The association between estimated glucose disposal rate and self-reported diabetic retinopathy: evidence from two independent cohorts and machine learning - Report - MDSpire

The association between estimated glucose disposal rate and self-reported diabetic retinopathy: evidence from two independent cohorts and machine learning

  • By

  • Chaofeng Yuan

  • Yue Hao

  • Jianghui Wang

  • Chuanxi Wang

  • Zhengxuan Jiang

  • June 26, 2026

  • 0 min

Share

Clinical Report: Examining the Link Between Estimated Glucose Disposal Rate and Self-Reported Diabetic Retinopathy

Overview

This study investigates the relationship between estimated glucose disposal rate (eGDR) and the prevalence of self-reported diabetic retinopathy (DR) using data from NHANES and a clinical cohort. Findings indicate a significant negative correlation between eGDR and DR prevalence.

Background

Diabetic retinopathy is a leading cause of vision impairment among individuals with diabetes, affecting approximately 160 million people globally. Insulin resistance is a critical factor in the pathogenesis of DR, yet the association between eGDR and DR prevalence has not been fully characterized.

Data Highlights

MeasureValue
Participants (NHANES)1,536
Participants (Clinical Cohort)297
Odds Ratio (eGDR and DR prevalence)0.79
95% Confidence Interval0.67–0.93
P-value0.0049
XGBoost AUC0.773
Random Forest AUC0.764

Key Findings

  • eGDR shows a significant negative correlation with self-reported DR prevalence (OR = 0.79, P = 0.0049).
  • Subgroup and sensitivity analyses confirm the stability of the negative association between eGDR and DR.
  • The Boruta algorithm identifies eGDR as a robust feature in predicting DR prevalence.
  • XGBoost and random forest models demonstrate moderate predictive performance for DR prevalence estimation.
  • SHAP analysis indicates eGDR, body mass index, and income poverty as key determinants of self-reported DR prevalence.

Clinical Implications

The findings indicate a significant negative correlation between eGDR and DR prevalence.

Conclusion

This study highlights the association between lower eGDR and higher prevalence of self-reported diabetic retinopathy. Further prospective research is needed to explore the causal relationship between these factors.

Related Resources & Content

  1. Frontiers in Medicine, 2026 -- Detection of Referable Diabetic Retinopathy using Machine Learning on Routine Clinical Data
  2. conexiant, 2026 -- Can Diabetic Eye Testing Be Simplified?
  3. Frontiers in Endocrinology, 2026 -- A clinically interpretable machine learning model for early detection of diabetic retinopathy in multiple community health centers
  4. Standards of Care in Diabetes | ADA Clinical Guidelines, 2026
  5. Frontiers in Endocrinology — Development of a risk stratification tool for rapidly progressive diabetic retinopathy in type 2 diabetes
  6. Standards of Care in Diabetes | ADA Clinical Guidelines
  7. Effect of Intensive Diabetes Therapy on the Progression of Diabetic Retinopathy in Patients With Type 1 Diabetes: 18 Years of Follow-up in the DCCT/EDIC - PMC
  8. Frontiers | Estimated glucose disposal rate and severe abdominal aortic calcification: evidence from a nationally representative study with external validation

Original Source(s)

Related Content