Smoking behaviour and compensation: a review of the literature

Rationale: Compensation or compensatory smoking, accurately defined, deals with the question of whether switching to cigarette brands with different s moke yields is associated with a change in smoke uptake proportional to the change in machine-derived yields. The issue of compensation is important b ecause it bears on whether switching to "lighter" brands means lower overal l smoke intake or not. Objectives: The present review investigated whether and to what extend low yield cigarettes are smoked more intensively. In add ition, published data on whether nicotine, "tar", or any other smoke consti tuent or property influence compensational smoking are summarized. Methods: The studies on compensation were classified as follows: (1) studies on smo king behaviour in relation to cigarette yields (with and without brand swit ching); (2) studies on compensation for nicotine (switching between cigaret tes which differ "only" in their nicotine yield, nicotine supplementation, manipulation of renal nicotine excretion, administration of nicotine agonis ts or antagonists); (3) studies on compensation for other factors (influenc e of tar, taste, irritation, draw resistance). In order to quantify the deg ree of compensation, an index is defined and applied to selected brand swit ching studies. This compensation index determines, in relative units, the d egree to which a smoker responds to a change in smoke yields with a change in smoke uptake measured by suitable biomarkers. The role of vent blocking is also briefly discussed. Results: Most of the studies which compare the s moking behaviour when smoking cigarettes with different smoke yields supply evidence for "partial" compensation, suggesting that cigarettes with lower yields are smoked more intensively than those with higher yields. These st udies also show that a change in the daily number of cigarettes is not a co mmon mechanism of compensation. Effective vent blocking during smoking is a rare event and can therefore also be regarded as an uncommon mechanism of compensation. Evaluation of a suitable subset of brand-switching studies re vealed an average compensation of 50-60% of the nicotine yield. Compensatio n tended to be more complete when changing to cigarettes with higher yields than when changing to cigarettes with lower yields. In general, brand-swit ching studies do not supply information on the underlying causal factors re sponsible for compensatory smoking. Results of the nicotine supplementation studies are not conclusive: some report evidence of nicotine titration, ot hers do not. A general problem with this type of investigation is that cont inuous nicotine application does not mimic the spike-wise application with cigarette smoking, and may lead to nicotine tolerance. There is limited evi dence that cigarettes were smoked more intensively when the urinary clearan ce of nicotine was increased. A small number of studies provide some eviden ce that smoking intensity increased after smokers were administered a nicot ine antagonist. Several reports indicate that tar, taste and sensory proper ties of the smoke as well as the draw resistance of the cigarette may play a role in compensatory smoking. Low-yield cigarettes usually have reduced p ressure drops which smoke researchers have suggested leads to increased puf f volume. This effect seems to be independent of the smoke yield of the cig arette. There is also some evidence that some smokers maintain a consistent pattern of smoking which works independent of any changes in nicotine or t ar yields, taste or design features of the cigarette ("functional autonomy" ). Conclusions: The available data suggest that smokers partially compensat e for a different smoke yield. While the factors and their interaction responsible for compensational smok ing are not fully understood, there are data suggesting that a subgroup of smokers may partially compensate for nicotine. Even in this subgroup of smo kers, however, the relative importance of the pharmacological versus the se nsory effects of nicotine in smoke remains to be determined.