Saliva as an alternative specimen for alcohol determination in the human body.

Saliva, breath, and blood samples were collected from 49 volunteers and over 700 values of ethanol concentration were obtained. The profiles of time-dependent changes in saliva and expired air-breath ethanol concentration were similar. The average difference between the respective values determined in blood and saliva amounted to -0.031 +/- 0.096 g/l, whereas the difference between the results for breath and saliva was -0.034 +/- 0.080 g/l. These differences in ethanol concentrations do not exceed those which occur between blood and breath (0.003 +/- 0.093 g/l). Introducing a correction value of 1.08, stemming from the varying water content in saliva and blood, results in a good agreement between the results for saliva and breath (0.005 +/- 0.077 g/l). The headspace gas chromatographic method applied for ethanol determination in saliva is specific (resolution > 1), shows good accuracy (recovery = 100.7%) and precision (SD = 0.0155 g/l). There is no matrix effect when water solutions are used for calibration instead of saliva.