[Factorial structure of Hamilton Anxiety Rating Scale]

Among subjects suffering from anxiety disorders, mixed anxiety-depression and dysthymia (DSM III R), severity of disease has been assessed in six centres, using the Hamilton Anxiety Rating Scale. Results from two centres have been rejected because of major difference in comparison with the other centres, in which there were no statistically relevant difference. Resting on 90 subjects' rates, the analysis of factorial structure and factor reliability has been conducted with the principal components method and varimax rotation. Four factors have been extracted: somatization, psychic anxiety, somatic anxiety and psychic depression. The rotated factors accounted from 13% to 19% of the total variance (all four--60%). Unlike many previous investigators, we found that psychic factors of depression and anxiety are clearly separated, what must be due to heterogeneity of the group. The factorial structure in case of a homogenic group could be reduced to two factors: somatization and psychic anxiety or depression--depending on cases selection. Results explain the ability of Hamilton Anxiety Rating Scale to assess severity of anxiety as well as depressive state, and point out that the scale is not diagnostic.