Rating depression severity in the elderly physically ill patient: reliability and factor structure of the Hamilton and the Montgomery-Asberg Depression Rating Scales.

OBJECTIVES: To assess the appropriateness of the Hamilton Depression Rating Scale and the Montgomery-Asberg Depression Rating Scale in depressed elderly physically ill patients.
DESIGN: Depression scale scores from depressed medical inpatients were assessed for internal consistency using Cronbach's alpha, and subjected to exploratory principal components factor analyses.
SUBJECTS: 100 medical inpatients, aged 65 years and over (median age 80.5 years, range 66-99), 74 female, with Geriatric Mental State Schedule-AGECAT case level diagnoses of depression. MATERIALS: The 17-item Hamilton Depression Rating Scale (HDS) and the Montgomery-Asberg Depression Rating Scale (MADRS).
RESULTS: Coefficient alpha for the HDS was 0.46; for the MADRS 0.61. Successive deletion of HDS items to maximize alpha resulted in a six-item scale (alpha = 0.60); after deletion of five MADRS items, alpha was 0.77. Factor analysis of the HDS yielded a four-factor solution accounting for 57% of the variance, the majority due to anxiety and insomnia items; the MADRS yielded a two-factor solution explaining 60% of the variance.
CONCLUSIONS: Coefficient alpha for both scales is well below the minimum necessary for the total score to be used to represent a single construct. The HDS appears to be an unreliable measure of depression severity in elderly people with physical illness, as the major variance in the scores is due to anxiety and insomnia. The MADRS performs better, and with modification may be an appropriate measurement of depression severity in this population.