Validity and reliability of a shorter version of the Geriatric Depression Scale in institutionalized older Portuguese adults

Abstract

Objectives

Depressive symptoms are common in older adults in institutional contexts; however, there is a lack of validated measures for these settings. Identifying depressive symptoms can help clinicians to manage them and to prevent or delay their complications. This study aimed to validate the Geriatric Depression Scale (GDS) in an institutionalized sample of older adults.

Method

493 institutionalized older people (73% women) aged 60 or over were evaluated through the GDS, the Mini International Neuropsychiatric Interview (MINI) (depression vs. no depression = 11% vs. 89%), the Geriatric Anxiety Inventory (GAI), the Positive Affect (PA) and Negative Affect (NA) Schedule, and the Satisfaction with Life Scale (SWLS). Test-retest reliability was assessed with 57 older adults.

Results

An 8-item version presented a Cronbach's alpha value of .87 with a single factor explaining its variance. The correlations (p < .01) attested the concurrent validity (GAI: r = .76; PA: r = -.22; AN: r = .62; SWLS: r = -.32). Test-retest reliability (6.51 months) was adequate (r = .52). ROC analysis (AUC = .82; sensitivity = 80%; specificity = 77%) and Youden index revealed a cutoff of 5/6 for the diagnosis of depression.

Conclusion

Results support the validity and the screening capacity of a short version of GDS in institutional contexts. Short screening instruments for depressive symptoms may facilitate their identification, allowing for timely clinical interventions in institutional settings.

Keywords:

• Depression
• institutionalization
• assessment
• Geriatric Depression Scale

Acknowledgements

The authors wish to thank older people for their participation.

C.D-F. and H.E-S. contributed equally to this work.

Portions of this work have been submitted in partial fulfillment of the requirements for a master degree, at Miguel Torga Institute of Higher Education, Coimbra, Portugal.

Disclosure statement

The authors report no conflicts of interest.