Role of Systemic Factors in the Progression to Vision‐Threatening Diabetic Retinopathy in a New Mexican Patient Population with Type 2 Diabetes: Glycemic Control Is Not Enough

TL;DR

This study finds that glycemic control alone is insufficient to predict vision-threatening diabetic retinopathy in a New Mexican population with type 2 diabetes.

Contribution

The study identifies systemic factors beyond glycemic control that influence diabetic retinopathy progression in a diverse ethnic population.

Findings

Glycemic control (HbA1c) poorly predicts progression to proliferative diabetic retinopathy or diabetic macular edema.

Microalbuminuria, macroalbuminuria, Hispanic ethnicity, diabetes duration, and medication use are significant predictors of diabetic retinopathy progression.

ROC analysis showed low predictive accuracy of HbA1c for diabetic retinopathy outcomes (AUC = 0.640).

Abstract

The purpose of this study is to identify associative factors of vision threatening complications of diabetic retinopathy (DR), including proliferative diabetic retinopathy (PDR) and diabetic macular edema (DME) in a New Mexican patient population with Type 2 diabetes featuring a high proportion of Hispanic and American Indian patients.

In a retrospective, case‐control, cross‐sectional study, we performed both univariate and multivariate logistic regression testing of systemic factors for association among patients with nonproliferative diabetic retinopathy (NPDR) without DME, NPDR with DME, and PDR. We also used receiver operating characteristic (ROC) curves to understand how reliable glycemic control is at predicting DR end‐complications.

Among our 584 patients, 172 were diagnosed with NPDR without DME, 293 with PDR, and 119 with NPDR with DME. A total of 25% of patients with NPDR without DME had poor glycemic control (HbA1c levels of ≥ 9.0%). Similarly, 10% of PDR and 7% of NPDR with DME patients had good glycemic control (HbA1c levels ≤ 7.0%), despite their advanced eye disease. For patients with PDR, we identified significant independent associations with microalbuminuria and macroalbuminuria, Hispanic ethnicity, duration of diabetes, and the use of insulin and beta‐blocker medications. In NPDR with DME patients, significant associations were noted with microalbuminuria and macroalbuminuria, HbA1c, GFR, and beta‐blocker medications. The hemoglobin A1c level was not significantly associated with PDR but was significantly associated with NPDR with DME. The ROC curve analysis indicated poor predictability for the HbA1c model concerning the presence of PDR (AUC = 0.640) and NPDR with DME (AUC = 0.640).

This study suggests that glycemic exposure alone is not a good enough predictor of DR progression. Other diabetic microvascular complications such as nephropathy, in addition to race/ethnicity with differing genetic associative factors, may also play a role in the multifactorial progression through DR phenotypes.

Figures