Fig. 2

a Hypotheses regarding walking patterns used by control (healthy young adults) and persons post-stroke to achieve step length symmetry. We hypothesized that healthy adults achieve step length symmetry using symmetric kinematics (as represented by small interlimb asymmetry (IA); left) and persons post-stroke achieve step length symmetry using asymmetric kinematics (large IA; right). b Limb orientations (blue = right, red = left, solid = nonparetic, dashed = paretic) during representative steps for each participant. c Persons post-stroke show marked IA during preferred walking and symmetric stepping (mean ± SEM). Control data shown for reference. d IA during symmetric stepping correlates strongly with IA during preferred walking in persons post-stroke (left). Net metabolic cost of transport correlates strongly with IA during preferred walking and symmetric stepping (right)