Could a fingerprick at home flag your Alzheimer's risk? New study says yes

By Pooja Toshniwal PahariaReviewed by Susha Cheriyedath, M.Sc.May 10 2026

A simple fingerprick taken at the kitchen table and dropped in the post could soon help identify who is most at risk of Alzheimer's disease, long before symptoms tighten their grip. In a new  study, researchers show that self-collected capillary blood, paired with online cognitive tests, matches the accuracy of traditional clinic-based sampling, opening the door to earlier, more equitable, and far more accessible dementia risk screening.

Study: Alzheimer’s Disease blood biomarkers measured through remote capillary sampling correlate with cognition in older adults. Image Credit: nito / Shutterstock

In a recent study published in Nature Communications, researchers found that self-administered fingerprick blood tests measuring plasma phosphorylated tau at amino acid 217 (p-tau217), along with glial fibrillary acidic protein (GFAP), closely aligned with conventional venous blood biomarkers. The markers were also associated with cognitive and functional performance across the Alzheimer's disease (AD) continuum. The findings suggest that combining these capillary blood tests with computerized cognitive assessments could support remote, at-home triage of individuals at risk of AD-related impairment, though the authors emphasize the approach is intended for risk stratification, not diagnosis. High patient acceptability further underscores its potential to improve early detection.

AD places a major burden on individuals, families, and healthcare systems, progressively impairing memory, independence, and quality of life. Despite advances in biomarker research, many people with early cognitive impairment remain outside specialist care, delaying diagnosis and intervention. Blood biomarkers such as p-tau217 and GFAP have emerged as promising tools for detecting AD-related pathology. However, current venous blood testing still requires clinic visits, limiting accessibility and scalability. This underscores the need for scalable approaches that extend risk assessment beyond specialist settings and support earlier identification of high-risk individuals.

Capillary Sampling Study Design

Building on earlier work validating capillary sampling against venous blood and cerebrospinal fluid markers, researchers evaluated whether a self-administered capillary blood test could support scalable triage for AD when collected remotely and returned by post.

The researchers included 174 participants spanning cognitively healthy adults, individuals with mild cognitive impairment (MCI), and patients with mild to moderate AD dementia, though only 28 participants had dementia, which limits the robustness of group-discrimination analyses. They recruited individuals with dementia through primary and secondary care sites across the South-West United Kingdom using established diagnostic criteria. Cognitively healthy individuals were drawn from the Platform for Research Online To investigate gEnetics and CogniTion and ageing (PROTECT-UK) study. The cohort was predominantly white and UK-based, so generalizability to other ethnic, socioeconomic, and healthcare contexts remains unconfirmed.

Using written and video instructions, participants collected fingertip capillary blood samples at home and returned them by post. Additionally, 40 of them provided venous blood samples for comparison of sampling methods. Researchers then quantified GFAP and p-tau217 using ultrasensitive immunoassay platforms (highly sensitive lab tests capable of detecting low levels of these proteins in blood).

Participants also completed a comprehensive assessment of cognition, function, and health status. Using a computerized testing platform developed within the PROTECT study, the team evaluated executive function, attention, and memory through tasks assessing recall, associative memory, reaction speed, verbal reasoning, and working memory.

Further, the team measured functional ability using the Lawton Instrumental Activities of Daily Living (IADL) scale. They then used the informant questionnaire on cognitive decline in older adults (IQCODE) to assess cognitive decline. Participants also provided information on vascular health conditions through structured medical history questionnaires. In addition, they completed an online Likert-scale survey evaluating usability, user experience, and readiness to adopt the at-home sampling method, with caregiver support provided where necessary. The researchers assessed correlations between blood-based biomarkers and cognitive outcomes. Logistic regression models explored associations between biomarker status and vascular risk factors, adjusting for age and sex.

p-tau217 and GFAP Cognitive Findings

The study included 146 cognitively normal individuals and 28 with dementia. Participants had a mean age of 66 years, and 54% were female. Both capillary biomarkers showed significant associations with cognitive performance. Elevated p-tau217 levels were linked to poorer memory, attentional processing, and higher-order cognitive abilities. GFAP was similarly associated with deficits in working memory and executive processing.

Functional findings showed a comparable trend. Elevated p-tau217 levels were associated with both daily functioning and cognitive decline scores, whereas GFAP was linked only to functional ability measures. Both biomarkers also significantly discriminated between participants with and without dementia in receiver operating characteristic (ROC) curve analyses (a standard method for evaluating diagnostic test performance), supporting their role in triage rather than as standalone diagnostic tests, particularly given the modest sensitivity (35%) of the chosen thresholds.

Using predefined thresholds prioritising 85% specificity (favoring few false positives at the cost of missing some true cases), researchers identified distinct at-risk groups. Participants exceeding the biomarker threshold values showed significantly poorer cognitive and functional performance across most domains. A dual-threshold approach combining p-tau217 with episodic memory performance, flagging individuals who scored abnormally on both tests, further refined stratification, identifying a high-risk group comprising 9% of participants with marked impairment and a low-risk group with consistently better performance across measures.

Notably, researchers found minimal overlap between the two biomarkers, with only 6% of participants testing positive for both. GFAP positivity was strongly linked to heart disease (odds ratio 4.14, 95% CI 1.31–13.10), indicating GFAP-positive individuals had roughly four times higher odds of reporting a cardiac history, pointing to a potential association with vascular-related cognitive impairment, while p-tau217 showed no such relationship.

The team also found strong agreement between capillary and venous biomarker measurements (r = 0.71–0.79 across diagnostic subgroups, including cognitively healthy and dementia participants), supporting the validity of the home-based method. Eighty percent of participants completed the test independently, and nearly all (96%) indicated they would be willing to use it as part of routine healthcare, though stated willingness does not necessarily translate to successful independent use, as the 20% requiring assistance demonstrates.

At-Home Alzheimer’s Triage Implications

The findings support capillary p-tau217, particularly when combined with computerized cognitive testing, as a scalable, remote triage tool for identifying individuals at the highest risk of AD-related impairment, while GFAP may provide additional insight into vascular-related cognitive decline.

The study showed strong agreement between capillary and venous biomarkers and validated the use of self-collected home samples, reinforcing the feasibility of this approach outside clinical settings. High patient acceptability further supports real-world implementation. Rather than diagnosing disease, this approach could enable earlier risk stratification and improve clinical pathways. It may also streamline clinical trial recruitment, where screen failure rates in preclinical AD trials currently reach as high as 90%.

Future studies should validate findings in independent and more diverse cohorts, benchmark against amyloid and tau PET imaging, assess longitudinal predictive value, and refine self-testing instructions (since 20% of participants required assistance) to integrate this strategy into primary care and research settings.

Readers should also note that commercial immunoassay platforms were used in this work, and funding and competing-interest disclosures should be reviewed in the original publication.