Comparative Performance and Neuropathologic Validation of the AD8 Dementia Screening Instrument

Gabriela M. Morris, Timothy R. Holden, Hua Weng, Chengjie Xiong, Dean W. Coble, Nigel J. Cairns, John C. Morris

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Background/Objective:The AD8 informant-based screening instrument has been validated with molecular biomarkers of Alzheimer disease (AD) but not with the gold standard of neuropathologic AD. The objective of this study was to validate the AD8 with neuropathologic AD and compare its predictive performance with that of the Mini-Mental State Examination and both participant-derived and informant-derived subjective memory complaint (SMC) regarding the participant.Methods:This longitudinal cohort study at the Knight Alzheimer Disease Research Center at Washington University included 230 participants, ages 50 to 91 years, who later had a neuropathologic examination. Four dementia screening instruments from their baseline assessment were evaluated: the AD8, Mini-Mental State Examination, participant SMC, and informant SMC. The primary outcome was a neuropathologic diagnosis of AD.Results:The average participant age at baseline was 80.4 years, 48% were female. All 4 dementia screening tests were predictive of neuropathologic AD. There was no significant difference in the predictive performance of the AD8 compared with the other instruments, but the AD8 had superior sensitivity and combined positive and negative predictive values.Conclusion:The AD8 is a brief and sensitive screening instrument that may facilitate earlier and more accurate AD diagnosis in a variety of care settings.

Original languageEnglish
Pages (from-to)112-117
Number of pages6
JournalAlzheimer disease and associated disorders
Issue number2
StatePublished - Apr 1 2020


  • AD8
  • Alzheimer disease
  • dementia
  • screening


Dive into the research topics of 'Comparative Performance and Neuropathologic Validation of the AD8 Dementia Screening Instrument'. Together they form a unique fingerprint.

Cite this