ERP Indices and Neuropsychological Performance as Predictors of Functional Outcome in Dementia

Ronald A. Cohen, Brian R. O'Donnell, Mary Ellen Meadows, Majaz Moonis, William F. Stone, David A. Drachman

Research output: Contribution to journalArticle

15 Scopus citations


We compared the relative value of neuropsychological and event-related potentials (ERPs) obtained during both passive and active auditory oddball paradigm measures for determining functional outcome in dementia 4 years following initial assessment. Functional outcome was assessed by structured interview of family members of 29 patients with dementia, and patients' functional status was rated in seven areas: mortality, incontinence, institutionalization, ADL dependence, verbal responsiveness, recognition of family members, and capacity for social interaction. A total functional outcome score (ADLTOTAL) was obtained by summing across these individual outcome measures. Many of the neuropsychological measures correlated strongly with overall functional outcome, whereas P3 amplitude and latency on the active ERP condition were the only ERP indices to predict functional outcome. When ERP and neuropsychological measures were considered simultaneously using stepwise multiple regression analyses, the neuropsychological measures were better predictors of most functional outcomes, although P3 latency was the best predictor of mortality. However, neuropsychological performance and ERPs appear to be sensitive to different functional outcomes. Therefore, evaluation of both ERPs and neuropsychological performance may ultimately have prognostic utility in the assessment of patients with dementia.

Original languageEnglish (US)
Pages (from-to)217-225
Number of pages9
JournalJournal of Geriatric Psychiatry and Neurology
Issue number4
StatePublished - Oct 1995

ASJC Scopus subject areas

  • Clinical Neurology
  • Geriatrics and Gerontology
  • Psychiatry and Mental health

Fingerprint Dive into the research topics of 'ERP Indices and Neuropsychological Performance as Predictors of Functional Outcome in Dementia'. Together they form a unique fingerprint.

  • Cite this