Objective: The aim of the study was to assess the external and internal validity of the 6- and 17-item versions of the Hamilton Rating Scale for Depression (HAM-D(6) and HAM-D(17)), the Bech-Rafaelsen Melancholia Scale, the 15- and 30-item versions of the Geriatric Depression Scale, and the Cornell Scale for Depression in Dementia in a population of depressed demented and nondemented Danish elderly.
Method: Two clinicians performed independent, blinded assessments of the study population, which was drawn from psychogeriatric outpatient clinics, and a control group of elderly subjects. Concurrent and convergent validity were assessed using correlation coefficient analyses, and to evaluate the internal validity, item response analysis using the Mokken coefficient and Rasch analysis was performed. A coefficient of homogeneity of 0.40 or higher indicated scalability. Data collection took place between October 2001 and April 2004.
Results: 145 subjects were included; 102 were female (mean age = 78.6 +/-6.8 years), and 43 were male (mean age = 72.4 +/-5.6 years). In the study group (N = 109), 73 subjects had depression only, and 36 had both depression and dementia; in the control group (N = 36), 11 subjects had dementia. The item-response analysis made a clear distinction between the scales. The HAM-D(6) was the only scale that fulfilled the criterion of total scalability in both the cognitively intact and the impaired populations. In terms of standardization according to the Clinical Global Impressions-Severity of Illness scale (CGI-S), the HAM-D(6) had the most convincing external validity overall. In terms of general correlation to the CGI-S, only small differences were shown between the scales.
Conclusion: The HAM-D(6) should be separately considered even when longer HAM-D versions are used for the measurement of depression in elderly persons.