Investigating potential ways to diagnose Alzheimer's disease using a blood test

Research project: Creating and early diagnostic blood test for Alzheimer's disease

Lead Investigator: Dr Richard Dobson
Institution: King's College London
Grant type: PhD studentship
Duration: 3 years
Amount: £79,457
Start Date: October 2013
End Date: September 2016

What was the project, and what did the researchers do?

It is increasingly thought that any successful treatment for Alzheimer's disease would need to be given in the very early stages of the condition, before symptoms start to show. At the moment, there is no accurate way to identify people who don't yet have symptoms but who will definitely go on to develop Alzheimer's disease.  

Alzheimer's disease is characterised by the presence of toxic clumps of the proteins amyloid and tau in the brain. Detecting how much amyloid or tau a person has in their brain or body is not easy and involves expensive or uncomfortable tests, such as lumbar punctures and brain scans. However, it may be that signals from other proteins could shed light on the levels of amyloid or tau a person has. 

The researchers on this project wanted to understand whether there were signals that could be detected using a blood test that could accurately identify whether someone had altered levels of amyloid or tau due to Alzheimer's disease.  A blood test could provide e a quick, cheap and non-invasive way of diagnosing the condition.

To achieve this goal, the researchers compared four different aspects of biology in people with and without Alzheimer's disease: 

  • Genetics: There have been a number of recent studies to identify parts of the individual's genetic code that are associated with developing Alzheimer's disease
  • Gene expression: Every cell contains every gene, but only certain genes are 'switched on' in the cell. This test helps to identify which genes are switched on or off. 
  • Proteins: Genes make proteins, which control what happens in cells, and also form structures (such as hair or collagen) 
  • Metabolites: These are all the small molecules found in blood.  They include substances such as vitamins, minerals and fats.

Each of these was tested to see if it contained any information that could predict Alzheimer's disease more accurately than current methods of diagnosis.  

What were the key results and how will this benefit people with dementia?

  • Genetics: The researchers created a combined score of the genetic markers that have previously been found to be associated with Alzheimer's disease. However, this gave no added benefit over typical tests that identify amyloid and tau levels. 
  • Gene expression: The scientists looked genes were switched on or off in Alzheimer's disease, and also the pathways that the genes were part of. Again, this provided no benefit over the information already avaiable. 
  • Proteins: Levels of two proteins (out of 41 identified) seemed to be related to levels of amyloid, so might be useful in a test.
  • Metabolites: A group of 5 metabolites was found that correctly classified high/low amyloid with 72% accuracy. This rose to 79% if one protein, fibrinogen was also measured.

Overall, characteristics such as age, gender and the presence of the Alzheimer's risk gene ApoE4 remain the most predictive a person's risk of Alzheimer's disease. However, the combination of 5 metabolites and 1 protein showed promising results. 

What happened next? Future work and additional grants

This work was carried out by PhD student Nicola Voyle. Nicola received her doctorate and continues to work in dementia research.

How were people told about the results? Conferences and publications


Voyle N et al. (2015) Blood protein markers of Neocortical Amyloid-β Burden: A candidate study using SOMAscan technology. Journal of Alzheimer's Disease 46 (2), 947 - 962. 
Justin TW et al. (2015) No evidence to suggest that the use of acetylcholinesterase inhibitors confounds the results of two blood-based biomarker studies in Alzheimer's disease. Journal of Alzheimer's Disease 47 (3), 741 - 750. 

Voyle N et al. (2015) A pathway based classification method for analysing gene expression for Alzheimer's Disease diagnosis. Journal of Alzheimer's Disease 49 (3), 659 - 669. 

Voyle N et al. (2016) Blood metabolite markers of Neocortical Amyloid Burden: Discovery and enrichment using candidate proteins. Translational Psychiatry 6, e719. 


Nine presentations were made at research conferences.