Visual attention and eye movements in Parkinson's disease

Abstract

Saccades (fast eye movements) reveal the overt and covert deployment of visual attention, reflect more or less directly the influence of basal ganglia outputs, and can be simply and accurately measured. Therefore, saccades are often used to study the effects of Parkinson’s disease on the visual orienting system. This investigation attempts to resolve some apparent inconsistencies in the results of those studies.

Dopamine depletion in PD is thought to cause excessive inhibitory output from the basal ganglia to the visual orienting system and characteristic deficits in PD include prolonged saccade latencies and reduced saccade amplitude. In addition, there are reports of hyper-reflexivity and impaired suppression of unwanted eye movements. Studies of visual attention in PD have reported evidence of overactive top-down response inhibition but also of abnormal facilitation of bottom-up attention. The generally accepted model assumes that abnormal bottom-up facilitation reflects a failure of top-down (cortical) control. This leaves the question: how or why does excessive inhibition co-occur with abnormal facilitation in the visual orienting system in PD? The features of an existing experimental paradigm were exploited to measure top-down and bottom-up effects in the visual orienting system in a group of 20 subjects with Parkinson’s disease and 20 control subjects. This specific paradigm (Deubel, 2008) combines saccades with a perceptual discrimination task. Our results suggest that Parkinson’s disease may affect the visual orienting system globally, instead of affecting top-down response selection only. There were three ways in which the Parkinson’s disease group differed from the control group. Firstly, the Parkinson’s disease group made saccades that were abnormally hypometric. Secondly, saccade initiation in the Parkinson’s disease group was hypersensitive to top-down facilitation in response to the demands of the discrimination task. Thirdly, saccade initiation in the Parkinson’s disease group was hypersensitive to bottom-up facilitation by visual inputs. These results were as expected, and confirmed previous findings. Interestingly, however, the observed effects did not necessarily co-occur in patients. In the Parkinson’s disease group, patients with shorter mean saccade latencies had greater top-down facilitation and patients with longer mean saccade latencies had greater bottom-up facilitation. Abnormally reduced saccade gain was not associated with abnormal saccade latencies. The Parkinson’s disease group was also impaired in a set of neuropsychological tests compared to the control group. Cognitive problems were strongly associated with smaller mean saccade gain values, but also to some extent with longer mean saccade latencies and hyper-sensitivity to bottom-up visual inputs. The discrimination task proved a valuable addition to standard saccade tasks. The finding that in the Parkinson’s disease group, top-down and bottom-up effects were both exaggerated but not necessarily in the same subjects, may reflect the often-observed heterogeneity among Parkinson’s disease patients. Further research can exploit these effects to investigate the variability of symptoms and the time-course of progression in PD, potentially leading to better management and targeted treatments.