Non-invasive biomarkers for diabetic complications: insights from corneal and retinal imaging.

Background: Diabetes-related complications, such as diabetic peripheral neuropathy (DPN) and chronic kidney disease (CKD), severely affect quality of life. Early detection is crucial. This study investigates ocular imaging parameters as potential biomarkers for these conditions using corneal confocal microscopy (CCM) and optical coherence tomography angiography (OCTA).

Methods: This cross-sectional study included 76 type 2 diabetes patients (139 eyes) from Fujian Medical University Union Hospital. Participants underwent CCM to assess corneal nerve fiber density (CNFD), branching density (CNBD), and nerve fiber length (CNFL). OCTA and OCT were used to evaluate macular and peripapillary retinal vascular densities (VD) and retinal nerve fiber layer (RNFL) thickness. Laboratory tests measured sural nerve conduction velocity (SSNCV), urine albumin-to-creatinine ratio (UACR), and estimated glomerular filtration rate (eGFR). DPN and CKD were categorized using Toronto consensus criteria and UACR thresholds, respectively. Statistical analyses included Spearman correlation and ROC curve evaluations.

Results: Significant reductions in CNFD, CNBD, and CNFL were observed in the DPN + group compared to DPN- (P < 0.001, P = 0.005, P < 0.001). Corneal nerve parameters correlated positively with SSNCV (r = 0.419-0.430, P < 0.001). ROC analysis demonstrated CNFD as the most sensitive marker for detecting DPN (AUC = 0.7179, 95% CI: 0.6328-0.8031). Retinal superficial VD in the superior macular region showed the highest diagnostic performance for CKD (AUC = 0.7140, 95% CI: 0.6057-0.8223), with significant correlations between retinal VD parameters and UACR.

Conclusions: Corneal nerve parameters measured by CCM and retinal vascular parameters assessed by OCTA are promising non-invasive biomarkers for early detection and monitoring of diabetic neuropathic and microvascular disorders.