Background: This study examined psychosocial and clinical predictors of depression non-remittance among a sample of initially clinically depressed elders.
Methods: Incident and prevalent unipolar depression cases (n = 166) were enrolled into the MHCRC for the Study of Depression in Late Life and followed for 12 months while undergoing treatment using a standardized algorithm. The outcome was remission vs non-remission (<6 vs > 7 on the Montgomery-Asberg Depression Rating Scale (MADRS)) at one-year follow-up. Baseline predictor variables included psychosocial factors, such as four domains of social support, basic and instrumental activities of daily living (ADLs), and clinical factors, which included use of ECT, past history of depression, comorbidities, and antidepressant treatment.
Results: At one-year follow-up, 45% of the sample was in remission based upon MADRS scores. In bivariate analyses, non-remitted patients were more likely at baseline to use benzodiazepines, anxiolytic/sedatives, and/or MAO inhibitors than patients in remission, and have more depressive episodes. Among psychosocial factors, non-remitted patients had at baseline, more ADL and IADL problems and decreased subjective social support as compared to patients in remission. In logistic regression analyses more depression episodes, using anxiolytic/sedatives, more IADL problems and decreased subjective social support predicted poor depression outcome after one-year.
Conclusions: While clinical and diagnostic variables were related to improvement, baseline psychosocial factors were also important.
Copyright 2002 John Wiley & Sons, Ltd.