No abstract

Patients suffering from some psychiatric and neurological disorders demonstrate abnormally high levels of saccadic distractibility when carrying out the antisaccade task. This has been particularly thoroughly demonstrated in patients with schizophrenia. A large body of evidence has been accumulated from studies of patients which suggests that such eye movement abnormalities may arise from frontal lobe dysfunction. The psychopharmacology of saccadic distractibility is less well understood, but is relevant both to interpreting patient studies and to establishing the neurological basis of their findings. Twenty healthy subjects received lorazepam 0.5 mg, 1 mg and 2 mg, sertraline 50 mg and placebo in a balanced, repeated measures study design. Antisaccade, no-saccade, visually guided saccade and smooth pursuit tasks were carried out and the effects of practice and drugs measured. Lorazepam increased direction errors in the antisaccade and no-saccade tasks in a dose-dependent manner. Sertraline had no effect on these measures. Correlation showed a statistically significant, but rather weak, association between direction errors and smooth pursuit measures. Practice was shown to have a powerful effect on antisaccade direction errors. This study supports our previous work by confirming that lorazepam reliably worsens saccadic distractibility, in contrast to other psychotropic drugs such as sertraline and chlorpromazine. Our results also suggest that other studies in this field, particularly those using parallel groups design, should take account of practice effects.

Latent inhibition (LI) is a measure of reduced learning about a stimulus to which there has been prior exposure without any consequence. It therefore requires a comparison between a pre-exposed (PE) and a non-pre-exposed (NPE) condition. Since, in animals, LI is disrupted by amphetamines and enhanced by antipsychotics, LI disruption has been proposed as a measure of the characteristic attentional deficit in schizophrenia: the inability to ignore irrelevant stimuli. The findings in humans are, however, inconsistent. In particular, a recent investigation suggested that since haloperidol disrupted LI in healthy volunteers, and LI was normal in non-medicated patients with schizophrenia, the previous findings in schizophrenic patients were entirely due to the negative effects of their medication on LI (Williams et al., 1998). We conducted two studies of antipsychotic drug effects on auditory LI using a within-subject, parallel group design in healthy volunteers. In the first of these, single doses of haloperidol (1 mg. i.v.) were compared with paroxetine (20 mg p.o.) and placebo, and in the second, chlorpromazine (100 mg p.o.) was compared with lorazepam (2 mg. p.o.) and placebo. Eye movements, neuropsychological test performance (spatial working memory (SWM), Tower of London and intra/extra dimensional shift, from the CANTAB test battery) and visual analogue rating scales, were also included as other measures of attention and frontal lobe function. Haloperidol was associated with a non-significant reduction in LI scores, and dysphoria/akathisia (Barnes Akathisia Rating Scale) in three-quarters of the subjects. The LI finding may be explained by increased distractibility which was indicated by an increase in antisaccade directional errors in this group. In contrast, LI was significantly increased by chlorpromazine but not by an equally sedative dose of lorazepam (both drugs causing marked decreases in peak saccadic velocity). Paroxetine had no effect on LI, eye movements or CANTAB neuropsychological test performance. Haloperidol was associated with impaired SWM, which correlated with the degree of dysphoria/akathisia, but no other drug effects on CANTAB measures were detected. We conclude that the effect of antipsychotics on LI is both modality and pharmacologically dependent and that further research using a wider range of antipsychotic compounds is necessary to clarify the cognitive effects of these drugs, and to determine whether there are important differences between them.

No abstract