Predicting Systemic Disease

Every day, we capture some of the highest-resolution images of living neural tissue and microvasculature to diagnose and monitor ocular disease. There is growing evidence that ocular images contain measurable information associated with brain health, cardiovascular and kidney function, metabolic health, and biologic aging. AI is increasingly enabling these signals to be used for systemic disease detection and risk prediction.^1-4 This evolving field of practice is called oculomics, and this article discusses how the eye allows for this information, recent supporting studies, and how to implement it.

WHY THE EYE

The eye occupies a unique position in medicine because it allows clinicians to directly visualize living neural tissue and microvasculature without surgery or invasive testing.

Specifically, imaging captures retinal and corneal architecture at near histologic resolution, showing distinct biomarkers that reveal neural, vascular, and structural health.

As an example, alterations in retinal vessel density, capillary perfusion, vascular caliber, and branching geometry have been shown to correlate with cardiovascular disease, hypertension, chronic kidney disease, diabetes, and generalized endothelial dysfunction.^1,5,6 What’s more, these vascular changes often present before patients develop overt ocular manifestations of systemic disease.

Additionally, large population-based investigations show that retinal imaging may aid in identifying biomarkers associated with Alzheimer disease and Parkinson disease, among other neurodegenerative disorders, years before patients receive a formal diagnosis.^7,8

Essentially, the eye is an accessible, noninvasive biologic sensor capable of revealing some of the earliest signatures of human health with the help of technology.

Deep-learning models analyze ocular images as extraordinarily complex datasets containing millions of individual pixels and mathematical relationships. Rather than searching for predefined abnormalities, they identify statistical patterns that correlate with systemic characteristics, disease risk, and, in some cases, the future development of disease.^1,2

One study shows that deep-learning AI models were able to predict cardiovascular risk factors (age, gender, smoking status, systolic blood pressure, and major adverse cardiac events) from retinal images after the models were trained on data from 284,335 patients and validated on two independent datasets of 12,026 and 999 patients, respectively.² Additionally, the deep-learning AI models used anatomic features, such as the optic disc or blood vessels, to make each prediction.

This study demonstrates that retinal images act as high-dimensional data containing “latent” biomarkers for cardiovascular health—signals that are invisible to the human eye but clear to algorithms.²

Perhaps the most compelling concept in recent literature is the “retinal age gap.” By using deep learning to estimate a patient’s “retinal age” and comparing it with their chronological age, researchers found a strong correlation with mortality. A larger retinal age gap, in which the retina looks “older” than the patient, was associated with increased all-cause mortality risk. Each additional year of retinal aging reflected cumulative vascular and metabolic stress rather than overt ocular pathology. This suggests the retina acts as a cumulative “odometer” for the body, recording the wear and tear of vascular and metabolic stress over time.⁴

PREPARATION

Much work remains before oculomics becomes part of routine clinical practice.

Many biomarkers require further validation, prospective studies are ongoing, and appropriate clinical guidelines will continue to evolve. That said, the science clearly tells us that the future practice of optometry will include identifying the earliest ocular signs of systemic health, collaborating with all our medical colleagues, and counseling patients long before disease becomes clinically apparent. To prepare for this change in practice, I recommend following these four steps:

1. Look Beyond the Eye. When reviewing fundus photographs or OCT scans, switch your mindset to ask, “Could these findings reflect something occurring elsewhere in the body?” Not every unusual retinal finding represents glaucoma, diabetic retinopathy, or AMD. Oculomics reminds us that some imaging abnormalities may reflect broader vascular, neurologic, or metabolic processes that warrant a more comprehensive differential diagnosis.

2.Identify Patterns. Oculomics is less about diagnosing Alzheimer disease or cardiovascular disease from an OCT scan than it is about recognizing when retinal findings, as an example, do not fit the expected clinical picture. Diffuse retinal nerve fiber layer thinning in the presence of minimal optic nerve cupping, generalized vascular attenuation without obvious ocular pathology, or unexplained microvascular changes should prompt thoughtful clinical consideration rather than premature conclusions.

3.Strengthen Documentation and Interdisciplinary Communication. Objective ocular imaging provides an opportunity to communicate meaningful findings to primary care physicians and medical specialists using clear, descriptive language. Rather than documenting “normal OCT” or “mild vascular changes,” consider describing the specific imaging findings and recommending appropriate systemic correlation when clinically indicated. Thoughtful documentation strengthens referrals, while remaining well within optometric scope of practice.

4. Remain Intellectually Curious. Oculomics is advancing at an extraordinary pace. New biomarkers, imaging technologies, and AI applications continue to emerge, and many of today’s research discoveries are likely to become tomorrow’s clinical standards. Optometrists who understand these developments early will be well-positioned to responsibly integrate them into patient care as the evidence continues to evolve.

IN OUR WHEELHOUSE

Optometrists are uniquely positioned to lead this field: We possess the skillset, widespread infrastructure, imaging technologies, and patient volume. As technology evolves, the optometric exam will expand from saving sight, to potentially extending lives through early, noninvasive risk recognition. Get ready: As primary eyecare providers, we will no longer just look at and inside the eye; we will look through it.