Relationship Between Plasma Aβ Levels, Cognitive Function and Brain Volumetrics Interview with:

Dr Anne Poljak Leader of the Proteomics Group Centre for Healthy Brain Ageing (CHeBA) UNSW, Australia

Dr. Anne Poljak

Dr Anne Poljak
Leader of the Proteomics Group
Centre for Healthy Brain Ageing (CHeBA)
UNSW, Australia What is the background for this study? 

Dr Poljak: Amyloid-beta (Aβ) peptides are found in abundance in the plaque particles which build up in the Alzheimer’s brain and small blood vessels of the brain, and are therefore considered hallmark features of Alzheimer’s disease. However they are also found in blood which is a convenient body fluid for sampling purposes. We therefore wished to assay them in plasma samples from one of our longitudinal population based studies of older age individuals (70 – 90 years) – the Centre for Healthy Brain Ageing’s Sydney Memory and Ageing Study.

Other research groups had previously measured these peptides in plasma, but there was controversy in the area because of differences in outcomes across laboratories. So one of the main questions was whether plasma levels of Aβ peptides have any relationship with what is happening in the brain, or are they a red-herring? We wanted to see how the levels we found would compare with the findings of others in relation to Alzheimer’s disease and mild cognitive impairment. A further question was how our plasma levels would relate to other clinical measures including cognition and brain volumetrics. One of the precautions we took was to use an assay kit with very well characterized antibodies, so we could be confident that our method was specific for the two full length Aβ peptides (Ab1-40 and Ab1-42). What are the main findings?

Dr Poljak: Firstly, we found that baseline plasma levels of Aβ peptides and the Aβ1−42/1-40 ratio measure were lower in amnestic mild cognitive impairment (aMCI) and Alzheimer’s dementia (AD) (aMCI is often considered to be a prodrome of AD). Because the trajectory of soluble Aβ levels in the brain goes downwards over time in Alzheimer’s disease, as the deposited plaque builds up (called the sink model), we took our dropping plasma levels to be a reflection of this model.

Our second finding was that the plasma Aβ levels were associated with clinical measures of change in cognition and brain volume: they were positively associated with global cognition and hippocampal volume and negatively associated with white matter hyperintensities.

Interestingly these relationships played out a bit differently for the two peptides, in that associations with Aβ1-40 were predominantly observed in Apoε4 allele carriers whereas associations with Aβ1-42 were mainly in non-carriers of the Apoε4  allele (one of the most important genetic risk factors for late onset Alzheimer’s disease). Of the two peptides the most striking changes were in relation to the longer and more amyloidogenic Aβ1−42.
A longitudinal analysis of our data revealed greater decline in global cognition and memory for the highest quintiles of Aβ1−42 and the ratio measure, so the plasma Aβ1−42 measure is predictive of cognitive decline. What should clinicians and patients take away from your report?

Dr Poljak: This is principally a research study into the relevance of plasma Aβ peptides to brain function and disorders such as mild cognitive impairment and Alzheimer’s disease. We feel that we have taken a step forward in this area, as our data do suggest a relationship between plasma  Aβ peptides and brain anatomy and function, and also support recent studies from other groups – so a consensus is building in support of the usefulness of these measures. However from the perspective of a diagnostic marker applicable in clinical practice, we are not there yet. The changes we observe are significant on a statistical level, but too small and too overlapping with the normal range, to be useful in the clinic as single time point measures. There may be some value clinically to track changes over time in patients – however this approach needs validation with large longitudinal studies and across labs and cohorts. What recommendations do you have for future research as a result of this study?

Dr Poljak: This was a relatively small study, and only included baseline measures, so it would be of value to assay larger sample numbers and also repeat measures over time, to see if the sink model is indeed recapitulated in plasma longitudinally.

Alzheimer’s is mechanistically very complex, and in some respects more typical of a syndrome than a disease. It is rarely expressed in “pure form”, more often accompanied by vascular issues, has features which overlap with other diseases and is typified by a variety both intracellular and extracellular changes and to proteins from a broad range of functional groups. Consequently the best biomarker approach may be multifactorial. Therefore seeking out other plasma markers associated with cognition and brain changes would be of value in preparation of a sufficiently specific biomarker approach. Is there anything else you would like to add?

Dr Poljak: The importance of sufficiently robust quantitative methodologies cannot be overstated. One of the most vexed issues in the plasma Ab area has been the question of what is actually being measured. The  Aβ peptides exist in a diversity of forms over which there continues to be debate as to the most toxic element. Compounding this mechanistic complexity are the diverse variety of technical approaches to their assay, questions of antibody specificity and so on. In the end our conclusions can only be as good as the measures we use.


Curr Alzheimer Res. 2016;13(3):243-55.

The Relationship Between Plasma Aβ Levels, Cognitive Function and Brain Volumetrics: Sydney Memory and Ageing Study.

Poljak A1, Crawford JD, Smythe GA, Brodaty H, Slavin MJ, Kochan NA, Trollor JN, Wen W, Mather KA, Assareh AA, Ng PC, Sachdev PS.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.