Abstract
Individual uptake of tobacco smoke constituents by smoking is highly variable in cigarette smokers and cannot be predicted by smoking behaviour variables and machine-derived smoke yields. It is well established that uptake of smoke constituents is best described by a series of bio-markers of exposure (BOEs) such as metabolites of nico-tine, tobacco-specific nitrosamines (TSNAs), polycyclic aromatic hydrocarbons (PAHs), aromatic amines, benzene, 1,3-butadiene, acrolein, hydrogen cyanide, 2,5-dimethyl-furan and other smoke constituents.
The purpose of this review is to investigate the relationship between BOE levels and machine-derived smoking yields on the basis of published data. The influence of other smoking behaviour variables, in particular the number of cigarettes smoked per day (CPD) and smoking topography (puffing and inhalation patterns) is also considered, pro-vided suitable data are available.
Twenty eight (28) published studies, which report data on machine-derived smoke yields and biomarker concentrations in body fluids of smokers of these products were identified. In total, 33 different BOEs were applied in these studies. Important properties of the BOEs used in the further evaluation were described and discussed. In almost all studies selected, data for CPD were reported. In only a few studies, puffing and inhalation profiles have been determined so that no systematic evaluation of the association between smoking topography and BOE levels was possible. In the studies evaluated, no statistically significant association between daily cigarette consumption (CPD) and smoke yields was observed. This clearly indicates that low machine-derived yields were not com-pensated by increasing the daily cigarette consumption. As expected, positive and statistically significant relationships were found between CPD and BOE levels for most of the biomarkers investigated.
Bi- and multivariate linear regressions were calculated for the relationships between BOE levels (dependent variable) and machine-derived yields as well as CPD (independent variables). Whenever possible, results from various studies were combined (this was only possible, when identical biomarkers and yield types were available). Aggregation of the results from all studies independent of BOE and yield type used is feasible on the basis of relative BOE and yield levels. The multivariate linear regression models obtained reveal that both CPD and machine-derived yields are significant predictors of the measured BOE levels. The models predict that, on average, a 50% reduction in CPD or yield are accompanied by a 33 or 15% reduction, respectively, in smoke uptake, as measured by various BOEs. Taken together, the evaluated data from the literature show that lower machine-derived yields lead to a reduced uptake of smoke constituents. The reduction is statistically significant, but substantially lower than the decrease in machine-derived yields. [Beitr. Tabakforsch. Int. 26 (2014) 138-175]