Reliability and validity of the Geriatric Depression Scale in detection of poststroke minor depression

Abstract

Objective: The aim of this study was to assess the validity and reliability of the 30-item Geriatric Depression Scale (GDS) as a screening tool for minor depression in poststroke patients. Method: Literate patients older than 18 years of age, diagnosed to have stroke, were eligible for the study. Standardized Mini Mental Status Examination (S-MMSE) and GDS were applied to all patients. The GDS was readministered 7 days later for retest reliability. Results: A total of 85 participants-49 nondepressed and 36 with minor depression-were eligible for the study. Cronbach's alpha coefficient was .89 in internal consistency analysis. The GDS scores were significantly higher (p < .001) in the depressed participants reflecting a high discriminant validity. The highest sum of sensitivity and specificity values of 1.44 (sensitivity = .69, specificity = .75) and 1.45 (sensitivity = .66, specificity = .79) were obtained for cutoff scores of 10/11 and 11/12, respectively. The area under receiver operating characteristics curve was .82. The test-retest reliability analysis revealed a high Pearson correlation coefficient (r = .75). Conclusion: Our findings suggest that the 30-item GDS has high discriminant validity, internal consistency, and test-retest reliability and reasonably useful cutoff scores; thus it can be used as a screening tool for minor depression in the poststroke population.