Influence of Prosthetic Ankle-Foot Mechanics on Intact-Limb Biomechanical Demand during Ambulatory Functional Tasks

Abstract

Individuals with single (unilateral), below-knee (transtibial) amputation (TTA) often exhibit altered gait biomechanics that increase mechanical demand on the non-amputated (intact) limb, contributing to an elevated risk of secondary osteoarthritis, particularly in the medial compartment of the knee. Because prosthetic ankle-foot devices play a central role in restoring mobility, their mechanics may influence how individuals with TTA adapt to various walking activities. This dissertation describes multiple studies investigating how prosthetic ankle-foot mechanical metrics influence lower-limb biomechanical demand during functional tasks, with particular emphasis on intact-limb medial knee loading during load carriage and other ambulatory activities. Across the studies, there are three overarching objectives: 1) influence of prosthetic ankle-foot type on measures of biomechanical demand, including gait kinetics, mechanical energy (i.e., joint work), and motor control during load carriage; 2) influence of prosthetic ankle-foot type on medial knee loading during load carriage; 3) influence of prosthetic ankle-foot stiffness on medial knee loading across ambulatory activities.The first study examined the effects of side load carriage on gait kinetics during steady-state walking. Twelve individuals with TTA carried a moderate load while walking with a clinically prescribed, passive prosthetic ankle-foot. Prosthetic-side load carriage reduced intact-limb propulsive force, whereas intact-side load carriage was associated with smaller intact-limb hip flexor and knee flexor moments. Hip and knee abductor moments increased when the load was carried on the opposite limb, and hip extensor power increased when the load was carried on the same limb. These mixed findings suggest that the preferred side for carrying a load depends on the desired biomechanical effect, particularly on the intact limb, to inform targeted gait rehabilitation strategies. Using the same protocol, we investigated how various load carriage positions and prosthetic ankle-foot types affect intact-limb medial knee loading associated with osteoarthritis risk. Five load-carriage conditions were tested: unloaded, front (anterior), back (posterior), intact side, and prosthetic side. In addition to the clinically prescribed, passive prosthetic ankle-foot, four other commercially available types were compared, including the same type but one category stiffer, the same type with a heel-stiffening wedge, a dual-keel type, and a powered type. Intact-side load carriage resulted in the smallest first peak knee adduction moment and impulse (surrogate measures of medial knee loading), whereas prosthetic-side load carriage produced the highest values, increasing these measures by approximately 30%. Prosthetic ankle-foot type did not significantly affect these outcomes, and only a weak trend was observed between prosthetic push-off work and first peak knee adduction moment. These findings indicate that load carriage strategy plays a greater role than prosthetic ankle-foot selection in modulating intact-limb medial knee loading during load carriage. Given the more significant influence of load carriage than prosthetic ankle-foot type on intact-limb medial knee loading during load carriage, we further investigated how prosthetic ankle-foot selection influences lower-limb joint total energy (mechanical joint work). Across unloaded, anterior, posterior, intact-side, and prosthetic-side load conditions, most differences were observed in the prosthetic limb. Ankle joint work differed significantly between prosthetic ankle-foot types within each load condition, with the powered type producing the greatest net positive ankle work in late stance and the clinically prescribed and heel-wedge types producing the greatest net negative ankle work in early stance of the gait cycle. By contrast, knee and hip joint work did not differ substantially across types, suggesting that prosthetic ankle-foot primarily modulates prosthetic ankle mechanics without markedly altering proximal joint work under moderate load carriage. To determine whether these biomechanical changes translate to altered motor control, we investigated the effect of changes in prosthetic ankle-foot selection on muscle activation patterns and coordinated muscle groupings (motor modules) during load carriage. Prosthetic ankle-foot type influenced activation of several intact-limb muscles, with powered and stiffer types often associated with greater plantarflexor and knee extensor activity, compared with the clinically prescribed type. Differences in muscle co-contraction were limited and occurred primarily during unloaded walking. Despite these muscle-level changes, motor module complexity, composition, and activation timing remained largely conserved across prosthetic ankle-foot types and load carriage conditions. These results indicate that individuals with TTA accommodate altered prosthetic ankle-foot mechanics and added load primarily through modulation of muscle activation rather than through reorganization of motor control strategies. Subsequently, we focused on stiffness, an important mechanical property of prosthetic ankle-feet. We evaluated the influence of commercial prosthetic ankle-foot stiffness (variation in ±2 categories from the clinically prescribed stiffness category) on medial knee contact force, another surrogate measure of medial knee loading, across a range of ambulatory activities. Prosthetic ankle-foot stiffness affected peak medial knee contact force in an activity-dependent manner. Intact limb results found that stiffer prosthetic ankle-feet were generally associated with reduced peak medial knee contact force during level, faster, and loaded walking, whereas different patterns emerged during downhill walking and turning. Residual limb findings were inconsistent. Although several differences were statistically significant, they remained below the minimal detectable change threshold, indicating that prosthetic ankle-foot stiffness alone may contribute only moderately to changes in medial knee loading across ambulatory activities. We extended this work by examining how stiffness-related prosthetic ankle-foot biomechanical metrics, including forward push (push-off), dynamic mean ankle moment arm, and roll-over radius, are associated with intact-limb medial knee contact force. Surprisingly, greater prosthetic push-off was not consistently associated with reduced intact-limb peak medial knee contact force across ambulatory activities. Instead, roll-over radius demonstrated the most consistent inverse associations with intact-limb peak medial knee contact force, particularly during more demanding walking activities such as upslope walking, faster walking, and load carriage. These findings suggest that effective forefoot lever arm behavior may be more important than push-off alone for reducing intact-limb medial knee loading. Collectively, this dissertation demonstrates that lower-limb biomechanical demand, particularly on the intact limb of individuals with unilateral TTA, is influenced by both activity demands and prosthetic ankle-foot mechanics. Load carriage strategy is consistently associated with changes in intact-limb medial knee loading, while prosthetic ankle-foot stiffness and other mechanical metrics like push-off exhibit more complex, activity-dependent effects. Overall, this dissertation advances understanding of how prosthetic ankle-foot mechanics interact with users during various ambulatory activities and highlights the importance of activity-adaptive prosthetic ankle-foot systems, as well as specific local biomechanical metrics, for reducing increases in medial knee loading, potentially preventing elevated OA risks in the intact limb and ultimately improving mobility in individuals with unilateral TTA.