Natural history of cognitive impairment after stroke and factors associated with its recovery

OBJECTIVES: To describe the natural history of cognitive impairment following stroke over three years; determine factors associated with recovery of post-stroke cognitive impairment and examine the effect of this recovery on stroke outcomes. DESIGN: Observational study. SETTING: Population-based register of first-ever strokes. MEASURES: Mini-Mental State Examination (MMSE) for cognition (cognitive impairment: MMSE<24), Barthel and Frenchay activity indices for disability. METHODS: To describe its natural history, cognition was assessed at three months, 1, 2 and 3 years after stroke in 163 subjects registered in 1995. Recovery of cognitive impairment was examined using a larger cohort registered between 1995 and 1998 (n = 476) and 193 subjects cognitively impaired at three months were reassessed a year post stroke; 34 who had recovered cognitively (MMSE 24-30) were compared with 102 with persistent cognitive impairment in terms of demography, risk factors, initial impairments and stroke subtype. RESULTS: At three months, 1, 2 and 3 years post stroke, the prevalence rates of cognitive impairment were 39%, 35%, 30% and 32% respectively. Multivariable analyses showed that recovery was associated with smoking (OR 3.7; 95% CI 1.2-11.8), compromised by visuospatial neglect (OR 0.27; 95% CI 0.08-0.89), and had a near-significant association with right hemispheric lesion (OR 2.87; 95% CI 0.94-8.78). Cognitive recovery was associated with less institutionalization (p = 0.032) and being less disabled, on Barthel (p = 0.001) and Frenchay Activity Indices (p = 0.028). CONCLUSIONS: Cognitive impairment remains highly prevalent up to three years after first stroke. Recovery from cognitive impairment is associated with smoking and possibly right hemisphere strokes, but compromised by visuospatial neglect. These associations require further clarification.