Student Research: Gary E. Whitney

Evaluation of the Use of Carbon Dioxide as a Normalizing Factor when Measuring Organic Solvents in the Expired Breath of Humans
Faculty Advisor: 


Breath sampling has been proposed as a non-invasive method of biological monitoring of workers exposed to organic solvents. To correct breath sampling variability, CO2 has been used as a normalizing factor. This method is based on the assumptions that individual alveolar CO2 concentrations are reasonably close to the normalizing factor used (i.e. variability is low), and that CO2 and organic solvents reach equilibrium between blood and alveolar air in a similar manner (i.e. they are well correlated in breath samples). The goal of this thesis was to test the validity of these assumptions.

To determine intrasubject and intersubject variability in alveolar CO2 concentration, end-tidal breath CO2 concentrations were measured once a week in a group of 13 adult males for periods of eight to ten weeks. Results indicate that intrasubject end-tidal CO2 concentrations are very consistent from week to week, with an average coefficient of variation of less than 5%. The intersubject end-tidal CO2 levels differed considerably, ranging from approximately 3.5% to 5.5% with a mean of 4.8%.

To investigate the correlation between expired breath CO2 and solvent concentrations, eight adult males were exposed to 100 ppm toluene. Breath samples were collected from each subject using methods designed to produce different breath CO2 concentrations (hyperventilation, standard, and breath holding). Sample toluene concentrations were plotted against CO2 concentration and regression lines were calculated. Breath CO2 and solvent were well correlated when using the standard and breath holding methods. Hyperventilation produced breath samples that had relatively higher solvent concentrations.

It is concluded that CO2 is a reasonable normalizing factor when measuring breath solvent concentrations providing care is taken to avoid hyperventilation and to use the subject's own end-tidal CO2 concentration as the normalizing factor.