Multiplanar, Multivariate Analyses of Hallux Valgus Radiographic Parameters

Abstract

Hallux valgus is a complex condition understood to involve pathomechanics in all 3 of the cardinal planes. Despite this, the bulk of its historical evaluation has been in the transverse plane, and one might argue that the traditional and more commonly performed univariate and bivariate analyses within the literature do not comprehensively describe the potential interrelationships between the planes during perioperative assessment. Therefore this investigation aimed to evaluate relationships between common radiographic parameters measured in the three cardinal planes by means of a multivariate regression analysis. Serial analyses utilizing the first intermetatarsal angle, hallux valgus angle, tibial sesamoid position, proximal articular set angle, Engel's angle, first metatarsal inclination angle, and the sesamoid rotation angle were performed with varying dependent and independent variables. The tibial sesamoid position (p < .001) and proximal articular set angle (p=.014) were found to be independently associated with the first intermetatarsal angle, while the hallux valgus angle (p=.712), Engel's angle (p=.646), first metatarsal inclination angle (p=.097), and sesamoid rotation angle (p=.099) were not. The tibial sesamoid position (p=.003), proximal articular set angle (p < .001), Engel's angle (p=.006), and sesamoid rotation angle (p=.003) were found to be independently associated with the hallux valgus angle, while the first intermetatarsal angle (p=.712) and first metatarsal inclination angle (p=.400) were not. The first intermetatarsal angle (p < .001), hallux valgus angle (p=.003), and proximal articular set angle (p=.015) were found to be independently associated with the tibial sesamoid position, while Engel's angle (p=.400), the first metatarsal inclination angle (p=.088), and the sesamoid rotation angle (p=.649) were not. These findings appear to question a direct relationship between the first intermetatarsal angle and hallux valgus angle, as well as potentially infer a relationship between the frontal plane with the hallux valgus angle.