Abstract
The walking pattern of patients with Parkinson's disease (PD) can be profoundly affected by visual stimuli. Lines on the ground can improve walking whereas doorways can cause the patient to come to an involuntary standstill (‘freezing of gait’). This apparent paradox can be resolved by supposing that PD patients respond very strongly to visual affordance information: thus lines become readily stepped over but doorways, whose affordance is to constrain walking, can slow progression. In a population of PD patients we measured the relation of visual information to gait disturbances in doorways. First, we asked participants to judge the door width they could just pass through without turning their shoulders. Since the PD group made judgements as well as an age-matched healthy control (HC) group, we can conclude that misperception of the doorway's affordance at an explicit level does not cause gait disturbances. Second, we used a motion capture system to measure walking through a doorway whose width varied between trials. The HC group scaled walking speed and stride length to door width, with a magnitude inversely proportional to door width. The PD group also scaled gait parameters to door width, but with a greatly amplified pattern so that walking speed decreased dramatically before the doorway. These changes could not be explained by motor difficulties, but rather suggest that PD patients respond more strongly than controls to the slowing affordance of the door. Using a novel method to measure freezing of gait, we found that freezes occurred most often in the doorway, with a frequency inversely proportional to door width; and that freezes were preceded by exaggerated gait changes of the type we found. Thus we propose that freezing of gait in doorways results from gait changes, which in turn result from exaggerated responses to the doorway's affordance.
Medical Research Council Grant G0502136.