Seven years ago the authors examined 64 consecutive female out-patients with macrocytosis (erythrocyte mean cell volume ⩾ 100 fl). The cause remained undetermined in 23 (35.9%). Their patient histories were evaluated in 1992, and the new highly specific alcohol marker, serum carbohydrate-deficient transferrin, from the frozen sera taken during the initial study was examined. It was elevated in 6 of the 23 women. Furthermore, four of these six had visited the health centre after 1985 and they all had clinical records indicative of alcohol problems. Because early intervention has proved to be effective, information about the risks of alcohol abuse (intervention) should be given to women with suspect alcohol-induced symptoms or signs.

Earlier studies have failed to differentiate the unipolar manic subtype from bipolar patients with both manic and depressive episodes. This retrospective cohort study of 50 elderly manic in-patients identified six patients (12%) who met strict criteria for a course of unipolar mania. Significant differences emerged in age at onset, with a mean of 41.2 years for unipolar mania compared with 64.7 years for the others. Consequently, clinical course was significantly longer, 27.7 v. 7.4 years. Elderly patients pursuing a unipolar manic course are among the very few elderly ‘bipolars’ whose illness begins early in life. Recent neuroradiological investigations and data from geriatric studies suggest that the concept of unipolar mania is worthy of further investigation.

The rate of treatment-emergent switch into mania has been calculated from all available clinical trial data on the selective serotonin re-uptake inhibitors (SSRIs) fluoxetine, fluvoxamine, paroxetine, and sertraline, relative to comparative groups treated with tricyclic antidepressants (TCAs) or placebo. In predominantly unipolar depressives, the rate of manic switch is less than 1% and differences between drugs and placebo are statistically but not clinically significant. In bipolar depressives, manic switch occurs substantially more often with TCAs (11.2%) than with SSRIs (3.7%) or placebo (4.2%).

A patient with a 30-year history of somnambulism and night terrors is described. The use of a home ambulatory sleep electroencephalogram (EEG) recording in clarifying the diagnosis and in monitoring the results of treatment is illustrated and successful treatment using a selective serotonin re-uptake inhibitor is reported.

A 59-year-old man with chronic treatment-resistant depression developed severe cognitive impairment. The severity of depression and the mode of presentation led to difficulty in diagnosis between a dementing disorder and a depressive illness. The diagnostic conundrum was resolved by the use of sleep deprivation (SD), which demonstrated clear subjective and objective improvements in his condition. These changes were quantified by a range of psychometric tests, which showed that initial deficits in cognitive performance improved temporarily after SD, and these improvements were maintained after effective treatment of the depression with electroconvulsive therapy.

Three cases of Capgras' syndrome were identified in one psychiatric hospital within a 1-year period. These are presented and their psychopathology discussed. The traditional assumption that this is a rare syndrome is challenged.

A case report is presented of koro associated with a depressive illness in a 31-year-old male Briton. The specificity or otherwise of koro as culture bound and a distinct nosological entity is discussed.