Comments from scientists regarding the methods used to measure ETS and its impact on non-smoker exposure, smoker exposure, and their relation to lung cancer. the study critiqued analyzed plasma cotinine, plasma nicotine levels and other biochemical markers among smokers and non-smokers in contact with ETS.
Page 1: fif08e00
2. Dose - measurements in humans
Discussion: Rapporteurs Roger A. Jenkins and:
Theodor D. Sterling
The discussion on Pershagen's paper dealth with the accurary
of estimating prevalence and amount of smoking from
questionnaire studies. Sterling drew attention to the Warner
study (1) showing that estimates of tobacco products consumed
in US based on personal and on telephone interview, were 40
percent less than the official count based on sale of tax
stamps which have to be attached to every pack sold in the US.
Another point of view brought up was that reports of the
amount smoked were mostly in multiples of "five" or "packs"
(i.e. 10, 20 etc or one pack, 2 packs etc). Not being able to
estimate more accurately from questionnaires how much people
smoked, how accurately could questionnaire methods help
estimate the amount of smoke to which they were exposed by the
actions of others? Zober mentioned that in his experience,
reports of the amount smoked were in multiples of five.
Pershagen replied that in his study, there had been good
agreement between self reports and that of relatives. Lebowitz
added that epidemiolgocial studies of many populations showed
very high reliability of smoking questions, and that a study
of smelter workers has shown good recall of type and length of
exposure the worker had of his experience (2). Rylander
concluded that all this showed how difficult it might be to
study the relation between lung cancer and exposure to
environmental tobacco smoke (ETS) exposure in view of the
difficulties to obtain good exposure data.
In comments to Russell's presentation, Jenkins contributed
that Russell's findings in human patterned what Jenkins had:
observed with smoke exposed beagle dogs, namely that only 1-2%
of the nicotine retained by the individual can be accounted
for at any given time be measurement of plasma nicotine. First
inquired if Russell's subjects were smokers or non-smokers.
They'yiere mostly non-smokers although some had smoked. He
Page 2: fif08e00
suggested that Russell's approach could be used,to study
possible difference in metabolism of nicotine between smokers
and non-smokers. Russell further thought that the problem
might be differences in potency between mainstream and
sidestream smoke. If it can be shown that exposure to ETS is
related to an increased risk for lung cancer even though the
exposure level is very low, then perhaps what caused lung:
cancer in the smoker was not the mainstream inhaled actively
but the sidestream taken passively.
On the paper by Lynch, First inquired why plasma cotinine was
chosen as a marker, rather than plasma nicotine. Lynch
indicated that the primary reason involved the suspected rates
of nicotine and cotinine decay in the plasma. The study
reported was part of a larger smoking cessation experiment
thus had to work within the constriction of that experiment.
The design of the main experiment involved obtaining blood and
breath samples 24 hours apart. Thus nicotine, with a suspected
half life of less than one hour, would nearly completely
disappear from the bLood.after 24 hours. Plasma nicotine's
utility as a long lived indicator of smoke exposure is
Aviado indicated that the differences observed between the CO
half lives of the Group I and Group II females may be due to
differences in the way in which those values were determined.
The former was a two point determination, the latter, an
exponential decay curve fitted to four points.
Jarvis was puzzled by the magnitude of the observed CO halif
time value compared to other findings of 3-6 hours, and
wondered if ambient CO levels were obtained during the study
and if ethanol filters were employed with the Ecolyzer used to
assay the breath samples. Lynch answered in the affirmative to
both questions, but indicated that only ambient CO level~s were
measured in the room in which blood and breath samples were
Page 3: fif08e00
Cosentino queried if the breath samples were true alveolar
air. Lynch replied in the affirmative and described the
procedure for sampling and analysis of the expired air for
carbon monoxide levels..
Bake suggested that there may be another more accurate
approach to obtaining C0 half life, which involved making
detail~ed measurements of COHb after controlled exposure
conditions. Consideration should also be given to the possible
differences in COHb half lives following acute and chronic
exposure, due to different equilibrium compartments. Actually,
some of the differences in the literature may be accounted for
on this basis. Lynch agreed, but indicatedithat he had:been
interested in obtaining half life values under real life
conditions, a half life which represented the average of
values which could be obtained for the various activities that
an individual might undergo in a given day.
Cosentino indicated that different average Mhalf time values
might be obtained for different individuals. Lynch agreed.
Regarding Jarvis' presentation, First started by asking if
cotinine levels were observed in the physiological fluids of
smokers which were one hundred times larger than those found
in non-smokers exposed to ETS, would it be reasonable to
conclude that the latter received;about 1/100 of the dose of
an active smoker? Jarvis thought that that would be a
reasonable estimate of exposure level differences.
Pershagen wondered about the sensitivety and specificity of
the biochemical markers to distinquish between non-smokers and
smokers with usually light smoking,habits, e.g. those who had
not smoked a day or two before the measurements. Jarvis
responded that no biochemical marker is completely speci,fic in
discriminating between smokers and non-smokers, firstly
because of ETS exposure of non-smokers, and secondly, because
some smokers may smoke only frequently. Neverthel~ess, of the
Page 4: fif08e00
available markers, cotinine discriminates best, and is
sufficiently sensitive to detect the majority of even very
Rylander askediwhat levels of plasma cotinine might be
expected in smokers. Jarvis felt that 300-500 ng/ml plasma
would be a reasonable estimate.
1. Warner K.E. Possible increases in underreporting of
cigarette consumption. J Am Stat Assoc. 1978: 73:314-318'.
2. Higgins I, Welch K, Oh M, Bond G, Hurwitz P. Influence of
arsenic and smoking on lung cancer among smelter workers:
A pilot study. Am J Ind Med. 1981: 2:33-41.