Accuracy of self-reported depression in persons with dementia.

OBJECTIVES: To investigate the reliability and validity of self-reported depression in demented elderly patients. DESIGN: This is the first study to combine the use of multiple depression criteria with dementia and depression control groups. SETTING: Houston Veterans Affairs nursing home and geropsychiatric inpatient unit and outpatient clinic, Baylor College of Medicine's Alzheimer's Disease Research Center. PARTICIPANTS: The total sample included 121 volunteer participants, their informants, and their clinicians, forming four diagnostic groups (27 controls, 37 dementia only, 28 depression only, 29 dementia and depression). MEASUREMENTS: Scales assessing depression, deficit awareness, physical illness, functional disability, caregiver burden, and severity of cognitive impairment were used. RESULTS: Analyses of variance using informant- and clinician-rated depression categories found main effects on patient-reported depression scores for deficit awareness and depression diagnosis but not for dementia diagnosis. Chi-square analyses revealed that demented and depressed patients underreported depression at similarly high rates compared with clinician and informant reports. Regression analyses revealed that deficit awareness accounted for significant variance in self-reported depression accuracy, whereas clinician- or informant-rated depression, severity of cognitive impairment, physical disability, functional status, and caregiver burden were not significant. CONCLUSION: The results of this investigation suggest that the presence of dementia per se does not predict inaccurate depression self-reports. Deficit awareness, which covaries with dementia, appears to account for the majority of the variance in self-report accuracy. These findings were stable across informant- and clinician-rated depression criteria and multiple analyses.