Skip to main content

Corneal endothelial changes in type 2 diabetes mellitus: a systematic review and meta-analysis with subgroup comparisons by diabetic retinopathy status and disease duration

In brief

Across 5,878 eyes, diabetes linked to poorer corneal cells, especially with retinopathy

A meta-analysis of 22 studies found people with type 2 diabetes had fewer corneal endothelial cells and less regular cell shape than controls; changes were greater in eyes with diabetic retinopathy and with diabetes lasting at least 10 years. Several findings were sensitive to individual studies or publication bias, so the case for preoperative assessment is plausible but not definitive.

Journal
Frontiers in endocrinology (Q1)
Published
29 September 2026
Study design
Systematic review of cohort studies
Evidence level
Level 2, Moderate (CEBM 2a)
Authors
Yongran Li, Chenlin Li, Zhimin Cen, Dingwang Su, Tulong Lin
PMID
42870909
DOI
10.3389/fendo.2026.1982467

Why clinicians should know about it

  • Picked for Ophthalmology (paper of the day, 19 October 2026): T2DM associated with lower endothelial cell density, higher CV

Abstract

BACKGROUND: Type 2 diabetes mellitus (T2DM) affects multiple ocular structures, yet the extent of corneal endothelial alterations across different stages of diabetic retinopathy (DR) and disease duration remains unclear. This meta-analysis quantified corneal endothelial changes in T2DM across subgroups defined by DR status and diabetes duration. METHODS: This systematic review adhered to PRISMA and MOOSE guidelines (PROSPERO: CRD420261473313). Five databases were searched from inception through July 15, 2026. Cross-sectional studies comparing endothelial cell density (ECD), coefficient of variation (CV), and percentage of hexagonal cells (HEX) in T2DM patients versus controls or between T2DM subgroups were included. Study quality was assessed using the AHRQ checklist. Effect sizes were calculated as Cohen's d using fixed-effect or random-effects models based on heterogeneity. Sensitivity analyses used the leave-one-out method; publication bias was assessed with funnel plots, Egger's test, and trim-and-fill analysis. RESULTS: Twenty-two studies (5,878 eyes) were included. T2DM patients showed significantly lower ECD (d = -0.35, P < 0.001), higher CV (d = 0.25, P < 0.001), and lower HEX (d = -0.28, P < 0.001) versus controls. Eyes with DR had lower ECD (d = -0.22, P = 0.002), higher CV (d = 0.26, P = 0.005), and lower HEX (d = -0.23, P = 0.02) versus eyes without DR. No significant differences emerged between PDR and NPDR. Longer disease duration (≥10 years) was associated with lower ECD (d = -0.39, P = 0.04) and lower HEX (d = -0.18, P = 0.003). Sensitivity analyses showed that HEX (T2DM vs. controls), CV (PDR vs. NPDR), and CV (duration comparison) were not robust; trim-and-fill reversed the CV duration result. CONCLUSIONS: T2DM is associated with impaired corneal endothelial morphology, with greater changes in the presence of DR and longer duration. Several estimates were sensitive to individual studies or publication bias, warranting cautious interpretation. Preoperative endothelial assessment may be beneficial in diabetic patients, particularly those with DR or long disease duration. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/view/CRD420261473313, identifier CRD420261473313.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.