Predicting progression to dementia in persons with mild cognitive impairment using cerebrospinal fluid markers

Abstract

Introduction: We aimed to determine the added value of cerebrospinal fluid (CSF) to clinical and imaging tests to predict progression from mild cognitive impairment (MCI) to any type of dementia.

Methods: The risk of progression to dementia was estimated using two logistic regression models based on 250 MCI participants: the first included standard clinical measures (demographic, clinical, and imaging test information) without CSF biomarkers, and the second included standard clinical measures with CSF biomarkers.

Results: Adding CSF improved predictive accuracy with 0.11 (scale from 0-1). Of all participants, 136 (54%) had a change in risk score of 0.10 or higher (which was considered clinically relevant), of whom in 101, it was in agreement with their dementia status at follow-up.

Discussion: An individual person's risk of progression from MCI to dementia can be improved by relying on CSF biomarkers in addition to recommended clinical and imaging tests for usual care.

Keywords: Alzheimer's disease; Conversion; Dementia; Mild cognitive impairment; Predict; Prognosis; Progression; Reclassification; Risk; Risk prediction model.