Journal of Aging and Health, Ahead of Print.
Objectives: To examine predictors of membership in discordant groups identified by subjective and objective measures of cognitive function. Methods: Participants in the Study of Older Korean Americans (N = 2046) were classified according to their subjective cognitive ratings (excellent/very good/good vs. fair/poor) and Mini-Mental State Examination scores (normal cognition vs. cognitive impairment), yielding two discordant groups: (1) positive ratings but cognitive impairment and (2) negative ratings but normal cognition. Logistic regression models examined how the discordant group membership was associated with personal resources. Results: Among those with positive cognitive ratings, the odds of belonging to the discordant group were associated with low personal resources (advanced age and lower levels of education, acculturation, and knowledge about Alzheimer’s disease). However, an opposite pattern was observed among those with negative ratings. Discussion: The pattern of discordance suggests ways to promote early detection of cognitive impairment and close the gap in cognitive health care.