Relationship between time in range and corneal nerve fiber loss in asymptomatic patients with type 2 diabetes

Background:Corneal confocal microscopy (CCM) is a noninvasive technique to detect early nerve damage of diabetic sensorimotor polyneuropathy (DSPN). Time in range (TIR) is an emerging metric of glycemic control which was reported to be associated with diabetic complications. We sought to explore the relationship between TIR and corneal nerve parameters in asymptomatic patients with type 2 diabetes (T2DM).Methods:In this cross-sectional study, 206 asymptomatic inpatients with T2DM were recruited. After 7 days of continuous glucose monitoring, the TIR was calculated as the percentage of time in the glucose range of 3.9 to 10.0 mmol/L. CCM was performed to determine corneal nerve fiber density, corneal nerve branch density, and corneal nerve fiber length (CNFL). Abnormal CNFL was defined as <= 15.30 mm/mm(2).Results:Abnormal CNFL was found in 30.6% (63/206) of asymptomatic subjects. Linear regression analyses revealed that TIR was positively correlated with CCM parameters both in the crude and adjusted models (all P < 0.05). Each 10% increase in TIR was associated with a 28.2% (95% CI: 0.595-0.866, P = 0.001) decreased risk of abnormal CNFL after adjusting for covariates. With the increase of TIR quartiles, corneal nerve fiber parameters increased significantly (all P for trend <0.01). The receiver operating characteristic curve indicated that the optimal cutoff point of TIR was 77.5% for predicting abnormal CNFL in asymptomatic patients.Conclusions:There is a significant independent correlation between TIR and corneal nerve fiber loss in asymptomatic T2DM patients. TIR may be a useful surrogate marker for early diagnosis of DSPN.