Can continuous glucose monitoring-derived glycemic variability predict retinal neurodegeneration earlier than HbA1c in patients without clinically apparent retinopathy?

Early screening for diabetic retinopathy (DR) has long relied on the detection of fundus vascular lesions. However, accumulating evidence indicates that retinal neurodegeneration may occur before overt microvascular abnormalities become clinically visible. At the stage of no clinically apparent diabetic retinopathy (NDR), glycated hemoglobin A1c (HbA1c), by virtue of its averaging nature, is unable to adequately reflect the acute metabolic stress imposed by short-term glycemic variability (GV) on the retinal neurovascular unit (NVU). The increasing use of continuous glucose monitoring (CGM) has made dynamic glycemic metrics such as time in range (TIR) and coefficient of variation (CV) clinically accessible. This review summarizes the clinical and mechanistic evidence supporting CGM-derived GV metrics in the association with early retinal injury in patients with NDR. Mechanistic studies indicate that fluctuating hyperglycemia induces mitochondrial oxidative stress more readily than sustained hyperglycemia and can activate Müller cell- and microglia-mediated neuroinflammation while disrupting tight junctions within the blood-retinal barrier. Clinical studies further show that reduced TIR and elevated CV are associated with early thinning of the retinal nerve fiber layer and reduced microvascular density, and some of these associations remain significant after adjustment for HbA1c. On this basis, we propose a dual-dimensional risk assessment strategy combining HbA1c with TIR/CV and suggest that patients with NDR who exhibit marked glycemic variability on CGM should undergo intensified monitoring. Integrating dynamic CGM metrics with optical coherence tomography (OCT) and OCT angiography (OCTA) may enable risk stratification during the NDR stage and shift the intervention window for DR forward to a critical phase for neuroprotection.