Identifying high glucose variability using non-glycemic factors in low continuous glucose monitoring use settings.

Abstract

Aims: Identifying non-glycemic factors associated with high Glucose variability (GV).

Methods: A cross-sectional observational study recruited people with type 2 diabetes, who wore a Freestyle Libre Pro CGM.

Independent variables: Age, sex, BMI, diabetes medication, diabetes duration, HbA1c and estimated glomerular filtration rate (eGFR). CGM-derived variables calculated included Time-in-Range (TIR, 70-180 mg/dl), below-range 1 (TBR1, <70 mg/dl), -below-range 2 (TBR2, <54 mg/dl) and -above-range (TAR, >180 mg/dl), coefficient of variation (%CV). A logistic regression model examined independent variables associated with high GV (CV ≥36 %). All analysis was done on R version 4.3.1 RESULTS: T2D cohort (n = 403), 46 % women, had median age of 61 y, BMI of 26.5 kg/m², diabetes duration 14 y, HbA1c 7.8 %(62 mmol/mol) and creatinine of 75 µmol/L. Using sulphonylurea, premixed or basal-bolus insulin had an odds ratio (OR) of 4.7 - 5.2 for CV ≥ 36 %. Longer diabetes duration [OR 1.2], and lower eGFR [OR 1.2] were associated with higher odds and older age [OR 0.8]and higher BMI [0.8] were associated with lower odds of CV≥ 36 %. Sex and HbA1c had no association with high GV.

Conclusion: Nonglycemic-factors like medication type, diabetes duration and eGFR can aid in identification of high GV even in low-CGM use settings.