Determinants of gender differences in depressive disorders are far from being established and their combination into integrated aetiological models continues to be lacking.

The higher rate of depression in women is not explained by gender differences in negative life events, social support or vulnerability.

Depressive disorder is a highly prevalent condition in Europe. The major finding is the wide difference in the prevalence of depressive disorders found across the study sites.

The internal consistency, validity and factor structure of the 12-item General Health Questionnaire (GHQ-12) were investigated in a homogeneous sample consisting of 18-year-old males in Italy. The GHQ-12 proved to be a reliable instrument, as indicated by a Cronbach's alpha of 0.81. When the screening characteristics of the GHQ-12 (scored by the Likert method) were evaluated against the psychiatrist's ratings, the best balance between sensitivity and specificity was found at the GHQ cut-off score of 8/9: at this threshold, sensitivity was 0.68 and was paired to a specificity of 0.59 and an overall misclassification rate of 0.40. Validity coefficients based on a single severity score were rather low compared with those reported in other settings. When a principal components analysis with varimax (and oblimin) rotation was performed, two factors were identified: factor A (general dysphoria) was defined by 7 items related to anxiety and depression; factor B (social dysfunction) included 6 items testing the ability to perform daily activities and to cope with everyday problems. The identified factors revealed distinct ability in the discrimination between subjects with and without emotional disturbance according to the psychiatrist's ratings and correlated differently with 3 Minnesota Multiphasic Personality Inventory subscales (depression, D; conversion hysteria, Hy; psychasthenia, Pt). Thus, the factor structure of the GHQ-12 might provide useful information along with that offered by a single severity score, and the detection of cases might be improved by examining an individual's profile of scores on different subscales derived from factor analysis.