Predicting progression to Alzheimer's disease dementia using cognitive measures.

Abstract

It is important to determine if cognitive measures identified as being prognostic in dementia research cohorts also have utility in memory clinics. We aimed to identify measures with the greatest power to predict future Alzheimer's disease (AD) dementia in a clinical setting where expensive biomarkers are not widely available. This study utilized routine Memory Clinic data collected over 18years. From 2214 patients assessed in the clinic, we selected 328 patients with an initial diagnosis of subjective cognitive decline or mild cognitive impairment. We compared two types of statistical model for the prediction of AD dementia. The first model included baseline cognitive test scores only, while the second model also included change scores between baseline and the first follow-up. Baseline scores on tests of global cognitive function (Mini-mental state examination and Cambridge Cognitive Examination-Revised), verbal episodic memory and psychomotor speed were the best predictors of conversion to AD dementia. The inclusion of cognitive change scores over 1year of follow-up improved predictive accuracy versus baseline scores alone. We found that the best cognitive predictors of AD dementia in a clinical setting were similar to those previously identified using research cohorts. Taking change in cognitive function into account enabled the onset of AD dementia to be predicted with greater accuracy.