At-home blood and brain tests revolutionize early dementia screening

A blood sample small enough to come from a finger prick, paired with a set of online thinking tests, may offer a new way to sort who faces the greatest risk of dementia. This approach could do so without sending everyone into a clinic.

That is the promise behind new research from the University of Exeter, published in Nature Communications. The study tested whether blood markers collected at home and mailed to a lab matched changes in memory and thinking. The answer, the team found, was yes. At least, it was well enough to suggest a practical role as a triage tool.

The idea is not to diagnose dementia from your kitchen table. Instead, it is to help identify who may need a closer look, who might benefit from monitoring, and who is less likely to need follow-up right away.

Professor Anne Corbett of the University of Exeter Medical School, who led the work, said: “Our previous research has shown that a finger-prick blood test can effectively be taken at home and posted to labs, and that we can identify the biomarkers in blood linked to dementia. This new study builds on that to show that we can link these biomarkers with performance on brain tests, giving us a potential way to predict risk of dementia.”

The study was carried out through PROTECT, an online UK research program. In PROTECT, more than 30,000 people over age 40 regularly complete cognitive tests. These tests measure memory, attention, and executive function, the mental skills used in planning and decision-making.

For this analysis, researchers recruited 226 people and found that 174, or 77%, successfully completed the at-home blood test and mailed it back. Among those participants, 146 had normal cognition and 28 had dementia. Their average age was just over 66. Also, 54% were women.

The team measured two proteins. One, p-tau217, is closely tied to Alzheimer’s disease and is already considered one of the strongest blood-based markers of Alzheimer’s pathology. The other, GFAP, has been linked more broadly to brain decline. It may reflect processes such as astrogliosis or neuroinflammation.

Both proteins tracked with worse performance on cognitive tests, but p-tau217 had the clearest ties. It correlated significantly with episodic memory, attention, and executive function. In addition, it lined up with two measures of everyday function, including instrumental activities of daily living and IQCODE, a questionnaire used to assess cognitive decline. GFAP also showed significant links, especially with working memory, executive function, and daily functioning.

That matters because the field has been moving toward blood biomarkers as a cheaper and more scalable alternative to brain scans and spinal fluid tests. Venous blood draws are already entering the diagnostic pathway. However, they still require clinic visits. The Exeter team is asking whether that process can move even further out into the community.

The researchers found that both markers could distinguish between people with and without dementia, though not strongly enough to stand alone as a diagnostic test. P-tau217 reached an AUC of 0.656, while GFAP reached 0.688. This is modest but statistically significant performance.

So the team turned to triage rather than diagnosis.

Using a pre-set threshold designed to favor specificity over sensitivity, the analysis identified 46 participants, or 26%, as positive on the p-tau217 capillary test and 29 people, or 17%, as positive on GFAP. Those above the thresholds showed significantly worse function and cognition on most measures.

A second step sharpened the picture. Researchers combined p-tau217 levels with memory performance and created a dual-threshold model. People above the p-tau217 cut-off and below the expected memory range for age formed a high-risk group. People below the biomarker cut-off and well above average on memory formed a low-risk group. Everyone else fell in between.

That high-risk group accounted for 9% of participants and performed significantly worse not just on memory, but also on attentional speed, accuracy, executive function, and measures of everyday function.

Professor Corbett said: “This work raises the potential for screening people for their risk without the need for clinic visits or complex clinical assessments. It would ensure the people at highest risk could be prioritized for monitoring and diagnosis, unlocking the best support and treatment for those that need it most.”

One of the more interesting results involved GFAP.

Even though both markers were linked to cognitive problems, only 11 participants, about 6% of the cohort with both measurements available, tested positive for both p-tau217 and GFAP. That small overlap suggests the two markers may be capturing different kinds of risk.

The analysis pointed in that direction. People who were GFAP-positive were nearly five times more likely to report a history of heart disease than GFAP-negative participants. They were also more likely to report hypertension, although that result did not reach statistical significance. P-tau217 positivity, by contrast, was not significantly linked to heart disease or hypertension.

The authors say this could mean GFAP is flagging people whose cognitive decline is tied more to vascular problems or non-Alzheimer’s inflammation than to classic Alzheimer’s disease. In a clinic trying to spot all-cause progressive cognitive impairment, that could make GFAP a useful extra signal. For Alzheimer’s-focused clinical trials, however, p-tau217 may still be the better primary screening tool.

The finger-prick results also held up reasonably well when compared with standard blood draws in a subgroup of 40 participants. Correlations between capillary and venous blood were strong, above 0.7, in both people with dementia and those without.

A screening tool does not get far if people will not complete it. On that front, the study was encouraging.

Eighty percent of participants said they completed the home blood test without help. Seventy-eight percent said they would be willing to use this kind of remote testing as part of routine healthcare without supervision from a clinician. In written feedback, participants often described the process as easy and relatively painless. Moreover, some pointed to the possible savings in time and travel.

Professor Clive Ballard of the University of Exeter Medical School said: “Almost a million people in the UK are estimated to have dementia, yet at the moment, only one in a thousand people with the earliest signs of brain decline receives a specialist evaluation. Our approach of combining our robust cognitive testing with measuring proteins via a postal blood test could provide a straightforward, efficient and cost-effective method to reach large numbers of people in the community who would not otherwise be prioritized for the next steps of diagnosis or support and to optimise the clinical pathway to enable early detection of those at highest risk.”

Still, the researchers were careful not to oversell what they had built.

The work has limits. About 23% of recruited participants did not complete the blood test. The sample with venous blood for comparison was small. Cardiovascular risk factors were measured through a single self-report item, which leaves room for error. The study also needs replication in an independent group. There is a need for longer follow-up to show whether these home measurements can predict who will later decline.

What this study offers is a possible sorting system for an overwhelmed pathway.

A person with elevated p-tau217 and poor memory scores might be moved up for specialist assessment. Someone with very low biomarker levels and strong cognitive performance might avoid unnecessary referral. People in the middle could be monitored over time, possibly through repeat online testing and future home sampling.

That could matter for regular care, where early specialist access remains rare, and for clinical trials. Screen failure rates for preclinical Alzheimer’s studies often approach 90%.

Professor Marian Knight, NIHR Scientific Director for NIHR Infrastructure, said: “The potential of this combination of cognitive and blood tests, both of which can be done at home, is really exciting. Not only could it reduce the burden on the NHS by screening people in their own homes rather than in hospitals or clinics, but it also might mean we can identify people with dementia earlier, tailor treatments more effectively, and improve outcomes for patients.”

The technology is not ready to replace diagnosis. But it may help decide who should get to diagnosis first.

Research findings are available online in the journal Nature Communications.

The original story "At-home blood and brain tests revolutionize early dementia screening" is published in The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.