Can clinical data predict progression to dementia in amnestic mild cognitive impairment?

To determine whether clinical data obtained by history and physical examination can predict eventual progression to dementia in a cohort of elderly people with mild cognitive impairment.

A prospective, longitudinal study of a cohort of elderly subjects with amnestic Mild Cognitive Impairment (MCI). Ninety subjects meeting the criteria for amnestic MCI were recruited and followed annually for an average of 3.3 years. Main outcome measure was the development of dementia determined by clinical assessment with confirmatory neuropsychological evaluation.

Fifty patients (56%) developed dementia on follow-up. They were older, had lower Mini-mental status exam (MMSE) scores and a shorter duration of symptoms at the time of first assessment. Multivariate logistic regression analysis identified age at symptom onset as the only clinical parameter which distinguished the group that deteriorated to dementia from the group that did not. The odds ratio for age was 1.1 (confidence interval 1.04 - 1.18).

Patients presenting with amnestic MCI insufficient for the diagnosis of dementia are at high risk of developing dementia on follow-up. In our cohort, 56% were diagnosed with dementia over an average period of 5.9 years from symptom onset. The only clinical predictor for the eventual development of dementia was older age at symptom onset. Clinical features alone were insufficient to predict development of dementia.