Rating scales for depression in the elderly: external and internal validity

Area of Science:

• Geriatric Psychiatry
• Psychometric Validation
• Neuropsychology

Background:

• Accurate depression assessment in the elderly is crucial, complicated by cognitive impairment.
• Existing depression scales, including various Hamilton Rating Scale for Depression (HAM-D) versions, require validation in diverse elderly populations.
• Differentiating depression from dementia requires reliable and valid measurement tools.

Purpose of the Study:

• To evaluate the internal and external validity of multiple depression scales in elderly Danes.
• Specifically assessed were the HAM-D(6), HAM-D(17), Bech-Rafaelsen Melancholia Scale, Geriatric Depression Scale (15- and 30-item), and Cornell Scale for Depression in Dementia.
• The study focused on populations with and without dementia.

Main Methods:

• Two independent, blinded clinicians assessed participants from psychogeriatric clinics and a control group.
• Concurrent and convergent validity were analyzed using correlation coefficients.
• Internal validity was assessed via item response theory (IRT) analysis, including Mokken and Rasch analyses; scalability required a coefficient of homogeneity ≥ 0.40.

Main Results:

• 145 elderly subjects (109 with depression, 36 controls) were included; 102 female, 43 male.
• The HAM-D(6) was the only scale demonstrating total scalability in both cognitively intact and impaired elderly individuals.
• The HAM-D(6) exhibited the strongest external validity when standardized against the Clinical Global Impressions-Severity of Illness (CGI-S) scale.

Conclusions:

• The HAM-D(6) shows robust psychometric properties for assessing depression in the elderly.
• Its scalability and external validity make it a reliable tool, even in the presence of cognitive impairment.
• The HAM-D(6) warrants separate consideration for depression measurement in elderly populations, irrespective of longer HAM-D versions.