A vascular “fingerprint” in the retina of the eye can predict a person’s stroke risk as accurately as traditional risk factors, according to research published online in the journal Heart.

Stroke affects approximately 100 million people worldwide and kills 6.7 million of them each year. Most cases are caused by modifiable risk factors such as high blood pressure, high cholesterol, poor diet and smoking.

The investigation

The complex vascular network of the retina is divided common anatomical and physiological features with the vascular system of the brain. The researchers wanted to investigate – with the help of artificial intelligence – whether it is possible to identify biological markers that can accurately predict the risk of stroke without the need for invasive laboratory tests.

The research measured 30 markers related to the caliber, density, branching and complexity of retinal veins and arteries, through fundus images from 68,753 participants in the UK Biobank database. They also took into account potentially influential risk factors such as demographic and socioeconomic factors, lifestyle, and health parameters including blood pressure, cholesterol, blood glucose, and weight. The final analysis included 45,161 participants with a mean age of 55 years. The mean follow-up period was 12.5 years.

In total, 118 measurable markers of retinal vasculature were included, of which 29 markers of vascular health were significantly associated with risk of first stroke. Among other things, they found that each change in vessel density markers was associated with a 10-19% increased risk of stroke, while changes in diameter markers were associated with a 10-14% increased risk. Each decrease in the complexity and convolution indices was associated with an increased risk of 10.5–19.5%.

The scientists found that this retinal vascular “fingerprint” was just as good as using traditional risk factors alone in predicting future stroke risk. Of course, the researchers clarify that this is an observational study, so no firm conclusions can be drawn about cause and effect. They also point out that the findings may not apply to different ethnicities. Finally, they were not able to assess the risk associated with different types of stroke, however they conclude that since retinal parameters can be easily obtained, this model is a practical and easily applicable approach for estimating stroke risk. episode.