Deficits in smooth pursuit (SP) are one of the most extensively studied behavioural features of schizophrenia. It has been suggested that they might serve as a biological marker for the underlying genetic and neurobiological abnormalities which lead to the disease. And yet basic questions remain about the nature of deficits. Many studies have concentrated on the maintenance phase of SP. We have now examined SP initiation in a group of schizophrenic patients and have investigated the effect of introducing a temporal gap between the extinction of the fixation target and the illumination of the pursuit target.

With local ethical approval a group of 12 schizophrenic subjects (11 male, 1 female; age range 23 to 40 years, mean 32.1) were recruited and given full optometric and orthoptic assessments. Oculomotor testing was carried out in a quiet room, with lighting dimmed. We used a chin rest and cheek pads to stabilise subjects' heads. They viewed a stimulus monitor with their left eye from 57 cm; the right eye was occluded. A fixation target (contrast 25%), presented for a variable period (0.5 - 1.5 s) in the middle of the display, was replaced by the SP target (contrast 25%; speed: 14 deg/s) which moved from 5 deg to the left/right of fixation, through the centre of the display. In sets of four interleaved tasks (two leftward, two rightward) presented in random order, one task had no gap, the other three had gaps of 100, 200 or 400 ms. Eye position was recorded using infrared oculography. Pre-saccadic SP latency was measured off-line from velocity traces using a regression technique.

All but one of the schizophrenic subjects responded to stimuli in a manner qualitatively similar to control subjects. However, quantitative analysis revealed that in non-gap trials, for both leftward and rightward SP, the latencies in the schizophrenic group were significantly longer than the latencies of a control group (leftward: 234 ±51 ms vs. 198 ±41; t = 7.438, df 273, p < 0.0001; rightward: 266 ±58 vs. 210 ±39; t = 10.03, df 187, p < 0.0001). Whereas 100 ms gaps caused significant reductions in all control subjects, latency was only reduced significantly in three schizophrenic subjects for rightward and one for leftward SP. Longer gaps produced significant reduction in seven subjects for leftward and two subjects for rightward SP. Across all four gap durations for both directions of SP, the mean latencies for the schizophrenic group were significantly longer than the control group latencies.

Contrary to some previous reports, we find that SP latency is significantly longer in schizophrenic compared to control subjects. This complements recent reports that there may be a deficit in motion processing associated with schizophrenia. We also find that the gap effect on SP latency is altered, although a definitive conclusion requires larger data sets from schizophrenic subjects.