Relationship Between Gastric Emptying and Diurnal Glycemic Control in Type 1 Diabetes Mellitus: A Randomized Trial

Abstract

Context: In type 1 diabetes (T1D), delayed gastric emptying (GE) may predispose to a mismatch between insulin delivery and glucose absorption. Previous studies evaluated, only partly, the relationship between delayed GE and postprandial, but not diurnal, glycemia.

Objective: To assess the relationship between GE disturbances and glycemic control in T1D and the effects of accelerating GE on glycemic control.

Design, setting, and participants: This was a randomized placebo-controlled trial in 30 patients with T1D on an insulin pump at an academic medical center.

Intervention(s): GE was evaluated with a [13C]-Spirulina breath test at baseline (GEbaseline), during intravenous saline or erythromycin (2 or 3 mg/kg; GEiv), and after 7 days of oral erythromycin or placebo (GEoral). Weighed meals were provided throughout the study.

Main outcome measure(s): These were GE and continuous glucose monitoring (CGM).

Results: The baseline glycosylated hemoglobin was 7.6% ± 0.8% (60 ± 8.7 mmol/mol); 12 patients (40%) had delayed GE; faster GE was associated with a greater postprandial CGM-based glucose, but slower GE was not associated with postprandial hypoglycemia (<70 mg/dL). Intravenous (3 mg/kg) but not oral erythromycin accelerated GE. The relationship between GE and glycemia differed between the postprandial periods and the entire day. After adjusting for carbohydrate intake and insulin consumption, faster GE was associated with more hyperglycemia during the postprandial period but lower glucose values across the entire study.

Conclusions: In T1D, pharmacologically mediated acceleration of GE increases postprandial CGM-based glucose. In contrast, delayed GE is associated with greater CGM-based glucose values over the entire day.

Trial registration: ClinicalTrials.gov NCT02755064.

Copyright © 2017 by the Endocrine Society

Figures

Figure 1.

Median glucose value (top panel), carbohydrate consumption (middle panel), and insulin dose (bottom panel) in patients with normal (continuous line) and delayed (dotted line) GE. Each point represents the 6-hour median value for that parameter. Periods 1, 2, 3, and 4 are from midnight to 7 am, 7 am to 1 pm, 1 pm to 7 pm, and 7 pm to midnight, respectively. GE was categorized as normal or delayed based on the results of GEbaseline for days 1 and 2, GEiv for day 3, and GEoral for days 3 to 10.

Figure 2.

Effects of intravenous erythromycin and placebo on GE (ckPCD values) during GEiv (top panel) and GEoral (bottom panel). Intravenous erythromycin (3 mg/kg only) increased (P = 0.03) ckPCD values at 60 minutes, signifying accelerated GE. Dashed lines indicate patients who had delayed GE at GEbaseline.

Figure 3.

Relationship between the difference in estimated thalf and ckPCD value at 60 minutes for GEiv – GEbaseline. A greater difference (GEiv – GEbaseline) in ckPCD values at 60 minutes was inversely correlated (r = −0.59, P < 0.001) with the GEBT thalf. Greater ckPCD values and a lower thalf reflect accelerated GE.

Figure 4.

Association between rate of increase in CGM-based glucose value during the 4-hour postprandial period and GE (ckPCD values). (Top panel) During GEiv, the rate of increase from fasting to peak postprandial CGM-based glucose concentrations was greater in patients who received intravenous erythromycin 3 mg/kg than placebo (*P < 0.05). (Bottom panel) During GEoral, the rate of increase from fasting to peak postprandial CGM-based glucose concentrations was positively correlated with the ckPCD value at 60 minutes.