Key Takeaways

  • Common metabolic measures—including BMI, blood pressure, triglycerides, and waist-to-hip ratio—can help predict future risk of retinopathy, even in people without diabetes at baseline
  • A predictive model incorporating both baseline and follow-up data showed significantly better accuracy, supporting its use for individualized risk assessment
  • The findings support a broader concept of “metabolic retinopathy,” suggesting screening strategies should extend beyond diabetes to include patients with wider metabolic risk factors

A new prospective cohort study from China suggests that routine health measurements—such as blood pressure, body weight, and cholesterol levels—could be used to accurately predict an individual’s risk of developing retinopathy, offering a potential pathway toward more personalized screening.1

The study, published in Diabetes & Metabolic Syndrome, followed 2,447 adults from the SENSIBLE cohort who did not have retinopathy at the start of the study. Participants represented a mix of metabolic health states, including normal glucose levels, prediabetes, and diabetes.

Over the course of the follow-up period, 5.9% of participants developed retinopathy, a condition that can lead to vision loss if left untreated. Researchers found that several common clinical measures were strongly associated with future retinopathy risk, including:

  • Body mass index (BMI)
  • Waist-to-hip ratio
  • Triglyceride levels
  • Blood pressure (both systolic and diastolic)
  • History of hypertension
  • Ethnicity

These findings suggest that retinopathy risk is closely tied to broader metabolic health, not just blood sugar levels. Using these factors, the research team developed two predictive models, known as nomograms. One used baseline data alone, while the other incorporated both baseline and follow-up information.

The more comprehensive model showed significantly better predictive performance, with an area under the curve (AUC) of 0.75 compared to 0.64 for the baseline-only model—indicating improved accuracy in identifying individuals at higher risk. Importantly, the model achieved a balance between sensitivity and specificity, meaning it could identify at-risk individuals without generating excessive false positives.

The study’s authors say the findings could help clinicians move toward risk-based screening strategies, identifying patients who may benefit from earlier or more frequent eye exams. Because the model relies on widely available clinical data, it could be easily integrated into routine care without requiring specialized testing.

Beyond its predictive tool, the study introduces the concept of “metabolic retinopathy,” emphasizing that retinal disease may arise from a combination of metabolic risk factors—not solely from diabetes.

Reference

1. Xu X, Wang D, Alam U, et al. Prediction of retinopathy risk: A prospective cohort study in China. Diabetes Metab Syndr. 2025 May;19(5):103251. doi: 10.1016/j.dsx.2025.103251.