THE TAR REDUCTION STUDY - RANDOMIZED TRIAL OF THE EFFECT OF CIGARETTETAR YIELD REDUCTION ON COMPENSATORY SMELTING

Background - Observational and short term intervention studies have re ported that smokers of low tar cigarettes inhale more deeply (that is, compensate) than those who smoke high tar cigarettes. To quantify thi s effect a long term randomised trial was conducted on the effects of switching to low tar cigarettes.Methods - The trial was carried out be tween April 1985 and March 1988 among cigarette smokers in the British Civil Service, measuring blood carboxyhaemoglobin (COHb) levels and s erum cotinine levels as markers of tobacco smoke intake. Volunteers fi rst switched to a cigarette brand yielding around 10% less tar than th eir usual brand to identify smokers able to change brand. The 434 subj ects who successfully switched were then randomly allocated to one of three groups: (a) ''fast reduction'' group which changed to a brand of cigarettes with a tar yield of about half that of their usual brand; (b) ''slow reduction'' group which reduced to the same level in steps over several months; and (c) a control group which continued smoking c igarettes with a tar yield 10% lower than their usual brand. Results - Over the course of the trial cigarette consumption declined slightly in all three groups. In both the ''fast reduction'' and the ''slow red uction'' groups, intake of COHb and cotinine was reduced, though not t o the same extent as the yield reduction. Comparison of the results be fore randomisation with those at the end of the trial showed that a re duction in carbon monoxide yield of 45% was associated with a decrease in carbon monoxide intake of 19% (95% confidence interval 14% to 24%) and that a reduction in nicotine yield of 40% was associated with an 11% (6% to 16%) reduction in nicotine intake, reflecting relative inta kes of about 1.5 for both carbon monoxide and nicotine in the ''fast r eduction'' group. Results were similar in the ''slow reduction'' group with a 42% reduction in carbon monoxide yield, a 16% (11% to 22%) red uction in carbon monoxide intake, a 37% reduction in nicotine yield, a nd a 6% (0% to 13%) reduction in nicotine intake. Estimates of compens ation derived from these results were 65% for carbon monoxide, 79% for nicotine, and 62% for tar. Conclusions - Compensation, demonstrated w hen switching from a high tar cigarette to a low tar one, was incomple te. Advising people who have failed to give up smoking to switch to lo w tar cigarettes will reduce the intake of smoke constituents to a sma ll extent. This would be expected to decrease their risk of smoking-re lated diseases, although by a smaller amount than would be achieved by giving up smoking altogether.