The validity of new memory complaints in the elderly

To determine the validity of new subjective memory complaints (MCs) from individuals who previously, when without dementia, denied having MCs.

Prospective cohort.

Longitudinal, community-based study of aging and dementia.

One hundred thirty-three community-dwelling elderly individuals who were part of a registry for the study of conditions related to aging in North Manhattan, NY. Patients were selected if they were initially without dementia and had completed at least 2 successive annual clinical and neuropsychological evaluations and provided their own medical history.

Performance on memory tests--the Buschke Selective Reminding Test and a visual memory task--and global performance on a neuropsychological test battery and clinical evaluation, by which questionable dementia or dementia was diagnosed according to a well-defined paradigm.

Fifty-three subjects with MCs at the initial evaluation performed no worse on the memory test than the 80 subjects who denied MCs initially. There was a weak association between MCs and the diagnosis of questionable dementia at baseline (P = .04), but this was nonsignificant after adjusting for age and education. At 1-year follow-up, 21 of the 80 without baseline MCs now reported MCs. At the follow-up evaluation, these 21 subjects performed significantly worse on the memory tests, were 5 times more likely to have significant cognitive impairment, and had shown significantly greater decline over the preceding year on several of the cognitive measures than the 59 who continued to deny MCs.

New MCs from individuals, who when without dementia recently denied MCs, may suggest the presence of significant impairment of memory or cognition.