, as there is an increased incidence of symptomatic hallux valgus in women.12
Patients who met the following criteria were excluded: previous foot surgery, underlying ankle deformity or pathologic hindfoot or midfoot deformities, or any deformity of the hallux other than valgus deformity. Patients with systemic pathologic disorders such as rheumatoid arthritis and gout were also excluded. Patients who were using anti-inflammatory drugs or analgesics or in whom manual therapy was contraindicated were excluded.13 Pregnant women were excluded because plain film radiography is contraindicated during pregnancy.
All of the measurements were made by an examiner with 20 years of experience, who is not the author. The measurements included 1) pain using a visual analog scale, 2) functional assessment using the metatarsophalangeal-interphalangeal score of the American Orthopedic Foot and Ankle Society (AOFAS), 3) radiographic measures of the hallux valgus angle and the first-second intermetatarsal angle, 4) goniometric measurement of passive range of motion of ankle dorsiflexion, 5) first metatarsophalangeal joint plantarflexion and abduction muscle strength, and 6) toe grip strength.
All of the measurements were made 2 days before the first treatment session, at the end of the last session, and 1 year after the treatment for the treatment and control groups.
Foot Pain and Functional Disability
The evaluation was performed on all of the patients. Pain was quantified with a 10-point visual analog scale (with 0 indicating no pain and 10 indicating severe pain). A score was determined by using the hallux interphalangeal joint scale of the AOFAS.14 This 100-point scale assesses subjective and objective factors. The numerical score is composed of a separate section for pain (0–40 points), function (0–45 points), and alignment (0–15 points). These scores were calculated before treatment, after treatment, and at 1-year follow-up.
Patient Demographic Data
The patient demographic data include women, mean ± SD age of 45.7 ± 6.8 years for the treatment group and 45.5 ± 6.2 years for the control group, height, weight, and mean ± SD body mass index (calculated as weight in kilograms divided by height in meters squared) of 25.6 ± 1.9 for the treatment group and 25.0 ±2.4 for the control group (Table 1).
Standardized anteroposterior weightbearing radiographs of all of the feet were taken. The hallux valgus angle and first-second intermetatarsal angle were measured according to the guidelines set forth by the AOFAS Committee on Angular Measurements.2
Goniometric Measurements of Passive Ankle Dorsiflexion
Patients were in the prone position with the knees extended and the feet over the edge of a plinth. The neutral position of the subtalar joint was identified and then the foot was dorsiflexed until a firm end point was felt. The goniometer, which was marked in 1° increments, was positioned on the lateral side of the ankle with the axis over the lateral malleolus.15
Hallux Plantarflexion and Abduction Strength Measurements
Hallux plantarflexion and abduction strength were evaluated using 50-kg load cells (GK 2126-50; Gedge Systems, Melbourne, Australia) mounted in a custom-built frame. This testing protocol has been previously described.16 While the patient was seated with the knee in 30° of flexion and the lower leg and foot stabilized using Velcro straps, the patient performed three isometric maximum voluntary contractions in hallux plantarflexion and abduction, and the maximum force achieved over three trials was used for analysis.
Measurement of Toe Grip Strength
The dynamometer (T.K.K. 3362; Takei Scientific Instruments, Niigata, Japan), which was used to measure toe grip strength, has been reported to be reliable.17 The patients sat upright on the chair without leaning on the backrest. Throughout toe grip strength measurement, the hips and knees were both flexed approximately 90°, and the ankles were placed in the neutral position and fixed with a strap. The first proximal phalanx was positioned at the grip bar, and the heel stopper was adjusted to fit the heel of each patient. The first toe was used as a benchmark to set up the testing position; the bar was then gripped with maximal effort using all of the toes. The examiner stabilized the toe grip dynamometer, and two toe grip strength measurements were made for each foot.
After the baseline examination, patients were randomly assigned to receive treatment or no intervention (natural history of the disease) using a computer-generated table of randomized numbers created before the start of data collection by an assistant to the author who was not involved in the recruitment or treatment of patients. The randomized group assignments' sequentially numbered index cards were placed in sealed, opaque envelopes. The author was blinded to the baseline examination findings. The envelope was opened and the treatment was performed according to the assigned groups.
Patients in the control group were asked to avoid surgical and foot orthotic therapy during follow-up, and they did not receive the intervention, so that the natural course of the condition could be determined. These patients were asked to continue taking their anti-inflammatory medications and performing normal activities without exacerbating their symptoms or seeking additional treatment during the study period.
All of the patients in the treatment group wore a toe separator made of silicon material (Voberry, Spain). The patients were required to wear the toe separator for more than 8 hours per day. The physical therapy program consisted of three sessions per week for 12 weeks. The author performed this treatment.
The physical therapy program was started with manual therapeutic interventions, which were performed for all of the metatarsophalangeal joints. These manipulations focused on an improvement of flexion and included caudal sliding of the proximal phalanx to improve extension. In addition, oscillating traction was performed to activate the mechanoreceptors that inhibit the afferent pain sensors. The treatment program also included mobilization of the first metatarsophalangeal Lisfranc, transverse tarsal, subtalar, and ankle joints.7
Nonweightbearing tendo Achillis stretching was performed manually by the therapist. The patients were instructed to lay supine, and the therapist did the stretching manually; the stretching was repeated five times during each session and was held for 15 sec each time. The patients were advised that the stretching should be pain free, although a small degree of unpleasantness was allowed.18
Hallux plantarflexion strengthening exercises consisted of ten isometric contractions, performed with a hold time of 10 sec each. The physical therapist provided resistance manually to the entire phalanx with the metatarsophalangeal joint in neutral. Patients were instructed not to flex the interphalangeal joint during the exercise. The isometric contraction occurred at the metatarsophalangeal joint. We increased the repetition by one set every week, and by the last week it had increased to eight sets of ten repetitions with 12 sec of rest between each set.19
For the hallux abduction strengthening exercise, the patients maintained a long sitting position, the heel was held, and pressure was exerted on the first metatarsal and proximal phalanx of the medial axis to create an abduction of the big toe.20 Ten isometric contractions were performed, with a hold time of 10 sec each. We increased the repetition every week by one set, and by the last week the exercise had increased to eight sets with ten repetitions, allowing 12 sec of rest between each set.
For the towel curl exercise, the patients were instructed to place a towel on the hardwood floor and place their toes on the edge of the towel. They were then instructed to drag the towel under their foot by flexing their toes, generating a strong grip on the fabric for 5 sec per repetition. The patients were instructed to perform ten repetitions, with 5 sec of rest between each set.21
Data were collected before and after treatment (after 24 sessions) and at 1-year follow-up. Therefore, three sets of data were available for analysis for each intervention and for both groups. Patients with hallux valgus consult with musculoskeletal practitioners such as podiatric physicians and physiotherapists, mainly because of pain and functional disability. Therefore, the outcome measures selected were pain, measured using a visual analog scale; functional disability, measured with the AOFAS scale; radiographic measures (hallux valgus angle and the first-second intermetatarsal angle); range of motion of ankle dorsiflexion; plantarflexion and abduction hallux muscle strength; and toe grip strength.
Statistical analysis was performed with IBM SPSS Statistics for Windows, Version 18.0 (IBM Corp, Armonk, New York). Data are statistically described in terms of mean, standard deviation, and 95% confidence interval, which were evaluated for each variable.
Sample size was based on a previous power calculation using standard deviations obtained from preliminary data analysis (n = 56). We determined that 56 patients in two groups would provide 80% power to detect a difference of 11 mm between patients on the 100-mm pain visual analog scale (α = 0.05).22
The Kolmogorov-Smirnov test showed a normal distribution of quantitative data. Because the participants received intervention only for one foot, and it was the worse affected limb, we used a 3 × 2 mixed model of variance (analysis of variance), with time (baseline, 3 months after therapy, and 1 year after therapy) as the within-subject factor and group (treatment, control). We also used the unpaired Student t test to compare the treatment and control groups before the intervention. Statistical significance was defined as P < .05.
Fifty-six female patients (56 feet) with symptomatic moderate hallux valgus screened by eligibility criteria agreed to participate in this study. They were randomized to either the treatment group (a program of exercises combined with use of a toe separator) (n = 28) or the control group (n = 28). Baseline features between the two groups were similar for all of the variables (Table 1). The 3 × 2 mixed-model analysis of variance revealed significant group × time interactions (P < .001) for all of the variables measured, in which patients were treated with the foot mobilization and exercise program combined with use of a toe separator (Table 2).
The mean ± SD hallux valgus angle before the treatment was 32.7° ± 4.2°. After the treatment, the mean ± SD angle decreased to 23.8° ± 3.1° (P < .001), and the improvement remained stable; at 1-year follow-up, the angle was 25.8° ± 2.1° (P < .001).
The mean ± SD first-second intermetatarsal angle before the treatment was 14° ± 1° and after the treatment was 11.8° ± 0.5° (P < .001), and the improvement remained stable; at 1 year it was 12° ± 0.9° (P < .001).
Pain Measurements Using the Visual Analog Scale
The mean ± SD pretreatment pain score was 5.6 ± 1 points (range, 0–10 points). The mean ± SD score on the 10-point visual analog scale after treatment was significantly decreased to 2.2 ± 1 points (P < .001), and the improvement remained stable (2.4 ± 1 points at 1 year; P < .001).
Functional Disability Measured by AOFAS Score
The mean ± SD pretreatment AOFAS score was 46.1 ± 1.4 points and after the treatment improved significantly to 76.2 ± 1.5 points (P < .001). The improvement remained stable; at 1 year it was 74.5 ± 2 points (P < .001).
Ankle Dorsiflexion Passive Range of Motion
Mean ± SD ankle dorsiflexion passive range of motion was 9.5° ± 1.2° before the treatment and 15.2° ± 2.1° after the treatment, which was significant (P < .001). The improvement remained stable; at 1 year it was 13.2° ± 2.1° (P < .001).
Hallux Plantarflexion Strength
Mean ± SD hallux plantarflexion strength before the treatment was 50.4 ± 2.8 N and after the treatment significantly increased to 65.9 ± 5.6 N (P < .001). It remained stable (62.9 ± 2 N at 1 year; P < .001).
Hallux Abduction Strength
Mean ± SD hallux abduction strength before the treatment was 6.4 ± 1 N and after the treatment was 10.5 ± 1.6 N (P < .001). The improvement remained stable; at 1 year it was 8.8 ±1.3 N (P < .001).
Toe Grip Strength
Mean ± SD toe grip strength before the treatment was 65.2 ± 10.4 N and after the treatment increased to 98.1 ± 9.2 N (P < .001). The improvement remained stable; at 1 year it was 93.1 ± 5.2 N (P < .001).
In the control group there were no statistically significant differences during the 1 year of waiting, despite increases in pain intensity, hallux valgus angle, and first-second intermetatarsal angle. There were also decreases in AOFAS scores, ankle dorsiflexion, hallux plantarflexion and abduction strength, and toe grip strength (Table 3).
Comparison of the Treatment and Control Groups After the Treatment and at 1-Year Follow-up
These comparisons showed significant differences for hallux valgus angle, first-second intermetatarsal angle, pain and AOFAS scores, ankle dorsiflexion, hallux plantarflexion and abduction strength, and toe grip strength (P < .001 for all) (Table 4).
The findings from the present study support the hypothesis that a foot mobilization and exercise program, combined with use of a toe separator, improves pain, functional disability, ankle dorsiflexion range of motion, hallux plantarflexion and abduction strength, and toe grip strength, in addition to decreasing the hallux valgus and first-second intermetatarsal angles.
Conservative treatment for symptomatic hallux valgus, such as manual and manipulative therapy, compared with the outcomes of surgical approaches, provided an approximately 15-mm change in visual analog scale scores at follow-up.23,24 This finding supports the development and testing of conservative treatments in the future. The cost savings over surgery and avoidance of postsurgical complications is expected to be significant.
The present study tested the results of an exercise program combined with use of a toe separator. Relief has been reported after manipulative therapy25 and with the use of an insole accompanied by a toe separator for painful hallux valgus.26 In the present study, the mean ± SD pain scores were significantly reduced from 5.6 ± 1 to 2.2 ± 1 points after treatment and continued to be stable at 1-year follow-up.
There were statistically significant associations between hallux valgus and pain, worsening function, and worsening foot health in patients with moderate hallux valgus deformity.27 The reported mean ± SD AOFAS scores after treatment increased from 46.1 ± 1.4 to 76.2 ± 1.5 and remained higher than the baseline measurement at 1-year follow-up.
Restricted ankle dorsiflexion may be associated with the development of hallux valgus.6 We acknowledge the limitation of using a goniometer to assess ankle dorsiflexion in this study. Measurements performed before and after the treatment showed that manual tendo Achillis stretching was efficient in increasing the mean ± SD range of motion from 9.5° ± 1.2° to 15.2° ± 2.1°. We performed this stretching because the muscles of the leg contribute to supporting the alignment of the hindfoot, midfoot, and first ray.28
The first ray is an inherently unstable axial array that relies on a fine balance between the static stabilizers (capsule, ligament, and plantar fascia), and dynamic stabilizers (peroneus longus and small muscles of the foot) to maintain its alignment.29 People with moderate-to-severe hallux valgus had significantly less hallux flexor strength than those without hallux valgus.30 We performed exercises to strengthen the plantarflexion of the hallux, and they were effective in increasing the mean ± SD strength from 50.4 ± 2.8 N to 65.9 ± 5.6 N.
The abductor hallucis muscle, in addition to its action in maintaining the alignment of the toe, also has a splitting effect on the first metatarsal head. Acting in a line parallel to this bone and using the head of the first metatarsal as a fulcrum, the abductor hallucis pushes the first metatarsal toward the second metatarsal.6 In measurements of the hallux abduction strength, there was a significant mean ± SD increase after the treatment from 6.4 ± 1.1 to 10.5 ± 1.6 N. Many studies used 50-kg load cells to evaluate the strength of hallux plantarflexion and abduction.16,26
The toe grip strength of individuals with hallux valgus has been reported to be weaker than that in healthy individuals.31 Toe strength grip was measured before the treatment, and there were significant increases in mean ± SD toe strength grip from 65.2 ± 10.4 N to 98.1 ± 9.2 N after treatment. A toe grip dynamometer, which we used in the present study, has been widely used in many studies.17,31,32
Toe curl exercises were successful in increasing toe grip strength because they are used to strengthen the flexor digitorum longus and brevis, lumbrical, and flexor hallucis longus muscles.33
There is an association between the magnitude of the hallux valgus angle and the first-second intermetatarsal angle,34 and these findings were confirmed in the present study. Before the treatment, the mean ± SD value of the hallux valgus angle was 32.7° ± 4.2° and of the first-second intermetatarsal angle was 14° ± 1°, and after the treatment, there were concomitant decreases in both angles; the hallux valgus angle was 23.8° ± 3.1° and the first-second intermetatarsal angle was 11.8° ± 0.5°.
In the present study, we confirmed that a toe separator can correct the alignment of the toes, and this functional realignment can help the foot to propel better in the gait cycles.11
In this trial we were able to evaluate the effectiveness of the treatment program. The treatment program showed considerable effectiveness in the outcome measures at 1-year follow-up, although we were not able to correct the deformity, and can be considered an option while waiting for surgery.
We selected the sample from patients with moderate hallux valgus. The sample did not include patients with mild hallux valgus because it was reported that there were no significant differences between those with mild hallux valgus and normal halluces in terms of muscle strength.16
The main limitation of this study is that the AOFAS scoring system is yet to be validated; however, the AOFAS scoring system is being widely used. Reliability of measurement methods must be considered as a potential limitation in any clinical research. The same examiner, who has 20 years of experience, performed all of the measurements in this study. We need further studies to compare different combinations of treatments to determine which combination is most effective.
None of the present patients had a complaint about the toe separator, and in the treatment group they continued to wear it after the 3 months of treatment till the last follow-up. No patients were lost from the study, and this was a strength of the study. All of the patients were instructed to wear shoes with a lower heel and a broader toe box.
The results of this randomized clinical trial suggest that the foot mobilization and exercise program, combined with use of a toe separator, decreases the pain intensity, functional disability, and radiographic measurements. It also increases the hallux plantarflexion and abduction strength, toe grip strength, and ankle dorsiflexion range of motion after the treatment and at 1-year follow-up in female patients with moderate hallux valgus deformity compared with a control group that did not receive any intervention. Further research is needed to evaluate the program with longer follow-up.
Financial Disclosure: None reported.
Conflict of Interest: None reported.