Mortality was high and symptomatic recovery protracted. Impulsivity, symptom severity, and chronicity were the important factors for predicting the 12-year outcome.

Detecting the effects of latent depressive schemas constitutes an important step toward validating A. T. Beck's (1967) tenet of the depressive cognitive style as a causal factor in the onset and relapse of major depression. The authors examined whether a sample of formerly clinically depressed individuals continued to exhibit a negatively biased information-processing style in the absence of concurrent depressed mood. As predicted, the scores of formerly depressed individuals on 2 questionnaires tapping dysfunctional thinking did not differ from those of never depressed individuals, but responses of formerly depressed individuals were negatively biased on 2 of 3 information-processing measures administered following a self-focus manipulation. These results provide evidence for the persistence of a depressive schema in individuals who have recovered from major depression.

Course, outcome, and mortality were similar for BED and BN-P. Both disorders had psychiatric comorbidity as the main predictor of outcome, and there was a diagnostic shift between BED and BN-P over time, pointing to their nosological proximity. Data are relevant for the formulation of DSM-V and ICD-11 diagnostic criteria.

T im e-b a sed p ro sp ective m em o r y, t h e ab ility t o ca r r y o u t a fu tu r e in t en tio n a t a sp eci® ed t im e, wa s fo u n d t o b e im p a ired in a co m m u n ity sa m p le o f clinica lly d ep ressed a d u lts, rela t ive to a n o n d ep ressed sam p le. N o n d ep ressed p ar t icipa n t s m o n ito red t h e t im e m o r e freq u en t ly an d , in th e ® n a l b lo ck o f t h e ta sk , a ccelerat ed t im e-m o n ito r ing a s th e t a rget t im e fo r th e p ro sp ect ive m em o ry r esp o n se a p p ro a ch ed . T h ese r esu lts a re co n sisten t wit h p r evio u s ® n d in gs o f d ep ression -rela ted im p a irm en t s in ret ro sp ect ive m em o r y t a sk s th a t req u ire co n t ro lled, self-initiat ed p rocessing.

As part of a prospective, long-term treatment study, 30 in-patients with bulimia nervosa (BN) were divided into groups with high and low expressed emotion (EE) family backgrounds according to the Camberwell Family Interview, and followed for a period of six years. The high EE group initially showed significantly more psychopathology than the low EE group. No group x time interactions were found, but the high EE group showed a worse outcome on the "conflict" and "organisation" subscales of the Family Environment Scale. They also showed significantly more eating disorder pathology according to the Eating Disorder Inventory (EDI) and the Structured Interview for anorexia nervosa (AN) and BN before treatment at discharge, after two years and, to some degree, even after six years. Depth of depression (Beck Depression Inventory) was significantly higher in the high EE group at admission (moderate depression), discharge and after the 6-year follow-up (still slight depression). The Parental Bonding Instrument (PBI) showed no differences between the high EE and low EE groups, but the individuals with "affectionless control" according to the PBI had more negative scores on three of the subscales of the Family Environment Scale (FES). In brief, the high EE individuals with BN were initially sicker and did not fully catch up over time in comparison with the symptomatic recovery of the low EE individuals. These data suggest that EE status upon admission to in-patient treatment is a relevant predictor of the severity and course of BN and depressive symptoms.