Adaptive control for backward quadrupedal walking. III. Stumbling corrective reactions and cutaneous reflex sensitivity.

1. Four cats were trained to walk backward (BWD) and forward (FWD) on a motorized treadmill. Mechanical (taps) or electrical (pulses) stimuli were applied to the dorsal or ventral aspect of the hind paw during swing or stance. Hindlimb kinematic data, obtained by digitizing 16-mm high-speed film, were synchronized with computer-analyzed electromyograms (EMG) recorded from anterior biceps femoris (ABF), vastus lateralis (VL), lateral gastrocnemius (LG), tibialis anterior (TA), and semitendinosus (ST). Responses to taps and pulses, as well as the modulation in cutaneous reflex sensitivity to pulses, were described for both walking directions and stimulus locations. 2. After dorsal taps that obstructed FWD swing, the hindlimb initially drew back away from the obstacle with knee flexion and ST activation, ankle extension with TA suppression and LG activation, and hip extension with ABF facilitation. Next, the limb was raised over the obstacle with resumed TA activity and enhanced knee and ankle flexion, and then compensatory knee and ankle extension positioned the limb for the ensuing stance phase. 3. For ventral taps that obstructed BWD swing, the initial response also tended to draw the limb away from the obstacle with hip and ankle flexion and TA facilitation and reduced knee flexion with weak VL facilitation and suppression of ST activity. Next, ST activity resumed as knee and ankle flexion raised the limb over the obstacle, and then compensatory extension completed the swing phase for BWD walking. Thus the initial kinematic and EMG responses to obstacles were opposite for BWD versus FWD swing, and these responses were consistent with active avoidance of the obstacles. Responses during BWD walking were subtle, however, compared with those for FWD. 4. After nonobstructing taps (ventral FWD, dorsal BWD), ST and TA activation and knee and ankle flexion were coincident, demonstrating that the aforementioned differences in responses to obstructing obstacles were not simply location dependent. Regardless of the direction of walking or the location of stimulation, taps applied during stance had little immediate kinematic effect, but the subsequent swing phase was usually exaggerated, as if the response was programmed to avoid any lingering obstacle. 5. Electrical pulses did not elicit the full-blown responses typically evoked by taps. The sequencing in activation of ST and TA characteristic after laps was absent after pulses, and there were rarely dramatic kinematic responses to pulses like those easily elicited by taps. There were, in fact, few differences in responses to electrical stimulation for BWD versus FWD walking.(ABSTRACT TRUNCATED AT 400 WORDS)