A blood test can identify apparently healthy adults on the road to dementia as much as a decade before symptoms appear, according to the largest study of its kind, published in the Journal of the American Medical Association on 14 July and drawing on cohorts from North America, Japan and Australia.
The research, led by Rachel Buckley of Harvard Medical School and the Mass General Brigham Neuroscience Institute, pooled data on 2,684 cognitively unimpaired older adults, median age just under 70, from six long-running studies, including the A4 trial that recruited Australians through a Melbourne site. Participants were followed for a median of 5.4 years and as long as 13.5; over that time 478 of them progressed to cognitive impairment.
The predictive marker is p-tau217, a protein measurable in ordinary blood samples that until now has mainly been used to assess people who already have symptoms. Participants with very high levels had a 38 per cent chance of developing cognitive impairment within five years, and 78 per cent within ten. Those with very low levels faced roughly the same risk as the general ageing population, and the association held regardless of age, sex or whether a person carried the APOE e4 risk gene.
“Our findings provide some of the clearest evidence yet that elevated p-tau217 levels may help detect dementia risk years earlier," Buckley said. Senior author Reisa Sperling, a neurologist at Mass General Brigham, told the Alzheimer's Association International Conference in London the test could pick up disease "five to 10 years before a healthy person develops symptoms", and said it was "especially encouraging that the findings were so consistent across different groups".”
In Australia the stakes are measurable. Dementia is the nation's leading cause of death, responsible for 9.4 per cent of all deaths in 2024, more than 17,500 in the year, according to the Bureau of Statistics, and it has been the leading cause of death for Australian women since 2016. Deaths have risen 39 per cent over the past decade as the population ages.
The caveats are real: the test is a research tool, not something a GP can order as a dementia screen, there is no cure, and the value of a ten-year warning depends on prevention trials that are still running. The researchers are clear about which way they are betting. "This is the future of Alzheimer's care, targeting the earliest stages of the disease," the Alzheimer's Association's chief science officer Maria Carrillo said.
