Background. Prisoners in Nazi concentration camps lived through extreme situations that included starvation. We test our hypothesis that there is a greater lifetime presence of binge eating among survivors from concentration camps than in a control group.

Methods. The subjects were 51 political prisoners who survived Nazi concentration camps and 47 ex-partisans of similar age and sex. A clinical interview investigated the lifetime occurrence of binge eating. The Eating Attitudes Test was also administered.

Results. The mean reported loss of weight among survivors was 27·3 kg. Thirty-three per cent of them and 4% of the ex-partisans reported going on eating binges at some time in their lives (P < 0·0007). There was no significant difference in the Eating Attitudes Test scores of survivors and ex-partisans, but, among survivors, the Bulimia subscale significantly discriminated subjects who reported current binge eating.

Conclusions. Our study confirms that subjects who have survived a period of extreme food deprivation are more likely to develop binge eating behaviour.

Background. No study has simultaneously explored key components of Beck's model of cognitive vulnerability to depression in people with bipolar disorders.

Methods. We compared 41 euthymic bipolar patients with 20 healthy control subjects. All subjects were assessed on the Hamilton Rating Scale for Depression, the Autobiographical Memory Test and the Mean Ends Problem-Solving procedure and also completed the Beck Depression Inventory, the Dysfunctional Attitude Scale, the Sociotropy Autonomy Scale and the Rosenberg Self-Esteem Questionnaire.

Results. In comparison to control subjects, patients with bipolar disorder demonstrated significantly higher levels of dysfunctional attitudes (particularly perfectionism and need for approval) and sociotropy, significantly greater over-general recall on an autobiographical memory test and significantly less ability to generate solutions to social problem-solving tasks. These between group differences remained significant when age, intelligence, latency to respond to autobiographical memory test cue words, and subjective mood ratings were included as co-variates in the statistical analysis. Within the patient group, cognitive dysfunction was significantly correlated with level of morbidity (as measured by number of previous illness episodes).

Conclusions. This study suggests that cognitive vulnerability in patients with bipolar disorder is similar to that described in unipolar disorders. It is not clear whether this dysfunction is a cause or an effect of repeated episodes of bipolar disorder. However, the findings may have implications for clinical treatment as well as suggesting a number of important new avenues of research into psychological models of affective disorder.