A photo at the back of the eye may hold early clues about Alzheimer’s risk.

Researchers say retinal images cannot diagnose Alzheimer’s years before symptoms appear, but they can pick up subtle signs tied to factors that may raise a person’s risk of developing the disease. The new study adds to growing evidence that imaging tiny blood vessels and nerves in the eye could offer a heads-up for people at higher risk of cognitive decline.

“We know that Alzheimer’s disease develops over decades, but most of the diagnostic tools focus on late-stage pathology when it is too late to intervene,” said biomedical engineer Ruogu Fang from the University of Florida, who led the study.

“By looking at novel biomarkers, like retinal health, we offer new opportunities to identify patients at risk, offer appropriate tests, and encourage them to develop healthy lifestyles to mitigate their risk.”

The retina is the light-sensitive tissue at the back of the eye and contains several layers of neurons. Researchers have already shown that retinal photos can identify people with active and ongoing cases of Alzheimer’s disease. In the new work, some of the same researchers found the images could also help estimate a person’s risk of developing Alzheimer’s before diagnosis.

The team used machine learning to analyse 62,876 retinal photographs from more than 40,000 UK Biobank participants. Their AI program was designed to predict 12 factors linked to a higher risk of Alzheimer’s disease, including sex, smoking, sleep, alcohol use, depression, age, body mass and blood pressure.

The deep learning models identified several subtle markers associated with the development of Alzheimer’s disease. The study points to vascular stiffening, reduced blood vessel density and thinning of the optic nerve as signs of retinal ageing that are also associated with Alzheimer’s.

Models predicting who would later develop Alzheimer’s used demographic, vascular, metabolic and lifestyle risk factors, along with retinal images and their corresponding risk factors. Patients who later developed Alzheimer’s showed constriction of small arteries and arterioles in the retina, matching previous research.

The researchers said these are associations that need more study.

“Existing work has largely focused on predicting individual risk factors or disease states in isolation, without systematically examining how retinal structural patterns relate to a broader spectrum of Alzheimer’s disease-related risk domains,” Fang and colleagues wrote.

Retinal photographs are already routinely taken for people with diabetes, glaucoma or cataracts. Fang said the images may act as a broad biological measure of risk.

“In this sense,” says Fang, “retinal imaging functions less as a surrogate questionnaire and more as an integrated biological sensor of cumulative risk.”

The study is published in the Journal of Alzheimer’s Disease.

Read more from Science Alert.