RESUMO
BACKGROUND: The prevalence of dementia is increased after stroke. Medial temporal lobe atrophy (MTLA) is associated with Alzheimer disease, and with prestroke dementia in patients who have had a stroke. OBJECTIVE: To determine the influence of MTLA on the long-term risk of dementia after stroke, after excluding the patients who had prestroke dementia. METHODS: The study was conducted in 144 consecutive patients who had a stroke, who were aged 40 years or older (66 women and 78 men; median age, 72 years), and who had an Informant Questionnaire on Cognitive Decline in the Elderly score lower than 104. On admission to the hospital all patients underwent a noncontrast computed tomographic scan including temporal lobe-positioned slices. A cut-off of 11.5 mm was used to differentiate patients with MTLA from those without MTLA. Patients were followed up with clinical and cognitive assessments over a 3-year period. RESULTS: Three years after stroke, 34 patients (23.6%) had developed new-onset dementia. The cumulative proportion of survivors without dementia was 57.6% in patients with MTLA and 80.8% in patients without MTLA (P =.02). The unadjusted relative risk of poststroke dementia associated with MTLA was 2.3 (95% confidence interval, 1.1-4.7). However, using the Cox proportional hazards model, MTLA did not seem to be an independent predictor of poststroke dementia. Independent predictors of poststroke dementia were increasing age, diabetes mellitus, severity of the clinical deficit at admission, and severity of leukoaraiosis on computed tomography. CONCLUSIONS: Patients who had a stroke and MTLA more frequently develop dementia than patients without MTLA, but our study does not suggest that MTLA independently contributes to dementia. A longer follow-up may be necessary to reevaluate the influence of MTLA.