Accuracy of the 15-item geriatric depression scale (GDS-15) in a community sample of the oldest old

Background Cognitive impairment is common in the oldest old. This might influence the sensitivity and specificity of the 15‐item Geriatric Depression Scale (GDS‐15). Few studies, however, have included subjects older than 85 years to evaluate the GDS‐15 as a screening instrument for depression. Objective To assess the sensitivity and specificity of the GDS‐15 in a community sample of the oldest old. Methods Seventy‐nine subjects aged 85 and over were enrolled in the study. The GDS‐15 and the Mini‐Mental State Examination (MMSE) were administered by a trained interviewer. Within two days the Geriatric Mental State (GMS)/AGECAT, was administered to obtain a clinical diagnosis of depression. Results Eight subjects (10%) were diagnosed with clinical depression. At a cut‐off point of 3/4 the sensitivity and specificity of the GDS‐15 were 88% and 76% respectively. In the group with MMSE scores of 28 and higher sensitivity was unaffected at all cut‐off points while specificity increased. In the group with MMSE scores below 28 sensitivity was also unaffected at all cut‐off points while specificity decreased. Conclusion The GDS‐15 is a suitable instrument to diagnose depression in the general population of the oldest old. The optimal cut‐off point depends on its intended use. In subjects with cognitive impairment the accuracy should be investigated further. Copyright © 2002 John Wiley & Sons, Ltd.