New research has raised hopes that a blood test may one day be able to predict the risk that a person with mild cognitive impairment will develop Alzheimer’s.
Memory loss becomes more common with age, with stress, medicines and mental-health disorders often to blame.
Alzheimer’s – the most common form of dementia – has no set diagnostic test, with specialists generally assessing an individual according to their memory, attention span and communication skills.
Expensive brain scans may also be carried out to look for the tell-tale protein plaques and tangles that cause tissue to die and the vital organ to shrink.
Investigating the potential of a blood test, scientists from Lund University in Malmö, Sweden, analysed 573 people with minor cognitive impairment.
Read more: Anxiety may accelerate Alzheimer's onset
The team found those with the proteins P-tau181 and NfL in their blood were more likely to develop the memory-robbing disease over the next four years, with the combined test being up to 89% accurate.
Experts were generally enthused by the result, with one describing it as a “potential game-changer”.
Many stressed further research is required, however, with one adding that unless the government follows through on its promise to double dementia research funding, patients “won’t benefit from these new breakthroughs”.
Of the around 850,000 people in the UK with dementia, between half (50%) and three quarters (75%) have Alzheimer’s. In the US, more than 5 million people have the disease.
Dementia tends to be progressive, starting with mild cognitive impairment.
While Alzheimer’s is incurable, identifying the disease early on means drugs can be prescribed that temporarily ease symptoms.
Read more: Deep sleep protects against Alzheimer’s
Cerebrospinal fluid can show up Alzheimer’s biomarkers, but taking samples from the lower spine has “perceived invasiveness”, according to the scientists.
“Blood-based biomarkers could overcome these hurdles,” they added.
Watch: Spotting dementia signs while socially distant from loved ones
To learn more, the scientists analysed patients with mild cognitive impairment who took part in the Swedish BioFINDER study and the Alzheimer’s Disease Neuroimaging Initiative.
After testing several models on how blood biomarkers may predict Alzheimer’s, results – published in the journal Nature Aging – suggest the proteins P-tau181 and NfL were the most accurate indicators of the disease up to four years later.
P-tau181 is a form of tau protein. An abnormal build-up of tau can form tangles within nerve cells, triggering Alzheimer’s.
NfL “reflects the presence of neuronal death and injury”, according to the scientists.
Looking at both biomarkers, the test was found to have a sensitivity of 89%, which describes a device’s ability to correctly identify those with a disease.
Its specificity, ability to spot those without a condition, came in at 88%.
Perhaps surprisingly, Alzheimer’s onset was less linked to the presence of the protein amyloid-beta in the blood.
Abnormal clusters of amyloid-beta are known to form damaging plaques between nerve cells in the brain, leading to the memory-robbing disease.
‘Transformative step to earlier diagnosis’
Speaking of the results, Professor Masud Husain from the University of Oxford said: “This is a potential game-changer.
“For the first time, we have a blood test that can predict well the risk of subsequent development of Alzheimer’s disease in people who have mild cognitive symptoms.
“We need further validation but in the context of other recent findings this could be a transformative step to earlier diagnosis, as well as testing new treatments at earlier stages of the disease.”
Read more: Eating soy may help ward off dementia
With testing for several biomarkers expected to be more accurate than looking for just one, “this paper appears to be a contribution in that direction”, according to Professor Tom Dening from the University of Nottingham.
“However, as the authors have acknowledged, even a sample size of over 500 is quite small for this kind of predictive work and, besides, longer periods of follow-up are needed to test how useful the predictions turn out to be.”
While the sensitivity and specificity results may sound impressive, Professor Dening stressed “in large-scale practice it could [lead] to a lot of misdiagnosis, both people wrongly diagnosed as having Alzheimer’s and people who in fact have [Alzheimer’s] being missed”.
“The average memory clinic will see hundreds of people with possible dementia in a year, so these numbers would be considerable,” he said.
Alzheimer’s is an incurable disease, with clinical trials into promising drugs often ending in disappointment.
“A big factor is the struggle to find people early enough before Alzheimer’s has progressed too far to test new drugs on,” said Dr Richard Oakley from the Alzheimer’s Society.
“This study only looked at a few hundred people, but if these blood biomarkers can predict Alzheimer’s in larger, more diverse groups, we could see a revolution in how we test new dementia drugs.
“Blood tests to predict dementia are moving at a breakneck speed, but if [the] government doesn’t double dementia research funding as they promised, people with dementia won’t benefit from these new breakthroughs.”
The effectiveness of treating people with mild cognitive impairment to ward off Alzheimer’s is also up for debate, however, it is increasingly emerging as a useful strategy.
“Researchers are coming round to the idea that effective treatments to prevent Alzheimer’s disease will need to be started before there is any sign of problems and in order to do this it will be essential to recognise who is at high risk,” said Professor David Curtis from University College London.
“Being able to recognise high-risk subjects will also be necessary so that they can be recruited into drug trials of such preventative therapies. However, all this is in the future.
“The hope is that one day people will be able to have a routine check for Alzheimer’s risk and then be started on treatment, just as nowadays a cholesterol test can lead to a prescription for a statin.
“However, at present such treatments are not available, so the usefulness of tests such as this will just be for research purposes.”
Watch: Deep sleep may prevent Alzheimer’s