Get access

A proposed metabolic strategy for monitoring disease progression in Alzheimer's disease

Authors

  • Nicola Greenberg,

    1. Pharmaceutical Sciences Research Division, King's College London, London, UK
    Search for more papers by this author
  • Antonio Grassano,

    1. Pharmaceutical Sciences Research Division, King's College London, London, UK
    Search for more papers by this author
  • Madhav Thambisetty,

    1. Department of Neuroscience, Medical Research Council Centre for Neurodegeneration Research, Institute of Psychiatry, Kings College London, Denmark Hill, London, UK
    Current affiliation:
    1. National Institute on Aging, Intramural Research Program, National Institutes of Health, Washington DC, USA.
    Search for more papers by this author
  • Simon Lovestone,

    1. Department of Neuroscience, Medical Research Council Centre for Neurodegeneration Research, Institute of Psychiatry, Kings College London, Denmark Hill, London, UK
    Search for more papers by this author
  • Cristina Legido-Quigley

    Corresponding author
    1. Pharmaceutical Sciences Research Division, King's College London, London, UK
    • Pharmaceutical Sciences Research Division, King's College London, London, UK Fax: +44-20-78484800
    Search for more papers by this author

Abstract

A specific, sensitive and essentially non-invasive assay to diagnose and monitor Alzheimer's disease (AD) would be valuable to both clinicians and medical researchers. The aim of this study was to perform a metabonomic statistical analysis on plasma fingerprints. Objectives were to investigate novel biomarkers indicative of AD, to consider the role of bile acids as AD biomarkers and to consider whether mild cognitive impairment (MCI) is a separate disease from AD. Samples were analysed by ultraperformance liquid chromatography–MS and resulting data sets were interpreted using soft-independent modelling of class analogy statistical analysis methods. PCA models did not show any grouping of subjects by disease state. Partial least-squares discriminant analysis (PLS-DS) models yielded class separation for AD. However, as with earlier studies, model validation revealed a predictive power of Q2<0.5 and indicating their unsuitability for predicting disease state. Three bile acids were extracted from the data and quantified, up-regulation was observed for MCI and AD patients. PLS-DA did not support MCI being considered as a separate disease from AD with MCI patient metabolic profiles being significantly closer to AD patients than controls. This study suggested that further investigation into the lipid fraction of the metabolome may yield useful biomarkers for AD and metabolomic profiles could be used to predict disease state in a clinical setting.

Get access to the full text of this article

Ancillary