Premium
P2‐265: Daily‐life functional instruments to identify mild cognitive impairment
Author(s) -
Memoria Claudia,
Aprahamian Ivan,
Forlenza Orestes
Publication year - 2012
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2012.05.973
Subject(s) - dementia , receiver operating characteristic , neuropsychology , cognitive impairment , psychology , cognition , audiology , disease , gerontology , medicine , psychiatry
complaints (a positive response to either the EAS health self-assessment, the CERAD or a CDR rating of 0.5). Incident AD occurred in 49 cases (average follow-up 4.3 years). Clinic evaluations were completed annually and AD was diagnosed using standard criteria. We applied time-dependent receiver operating characteristics (ROC) analysis to compute AUC for predicting incident AD over 2-4 years and differences in AUCs between tests. Youden’s index, (the sum of sensitivity and specificity minus 1), was used to select the cut-point that optimizes correct prediction of incident AD. Results: For prediction of incident AD over 2-4 years, the FCSRT-FR has better operating characteristics and the AUCs for FCSRT-FR are consistently greater than those for LM. The cut-point that optimized prediction did not vary significantly with duration of follow-up for either test. Optimal cutpoints were similar regardless of APOE4 status. Positive predictive values were higher among APOE4 positive individuals, most likely due to higher disease prevalence. Time dependent ROC curves are shown in the attached figure. Conclusions: When the planned follow-up interval is 2-4 years, FCSRT-FR is preferable to the LM for identifying which individuals with memory complaints will develop incident AD. APOE4 status improves the positive predictive value of the test, but the cut scores that optimize correct prediction of incident AD dementia do not differ substantially by APOE4 status.