Effects of ventilation on the collection of exhaled breath in humans.

A computerized system has been developed to monitor tidal volume, respiration rate, mouth pressure, and carbon dioxide during breath collection. This system was used to investigate variability in the production of breath biomarkers over an 8-h period. Hyperventilation occurred when breath was collected from spontaneously breathing study subjects (n = 8). Therefore, breath samples were collected from study subjects whose breathing were paced at a respiration rate of 10 breaths/min and whose tidal volumes were gauged according to body mass. In this "paced breathing" group (n = 16), end-tidal concentrations of isoprene and ethane correlated with end-tidal carbon dioxide levels [Spearman's rank correlation test (r(s)) = 0.64, P = 0.008 and r(s) = 0.50, P = 0.05, respectively]. Ethane also correlated with heart rate (r(s) = 0.52, P < 0.05). There was an inverse correlation between transcutaneous pulse oximetry and exhaled carbon monoxide (r(s) = -0.64, P = 0.008). Significant differences were identified between men (n = 8) and women (n = 8) in the concentrations of carbon monoxide (4 parts per million in men vs. 3 parts per million in women; P = 0.01) and volatile sulfur-containing compounds (134 parts per billion in men vs. 95 parts per billion in women; P = 0.016). There was a peak in ethanol concentration directly after food consumption and a significant decrease in ethanol concentration 2 h later (P = 0.01; n = 16). Sulfur-containing molecules increased linearly throughout the study period (beta = 7.4, P < 0.003). Ventilation patterns strongly influence quantification of volatile analytes in exhaled breath and thus, accordingly, the breathing pattern should be controlled to ensure representative analyses.