Lower limb prosthesis prescription is based on anecdotal evidence and manufacturer claims, with little research-based evidence of their effect on the musculoskeletal system. Amputees are at a high risk of musculoskeletal health complications, which may be avoided through adequate prosthesis design and prescription. This thesis aims to understand the bilateral above/through-knee (BTF) and unilateral above/through-knee amputee (UTF) function and loading with various prostheses in order to support informed decision-making in prosthesis prescription.
The thesis presents a method to create amputee musculoskeletal models based on anatomical datasets derived from magnetic resonance scans, where linear scaling to a model with the most similar pelvis width, body mass index and stump length to pelvis width ratio resulted in predictions of hip joint contact forces with minimal errors (average 0.4 BW). This method then enabled musculoskeletal modelling to be used in the remainder of the thesis, where new datasets of amputee gait were obtained from a number of different prostheses. Analysis of the biomechanical outcomes found that BTF with full-length articulated prostheses presented higher hip contact forces by 3.5 BW than non-articulated stubby prostheses and that reduced flexor muscle volume and residual limb length might limit the use of full-length articulated prostheses. When investigating UTF function with microprocessor knees (MPK), the analysis identified higher lateral intact knee impulse by 27.7 BW.s than able-bodied persons, and this may explain the incidence of intact knee osteoarthritis in the UTF population. As the vast majority of amputees are in low-to middle-income countries, it was important to also assess the biomechanical effect of more affordable prostheses. Therefore, this thesis analysed UTF function with the International Committee of the Red Cross (ICRC) polycentric knee, where amputees presented with ICRC lower walking speeds (by 0.2 m/s) and higher loading of the intact knee (by 14.9 BW.s/m) and intact ankle (by 24.1 BW.s/m) with ICRC than MPK. The spring extension assist mechanism in polycentric knees is a key feature that influences amputee function, and this thesis showed that a novel extension assist, which links the top and bottom knee parts, improves walking speeds by 0.1 m/s and symmetry between the intact and prosthetic knee flexion angles by 5.5% compared to the conventional extension assist used by ICRC.
Finally, regression analysis was used to generalize learning of amputee function and loading, and identified the prosthesis type, age, stump length and body mass index to be significant predictors of cadence and speed. Prosthesis type was the only significant predictor of BTF hip and UTF lateral intact knee loading. Appropriate prosthetic design and prescription may reduce the risk of musculoskeletal complications, whilst maintaining appropriate functional levels. The studies in this thesis form the basis for future improvements of prosthesis design.