AIMS: The purpose of this secondary analysis of the StenoABC Study was to identify determinants of the changes in HbA1c observed after training of people with type 1 diabetes in advanced carbohydrate counting (ACC) and automated bolus calculator (ABC) use, and further to investigate psychosocial effects of these insulin dosing approaches.

METHODS: Validated diabetes-specific questionnaires were used to assess diabetes treatment satisfaction, problem areas in diabetes, fear of hypoglycemia and diabetes dependent quality of life before and one year after the training. In addition, numeracy was tested (using a non-validated test developed specifically for this study) and behavioral measures (number of daily blood glucose measurements and self-reported use of ACC) were obtained. Associations between change in HbA1c and these measures plus sex, age, diabetes duration and BMI were tested.

RESULTS: Numeracy was the only baseline predictor of yearly change in HbA1c identified. Higher levels of numeracy were associated with greater reductions in HbA1c (P=0.031). No associations between change in HbA1c and the behavioral measures investigated were found, nor were any clinically relevant associations between changes in HbA1c and questionnaire scores. Treatment satisfaction increased in all users of ACC (P<0.001). People who also used an ABC reported significantly lower levels of fear of hypoglycemia than people who practiced ACC without such device (P=0.005).

CONCLUSIONS: Improvements in HbA1c after training in ACC were inversely related to numeracy. Use of an ABC did not compensate for poor numeracy skills. However, device use reduced fear of hypoglycemia compared with ACC without ABC use.