The miscibility of human semisynthetic regular and lente insulins (Actrapid human/Monotard human) and human biosynthetic regular and NPH insulins (Humulin regular/Humulin NPH) was studied in vitro and in 16 insulin-dependent diabetic patients. In vitro a decrease of regular insulin was found immediately after mixture with semisynthetic lente insulin in proportions 1:4, 1:3, 1:2, 1:1, and 2:1. This was not found after mixture with human biosynthetic regular and NPH insulins. Free insulin and blood glucose were determined over 8 h after the morning injection of regular/lente (n = 9) or regular/NPH (n = 7) insulins in proportions used by the patients (approximately 1:2) mixed in one syringe or in separate syringes. Mixing the regular/lente preparations in one syringe immediately before injection resulted in a significant loss of the early rise in free insulin (P less than .01), with a tendency to a higher free insulin in the afternoon. A corresponding change in the blood glucose profile was seen. No changes were observed in free insulin or blood glucose after mixing regular and NPH preparations compared with separate injections. We conclude that mixing human semisynthetic regular and lente insulins (Actrapid human/Monotard human) instantly results in a decrease of regular insulin, probably due to formation of a longer-acting preparation, whereas no such changes occur with human biosynthetic regular and NPH insulins (Humulin regular/Humulin NPH).

Human ultralente insulin gives an insulin profile suitable for overnight substitution, but the great day-to-day variability limits its usefulness. It can be injected before dinner or at bedtime without any change in the insulin profile during the night.

Ten insulin-dependent diabetic patients were investigated from 2100 to 0700 h during treatment with either a bedtime injection (BI) of intermediate-acting insulin or continuous subcutaneous insulin infusion (CSII) at a constant basal rate. In the evening, blood glucose was slightly higher during treatment with BI than with CSII, whereas the metabolic control in the morning was equal on both regimens with a fasting blood glucose of 5.7 mM (4.2-7.1) (median and interquartile ranges) on BI and 5.4 mM (4.6-5.8) on CSII (NS). No rise in morning blood glucose was seen, but serum beta-hydroxybutyrate tended to rise (NS). There was a significant hyperinsulinemia at midnight during BI compared with CSII with a serum free insulin of 14.5 (11.7-16.0) vs. 9.6 (7.2-11.2) mU/L (P less than .05), respectively, and the area under the curve during the middle of the night (midnight to 0400 h) was greater with BI than CSII (P less than .02). A greater fall in blood glucose was seen with BI than with CSII during this period (P less than .02). Differences in blood glucose and serum free-insulin profiles between those using NPH or lente insulin at bedtime were registered. We conclude that, although the same metabolic control in the morning was achievable with CSII at a constant basal rate and BI, CSII is superior for overnight metabolic control due to less-pronounced hyperinsulinemia during the night and a steady-state level of free insulin in the morning.