Table 1.
Participant demographic information.
Fig 1.
Participants were first fitted with a new prosthetic foot that was classified as either High Activity (HA) or Low Activity (LA) according to the Medicare Functional Classification System.
Participants then wore the new prosthetic foot for a period of three weeks to acclimate to the new device. After the three-week period, participants underwent gait analysis during both treadmill and overground walking. Treadmill walking consisted of three minutes at subject-selected speed. Continuous ankle angle data from a single treadmill walking bout (obtained using infrared cameras in conjunction with retroreflective markers on the shoes, prosthesis, and legs) was used to calculate stride-to-stride fluctuations via Lyapunov exponent for both limbs. Overground trials were completed at self-selected walking speed, and force data were recorded using a Kistler force platform to calculate ankle-foot power.
Fig 2.
Prosthesis push-off work was 140% greater in the High Activity (HA) prostheses compared to the Low Activity (LA) prostheses (p < 0.001).
Sound limb push-off work showed no significant difference between the two prosthesis conditions (p = 0.576). The late-stance positive power was integrated (a, shaded regions) to determine push-off work from the sound (b) and prosthetic (c) ankle-foot structures during overground walking. This process was completed for each of four limb conditions: sound limb, contralateral to low activity prosthesis (blue, dashed line); sound limb, contralateral to high activity prosthesis (blue, solid line); prosthetic limb, low activity prosthesis (red, dashed line): prosthetic limb, high activity prosthesis (red, solid line).
Fig 3.
Stride-to-stride fluctuations at the ankle were not significantly different between high and low activity prostheses for the sound limb (a, p = 0.325) or the prosthetic limb (b, p = 0.652).
Lyapunov exponent is a method of characterizing the fluctuations of a repetitive action. Larger values of Lyapunov exponent indicate increased divergence in the gait pattern while smaller values indicate lower divergence.
Fig 4.
Changes in prosthesis push-off work did not correlate with stride-to-stride fluctuation changes within individuals at the sound ankle (a, R2 = 0.134, n = 22, p = 0.0903) or the prosthetic ankle (b, R2 = 0.0417, n = 22, p = 0.362). As stride-to-stride fluctuations result from the interaction of various systems including neural, mechanical, and (in this case) synthetic, we expected that changes to the prosthesis mechanics would result in changes in the stride-to-stride fluctuations. The lack of relation indicates that the mechanical contribution of the limb to stride-to-stride fluctuations may not be a significant contributor. (Δ = High Activity prosthesis value–Low Activity prosthesis value).