Objective: Current methods for measuring gastric emptying by breath test require sampling over several hours and are too inaccurate for clinical use. The aim of this study was to develop an office-based method for measuring gastric emptying of solids in patients with diabetes using a [13C]octanoic acid breath test.
Methods: In 22 symptomatic diabetic patients (17 insulin-dependent diabetes, 5 non-insulin-dependent diabetes) and 6 controls, we simultaneously measured gastric emptying of an egg meal (420 kcal) by scintigraphy and [13C]octanoic acid breath test. Conventional (nonlinear) methods for scintigraphic and [13C]octanoic acid breath test emptying and generalized linear regression method to predict scintigraphic half-life (t(1/2)) using four breath samples obtained during the first 3 h.
Results: Despite 8 h of breath sampling, the t(1/2) estimate using the conventional method was markedly different from the scintigraphic value (delta t(1/2): median, 113 min; range, 19-282 min). The generalized linear model (using samples at baseline, 30, and 120 or 150 min) yielded predicted scintigraphic tLAG and t(1/2) that were more accurate than the conventional method; mean standard deviations of differences were 16 and 27 min, respectively. Breath test correctly assessed normal or prolonged emptying in 21 of 22 patients.
Conclusions: The [13C]octanoic acid breath test can be simplified to measure gastric tLAG and t(1/2) and can be expected to correctly identify normal t(1/2) in symptomatic diabetics. Further refinement of the model will need to include studies of patients with markedly delayed t(1/2).