Dr Josef Huemer looking at John Brewer's eyes. Credit: Moorfields Eye Hospital
Thinner Retinas Tied To Higher Atrial Fibrillation Risk, Large Study Finds
In A Nutshell
- People with thinning in a specific retinal nerve layer faced a 27% higher risk of developing atrial fibrillation years later.
- The study drew on eye scans from more than 90,000 people, including a group of over 39,000 followed for about a decade.
- The retinal thinning was already detectable before atrial fibrillation was formally diagnosed, not just alongside it.
- Researchers say eye scans are not a replacement for existing heart monitoring, but could serve as a helpful early-warning signal.
People whose retinas showed thinning in a specific layer of nerve tissue faced a 27% higher risk of developing atrial fibrillation years later for every added step of thinning measured, according to a new study that examined eye scans from more than 90,000 people, including a group of over 39,000 who were followed for roughly a decade.
That heart condition, the most common form of irregular heartbeat, is notoriously hard to catch because it often comes and goes without warning, sometimes without any symptoms at all. That gap matters because atrial fibrillation raises the risk of stroke, heart failure and death. This new research suggests a simple eye scan, the kind already done in routine eye checkups, might flag the danger before a doctor ever suspects a heart problem.
Researchers analyzing those eye images found that people with atrial fibrillation had measurably thinner layers in the back of the eye, called the retina, compared with people who didn’t have the condition. The thinning showed up specifically in a layer made of nerve cells called the ganglion cell-inner plexiform layer, and it appeared consistently across both groups studied.
This isn’t the first time scientists have looked at the eye as a window into the rest of the body. The retina shares blood vessels and nerve tissue with the brain and heart, so changes on a scan can reflect problems elsewhere. This study stands out simply for its size, drawing on eye scans from tens of thousands of people rather than a small trial.
Two Large Patient Groups Link Retinal Thinning To Atrial Fibrillation
Scientists pulled data from two very different sources. The first, AlzEye, included more than 50,000 patients age 40 and older who visited a London eye hospital between 2008 and 2018; about 11% had atrial fibrillation. The second, UK Biobank, is a long-running health study that recruited people ages 40 to 69 between 2006 and 2010 and tracked their health over time.
For both groups, researchers used two kinds of eye imaging already common in eye care: color photographs of the retina and optical coherence tomography, a scanning technique that produces detailed cross-section images of the retina’s layers, somewhat like an ultrasound for the eye. From these images, the team measured retinal layer thickness and features of the retina’s tiny blood vessels.
Atrial fibrillation status came from hospital records and, in UK Biobank, self-reported diagnoses. Statisticians then compared retinal measurements between people with and without the condition, adjusting for age, sex, ethnicity, blood pressure and diabetes so those factors wouldn’t skew results. The findings were published in the journal PLOS Digital Health.
One Retinal Layer Consistently Thinner In Atrial Fibrillation Patients
In the hospital-based group, people with atrial fibrillation had thinner retinas across several layers, including nerve fibers that carry visual signals to the brain. But only thinning in that same nerve cell layer also turned up in the UK Biobank group, making it the sole retinal change confirmed across both populations. People with the heart condition also had smaller retinal veins and altered blood vessel branching, though that vascular pattern showed up only in the hospital-based group.
To put the thinning in perspective, researchers noted that the difference seen in the hospital group was roughly equivalent to about 13 years of typical age-related thinning, and about four times the thinning normally associated with daily alcohol drinking. In the UK Biobank group, the difference was comparable to roughly eight years of aging.
Following people over time produced the most compelling evidence. Researchers tracked more than 39,000 UK Biobank participants free of atrial fibrillation at baseline for about 10 years; 838 developed the condition, typically about four years after their initial eye scan. After accounting for other health and lifestyle factors, each standard step of thinning in that nerve cell layer was tied to a 27% higher risk of a later diagnosis. That timing detail matters: the retinal changes were already detectable before atrial fibrillation was formally diagnosed.
Silent Strokes And Poor Blood Flow May Drive The Pattern
Researchers believe a couple of things could explain the pattern. Atrial fibrillation raises the risk of small, sometimes silent strokes a person might never notice, and nerve damage from those strokes can travel backward along shared pathways, showing up as thinning in the retina. Separately, some thinning might reflect reduced blood flow to parts of the retina that depend on a delicate web of small vessels, a vulnerability that could mirror broader circulation problems caused by the irregular heartbeat.
Researchers were careful to note that eye scans are not meant to replace tools currently used to diagnose atrial fibrillation, such as heart rhythm monitors. Instead, they described the retinal findings as a possible companion signal that could flag people who deserve a closer look at their heart health, particularly since the condition is so easy to miss with monitoring alone.
Atrial fibrillation is already common and expected to become more so in the coming decades, and untreated cases carry real risks of stroke and other complications. A routine eye exam pointing toward an unnoticed heart risk would mark a real shift in how early warning signs get caught. This research does not prove eye scans can diagnose atrial fibrillation on their own, but it makes a strong case that the eyes may hold clues about heart health that could complement the screening tools doctors already use.
Disclaimer: This article is for general informational purposes and is not a substitute for professional medical advice, diagnosis, or treatment.
Paper Notes
Limitations
The researchers relied on hospital records and self-reported diagnoses to identify who had atrial fibrillation, which can miss cases, especially since the condition is often silent or comes and goes. That means some people counted as not having the condition may have actually had undiagnosed atrial fibrillation, which the authors say likely means their results underestimate the true strength of the connection rather than overstate it. Neither of the two groups studied fully represents the general population. UK Biobank participants tend to be healthier than the average person, while the hospital-based group skews toward people with more socioeconomic hardship and existing eye or health conditions. The researchers also noted that about 20% of the people who developed atrial fibrillation during the follow-up period did so within about two years of their eye scan, meaning the findings may apply more clearly to detecting existing cases than truly predicting far-off future ones. Finally, as with any study of this kind, other unmeasured factors could be influencing both heart and eye health at the same time.
Funding and Disclosures
The research was funded by the National Institute for Health and Care Research through several Biomedical Research Centres, including those at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, Barts, Birmingham, Great Ormond Street Hospital, and University College London Hospitals. Several authors disclosed relationships with companies and organizations including Abbvie, Aerie, Google Health, Heidelberg Engineering, Novartis, Regeneron, Topcon, Bayer, Roche, Boehringer Ingelheim, and others, along with grants from organizations such as the British Heart Foundation and the European Society of Cardiology. The funders had no role in designing the study, collecting or analyzing data, or deciding to publish the results, according to the paper.
Publication Details
The study, titled “Atrial fibrillation and retinal imaging-based oculomics: A cross-sectional and longitudinal analysis of two large cohorts,” was authored by Josef Huemer, Kelsey V. Stuart, Yukun Zhou, Dominic J. Williamson, Peter Woodward-Court, Alastair R. Mobley, Arvind Chandratheva, Florian Berger, Nay Aung, Steffen E. Petersen, members of the UK Biobank Eye and Vision Consortium, Anthony P. Khawaja, Axel Petzold, Jugnoo Rahi, Alastair K. Denniston, Dipak Kotecha, Pearse A. Keane, Praveen J. Patel, and Siegfried K. Wagner. It was published in PLOS Digital Health (2026), DOI: 10.1371/journal.pdig.0001661.







