Smoking and Tobacco Control Monograph No. 7 
and inhaling could result in the extraction of substantial amounts of nicotine 
(Butschky et al., 1995). 
Compensatory Compensation is nicely described in the B&W documents (Brown and 
Behavior Williamson, 1984) as "the tendency for a smoker to obtain similar 
delivery, intake and uptake of smoke constituents on a daily basis from a 
variety of products with different standard (machine-smoked) deliveries." 
As the B&W researchers noted, if smokers are dependent, then the 
nicotine they receive from cigarettes can be supplemented by other forms, 
and this will reduce smoking. Likewise, cigarettes of different strengths are 
smoked differently; that is, smokers given low-delivery cigarettes smoke 
them more intensively and vice versa. 
In fact, this is what has been found in many studies (U.S. Department 
of Health and Human Services, 1988). Cigarette consumption increases 
in response to reduced nicotine, and most compensation occurs at the 
individual cigarette level, not by cigarettes per day. Whereas people given 
cigarettes of lower nicotine yield also may smoke a few more cigarettes 
per day, they smoke each of the cigarettes more intensely to obtain 
proportionately more nicotine than the rating of nicotine yield would 
suggest (Hill and Marquardt, 1980; Russell et al., 1980; Benowitz et al., 1983; 
Robinson et al., 1983). 
When people are given nicotine gum and their smoking is measured, 
smoking decreases as the nicotine gum dose increases (Nemeth-Coslett and 
Henningfield, 1986). When mecamylamine is administered to antagonize 
nicotine's effects, people smoke more cigarettes, take more puffs per cigarette, 
and take in more total smoke, as can be seen by increased carbon monoxide 
level (Nemeth-Coslett et al., 1986; Rose et al., 1989). Taste and other sensory 
factors are also important modulators of human smoking behavior (Butschky 
et al., 1995; Rose and Behm, 1987; U.S. Department of Health and Human 
Services, 1988). 
This finding addresses why the nicotine dependence issue is relevant to 
why the FTC method of measuring tobacco smoke constituents is seriously 
flawed. Simply put, the FTC method uses machines that do not change their 
behavior to self-administer a preferred nicotine dose or in response to the 
taste of the smoke, as human smokers do. It may be an accurate predictor of 
what smoking machines obtain under specifically programmed conditions, 
but it is not an accurate predictor of what people get from cigarettes. 
I'he dose-response relationship between FTC ratings and plasma nicotine 
levels is weak, except at low doses (Russell et al., 1980 and 1986; Rickert and 
Robinson, 1981; Benowitz et al., 1983 and 1986b; Robinson et al., 1983; 
Gori and Lynch, 1985; Maron and Fortmann, 1987; Coultas et al., 1993). 
I he relationship between cigarette dosage ratings and plasma nicotine levels 
may be better in studies using research cigarettes where nicotine content 
120 
