GERIATRIC DEPRESSION RATING SCALE (GDS)

The Geriatric Depression Scale was developed by Yesavage, Brink, Rose, Lum, Huang, Adey, and Leirer (1982). It was designed specifically for the aged, as a screening instrument for depression. The scale

• Originally contained 100 items, but was condensed to 30 Is a self-administered test, but can be used in observer-administered formats also
• The 30 items are yes/no questions.

Later, Sheik & Yesavage (1986) created a short form of the GDS (GDS-SF), which contained 15 items. The original can be referred to as the GDS- Long Form (GDS-LF). Literature is divided in terms of the short form being a suitable substitute (Aikman& Oehlert, 2001; Holroyd & Clayton, 2000).

Cut-off scores for different severities of depression are as follows:

For the long form: Normal 0 – 10, Mild 11 – 20, Moderate to Severe 21 – 30.

For the short form: Normal 0 – 4, Mild 5 – 9, Moderate to Severe 10 – 15.

[quads id=5]

Validation and Psychometric properties

The scale has a high degree of internal consistency, with a Chronbach’s alpha coefficient of .94, and split-half reliability score of .94. Test retest reliabilities of .85 (p < .001) for one week apart (Yesavage et. al., 1982) and .85 (p < .001) for one month apart (Parmelee, Lawton & Katz, 1989) show that within the time limits, scores reflect stable individual differences.

The GDS is a valid tool for discriminating symptom severity, and presence vs absence based on DSM-IV criteria, but not among different diagnostic groups. It should not be used as a single diagnostic measure (Watson, Zimmerman, Cohen, & Dominik, 2009).

The GDS has high correlations with the Zung Self-Rating Depression Scale (SDS) and the Hamilton Rating Scale for Depression (HRS-D) (.84 and .83 respectively); further evidence of validity.

Sensitivity (true positives) and specificity (true negatives) with a cutoff of 11 were 84% and 95% respectively, and they were 80% and 100% respectively at a cutoff of 14; providing evidence for scores of 11+ to be considered a possible indicator of depression.

Validity and reliability are unaffected by pertinent individual difference factors such as age, education, gender, race, and culture (Marc, Raue & Bruce, 2008; Rait et. al., 1999; Harralson et. al., 2002).

[quads id=5]

Critical Analysis

Overall, the GDS-LF is a reliable and valid measure of depression in aged individuals. It is…

• Easy to administer (self-administered or observer)
• A simple scale to complete (yes/no responses), especially for older adults
• Useful in a variety of settings; nursing homes and the community, with medical inpatients, medical outpatients, and day-treatment clients
• Shown to maintain it’s reliability and validity when administered by phone (Burke, Roccaforte, Wengel, Conley & Potter (1995)
• Adequate in screening mildly demented subjects (McGivney, Mulvihill & Taylor, 1994)

Its few weaknesses include the possibility of over-diagnosing depression (Lesher & Berryhill, 1994), the inclusion of items/terms that could be seen as western value judgments (Sansoni et. al., 2007) and that it is not a useful or valid tool for screening cognitively impaired patients (Holroyd & Clayton, 2000).