Aphasia after stroke: type, severity and prognosis. The Copenhagen aphasia study.

AIM To determine the types, severity and evolution of aphasia in unselected, acute stroke patients and evaluate potential predictors for language outcome 1 year after stroke. METHODS 270 acute stroke patients with aphasia (203 with first-ever strokes) were included consecutively and prospectively from three hospitals in Copenhagen, Denmark, and assessed with the Western Aphasia Battery. The assessment was repeated 1 year after stroke. RESULTS The frequencies of the different types of aphasia in acute first-ever stroke were: global 32%, Broca's 12%, isolation 2%, transcortical motor 2%, Wernicke's 16%, transcortical sensory 7%, conduction 5% and anomic 25%. These figures are not substantially different from what has been found in previous studies of more or less selected populations. The type of aphasia always changed to a less severe form during the first year. Nonfluent aphasia could evolve into fluent aphasia (e.g., global to Wernicke's and Broca's to anomic), whereas a fluent aphasia never evolved into a nonfluent aphasia. One year after stroke, the following frequencies were found: global 7%, Broca's 13%, isolation 0%, transcortical motor 1%, Wernicke's 5%, transcortical sensory 0%, conduction 6% and anomic 29%. The distribution of aphasia types in acute and chronic aphasia is, thus, quite different. The outcome for language function was predicted by initial severity of the aphasia and by the initial stroke severity (assessed by the Scandinavian Stroke Scale), but not by age, sex or type of aphasia. Thus, a scoring of general stroke severity helps to improve the accuracy of the prognosis for the language function. One year after stroke, fluent aphasics were older than nonfluent aphasics, whereas such a difference was not found in the acute phase.