Smooth pursuit performance in patients with affective disorders or schizophrenia and normal controls: analysis with specific oculomotor measures, RMS error and qualitative ratings

Abstract
Synopsis: Smooth pursuit performance in schizophrenia and affective disorders has generally been found to be abnormal using a variety of measures. The purpose of this study was to assess patients with these disorders and normal controls in order to compare the different measures across diagnoses. Smooth pursuit was assessed using quantitative specific measures (gain, catch-up saccade rate and amplitude, square-wave jerk rate, number of anticipatory saccades and total time scored), as well as two global measures: root mean-square error (RMS) and qualitative rating. As previously reported, patients with schizophrenia had low gain, increased catch-up saccade rate and spent less time engaged in scoreable smooth pursuit than normal controls. Patients with affective disorders were not statistically different from controls on any of these measures, and had significantly higher gain than patients with schizophrenia. RMS error and qualitative rating measures were highly correlated (r = 0·87). In linear regression analyses, the quantitative specific measures were highly significant predictors of both RMS error and qualitative ratings (P < 0·0001). Linear regression analyses and a modelling study indicated that one quantitative specific measure, the percent of time engaged in scoreable smooth pursuit (total time scored), was most related to global ratings. However, RMS error and qualitative ratings were less sensitive than total time scored to the difference between controls and patients with schizophrenia. These data indicate two smooth pursuit performance deficits in schizophrenia: patients spend less time engaged in scoreable smooth pursuit and have low gain (accompanied by increased compensatory saccades) when the smooth pursuit is engaged.