Premium
P3‐100: Predicting MCI or dementia at follow‐up: Using subjective memory and non‐memory complaints from both the participant and informant
Author(s) -
Slavin Melissa,
Brodaty Henry,
Kochan Nicole,
Crawford John,
Reppermund Simone,
Trollor Julian,
Draper Brian,
Sachdev Perminder
Publication year - 2011
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.2011.05.1540
Subject(s) - dementia , psychology , memory impairment , complaint , cognition , neuropsychology , clinical psychology , memory problems , memory clinic , psychiatry , cognitive impairment , medicine , disease , political science , law
Background: Subjective complaints, required for most definitions of mild cognitive impairment (MCI), have not been operationalised. We previously showed that if a number of questions are asked, nearly all individuals will endorse some cognitive complaint/s, thereby making the criterion redundant. The ideal criterion for MCI is one that best predicts decline. While some studies have found subjective complaints predict progression to dementia, to date there has been no evaluation of which type (participant/informant, memory/non-memory) or severity of complaints are most sensitive to progression to dementia, or indeed progression from normal to MCI. We aimed to assess whether cognitive complaints at baseline could predict classification at two year followup. Methods: Participants were drawn from the Sydney Memory and Ageing Study, a longitudinal study of community-dwelling older adults initially aged 70-90 years. A subsample of 637 participants who completed a comprehensive assessment including 15 memory and 9 non-memory complaint questions and neuropsychological testing at baseline were included. An informant answered 15 memory and 4 non-memory complaint questions, and rated their functional independence. Repeat testing was conducted two years later, with each participant classified into normal, MCI, or dementia groups. Results: Logistic regression showed that sum of participant memory complaints predicted MCI two years later, and sum of informant memory complaints predicted dementia two years later. Both participant and informant complaints (memory and non-memory combined) predicted MCI, but only informant complaints predicted dementia. The question “have you noticed difficulties with your memory?” was not sensitive to either MCI or dementia at follow-up. The areas under the curve of the Receiver Operating Characteristic curves were highest for informant memory complaints predicting dementia, and participant memory complaints predicting MCI after two years. Conclusions: For cognitive complaints to be useful to clinicians, they should indicate future progression to MCI or dementia. Participant memory complaints seem most useful for predicting MCI, and informant memory complaints for predicting dementia after two years. A subset of participant memory questions graded in severity also predicted progression to dementia. Non-memory complaints did not predict MCI or dementia. Thresholds for memory complaints are recommended to maximize sensitivity and specificity for predicting dementia after two years.