Prevent hypoglycaemia when using automated insulin delivery systems in type 1 diabetes requires near normal glycaemic...

Although newer technologies of insulin delivery in type 1 diabetes have facilitated an improvement in glycaemic control the risk of hypoglycaemia remains a threat. Therefore, it is important to define the thresholds of glycaemic variability below which the risk of hypoglycaemia can be eliminated or at least minimized.

Randomized controlled trials conducted from 2017 to 2023 comparing Sensor-Augmented-Pumps and Augmented Insulin Delivery Systems (n = 16 and 22 studies, respectively) were selected. A weighted linear model of regression was used to compute the relationship between glycaemic variability and times spent below glucose range. The intercepts of regression lines with the abscissa axis (time below range = 0 %) defined the glycaemic variability thresholds.

Positive relationships were observed between the 2 metrics. The scatter plots indicated that the times spent below range never reached the value of 0 % and that the glycaemic variability never fell below 28 %. By extrapolating the regression lines, the glycaemic variability at intercepts with time below range < 70 mg/dL of 0 % was 30.1 % with sensor augmented pumps and 18.9 % with automated insulin delivery. For a time below range < 54 mg/dL of 0 % the respective glycaemic variability values were 32.7 % and 19.9 % (with sensor augmented pumps and automated insulin delivery, respectively).

Importantly, glycaemic variability targets and ambient hyperglycaemia are interdependent. Users of automated insulin delivery need to reach a glycaemic variability of 18 % to 20 % to minimize or eradicate the risk of hypoglycaemia. Such values are those observed in healthy non-diabetic people.