What Is Retinal Age?
Retinal age is an estimate of age generated from a photograph of the retina using artificial intelligence.
The retina is the light-sensitive tissue at the back of the eye. A standard retinal photograph captures detailed information about its blood vessels, optic disc and surrounding tissue.
Artificial-intelligence models can analyse patterns within these images and learn which features tend to change with chronological age.
The result is an AI-predicted retinal age.
For example, a 60-year-old person might receive a predicted retinal age of 64.
This does not mean that the person has suddenly become 64 years old biologically. It means that, according to that particular algorithm, features within the retinal image resemble those typically associated with an older age.
What Is the Retinal Age Gap?
The most interesting information may come from comparing retinal age with actual chronological age.
Researchers call this difference the retinal age gap:
Retinal Age Gap = Retinal Age − Chronological Age
If a 60-year-old has a retinal age of 65, the retinal age gap is +5 years.
If the predicted retinal age is 56, the gap is −4 years.
A positive gap therefore means that the retina appears older than expected for chronological age, while a negative gap means that it appears younger.
Why Could This Difference Matter?
Researchers have investigated whether retinal age gap is associated with health outcomes.
In a large UK Biobank study, more than 35,000 adults were followed for approximately 11 years.
People with larger positive retinal age gaps tended to have higher all-cause mortality during follow-up, even after researchers accounted for factors including smoking, physical activity, general health and several major medical conditions.
The relationship was not perfectly linear. A retinal age gap should therefore not be interpreted as a simple scale in which every additional retinal year produces exactly the same increase in risk.
What the study showed is an association: an older-appearing retina was linked with higher subsequent mortality at the population level.
It did not show that an older retinal age causes shorter life.
Is Retinal Age Your Biological Age?
Retinal age can provide information relevant to biological aging, but the number needs context.
It is derived from one tissue — the retina — using a specific artificial-intelligence model.
The body does not necessarily age uniformly. Blood vessels, brain, heart, muscles and metabolic systems can show different patterns of age-related change.
This means that a retinal age of 65 should not simply be translated into:
“My biological age is 65.”
A more accurate interpretation is:
“The age-related features detected in my retinal image resemble those the algorithm associates with approximately age 65.”
Whether that information improves assessment of an individual’s overall biological aging is an important area of continuing research.
Can Retinal Age Change?
This is an important question, but current evidence does not yet give us a simple answer.
A single retinal photograph provides a snapshot at one point in time.
To determine whether someone’s retinal age is changing faster or slower, retinal imaging would need to be repeated over time using a validated and consistent method.
It also remains to be established whether lifestyle changes or medical treatment can meaningfully alter retinal age — and whether such a change would translate into better long-term health.
For this reason, retinal age should not currently be treated like a score that needs to be “lowered.”
How Should Retinal Age Be Interpreted?
Retinal age is most useful when understood as additional information, rather than as a diagnosis or a definitive biological-age score.
Its potential strength is that retinal photography is fast, non-invasive and capable of capturing detailed information about vascular and neural tissue.
As research develops, retinal age may contribute to a broader assessment that combines information from different aspects of health and aging.
The number itself is only part of the picture. What it predicts, how consistently it performs across different populations, and whether changes over time reflect meaningful changes in health are ultimately more important.
