On the validity of the 13C‐acetate breath test for comparing gastric emptying of different liquid test meals: a validation study using magnetic resonance imaging

Abstract

(13) C-acetate breath testing (BT) is applied to assess and compare gastric emptying of liquid meals. Gastric half-emptying times (t50 ) from BT show offsets compared to t50 values from γ-scintigraphy and ultrasonography. Linear transformations have been proposed to correct these offsets. This investigation critically validates the BT for the assessment of liquid gastric emptying by using simultaneously recorded meal and total gastric content volume emptying data from magnetic resonance imaging (MRI). Data were collected during a recently published double-blind, randomized, cross-over MRI gastric emptying study of three (13) C-labeled enteral formulas differing in protein sources (PMID: 24699556). Breath testing-derived t50 was computed with the analysis methods commonly applied in gastric emptying research, i.e., the exponential-beta function and the Wagner-Nelson (WN) method, respectively. Breath testing t50 values from exponential-beta function and WN method showed a positive and negative offset to MRI data, respectively. Linear regression detected low concordance between MRI and both BT methods revealing meal specific and emptying rate-dependent offsets. The WN method showed worse agreement and correlation with MRI emptying data. Breath testing rather reflected meal volume than total gastric content volume emptying. This validation study indicates that the (13) C-acetate breath test may not be applied to compare gastric emptying of arbitrary liquid meals without prior validation by imaging methods. t50 values from BT are biased by (i) the properties of the meal and (ii) the selected method used for (13) CO2 exhalation analysis. No linear transformation common for all meals was applicable to correct the offsets between BT and MRI.