Optimising the diagnostic performance of the Geriatric Depression Scale

Area of Science:

• Geriatric Psychiatry
• Psychometric Evaluation
• Clinical Psychology

Background:

• The Geriatric Depression Scale (GDS) is widely used for depression screening in older adults.
• Existing shortened versions may compromise diagnostic accuracy.
• Optimizing scale length while maintaining performance is crucial for efficient clinical use.

Purpose of the Study:

• To develop and validate a shortened version of the Geriatric Depression Scale (GDS).
• To maximize diagnostic performance in detecting depression among older adults.
• To create a more efficient screening instrument than existing versions.

Main Methods:

• A 30-item GDS was administered to 233 adults aged 60+.
• Stepwise logistic regression identified 10 key items from the GDS in a Day Hospital sample.
• Diagnostic performance of the 10-item scale (GDS-R) was validated using ROC analysis in a community sample.

Main Results:

• The 10-item GDS-R correctly classified 97.4% of Major Depression cases in the initial sample.
• Validation in a community sample showed high sensitivity (100%), specificity (97.2%), and AUC (0.994).
• The GDS-R demonstrated superior diagnostic performance compared to GDS-15, GDS-10, and GDS-5, and excellent unidimensionality (CFI=0.99).

Conclusions:

• The GDS-R is a valid and reliable 10-item scale for detecting depression in older adults.
• It retains the diagnostic accuracy of the full GDS-30 while offering improved efficiency.
• The GDS-R presents enhanced sensitivity and predictive values over other abbreviated GDS versions.