Chapter 12 
Compensation for Nicotine by Smokers 
of Lower Yield Cigarettes 
Lynn T. Kozlowski and Janine L. Pillitteri 
BACKGROUND The question has been asked whether brand-switching smokers 
oversmoke lower nicotine cigarettes. The Federal Trade Commission 
(FTC) testing method is a per-cigarette test and should be judged as such. 
(Forty truly low-calorie candy bars together could be high calorie and still, 
individually, be low calorie.) The FTC test cannot be blamed because 
smokers smoke more cigarettes when they switch to those having a lower 
5 deld. Therefore, for this review compensation data were adjusted to per- 
cigarette values. However, such per-cigarette adjustments only approximate 
what would happen if the number of cigarettes were fixed for smokers. 
If smokers have already compensated by smoking many more cigarettes, 
then presumably they would have less need to smoke more of each cigarette. 
In the five studies included in the authors' main review, the compensatory 
percentage change in cigarettes per day averaged 15 percent (±6, 95-percent 
confidence interval). No studies showed a decreased number of cigarettes 
smoked with a lower yield brand of cigarettes. 
Experimental brand-switching studies offering measures of nicotine 
and cotinine were reviewed. An index of compensation was calculated 
using a sequence of formulas developed by Russell and colleagues (1982). 
Calculation of these formulas first requires information on the machine- 
smoked nicotine yields of cigarettes to calculate (a) the percentage change 
in nicotine yields. Information on the measured level of nicotine (or 
cotinine) in body fluids is then used to calculate (b) the percentage change 
in nicotine (or cotinine) intake. Finally, three consecutive formulas are 
used to calculate (c) the actual compensatory increase in smoke intake 
[(b/a - 1) X 100]; (d) the increase in smoke intake necessary for complete 
compensation [(1- a)/a x 100]; and (e) using the values obtained in (c) and 
(d) above, the degree of compensation [(c/d) - 100]. 
CIGARETTE BRAND Research on brand switching makes use of repeated-measures 
SWITCHING IN designs. With these designs, the same smokers get different 
EXPERIMENTAL cigarettes. This controls for individual differences in drug 
RESEARCH metabolism (Benowitz et al., 1982) and for important biases 
in brand selection, which usually are not controlled for in cross-sectional 
research. This issue has been discussed by others (e.g., Giovino et al. [this 
volume]; Cohen [this volume]). Wynder and coworkers (1984) explored the 
demographics of smokers of the low-yield cigarettes and showed that age, 
sex, race, education, and religion were strongly related to the selection of 
low-tar cigarettes. Wynder and colleagues (1984) reported that education 
is negatively associated with tar for males, but not for females. (Tar and 
161 
