Background: Agreement on the factor structure of the Hamilton Depression Rating Scale (HDRS) has not been consistent among studies, and some investigators argued that the scale's factor structure is not reliable. This study aimed at shedding more light on this debated issue.
Methods: We studied 186 adults with unipolar depression (Major Depressive Disorder, n=80; Dysthymic Disorder, n=71; Depressive Disorder Not Otherwise Specified, n=25; Adjustment Disorder, n=10). They had no comorbid DSM-IV axis I or axis II disorders, and had received no treatment with antidepressant drugs in the previous 2 months. The factor structure of the scale was studied using the principal factor method, followed by oblique rotation. Factor scores were computed for each subject using the regression method.
Results: Using the scree-test criterion for factor extraction, we obtained a four-factor solution, explaining 43.8% of total variance. The four factors extracted were identified as (1) somatic anxiety/somatization factor; (2) a psychic anxiety dimension; (3) a pure depressive dimension; and (4) anorexia factor. Patients with Major Depressive Disorder scored significantly higher than patients with other diagnoses on the pure depressive dimension.
Limitations: These results need to be replicated in different cultures, using analogous factoring techniques.
Conclusions: Though not exhibiting factorial invariance in the stricter sense of the term, the 17-item HDRS did exhibit a relatively reliable factor structure. Our analysis provides further evidence that the scale is multidimensional. However, as long as the multidimensional character of the scale is taken into account the scale should be able to play a useful role in clinical research.