This study was carried out to investigate the significance of Beck Depression Inventory (BDI) scores and the influence of somatic items in a non-psychiatric population presenting with somatic complaints. Internal consistency, item frequency, factor dimensions, and correlation with Eysenck's ‘N’ were analysed in a sample of 882 non-psychiatric pregnant women. The BDI showed a high level of internal consistency (Cronbach alpha coefficient was 0·85). Factor analysis yielded two major factors. Low BDI scores were determined mainly by items related to somatic complaints or decreased activity without unequivocal depressive content. In contrast, it was unambiguous depressive items that were endorsed in high scores, and these showed higher correlations with neuroticism. It is concluded that in non-psychiatric populations the BDI may measure conditions other than depression and that low or medium range scores should not be interpreted as indicators of mild depressive symptomatology.

This paper describes a patient whose amnesia for an offence (fraud) and two fugue episodes occurred against the background of an underlying organic amnesia. The fugue states conformed in their duration and precipitating factors to previous accounts in the literature. The organic, anterograde memory impairment was attributed to multiple small infarcts and a larger infarction in the left medial temporal lobe, which were evident on MRI and PET scans after the patient had developed transient neurological signs. At follow-up, the anterograde amnesia had persisted, and the patient also showed some difficulty in retrieving autobiographical memories of past incidents or events, although other aspects of his retrograde memory were intact (including his knowledge of facts about his past life and his general knowledge of public events). The difficulty in retrieving autobiographical incidents may have resulted from the presence of a moderate degree of frontal lobe dysfunction or, just possibly, from ischaemia in the left anterior temporal lobe. The persistence of the organic memory impairment and the importance of both the clinical history and neuropsychological testing in assessment are discussed, as well as the need to examine for possible organic factors in patients who may initially appear to manifest purely ‘psychogenic’ memory loss.