Comparison of three sampling and analytical methods for measuring m-xylene in expired air of exposed humans

Three breath sampling and analytical methods were tested following exposure of 12 subjects for 4 hr to 75 ppm m-xylene in a controlled environmental chamber. Mixed-expired breath was sampled for m-xylene from all 12 subjects with a new stainless steel device that permits continuous mainstream or sidestream sampling of the solvents present. The m-xylene was sampled from the mainstream using charcoal cloth and from the sidestream using Tenax TA. Alveolar breath also was sampled for m-xylene from 6 of these subjects using bags. The carbon dioxide concentrations of the mixed and alveolar samples, obtained from these 6 subjects, were also determined and used to assess the accuracy of the mixed-expired sampling and analytical procedures. Breath sampling was conducted over the immediate 240-min postexposure period. All m-xylene samples were analyzed using gas chromatography with flame ionization detection. Carbon dioxide concentrations were determined with an infrared analyzer. Nonlinear regression analysis was used to model the desaturation of m-xylene via the breath. Overall, the desaturation of m-xylene from all subjects by all methods was best described using three-compartment pharmacokinetic models. The precision of each sampling and analytical method, estimated from the residual variabilities of the desaturation curves were 0.13 for alveolar sampling, 0.14 for mainstream-mixed sampling (12 subjects), and 0.23 for sidestream-mixed sampling (12 subjects). For all 12 subjects, the breath m-xylene concentrations determined by sidestream-mixed sampling averaged 83% of those determined by mainstream-mixed sampling; this bias was significant. For the 6 subjects from whom both mixed-expired and alveolar breath samples were obtained, the average m-xylene desaturation rates determined by both mainstream-mixed and alveolar sampling were comparable but substantially different from those determined by sidestream-mixed sampling. For these subjects, comparison of the average and individual mixed to alveolar ratios of m-xylene and carbon dioxide showed that mainstream-mixed sampling was accurate and that sidestream-mixed sampling was not.