Eyes, the high-resolution biological devices that help us visualize the outside world, are now being used as a portal to assess our internal health. Scientists have found that a closer evaluation of how one’s retina is aging can provide crucial hints about bone health, especially in conditions such as osteoporosis, which makes bones weaker and more prone to fractures.
A recent study conducted in Singapore and the UK collected over 45,000 retinal images and used an artificial intelligence (AI) tool called RetiAGE to estimate a person’s retinal biological age. When researchers compared retinal age with bone mineral density, they found an inverse relationship between the two.
Among participants in Singapore, people with older-looking retinas tended to have lower bone mineral density and higher fracture risk scores. Meanwhile, the UK-based cohort, where participants were studied for over a decade, revealed that a higher retinal biological age at the start of the study was a predictor for a greater chance of developing osteoporosis by the end of it.
The findings are published in PLOS Digital Health.
Overburdened yet underdiagnosed
Osteoporosis is a degenerative bone disease that causes bones to lose mass and density, while also altering the internal architecture that gives them strength. Over time, these changes weaken the bones, making them increasingly prone to fractures. Today, one in five people lives with osteoporosis, and the risk of developing the condition rises significantly with age.
Statistics from the International Osteoporosis Foundation highlight the staggering impact osteoporosis has on both people and health care systems worldwide.
Hip fractures due to osteoporosis have the most dangerous consequences, with mortality rates reaching 20–24% within the first year. Even among survivors, the toll is severe: nearly 40% lose the ability to walk independently, while one in three becomes fully dependent on nursing home care within a year.
Despite the seriousness, this condition remains highly underdiagnosed, as many people remain asymptomatic up until a bone breaks for the first time. Diagnosing osteoporosis today largely relies on measuring bone mineral density (BMD) through Dual-energy X-ray Absorptiometry (DEXA) scans, a test that is out of reach for most people, both in terms of finance and access.
Eyes give a glimpse of within
The researchers behind the study wanted to find a non-invasive way to predict osteoporosis risk, leading them to explore the retina—a part of the eye already known to reveal clues about several systemic conditions, including cardiovascular and kidney diseases. To study if the retina can reflect how the bones are doing, they collected data from two large study groups.
The first was the Singapore PIONEER study that looked at 1,965 older adults and compared their retinal age to their actual bone density, measured using DEXA scans. The second was a long-term, large-scale UK Biobank study that took retinal photos of 43,938 participants at baseline and then tracked their medical records for 12 years to see whether they developed osteoporosis.
The AI tool RetiAGE, previously trained on over 100,000 eye images to recognize patterns of aging, was then used to predict the participants’ retinal age.
On analyzing both the parameters, the data indicated that the retina’s biological age is a good predictor of the onset of osteoporosis years in advance.
The researchers identified that participants whose eyes looked older tended to have lower bone mineral density, faced a higher risk of osteoporosis, and were more likely to suffer major bone or hip fractures later in life. This association was consistent for both men and women.
They also found that adding RetiAGE analysis to existing tools improved their ability to screen people at risk for osteoporosis and fracture. The C-index, which is a measure of the predictive power, increased from 0.585 to 0.635.
Retinal imaging is simple, quick, and non-invasive, making it a promising low-cost tool for screening large populations for osteoporosis risk before painful fractures occur. However, the approach still has limitations. Since retinal structure and appearance can vary across ethnicities and eye conditions, the AI model, which was trained on the Korean population, may not work equally well for everyone.
To ensure the wider use of this technique, future studies will need to train the AI tool using data from more diverse populations.
Written for you by our author Sanjukta Mondal, edited by Sadie Harley, and fact-checked and reviewed by Robert Egan—this article is the result of careful human work. We rely on readers like you to keep independent science journalism alive.
If this reporting matters to you, please consider a donation (especially monthly). You’ll get an ad-free account as a thank-you.
Publication details
Qingsheng Peng et al, Retinal biological age correlates with bone mineral density and fracture risk score and predicts incident osteoporosis, PLOS Digital Health (2026). DOI: 10.1371/journal.pdig.0001360
Journal information:
PLOS Digital Health
The content is provided for information purposes only.