SNIPE score can capture prodromal Alzheimer's in cognitively normal subjects

Abstract

Capturing early changes in the brain related to Alzheimers disease may lead to models that successfully predict cognitive decline and the eventual onset of dementia, well ahead of onset of clinical symptoms. In this study we used both hippocampal volume and our hippocampal driven SNIPE score to show which marker better captures Alzheimers related changes in a large dataset of normal controls (N=515) from the ADNI study, comparing controls that remain cognitively stable and controls that progress to either MCI or Alzheimers dementia during 10 years of follow-up (median follow-up: 30 months). We measured hippocampal volume and SNIPE score and found that the effect size to differentiate between stable and progressor groups was significantly larger for SNIPE score than for volume. Our results also show that there is a significant age-related difference between groups for both markers, and the difference is greater with the SNIPE score. Our experiments show that considering high sensitivity of our SNIPE score regarding to early AD-related brain changes, this marker is a better candidate in comparison to hippocampal volume for predicting the future onset of dementia.