Postprandial insulin profiles with implantable pump therapy may explain decreased frequency of severe hypoglycemia, compared with intensive subcutaneous regimens, in insulin-dependent diabetes mellitus patients.

PURPOSE: To examine the mechanism of the decreased frequency of severe hypoglycemia with implantable pump therapy compared with subcutaneous intensive therapy. PATIENTS AND METHODS: Eight subjects with insulin-dependent diabetes mellitus (IDDM), enrolled in an implantable insulin pump study, were admitted to the General Clinical Research Center and on 2 separate days were given either a dose of preprandial insulin chosen to maintain normoglycemia for a standard (450 kcal, 50% carbohydrate) breakfast or 1.75 times the dose. The two doses were administered subcutaneously (by syringe or with an external pump) during one inpatient admission and by implantable pump (intraperitoneally, n=6; or intravenously, n=2) during a separate admission. Blood glucose, plasma-free insulin, and neurocognitive function were measured for 4 hours after the meal. RESULTS: Subcutaneous administration resulted in 7 episodes of hypoglycemia (2 with the usual dose and 5 with the 1.75-fold dose), defined as blood glucose less than 50 mg/dL; implantable pump treatment resulted in only 2 episodes, both with the 1.75-fold dose (P <0.05, Fisher's two-tailed test for implantable versus subcutaneous). Compared with subcutaneous delivery, implantable pump therapy provided significantly lower insulin levels during the final 2 hours after administration of the usual dose and the 1.75-fold dose (P <0.005). In addition to the decreased frequency of hypoglycemia, implantable pump therapy resulted in significantly lower area under the glycemia curve during the first 120 minutes with the 1.75-fold dose compared with subcutaneous administration. CONCLUSIONS: The lower frequency of severe hypoglycemia with intensive therapy administered by implantable pump therapy is explained by the more rapid clearance of insulin delivered intraperitoneally or intravenously compared with intensive subcutaneous injection regimens. The lower frequency of severe hypoglycemia with implantable pump therapy compared with subcutaneous therapy demonstrated in clinical trials is confirmed by this study, in which we attempted to induce hypoglycemia.

Glycemic control in diabetes mellitus: have changes in therapy made a difference?

PURPOSE: New methods of measuring and controlling glycemia in diabetes mellitus have been developed and implemented in the past 10 years. We examined whether glycemia, as measured by glycosylated hemoglobin, changed in outpatient insulin-dependent diabetes mellitus (IDDM) and noninsulin-dependent diabetes mellitus (NIDDM) populations between 1985 and 1993 and whether contemporaneous changes in therapy could account for observed changes in glycemia. PATIENTS AND METHODS: Outpatients were selected based on having glycated hemoglobin (HbA1c) measured in the Massachusetts General Hospital laboratory during March 1985 (IDDM n = 94 and NIDDM n = 137) or during March 1993 (IDDM n = 89 and NIDDM n = 118). Chart reviews established demographic and clinical characteristics, including frequency of blood glucose self-monitoring, insulin injections, office visits, and HbA1c measurements during the year prior to the HbA1c result. RESULTS: Mean HbA1c level was significantly lower in the 1993 IDDM cohort compared with the 1985 cohort (8.77% +/- 1.7% versus 9.47% +/- 2.1%, P = 0.014). In the NIDDM cohorts, the difference in mean HbA1c did not achieve statistical significance (8.35% +/- 1.6% in 1993 versus 8.75% +/- 2.1% in 1985, P = 0.09); however, when adjusted for differences in NIDDM duration, HbA1c in the 1993 cohort was significantly lower than that in the 1985 cohort. The largest decrease in HbA1c in NIDDM was in patients treated with insulin (9.53% +/- 2.0% versus 8.54% +/- 1.5% in 1985 and 1993, respectively, P = 0.004). Multiple linear regression analyses demonstrated that increased frequency of self-monitoring and of insulin injections were associated with lower HbA1c in IDDM. CONCLUSIONS: The level of average glycemia has decreased in IDDM patients over the past 8 years, attributable, at least in part, to an increased frequency of monitoring and of insulin injections. Glycemia decreased in NIDDM, especially in the subset of patients treated with insulin. This temporal shift in glycemic control should have a salutary effect on the development of long-term microvascular and neurologic complications.