A nasal swab detects Alzheimer's biological signals before memory loss begins

Most of what we know about the cellular biology of Alzheimer's disease comes from tissue collected after patients have died. By then, the brain has been ravaged by decades of progressive damage. The information is valuable but backward-looking: it tells us where the disease ended, not where it began.

A study published March 18 in Nature Communications from Duke Health researchers offers a different approach - and a strikingly simple one. A gentle brush inserted high into the nose collects living nerve and immune cells whose gene activity patterns can distinguish people with early or diagnosed Alzheimer's from healthy controls. The procedure takes minutes. It requires no imaging equipment, no spinal tap, no radioactive tracers. And it may detect biological changes before any memory problems appear.

Living nerve cells, accessible without surgery

The upper reaches of the nasal cavity contain olfactory sensory neurons - nerve cells responsible for detecting odors. These neurons are unique among nerve cells in the body because they are directly exposed to the external environment and are accessible without invasive procedures. A clinician applies a numbing spray, then guides a small cytology brush into the upper nose to gently collect cells from the olfactory epithelium.

Bradley J. Goldstein, M.D., Ph.D., professor in the departments of Head and Neck Surgery and Communication Sciences, Cell Biology, and Neurobiology at Duke University School of Medicine, leads the team behind the approach. The core insight driving the work is that these nasal neurons are part of the nervous system and may reflect disease processes occurring in the brain - providing a window into neural biology without requiring direct access to brain tissue.

The study compared nasal swab samples from 22 participants: some with clinically diagnosed Alzheimer's, some showing laboratory-based biomarker evidence of Alzheimer's but without symptoms, and healthy controls. Using single-cell RNA sequencing, the team measured the activity of thousands of genes across hundreds of thousands of individual cells, generating millions of data points per participant.

Patterns that emerge before symptoms do

The analysis revealed distinct gene activity signatures in both nerve cells and immune cells collected from the nose. People with Alzheimer's - including those who showed biomarker evidence of the disease but had no cognitive symptoms yet - displayed patterns of gene expression that differed systematically from healthy controls.

A combined tissue gene score, integrating signals from multiple cell types, correctly separated early and clinical Alzheimer's cases from healthy controls approximately 81% of the time. That is not perfect, but for a screening tool based on a minimally invasive procedure, it represents a potentially useful level of discrimination.

The finding that the nasal swab picked up changes in pre-symptomatic individuals is particularly important. Alzheimer's pathology - the accumulation of amyloid plaques and tau tangles in the brain - begins years or even decades before memory loss and cognitive decline become apparent. A growing class of new therapies, including anti-amyloid antibodies like lecanemab and donanemab, appear to work best when administered early in the disease process. But early administration requires early detection, and that has been a persistent bottleneck.

How this compares to current diagnostic tools

Current methods for detecting Alzheimer's before symptoms appear are either expensive, invasive, or both. PET brain scans using radioactive tracers can visualize amyloid plaques and tau tangles, but they cost thousands of dollars and require specialized imaging facilities. Cerebrospinal fluid analysis requires a lumbar puncture - a spinal tap - which many patients and physicians prefer to avoid for routine screening.

Blood tests for Alzheimer's biomarkers have advanced considerably in recent years, measuring proteins like phosphorylated tau (p-tau) and amyloid beta ratios. These tests are less invasive and increasingly accurate, but they detect markers that appear later in the disease process - downstream products of pathology that has already been underway for some time.

The nasal swab approach captures something different: the activity states of living nerve and immune cells. Rather than measuring proteins that accumulate as a consequence of disease, it reads the cellular machinery itself - which genes are turned on, which are turned off, and how those patterns differ between health and disease. This could, in theory, provide a more direct and earlier look at disease-related changes in the nervous system.

Vincent M. D'Anniballe, the study's first author and a student in Duke's Medical Scientist Training Program, emphasized the significance of being able to study living neural tissue. Much of what researchers know about Alzheimer's comes from autopsy tissue. This approach opens new possibilities for both diagnosis and treatment monitoring in living patients.

A participant's perspective

Mary Umstead volunteered for the study after her sister Mariah was diagnosed with young-onset Alzheimer's at age 57. The family had noticed signs of the disease long before the formal diagnosis came. Mary described being motivated by the desire to spare other families the experience hers went through - and to spare other patients what Mariah endured.

Stories like Mary's underscore why early detection matters so much. By the time Alzheimer's is clinically diagnosed through cognitive testing and imaging, substantial brain damage has typically already occurred. A simple, affordable screening tool that catches the disease in its earliest biological stages could change the timeline dramatically - particularly as treatments that target early pathology continue to reach the market.

Twenty-two participants and the road ahead

The study's most significant limitation is its size. Twenty-two participants is a very small cohort, and the 81% accuracy figure comes with wide confidence intervals at that sample size. The gene expression patterns identified here need to be validated in much larger, more diverse populations before any clinical conclusions can be drawn.

The study population was also drawn from a single academic medical center, raising questions about how well the results would generalize across different demographics, ethnicities, and clinical settings. Alzheimer's prevalence and presentation vary across populations, and a diagnostic tool must perform reliably across these differences to be clinically useful.

Single-cell RNA sequencing, while powerful, is currently expensive and labor-intensive. For the nasal swab to become a practical screening tool, the analysis pipeline would need to be simplified and the costs reduced substantially. The researchers are working toward identifying a smaller panel of key genes that could be measured with cheaper, faster assays rather than full single-cell sequencing.

The Duke team, in collaboration with the Duke and UNC Alzheimer's Disease Research Center, is now expanding the research to larger groups and exploring whether the swab could also track treatment response over time - allowing clinicians to monitor whether a therapy is actually changing the biological trajectory of the disease in an individual patient. Duke has filed a U.S. patent related to the approach.

The concept is compelling: a test that takes minutes, requires no specialized equipment beyond a brush, and reads biological signals from nerve cells that are part of the same nervous system under attack in Alzheimer's disease. Whether it can scale from a 22-person proof of concept to a reliable clinical tool remains the central question. But the potential, if validated, is a fundamentally different entry point for detecting a disease that currently reveals itself far too late.