RSS Feeds
Effect of Inhaled Insulin on Fasting and Postprandial Plasma Glucose
- Lisa Ceglia, MD;
- Joseph Lau, MD; and
- Anastassios G. Pittas, MD, MSc
IN RESPONSE:
We thank Dr. Mikhail and colleagues for their comments, and we appreciate the opportunity to respond. We chose not to present data on fasting plasma glucose because this measure reflects the effectiveness of a long-acting basal insulin regimen or oral medication rather than that of inhaled insulin therapy, which has a short time of action. Furthermore, fasting blood glucose is a less reliable measure of glycemia in open-label trials because it can be affected by short-term changes in various factors. After combining data for fasting blood glucose available from 8 of the 16 trials (1–8) included in our meta-analysis, we found statistically significant differences in fasting blood glucose levels from baseline favoring the inhaled insulin group (weighted mean difference vs. subcutaneous insulin, −1.43 mmol/L [−25.8 mg/dL] [95% CI, −2.33 to −0.54 mmol/L {−41.9 to −9.7 mg/dL}]; weighted mean difference vs. oral agents, −1.23 mmol/L [−22.1 mg/dL] [CI, −2.44 to −0.02 mmol/L {−44.0 to −0.3 mg/dL}]) but with significant heterogeneity among studies.
We agree with Dr. Mikhail and colleagues that the effect of inhaled insulin on postprandial hyperglycemia is of interest, given the premeal indication for inhaled insulin therapy. Only 5 of the 16 trials (3–7) in our meta-analysis reported postprandial blood glucose levels using the same standard meal (16 oz. of Boost liquid meal [Mead Johnson Nutritionals, Evansville, Indiana]). No trial specified whether insulin (inhaled or subcutaneous) was administered shortly before the liquid meal. When we combined the data from these trials, we found no statistically significant differences in postprandial blood glucose levels (weighted mean differences, −0.82 mmol/L [−14.8 mg/dL] [CI, −1.84 to −0.20 mmol/L {−33.2 to 3.6 mg/dL}] and −2.49 mmol/L [−44.9 mg/dL] [CI, −5.12 to −0.13 mmol/L {−92.2 to 2.3 mg/dL}], respectively).
The results of the fasting and postprandial glucose concentrations, which tend to favor inhaled insulin over subcutaneous insulin, contrast with the hemoglobin A1c results, which slightly favored subcutaneous insulin.
In our analysis, we elected to present only the change in hemoglobin A1c concentration because it is a measure that captures both fasting and postprandial glycemia and is the most reliable and least-biased glycemic measure (9). Hemoglobin A1c is also the best predictor of diabetic complications. Hemoglobin A1c is, therefore, the preferred outcome when evaluating the glycemic efficacy of new diabetes therapies (10).
Lisa Ceglia, MD
Joseph Lau, MD
Anastassios G. Pittas, MD, MSc
Tufts-New England Medical Center
Boston, MA 02111
Article and Author Information
-
Potential Financial Conflicts of Interest: None disclosed.
