Mark J. Smith

Memory impairment and schizophrenia: the role of processing speed

A link between slowing of processing speed and cognitive disorders, including memory, has repeatedly been found in research on aging, and suggested in other cognitively impaired populations. We tested the hypothesis that a link between memory impairment and slowing of processing speed would also be observed in schizophrenia. Forty-four schizophrenic patients and 40 normal controls were administered a memory task involving free recall and recognition. Processing speed was assessed by the Digit Symbol Substitution Test. Working memory span was assessed as well. The measure of processing speed was consistently correlated with the various memory measures in patients, including efficiency of encoding. These correlations remained significant, or tended to be significant, when working memory span was partialled out. Memory deficits observed in schizophrenia may thus be partly accounted for by a slowing of processing speed. It is suggested that research on cognitive deficits in this and other mental diseases focus more on processing speed.

Abnormal luteal phase excitability of the motor cortex in women with premenstrual syndrome

BACKGROUND Premenstrual syndrome (PMS) involves an aberrant behavioral response to normal hormone secretion. Pathogenetic theories posit abnormal modulation of gamma-aminobutyric acid (GABA) transmission in the brain by neuroactive metabolites of progesterone (neurosteroids). In earlier transcranial magnetic stimulation (TMS) studies of the motor cortex, we showed that inhibition increases in the luteal phase, consistent with neurosteroid action at the GABA(A) receptor. Here, we studied women with PMS to see if their response to endogenous progesterone differed from that of control subjects. METHODS We studied nine women with PMS and 14 control subjects during the midfollicular and luteal phases with paired TMS. Subthreshold conditioning TMS was followed by test stimulation that produced a motor evoked potential (MEP) in a hand muscle. We gave pairs at each of seven intervals (2-10 msec) and unconditioned stimuli, measuring the amplitude ratio of the average MEP from the pairs at each interval to that from the unconditioned stimuli (ratio < 1 = inhibition). RESULTS Both groups showed the same follicular phase response to paired TMS. Control subjects showed more inhibition in the luteal phase. Women with PMS showed relative facilitation. CONCLUSIONS This is the first physiological evidence for an abnormal brain response to progesterone in PMS.

Memory and schizophrenia: differential link of processing speed and selective attention with two levels of encoding

The purpose of this study was to investigate how underlying cognitive deficits such as a defect in processing speed or in selective attention contributed to different types of memory impairment observed in schizophrenia (superficial vs deep encoding). 49 schizophrenic patients and 40 normal controls were administered a verbal memory task. Superficial encoding was assessed by the ability to recall items in their serial order. Deep encoding was assessed by the ability to organise words into semantic categories. Two measures of processing speed (Digit Symbol Substitution Test and Stroop colour time) and one measure of selective attention (Stroop test) were used. Regression analyses were carried out. In the patient group, processing speed contributed to both superficial and deep encoding, and to a global verbal memory score. Selective attention only contributed to the superficial encoding processes. Thus, slowing of processing speed in schizophrenia seems to be more crucial for memory performance, since it affects memory in a pervasive way.

Discrimination accuracy and decision biases in different types of reality monitoring in schizophrenia.

A reality monitoring task was administered to 31 schizophrenic patients and 31 normal controls. Twenty-four items were produced, either orally by the experimenter, orally by the subjects, or seen as pictures. Subjects were later read a list of 48 items and were asked to indicate if each item was new, self-generated, experimenter-generated, or presented as a picture. Results showed that schizophrenic patients were impaired in discriminating old items from new, with a higher bias than controls toward reporting new items as if they were old (false alarms). In addition, patients were impaired in discriminating self-generated items from externally generated items, with a higher bias than controls toward attributing self-generated items to an external source. Lastly, they were significantly impaired in discriminating the modality (auditory versus visual) in which the event was presented. The bias toward remembering orally produced items as pictures was correlated with positive symptomatology and was significantly higher than controls in patients with high levels of positive symptoms. This suggests that mental imagery may play a role in positive symptomatology. These results demonstrate the relevance of studying decision biases along with discrimination performance for the understanding of the mechanisms of reality monitoring impairment in schizophrenia.

Word recognition, discrimination accuracy, and decision bias in schizophrenia: association with positive symptomatology and depressive symptomatology.

The purpose of this experiment was to replicate and extend to a memory task Bentall and Slades (1985) finding that hallucinations in schizophrenic patients were linked to a liberal decision bias. A word recognition task was administered to 40 schizophrenic patients and 40 normal controls that yielded two indices of performance: an index of discrimination accuracy (Pr) and one of decision bias (Br). Patients obtained a lower Pr than controls, whereas Br was similar in both groups. In patients, Br was selectively correlated with positive symptomatology: the more the positive symptoms, the more liberal the bias. In particular, there was a specific correlation between decision bias and hallucinations. Conversely, Pr was inversely correlated with severity of depression, but not with either positive or negative symptoms. Thus, positive symptomatology may be linked more to difficulties in distinguishing between representations of internal versus external events than to deficits in encoding external events.

Opposite links of positive and negative symptomatology with memory errors in schizophrenia

We wished to confirm and extend a previous correlational study of our group, suggesting that positive symptoms in schizophrenia were linked to an increase in certain types of memory errors, and negative symptoms to a decrease in other types of errors. A post-hoc analysis was conducted in 33 schizophrenic patients and 40 normal control subjects on memory errors collected in a free recall task and two types of recognition tasks. The memory errors were intrusions and list errors in free recall, and decision bias towards false alarms in recognition, all assumed to reflect a source-monitoring failure. In a first analysis, the patient sample was split along the median for positive symptoms as rated by the Scale for the Assessment of Positive Symptoms (SAPS). In a second analysis, it was split along the median for negative symptoms as rated by the Scale for the Assessment of Negative Symptoms (SANS). Patients with high ratings of positive symptoms made more memory errors (intrusions, list errors, false alarms) than those with low ratings, supporting the hypothesis of a link between positive symptomatology and source-monitoring failure. On the other hand, patients with high ratings of negative symptoms made fewer of these errors than the other patients. Fewer errors were specifically associated with more affective flattening, alogia and anhedonia, whereas avolition was entirely unrelated to them.