ABSTRACT
BACKGROUND AND METHODS: We investigated the factors that determine the best glycaemic control on multiple daily insulin (MDI) injections and continuous subcutaneous insulin infusion (CSII), and the hypothesis that blood glucose variability is a major determinant of control and that the resultant HbA(1c) on MDI correlates with the improvement achieved by CSII. We studied 30 type 1 diabetic subjects already receiving MDI. Renewed attempts to improve control on MDI were made for a median of five months, and then the subjects were switched to CSII. The variability of within-day and between-day blood glucose concentrations was calculated from blood glucose self-monitoring data. RESULTS: HbA(1c) during MDI varied from 5.7 to 11.7% (mean +/- SD, 8.5 +/- 1.4%). Within- and between-day blood glucose variability correlated with HbA(1c) on MDI (r = 0.59, p < 0.001; r = 0.48, p < 0.03). Within-day variability remained an independent predictor of HbA(1c) on MDI. Mean HbA(1c) improved with CSII (to 7.3 +/- 0.9%, p < 0.001), but reduction in HbA(1c) was variable and was related to the HbA(1c) on MDI (r = 0.79, p < 0.001) and within-day variability (r = 0.56, p < 0.01). Similar results were observed for subjects treated only with glargine-based MDI. CONCLUSIONS: The best glycaemic control achievable on MDI is related to blood glucose variability-those with the largest swings in blood glucose retaining the highest HbA(1c). The improvement in control achieved by CSII is related to HbA(1c) and blood glucose variability on MDI. Pump therapy is most effective in those worst controlled on MDI.
