Ankle Arthritis Etiology Predicts Patterns of Gait Dysfunction: A Prospective Multivariate Gait Analysis

CATEGORY: Ankle; Ankle Arthritis; Basic Sciences/Biologics; Hindfoot INTRODUCTION/PURPOSE: Preoperative factors influencing functional disability imparted on the patient by ankle arthritis have not previously been assessed with gait analysis. The purpose of this study was to assess the influence of ankle arthritis etiology and deformity, measured radiographically, on gait performance in a dedicated gait lab utilizing a multisegment foot model. With three calcaneal and four metatarsal markers in addition to standard lower extremity markers, the modified Helen Haynes model allows for the evaluation of range of motion (ROM) within the 'ankle-hindfoot segment.' The primary hypothesis was that three- dimensional ankle-hindfoot segment ROM would be more restricted in patients with post-traumatic ankle arthritis than other etiologies. The secondary hypothesis was that temporospatial and kinetic measures would not vary by etiology. METHODS: A longitudinal cohort of 183 patients with end-stage ankle arthritis were prospectively enrolled from 2008-2018. Mean age was 61, BMI 29, and 56% were male. Four etiologic groups were defined: Post-fracture (100), arthritis caused by planovalgus foot deformity (23), chronic instability associated with cavovarus (32), and miscellaneous (28), comprised of inflammatory (7), idiopathic (6), instability without deformity (5), septic (2), and avascular necrosis (3) as causes. The four-segment Milwaukee foot model was used in a dedicated gait lab with a 12-camera motion capture system. Gait data was collected over a minimum 20 gait cycles across a 10-meter walkway. Kinetic data was simultaneously collected with two force plates embedded in the walkway operating at 1 MHz. AP and lateral tibiotalar angles, lateral talus-first metatarsal angles, calcaneal pitch, and tibiotalar ratio were measured. Multivariate regression analyzed the effect of etiology and radiographic measures on gait function, controlling for age, gender, and BMI. RESULTS: The primary hypothesis was confirmed. Sagittal plane ankle-hindfoot segment ROM was lower in post-traumatic and higher in valgus patients compared to other etiological groups (P<0.0001) (Figure 1). Sagittal plane ankle-hindfoot segment ROM restriction relative to the contralateral limb was also more severe in the post-traumatic group than others (P=0.0005). Valgus AP tibiotalar angles were associated with greater sagittal plane ankle-hindfoot ROM (P=0.0016). The secondary hypothesis was disproven. Post-traumatic patients ambulate with greater maximum ankle moment than other groups (P=0.0043). Valgus patients ambulate with a comparatively longer step length (P<0.0001). Significant reductions in affected limb walking speed (P<0.0001), step length (P<0.0001), and maximum ankle moment (P=0.036), as well as increases in double limb (P=0.0007) and total support percentage (P<0.0001) were found among the miscellaneous etiology group. CONCLUSION: Of the four groups, patients with post-traumatic ankle arthritis ambulated with the greatest ankle and hindfoot stiffness, but also the greatest ankle moment. Patients with valgus ankle arthritis had the greatest ROM through the ankle and hindfoot and the longest step length. In addition to diminished ROM, patients in the miscellaneous group had the lowest cadence, symmetry, and torque of gait. The etiology of severe ankle arthritis can predict the pattern of gait dysfunction, which, in turn, may inform choices of surgical reconstruction.