Background
Breath hydrogen testing is used to assess colonic fermentation of malabsorbed carbohydrates and their role in symptom genesis. Lactulose and fructose breath responses are used to guide dietary restriction of FODMAPs (fermentable carbohydrates) in patients with irritable bowel syndrome (IBS). Parameters of lactulose breath tests have also been used to indicate the presence of small intestinal bacterial overgrowth (SIBO).
Objective
To investigate (i) the reproducibility of lactulose and fructose breath hydrogen testing and (ii) the association between fructose breath response and symptom induction in IBS.
Design
A retrospective audit was conducted in 21 IBS patients completing a 15g lactulose breath test and in 30 patients completing a 35g fructose breath test. A repeat test was performed ≥2 weeks for lactulose and fructose. Changes in breath hydrogen responses between test and retest were analysed qualitatively (positive response: 2 x ≥10ppm hydrogen rise) and quantitatively as area-under-curve (AUC)) and oro-caecal transit time (OCTT). Symptoms were measured on a 100mm visual analogue scale (VAS) in another IBS cohort (n=36) and examined for correlations with fructose breath testing results.
Outcomes
A positive lactulose response was maintained in 96% subjects, but 30% (p=0.0006) lost a positive fructose response upon retest. Initial AUC to lactulose and fructose were poorly correlated with AUC values on repeat testing (lactulose: r2=0.08, p=0.16; fructose:r2=0.07, p=0.18; regression analysis). Lactulose OCTT (r=0.29; p=0.15; Spearman's correlation) or fructose (r=0.29; p=0.31) were not correlated between test-retest. Despite a higher proportion of patients reporting symptoms with a positive fructose test (56 vs 17%; p=0.04), no differences were observed for changes in VAS scores between those with a negative vs positive result.
Conclusion
Poor reproducibility of lactulose and fructose breath testing was demonstrated. The application of these tests (e.g. SIBO diagnosis or to guide dietary fructose restriction) beyond determination of hydrogen-producing status is not justified.