Carbohydrate counting accuracy and blood glucose variability in adults with type 1 diabetes

Abstract Aims Carbohydrate counting is an established approach used by patients with type 1 diabetes to improve their glycemic control. The aims of this study were to evaluate, in real life conditions, the accuracy of meal carbohydrate estimate and its impact on glycemic variability. Methods In this cross-sectional study, we observed the ability of 50 adults (48% women) with type 1 diabetes (age: 42.7 ± 11.1 years); diabetes duration: 21.4 ± 12.7 years; HbA1c: 7.2 ± 1.2% (60 ± 10 mmol/mol) to accurately estimate carbohydrates by analyzing 72-h food records and their corresponding 72-h blood glucose excursions using a continuous glucose monitor. Results The mean meal carbohydrate difference, between the patients’ estimates and those assessed by a dietitian using a computerized analysis program, was 15.4 ± 7.8 g or 20.9 ± 9.7% of the total CHO content per meal (72.4 ± 34.7 g per meal). Sixty-three percent of the 448 meals analyzed were underestimated. Greater differences in CHO's estimates predicted higher glycemic variability, as measured by the MAGE index and glucose standard deviation, and decreased time with glucose values between 4 and 10 mmol/L ( R2 = 0.110, 0.114 and 0.110, respectively; P < 0.05). Conclusion Inaccurate carbohydrate counting is frequent and associated with higher daily blood glucose variability in adults with type 1 diabetes.