With a validated GC/MS method, the tobacco smoke biomarker cotinine has been estimated in urine for 148 non-smokers (male; 43 +/- 13 years; median 5.0 micrograms/g creatinine; 95th percentile 104 micrograms/g) and 96 smokers (male; 39 +/- 12 years; 1002 micrograms/g; 2993 micrograms/g). For a subgroup of 50 persons, the GC/MS results were compared with those by a commercially available radio immunoassay. Both methods identified the same persons as non-smokers and smokers, respectively, and were closely related. For smokers, the relationship was distinctly closer than for the non-smokers (r = 0.90, p <0.001, n = 14 vs. r = 0.41, p <0.02, n = 36). The RIA values were 2.4 times (smokers) and 2.9 times (non-smokers) higher than the GC/MS results. This was probably caused by the cross reactivity of the RIA antibodies against other urinary nicotine metabolites, e.g. trans-3'-hydroxycotinine, and has to be taken into account to correctly compare results of studies obtained with different analytical techniques and for choosing cut-off points to discriminate between active smokers and non-smokers of between non-smokers with higher or lower exposure to environmental tobacco smoke.
