Abstract

During the clinical investigation of patients with gastrointestinal disease by exhaled breath hydrogen measurement, the occurrence of inexplicable variations in recorded hydrogen values led to a search for extra intestinal factors which were capable of adversely influencing breath hydrogen concentration and impairing the diagnostic accuracy of the test. Serial breath samples were collected from normal subjects under a variety of conditions which might occur during routine clinical study, including, hyperventilation, exercise, cigarette smoking, and carbohydrate ingestion. Breath hydrogen concentrations were consistently reduced by hyperventilation (p less than 0.01) and exercise (p less than 0.05). Cigarette smoking, in contrast, caused a marked rise in measured breath hydrogen (p less than 0.01), as did oral carbohydrate (p less than 0.05). Prior bactericidal mouthwash abolished this carbohydrate associated rise, suggesting that the hydrogen was the result of fermentation by oropharyngeal bacteria. Because, in all instances, the changes in breath hydrogen were of sufficient magnitude to interfere with data interpretation, it is recommended that these factors are eliminated, whenever possible, from conditions of study.