New blood test predicts Alzheimer’s 3.5 years in advance

An international team of European researchers led by scientists at King’s College London in the UK has discovered a blood-based test that can accurately predict 3.5 years in advance whether individuals diagnosed with mild cognitive impairment will go on to develop Alzheimer’s Disease.

Characterized by forgetfulness, slightly impaired judgment, and minor language problems – which together generally don’t impact daily life – mild cognitive impairment (MCI) represents the gray area between the normal cognitive slowdown we all experience in old age and much more serious dementia. Worldwide, about 15% of adults fifty and older experience MCI. Over age 65, one in nine American adults have developed Alzheimer’s, but the transition often takes years.

That makes an MCI diagnosis a solid impetus to take action to lower one’s risk of Alzheimer’s. Being bilingual, exercising, and eating a healthy diet are associated with slower progression of Alzheimer’s symptoms.

New pharmaceutical treatments can also slow the decline into dementia, especially if initiated early. But since the majority of people who experience MCI don’t progress to Alzheimer’s, and since the treatments are not cheap or without side effects, who should undertake these interventions? That’s where a predictive test can be extremely useful. 

As they described in a study published to the journal Brain, the scientists took numerous serum samples (blood with all of the cells and clotting factors removed) from 56 individuals diagnosed with MCI over a 3.5 year timespan and observed whether those individuals went on to develop Alzheimer’s. The researchers mixed the serum samples with human brain cells from the hippocampus – one of the first brain regions damaged from Alzheimer’s – and watched what happened.

The test correctly identified 90% of people with MCI who would go on to develop Alzheimer’s.

They found that the serum of individuals who later went on to have Alzheimer’s resulted in decreased proliferation, increased cell death, and boosted neurogenesis (development of neurons) amongst the cultured brain cells. Then, coupling cellular measurements of these factors with a patient’s years of education, they formulated a test to predict whether someone would develop Alzheimer’s. 

They subsequently validated this model with machine-based learning, finding that it correctly identified 90% of people who would go on to develop Alzheimer’s and 79% of people who would not.

“Our findings are extremely important, potentially allowing us to predict onset of Alzheimer’s early in a non-invasive fashion,” Dr. Edina Silajdžić, a clinical neuroscientist at King’s College London and joint first author of the study, said in a statement.

“It is now essential to validate these findings in a bigger and more diverse group of people,” Dr. Hyunah Lee, also a first author, added.

The present work isn’t the first to attempt to formulate a predictive test for Alzheimer’s, but it is one of the best so far. In 2021, researchers with IBM and Pfizer designed a test based on recognizing written language patterns, such as short and simple phrases, repeating and misspelling words, and skipping punctuation, which was 59% accurate at predicting a diagnosis within seven years. 

Later that year, scientists at Lund University in Sweden described a tool that blended a blood test with a 10-minute cognitive assessment that was 90% accurate at predicting an Alzheimer’s diagnosis within four years. And late last year, a University of Washington team announced solid steps towards making a blood test that could predict Alzheimer’s disease a decade or more in advance.

We’d love to hear from you! If you have a comment about this article or if you have a tip for a future Freethink story, please email us at [email protected].