Abstract

Hypothesizing that agraphia in Alzheimer's disease (AD) reflects disturbances in multiple cognitive domains, we evaluated writing samples from 33 patients meeting strict criteria for probable AD. We found agraphia to be common on a standard narrative writing task. When compared with 41 education- and age-matched normal control subjects, AD patients had significantly lower writing scores, wrote significantly fewer words, mentioned significantly fewer categories of information, and were significantly more likely to make writing errors. On stepwise regression procedures, neuropsychological measures of visuoperceptual impairment and disease severity were the strongest predictors of agraphia, but other analyses indicated that measures of language, praxis, and attention could also contribute significantly to agraphia. On two writing tasks, we failed to confirm the previous contention that agraphia is a marker for familial AD. However, there was a highly significant interaction between family history, oral naming, and writing: patients with nonfamilial AD, but not those with a family history of dementia, showed a strong correlation between naming and writing performance. We conclude that agraphia in AD can be variously determined and that agraphia is not a reliable marker for familial disease.

View details for PubMedID 1565231