Eye scans and heart health: retinal OCT features linked to atrial fibrillation risk in oc- ulomics research

The Conversation explains how optical coherence tomography (OCT) retinal scans might reveal signals of cardiovascular health beyond vision, a field known as oculomics. In a large observational study using Moorfields eye hospital data linked with health records, researchers found that people with atrial fibrillation tended to have thinner retinal nerve layers, and that thinning predicted higher future risk of AF in participants from UK Biobank. The study demonstrates a promising but limited view of eye-based health clues for heart conditions.

• Retinal OCT scans can capture information about systemic health, not just eye health.
• AF patients showed differences in retinal layers and blood vessels, particularly a thinner macular ganglion cell-inner plexiform layer.
• Among participants without AF at imaging, thinner retinal layers were linked to a higher later AF risk after accounting for age, sex, ethnicity, and risk factors.
• Limitations: observational design, AF may have been undiagnosed earlier, and eye scans are not diagnostic tests for AF.

What do the findings mean for eye and heart health?

The retina has a high demand for oxygen and an extensive blood supply, linking eye health to systemic processes. AF—an irregular heart rhythm that can be intermittent and asymptomatic—poses diagnostic challenges. The imaging findings in the retina could provide a non-invasive clue prompting cardiovascular health assessment, but they do not diagnose AF on their own and cannot replace rhythm monitoring such as ECG. The results contribute to the growing field of ocualomics, which seeks to connect retinal biomarkers with cardiovascular risk, and highlight the potential role of retinal imaging in risk stratification when integrated with other clinical data and AI analysis.

Limitations and caveats

The study is observational and cannot establish causality. It remains unclear whether retinal thinning contributes to AF development, or whether both retinal changes and AF reflect shared cardiovascular risk factors or systemic processes. Additionally, some AF cases may have been undiagnosed at baseline, and the longitudinal analysis identified AF when hospital records captured the diagnosis, not through continuous rhythm monitoring. While retinal imaging is non‑invasive and already common in eye care, whether retinal biomarkers will be accurate and specific enough for screening remains uncertain. Future work will need to determine whether adding retinal imaging to existing cardiovascular risk assessment improves patient health outcomes and whether AI tools can reliably extract clinically useful information from retinal images.