Background: Intensive insulin treatment of type 1 diabetes mellitus increases the risk for nocturnal hypoglycemia.
Objective: To demonstrate that splitting the evening insulin regimen reduces the risk for nocturnal hypoglycemia in intensive treatment of type 1 diabetes mellitus.
Design: Randomized, open, two-treatment crossover trial in two 4-month periods.
Setting: University research center in Italy.
Patients: 22 C-peptide-negative persons with type 1 diabetes mellitus (mean age [+/-SD], 29 +/- 3 years).
Interventions: Each patient was randomly assigned to one of two insulin regimens for 4 months and then switched to the other regimen for another 4 months. The two treatment regimens were 1) mixed treatment--a mixture of human regular and neutral protamine Hagedorn (NPH) insulin administered before dinner and 2) split treatment--human regular insulin administered at dinner and NPH insulin administered at bedtime.
Measurements: Frequency of nocturnal hypoglycemia. Secondary end points were levels of fasting blood glucose and hemoglobin A1c and responses to experimental hypoglycemia.
Results: During the split-regimen treatment period, patients had fewer episodes of nocturnal hypoglycemia (mean [+/-SE], 0.10 +/- 0.02 episode/patient-day vs. 0.28 +/- 0.04 episode/patient-day; P = 0.002), a lower fasting blood glucose level (mean [+/-SE], 7.6 +/- 0.2 mmol/L vs. 8.3 +/- 0.5 mmol/L [137 +/- 4 mg/dL vs. 160 +/- 8 mg/dL]; P = 0.030), less variable fasting blood glucose levels (SD range, 2.0 +/- 0.4 vs. 3.5 +/- 0.6; P = 0.001), and lower hemoglobin A1c value (mean [+/-SE], 7.0% +/- 0.11% vs. 7.5% +/- 0.15%; P = 0.004) than during the mixed regimen. Responses to experimental hypoglycemia were better preserved with the split regimen than with the mixed regimen.
Conclusion: When the goal of insulin therapy in type 1 diabetes mellitus is near-normoglycemia, splitting the evening insulin treatment regimen into short-acting insulin at dinner and NPH insulin at bedtime reduces the risks for nocturnal hypoglycemia and hypoglycemia unawareness and decreases the hemoglobin A1c value compared with mixing short-acting insulin and NPH insulin at dinner.