Journal of NeuroEngineering and Rehabilitation

step length improved step length symmetry ( p < 0.001). The difference in step length between the intact and prosthetic step, reduced from 0.05 ± 0.04 m while walking with the SACH foot, to only 0.01 ± 0.04 m while walking with the ESAR foot. Center of mass velocity decreased significantly more during the double support phase when walking with the SACH foot compared to walking with the ESAR foot (interaction effect foot x time p < 0.000) (Fig. 4 ) . At the instant of toe-off, center of mass velocity was higher when walking with the ESAR foot compared to the SACH foot (1.22 ± 0.05 vs 1.19 ± 0.04 m ∙ s − 1 , p = 0.03, Fig. 3 ) , while there was no difference at heel strike. Concurrently, backward margin of stability was lower at heel strike when walking with ESAR compared to SACH (0.137 ± 0.022 vs 0.153 ± 0.023 m, p = 0.001), but showed a larger increase during double support (inter- action effect foot x time p = 0.001) (Fig. 4 ) . Hence, at toe off the backward margin of stability did not differ significantly between foot conditions (0.271 ± 0.022 vs 0.272 ± 0.022 m, p = 0.36) (Fig. 3 ) . Discussion In this study, we investigated a potential functional benefit of energy storing and return (ESAR) prosthetic feet. Specifically, we investigated whether ESAR feet could preserve gait stability while restoring gait sym- metry. Such benefit may possibly contribute to the general preference of people with a lower limb amputa- tion for these types of prosthetic feet, considering the previously observed absence of improvements in gait economy. Both gait stability and symmetry are fre- quently mentioned objectives for people with a lower limb amputation who need to regain walking ability with a prosthesis. Based on a simple biomechanical model of human gait (Fig. 1 ) and the known increase in push-off power, we hypothesized that compared to the conventional SACH feet ESAR feet would increase center of mass velocity at toe-off, increase extrapolated center of mass forward projection and as such potentially enhance the backward margin of stability. This would allow the prosthetic user to increase intact step length and restore step length sym- metry, without reducing the backward stability margin. Conditional for this hypothesis is an increase in push-off power of ESAR feet relative to SACH feet. In- deed for the type of ESAR foot used in this study the Fig. 3 Difference in push-off work of the prosthetic foot (Work), center of mass velocity (v com ), intact step length (SLintact), step length symmetry (SLsymm) and backward margin of stability (MOS BW ) between walking with the SACH foot (green) and ESAR foot (red). * denotes significant difference between foot conditions Fig. 4 Change in center of mass velocity (v com ) and backward margin of stability (MoS BW ) during the double support phase, from heel strike (HS) of the intact leg until toe-off (TO) of the prosthetic leg during walking with the SACH (green) and ESAR (red) foot Houdijk et al. Journal of NeuroEngineering and Rehabilitation 2018, 15 (Suppl 1):76 Page 45 of 72