Background: Fructose malabsorption is commonly diagnosed by the hydrogen fructose (H₂) breath test. However, the mechanisms behind fructose malabsorption in humans are not well understood and the clinical relevance of this test is considered controversial. Hence, the main aim of this study is to evaluate the predictive value of the H₂ breath test.

Methods: Regarding exclusion criteria, the study enrolled 562 consecutive patients, enlisted to a gastroenterology clinic between 2009 and 2011 for testing malabsorption. In the final data analysis, 246 patients were included. Ecotrophologists used 3 categories to rate dietary success: complete response, partial response and no response to the diet. They also rated the occurrence of abdominal pain, diarrhoea and bloating during the H₂ breath test. Ordinal regression analysis using SPSS was performed to evaluate whether H₂ breath test results – measured as the maximum H₂ level, the maximum increase in H₂, and the area under the curve (AUC) – predicted dietary success or failure. Correlation analyses were applied to test whether symptoms of fructose malabsorption correlated with the H₂ breath test measures. Finally, we evaluated whether cut-off-values of 40 or 60 parts per million (ppm) serve better than the test measure of 20 ppm to diagnose fructose malabsorption.

Results: When a fructose-free diet was administered it was found that 103 patients (41.9%) were complete responders, 116 (47.2%) were partial responders and 27 (11%) were non-responders. The H₂ breath test with the 20 ppm cut-off-value, that is, the maximum H₂ level, the maximum increase in H₂, and the AUC did not predict dietary response (all 95% CI ns). This was also the case when using 40 or 60 ppm as cut-off-values (all 95% CI ns). Abdominal pain during the test correlated significantly with the AUC. Diarrhoea and bloating correlated significantly with the AUC, the maximum H₂ level and the maximum increase in H₂ ( p < 0.05).

Conclusions: The H₂ breath test produced no predictive value for the fructose-free diet outcomes; its value as a predictive test is therefore questionable. However, the symptoms of fructose malabsorption correlated significantly with the H₂ breath test measures, and this is an indication that there is at least a degree of validity of the H2 breath test beyond the simple detection or exclusion of fructose malabsorption.