General and disease-specific risk factors for depression after ischemic stroke: a two-step Cox regression analysis.

Background: Post-stroke depression (PSD) frequently complicates stroke and is associated with an impaired functional outcome, more severe cognitive deficits, a reduced quality of life, and a higher mortality. The aim of this study was to assess whether general risk factors for major depressive disorder (MDD) in the community are also risk factors for PSD, and to identify additional, stroke-related risk factors. Methods: In a hospital setting, 190 consecutively admitted patients were assessed for MDD 1 month after stroke, and at follow-up after 3, 6, 9 and 12 months. A Cox model was created with four established risk factors for MDD in the community (female sex, prior personal history of depression, positive family history of depression, and somatic comorbidity other than stroke). Five potential disease-related risk factors (disability, cognitive deterioration, inter- and intrahemispheric lesion location, and generalized vascular damage on computed tomography (CT) scan) were then added individually to this model, to see whether these would improve the significance of the overall model. Results: The Cox model of four general risk factors for depression in the community was shown to be a valid model to predict depression in stroke patients. Of the disease-specific factors, only incorporation of "disability" in this model improved its significance. Conclusion: Established risk factors for depression in the community are also predictors of depression in the first year after stroke. Disability is a non-specific disease-related variable that is associated with PSD. The contribution of stroke-specific factors may be less than is generally assumed.