0198-0211/88/0806-0331 $0200/0 FOOT & ANKLE . Copyright © 1988 by the American Orthopaedic Foot and Ankle Society, Inc.

Relationship Between Adolescent Bunions and Flatfeet Vicki Kalen, M.D." and Allan Brecher, M.D.t Albuquerque, New Mexico

ABSTRACT Pes planus has been suggested as a factor in the development of bunions in adolescents. This study objectively evaluated the radiographs of 36 adolescents with 66 hallux valgus deformities for the presence of associated flatfeet. The bunion deformity in these adolescents was also characterized. Only 26.6% of these adolescents had an increased intermetatarsal angle and 8.1% an abnormal metatarsus varus angle as compared with the published norms. Measurements of pes planus including talar pitch, calcaneal plantar angle, dorsoplantar talonavicular angle, and lateral talocalcaneal angle showed an incidence of flatfeet in adolescents with bunions that was 8 to 24 times greater than expected.

and 1985 undertaken to identify cases clinically diagnosed as hallux valgus, 75 patients younger than 21 years of age were found. Patients were exc~u~ed f~o~ the study if they had had surgery prior to their first VISIt. Because we needed objective radiographic verification of the hallux valgus for the planned study, those patients without standard weightbearing AP and lateral radiographs of their feet were also excluded. We assumed that those patients without radiograms probably had milder cases that may not have even satisfied the diagnostic radiographic criterion. Charts and weightbearing radiographs were reviewed. Hallux valgus was defined as an angle greater than or equal to 20° between the long axis of the proximal phalanx and the long axis of the first metatarsals of the great toe (hallux valgus angle) (Fig. 1)6,9 This definition excluded an additional 11 patients who although described as having hallux valgus by the examining physician, did not meet the radiographic criterion. A balance of 36 patients with 66 affected feet remained for further evaluation. Two additional measurements commonly used to define a bunion deformity were obtained. These were the metatarsus varus angle (the angle between the longitudinal axis of the medial cuneiform and the longitudinal axis of the first metatarsal in the AP projection) and the intermetatarsal angle (the angle between the longitudinal axes of the first and second metatarsals) (Fig. 1). Abnormal values (more than 2 standard deviations from the norm) were greater than 25° and greater than 14 0, respectively. 5,6,9 Each of these measurements is used in the literature to define "metatarsus primus varus," although the metatarsus varus angle is the more appropriate. Pes planus was radiographically defined by five different measurements: the calcaneal plantar angle (the angle the plantar surface of the calcaneus makes with the horizontal in a lateral projection). dorsoplantar talonavicular angle (the angle between a line parallel to the distal articular surface of the navicular and a line through the longitudinal axis of the talus in an AP projection), talar pitch (the angle the long axis of the talus makes with the horizontal in a lateral projection),

The question of what causes hallux valgus remains partially unanswered. Flaccid ligaments, poor muscle tone bony anatomy, and joint configurations have been , . t he past. 2,5,15 Inman 10 implicated as causal factors In suggested that pes planus is a causative factor in hallux valgus, stating that he never saw a cavus foot with a bunion. Hohmann" asserted that hallux valgus is always seen in combination with pes planus and that pes planus is always the predisposing factor in hallux valgus. To date, no attempts have been made to objectively look at the relationship between pes planus and hallux valgus. In this study, we reviewed a group of young patients with hallux valgus deformity to determine whether pes planus was found more frequently than expected in this patient group. We hoped to verify this previously anecdotal correlation between the two. MATERIAL AND METHODS

In a search of the medical records of Carrie Tingley Hospital, Albuquerque, New Mexico, between 1937 • Assistant Professor, Department of Orthopaedics, University of Florida, Box J-246, JHM Health Center, Gainesville, FL 32610. To whom requests for reprints should be addressed. t Chief Resident, Department of Orthopaedics, University of Colorado Health Sciences Center, Box B-202, 4200 E. 9 Ave., Denver, CO 80262. 331

332

KALEN AND BRECHER

Foot & Ankie/Vol. 8, No. 6/June 1988 RESULTS

Fig. 1. HV, hallux valgus angle; MV, metatarsus varus angle; 1M, intermetatarsal angle.

lateral talocalcaneal angle, and the AP talocalcaneal angle. (Fig. 2). The AP talocalcaneal angle could not be accurately measured on many radiographs and was therefore deleted as a criterion. Abnormal values for these measurements were obtained from the literature, and, in each case, measurements exceed two standard deviations from the norm. Pes planus was thus defined by the following values: calcaneal plantar angle less than 15%, dorsoplantar talonavicular angle less than 60%, talar pitch greater than 37°, and lateral talocalcaneal angle greater than 50°.3.17 Charts were reviewed for epidemiologic data and any secondary diagnoses. Surgical procedures, when performed, were noted.

The patient group consisted of 24 females and 12 males, a female predominance of 2:1. The median age was 13 years and 4 months, with a range from 6 months to 20 years and 8 months. The age distribution showed the highest frequency between 11 and 15 years of age (Fig. 3). Other diagnoses in the 36 patients included 1 patient with Marfan syndrome, 1 patient with cerebral palsy, 2 patients with short tendo achilles, 1 patient with hereditary multiple exostoses, 1 patient with syndactyly, 1 patient with claw toes, 1 patients with pes cavus, and 3 patients with congenital dislocation of the hip. There were no patients with Down syndrome. The mean, standard deviation, median, minimum, and maximum values for each of the seven radiologic measures are seen in Table 1. The mean hallux valgus angle was 28.5° and the median 27.0°, whereas the normal range is 14 to 19°. 9 The mean intermetatarsal angle was 12.3° and the median 12.0°, both falling within the normal range of 5 to 14°. The metatarsus varus angle was also in the normal range, with a mean in our patients of 18.5° and a median of 19.0° (normal range 11 to 25°). Put more simply, our patients all had true hallux valgus by definition of the hallux valgus angle, but not many had metatarsus primus varus by definition of either the intermetatarsal angle or the metatarsus varus angle. The mean calcaneal plantar angle was just outside the normal range at 14.7°, as was the median value of 14.0° (normal values greater than 15°).17 The mean dorsoplantar talonavicular angle of 58.8° and median of 59.5° also were less than the normal range of 60 to 80° .3.17 The mean talar pitch was 30.1 ° and the median was 29.5°, with normal values of 16 to 37°.17 The mean lateral talocalcaneal angle also was within the normal range at 44.0° (normal 35 to 50°) as was the median of 42.5°.16 Thus, of our four measurements of pes planus, looking at only the overall averages, the means and medians of the calcaneal plantar angle and the dorsoplantar talonavicular angle were in the range defining an abnormally flatfoot. We evaluated the data on an individual case basis and then summarized the percentage of feet classified as abnormal for each radiological measure (Table 2). These data reiterate the findings in Table 1. Abnormal measurements were found in the calcaneal plantar angie and the dorsoplantar talonavicular angle in 63.3% and 50.0% of the patients, respectively. Again, the measures of metatarsus primus varus were abnormal in only 26.6% (intermetatarsal) and 8.1% (metatarsus varus) of the feet. The talar pitch and lateral talocalcaneal angles were abnormal in 18.3% and 21.7%, respectively. In Table 3, the percentage of patients who had pes

ADOLESCENT BUNIONS AND FLATFEET

Fl?ot & Ankle/Vol. 8, No. 6/June 1988

A

Fig. 2.

333

A. LTC. lateral talocalcaneal angle; CPo calcaneal plantar angle.

20

en

I-

z

w i=

15

-c c,

u.. 0

a: w

10

al

~

::::>

z

5

0-5

6-10

11-15

16-21

AGE (in years) Fig. 3.

Fig. 2. B. DPTN, dorsoplantar talonavicular angle; APTC. anteroposterior talocalcaneal angle.

Age distribution.

planus by anyone of the four different measures and additionally met the criteria for a bunion deformity is shown. The highest correlations were again seen in those cases with abnormal calcaneal plantar angles or dorsoplantar talonavicular angles. Of those patients with hallux valgus alone, 63.3% had an abnormal CP angle, whereas 61.1% with both hallux valgus and an abnormal intermetatarsal angle, and 50.0% with hallux valgus and metatarsus varus, had an abnormal CP angle. The results for the talonavicular angle were similar: abnormalities in this measurement were seen 50.0% of the time when there was an abnormal hallux valgus angle only, 61.1% of the time with abnormal hallux valgus and intermetatarsal angles, and 50.0'% of the time with abnormal hallux valgus and metatarsus varus angles. The talar pitch and lateral talocalcaneal angles correlated less strongly with the bunion measurements, ranging from 16.7% to 33.3% in any combination. We calculated the statistical agreement between abnormal values in the intermetatarsal angle and the

334

KALEN AND BRECHER

Foot & Ankle/Vol. 8, No. 6/June 1988 TABLE 1 Values for Radiologic Measures'

8

Variable angle

Mean

Standard deviation

Median

Minimum

Maximum

Hallux valgus Intermetatarsal Metatarsus varus Calcaneal plantar Talonavicular Talar pitch Lateral talocalcaneal

28.5 12.3 18.5 14.7 58.8 30.1 44.0

8.3 2.5 5.1 6.9 9.9 7.6 7.8

27.0 12.0 19.0 14.0 59.5 29.5 42.5

19.0 7.0 8.0 -6.0 34 12 23

54 17 30 33 78 50 64

Values are given in degrees. TABLE 2 Percentage of Feet Classified as Abnormal Variable angle

Abnormal

Intermetatarsal Metatarsus varus Calcaneal plantar Talonavicular Talar pitch Lateral talocalcaneal

>14 >25 50

o

Yo Abnormal

17/64 5/62 38/60 30/60 11/60 13/60

(26.6%) (8.1%) (63.3%) (50.0%) (18.3%) (21.7%)

95%

Confidence interval

15.6%,37.6% 1.2%,15.0% 50.9%, 75.7% 37.1%,62.9% 8.2%,28.4% 11.1%,32.3%

metatarsus varus angle and found no agreement at all (Table 4). It should be recalled that all cases had an abnormal hallux valgus angle by our definition, and therefore agreement between hallux valgus and intermetatarsal and between hallux valgus and metatarsus varus was 26.6% and 8.1%, respectively, as seen in Table 2. Agreement was defined as the number of cases that were abnormal on both measures divided by the number of cases abnormal on either measure. Agreement between the four pes planus measurements was insignificant for all pairs except the calcaneal plantar and talonavicular angles, which were in 40.9% agreement. In our final statistical analysis, correlations were used that defined the association between two measurements without dichotomizing the scale (Tables 5-7). The only significant correlations were between some of the various flatfoot measurements, as expected. There was a 0.36 correlation between the talar pitch and calcaneal plantar angles, i.e., as one increased, the other decreased (Table 6). A significant positive correlation was found between the calcaneal plantar and lateral talocalcaneal angles (0.36) and the talar pitch and lateral talocalcaneal angles (0.65). Intuitively, the correlation is what one might expect because of the bony relationships these measurements define. There were no significant correlations between the three measures of the bunion deformity and the four measures of the flatfoot deformity (Table 7). Of the 66 feet, 24 were surgically corrected; distal metatarsal osteotomies of the Mitchell type were performed on 12 feet; more proximal metatarsal osteotomies were done in 8 teet:" 3 feet were treated by

simple bunionectomy with soft tissue realignment; and 1 foot was corrected by osteotomy of the proximal phalanx. DISCUSSION

Hallux valgus is an adolescent orthopaedic problem. Hardy and Clapham" found that 46% of adults reported an age of onset of their bunions of less than 20 years, whereas 15% reported an age of onset of less than 15 years. Piggott 13 reported that 57% stated that hallux valgus began in their teen years or earlier, 38% did not know, and only 5% believed that the problem began after 30 years of age. Due to hospital policy, our patient population was within 21 years of age, yet 67% of our patients with bunions were 15 years of age or younger (24 out of 36). Much emphasis has been placed on whether lateral deviation of the proximal phalanx (hallux valgus) or medial deviation of the first metatarsal develops first. Piggott 13 likened the controversy concerning which deformity appears first to that of "the chicken or the egg." A study by Houghton and Dickson? showed that the structural abnormality in hallux valgus in the young is really a valgus disposition of the second and subsequent metatarsals rather than varus inclination of the first metatarsal. They did not find true metarsus primus varus more often in feet with hallux valgus than in a normal control group. Our data agree with the absence of metatarsus primus varus in adolescents. In addition, "metatarsus secundus valgus" was rarely found in our patients. Only 27% had a widened intermetatarsal angle. The order in which either deformity develops is a moot point, because our data suggest that there is only one deformity. The hallux is in valgus alignment, but the metatarsals are usually in normal alignment. Although the majority of our patients had no other diagnoses, 11 did have other conditions, many of which are known to be associated with an increased incidence of flatfeet and/or hallux valgus. One patient had Marfan syndrome, a disease in which connective tissue support is inadequate. Two patients had short tendo achilles with hypermobile flatfoot, an entity well documented by

Foot & Ankle/Vol. 8, No. 6/June 1988

ADOLESCENT BUNIONS AND FLATFEET

335

TABLE 3 Percentage of Patients with Pes Planus Hallux valgus> 19

Variable angle

(N = 60)

Calcaneal plantar < 16 Talonavicular < 60 Talar pitch> 37 Lateral talocalcaneal > 50

Hallux valgus> 19 and intermetatarsal > 14

Hallux valgus> 19 and metatarsus varus> 25

(N = 18)

(N = 6)

61.1 55.6 33.3 33.3

50.0 66.7 16.7 16.7

63.3 50.0 18.3 21.7

TABLE 4 Percentage of Agreement Variable angle

lnterrnetatarsal

Intermetatarsal Metatarsus varus Calcaneal plantar Talonavicular Tatar pitch Lateral talocalcaneal

100.0

Metatarsus varus

100.0

Hallux valgus Intermetatarsal Metatarsus varus

Hallux valgus

1.00

Intermetarsal

Metatarsus varus

0.03 1.00

0.21 -0.01 1.00

TABLE 6 Correlations Among Measures of Flatfoot Variable angle Calcaneal plantar Talonavicular Talar pitch Lateral talocalcaneal

Calcaneal plantar

Talonavicular

1.00

0.09 1.00

Talonavicular

Talar pitch

40.9 100.0

2.6 3.7 100.0

Lateral talocalcaneal

0.0 100.0

TABLE 5 Correlations Among Measures of Bunion Variable angle

Calcaneal plantar

Talar pitch

-0.36' -0.06 1.00

Lateral talocalcaneal

0.36' 0.00 0.65' 1.00

, Significantly greater than 0.00, P < 0.05.

TABLE 7 Correlations Between Measures of Both Bunion and Flatfoot Variable angle

Calcaneal plantar

Talonavicular

Talar pitch

Lateral talocalcaneal

Hallux valgus Intermetatarsal Metatarsus varus

0.02 0.17 0.03

-0.04 -0.08 -0.01

-0.18 0.00 0.06

-0.05 0.02 0.07

Harris and Beath." One patient had cerebral palsy. Hoffer and Seaquist" reported that an analysis of 100 consecutive cerebral palsy clinic patients showed 47% with a valgus heel and, of those, 18% had an associated hallux valgus of 30° or more. Of interest are the three patients with congenital dislocation of the hip, a condition known to be associated with increased ligamen-

8.5 9.3 0.0 100.0

tous laxity. There has not previously been a report of an increased incidence of hallux valgus in patients with congenital dislocation of the hip, however. One patient was clinically described as havinq a cavus foot. Her hallux valgus angle was markedly increased at 54°. Her talar pitch was abnormally low, consistent with cavus deformity, but her lateral talocalcaneal angle was within the normal range. Thus, it is unclear whether this truly was a cavus foot and if it refutes inman's" claim that bunions are never seen in a cavus foot. The increased percentage of associated other diagnoses (11/36 or 31%) underlines the fact that systemic diseases causing generalized ligamentous laxity and abnormal muscle tone appear to play a significant role in the development of pediatric hallux valgus. The possibility of any connective tissue or muscle dysfunction must be considered in the evaluation and treatment of adolescent bunions. It is already well accepted that generalized ligamentous laxity and hypotonia contribute to flatfeet. When discussing pes planus, it is essential to define the term. Unfortunately, Rosa" states that the term is meaningless in any scientific sense. No concensus exists on a definition. Bleck' says, when discussing the longitudinal arch of the foot, "how high is high and how low is low?" The term, valgus of the heel, does not describe the arch or forefoot and is not invariably present in flatfeet. Hyperpronation is probably a better term, but how does one measure that? Most authors":":" use a variety of radiographic measures and often a footprint measurement to objectively diagnose pes planus. Because this is a retrospective study, we had only radiographs for objective measurements.

336

KALEN AND BRECHER

We chose the five radiographic measurements used to depict different aspects of this complex, ill-defined entity, "pes planus." However, three of the measurements focus on the hindfoot and only one on the midfoot. The dorsoplantar talonavicular angle as defined by Giannestras" delineates the medial deviation of the talus in relation to the navicular, which Giannestras believed contributed to valgus flatfoot. The lateral talocalcaneal angle is the most commonly used measurement to detect hindfoot varus or valqus." Depression of the longitudinal arch can result from plantar and/ or valgus deviation of any or all three of the joints comprising the arch-talocalcaneal, talonavicular, and naviculocunertorrn." Both the talar pitch and calcaneal plantar angles should reflect this plantar deviation.v'" An increased AP talocalcaneal angle shows valgus deviation. Unfortunately, this measurement could not be accurately seen on many of our films and was not used. On the lateral radiograph, sagging of the normal straight line through the long axis of the talus, navicular, cuneiform, and first metatarsal indicates a depressed arch.":" No quantification has ever been applied to this "sag." Therefore, we did not use this linear relationship, having no data for normal range or standard deviations. Also, because we did not attempt to locate the particular joint(s) where the break in the longitudinal arch occurred, this type of radiographic analysis was not germane to our study. Our study is thus hampered by the lack of a standardized radiographic definition of a flatfoot. For the most part, with the measurements reported in the literature and those used by us only the hindfoot will be evaluated. The diagnosis of pes planus in practice is based to a greater extent on clinical appearance than on radiographic measurements and unfortunately, is often a subjective impression. Radiographic measurements for quantitative data are useful tools in undertaking any objective study of flatfeet. In our data, a significant incidence of abnormally low calcaneal plantar angles and dorsoplantar talonavicular angles are seen among patients with hallux valgus (50% to 60%). The additional two measurements of lateral talocalcaneal angle and talar pitch were considered abnormal only among approxirnately 20% of the bunion patients. Although this is a much lower percentage than the other two angles, it still represents a much higher percentage than should be ~een in the general population. Our measurements were considered abnormal if they were more than 2 standard deviations from the norm. In a random population only 2.5% of patients in that range should be seen. Thus, the 20% to 60% of our patients greater than 2 standard deviations from the normal values for these four flatfeet measurements exceed this expected 2.5% substantially. By radiographic measurements, adolescent patients with bun-

Foot & AnklejVol. 8, No. 6jJune 1988

ions have an 8 to 24 times higher than expected incidence of flatfeet. Certainly this finding confirms the existence of a relationship between the two, but does not explain which is cause and which is effect. The correlation data show no trends in which an increase or decrease in any of the individual flatfeet radiographic parameters correlate with an increase in the hallux valgus angle. This study objectively substantiates a relationship between adolescent bunions and flatfeet as defined by radiographic criteria for the two conditions. The longstanding theory in the literature that there is an association between the two is thus elevated beyond the merely anecdotal. Further prospective study is needed to clarify the relationship and ellucidate which is cause and which is effect.

REFERENCES 1. Bleck, E.E.: Personal communication, 1983. 2. Cholmeley, J.A.: Hallux valgus in adolescents. Proc. R. Soc. Med., 51:903,1958. 3. Giannestras, N.J.: Flexible valgus flatfoot resulting from naviculocuneiform and talonavicular sag: surgical correction in the adolescent. In Foot Science. Bateman, J.E. (ed.), Philadelphia, W.B. Saunders, 1976, pp. 67-105. 4. Giannestras, N.J., and Smith, R.W.: Grice arthrodesis in nonparalytic flexible flat feet. In The Foot and Ankle. Bateman, J.E., and Trott, A. W. (eds.), New York, Thieme-Stratton, 1980, pp. 116-126. 5. Goldner, J.L., and Gaines, R.W.: Adult and juvenile hallux valgus: analysis and treatment. Orthop. Clin. North Am., 7:863887,1976. 6. Hardy, R.H., and Clapham, J.C.R.: Observations on hallux valgus. J. Bone Joint Surg., 33B:376-391, 1951. 7. Harris, R.I., and Beath T.: Hypermobile flat-foot with short tendo achillis. J. Bone Joint Surg., 30A: 116-140, 1948. 8. Hoffer, M. M., and Seaquist, J.L.: Surgical correction of hallux valgus in cerebral palsy. In The Foot and Ankle. Bateman, J.E., and Trott, A.w. (eds.), New York, Thieme-Stratton, 1980, pp. 143-147. 9. Houghton G.R., and Dickson, R.A.: Hallux valgus in the younger patient. J. Bone Joint Surg., 61B:176-177, 1979. 10. Inman, V.T.: Hallux valgus: a review of etiologic factors. Orthop. Clin. North Am., 5:59-66, 1974. 11. Hohmann, G., cited by Kelikian, H.: Hallux Valgus Allied Deformities of the Forefoot and Metatarsalgia. Philadelphia, W.B. Saunders, 1965, pp. 58-61. 12. Mitchell, C.L., Flemming, J.L., Allen, R., Glenney, C., and Sanford, G.A.: Osteotomy bunionectomy for hallux valgus. J. Bone Joint Surg., 40A:41-60, 1958. 13. Piggott, H.: The natural history of hallux valgus in adolescence and early adult life. J. Bone Joint Surg., 42B:749-760, 1960. 14. Rose, G.K.: Pes planus. In Disorders of the Foot. Jahss, M.H. (ed.), Philadelphia, W.B. Saunders, 1982, pp. 486-520. 15. Scranton, P.E., Jr., and Zuckerman, J.D.: Bunion surgery in adolescents: results of surgical treatment. J. Pediatr. Orthop., 4:39-43,1984. 16. Simons, G.W.: Analytical radiography of clubfoot. J. Bone Joint Surg., 59B:485-489, 1977. 17. Tachdjian, M.O.: The Child's Foot. Philadelphia, W.B. Saunders, 1985, pp. 556-557.