Indirect vs direct assessment of gastric emptying : A randomized crossover trial comparing C-isotope breath analysis and MRI

Background: Indirect methods to assess gastric emptying (GE), such as 13C breath tests (BT), are commonly used. However, BT usually use a sampling time of 4+ hours. The current study aims to assess the validity of BT for four liquid meals differing in physicochemical properties. To this aim, we compared them to MRI GE-measurements. Methods: Fifteen healthy males (age 22.6 ± 2.4 years, BMI 22.6 ± 1.8 kg/m2) participated in a randomized 2 × 2 crossover experiment. Test foods were liquid meals, which were either thin/thick and 100/500 kcal, labeled with 100 mg of 13C-octanoate. GE was measured with MRI and assessed by 13C recovery from breath. Participants were scanned every 10 minutes and at six time points breath samples were collected up to t = 90 minutes. Two curves were fitted to the data to estimate emptying halftime (t50 Ghoos and t50 Bluck). T50 times were ranked per participant and compared between methods. Key Results: On average, MRI and BT showed similar t50 rankings for the four liquid meals. In comparison to MRI, t50 Ghoos overestimated, while t50 Bluck underestimated GE time. Moreover, more viscous foods were overestimated. In most participants individual t50 time rankings differed significantly between methods. Conclusions & Inferences: BT can assess relative emptying differences on group level and collecting breath data for 90 minutes constitutes a lower burden for participants and the research facility. However, BT has severe shortcomings compared to MRI for individual GE assessment. Notably, food matrix effects should be considered when interpreting the results of BT.