Background. Depression after stroke is common, affecting 185,000 persons per year. The majority of stroke patients suffer significant physical disability and inability to carry out activities of daily living, adversely affecting quality of life. Poststroke functional impairment is considered to be the major cause of disability in adults. While there has been research on the relationship between poststroke depression (PSD) and functional outcome at a fixed point in time, no study has looked at the pattern of change in depression following stroke longitudinally. The purpose of this study was to investigate the relationship between PSD and functional outcome over time.

Method. Using a prospective non randomized longitudinal design, a convenience sample of 23 ischemic stroke patients admitted to acute rehabilitation of University of Minnesota Medical center, Fairview, Riverside campus were recruited. Functional status (Functional Independence Measure, FIM), depression (Geriatric Depression Scale, GDS; Centers for Epidemiological Studies – Depression; CESD), cognition and severity of stroke were assessed in each participant on admission, and function and depression were reassessed at 2 weeks, 1 month and 3 months.

Results/conclusions. Stroke severity (p=.008), increasing age (p=.01), living alone (p=.06) and increased length of stay in rehabilitation (p=.06) were associated with a lower functional status at baseline. Older age was associated with lower functional status at baseline (p=.01), and less improvement from baseline to 2 weeks (p=.07). Living alone (p=.07) and greater numbers of comorbidities (p=.09) were correlated with increased baseline CES-D scores. Longer length of stay in rehabilitation was associated with less decrease in CES-D scores between 2 weeks and 3 months (p=.06). As expected, as functional status improved, depression scores also improved in a similar pattern. A majority of participants had one inflection point at week 2, with the greatest improvement in both function and depression (CES-D) at this time point. An important finding of this study was that the GDS did not reliably assess depression in this population; several questions assess somatic symptoms of depression, which can also be present as a result of stroke. More reliable measures of depression for the stroke population need to be developed.