Smoker Compensation Studies

Compensation, the tendency of a person to smoke a cigarette having a lower machine-measured nicotine delivery more vigorously than a higher-delivery product, was mentioned for the first time in documents from the mid-1970s. The studies reported at the research conference held at Duck Key, Florida, in 1974 show that the tobacco industry was years ahead of the general scientific community in examining and understanding this phenomenon. The first study of compensation listed in the Surgeon General's 1988 report on nicotine addiction (3) is from 1980 (24). The work was confirmed three years later (25).

Specifically, the report of the conference notes:

The Kippa study [a study of nicotine delivery at the German BAT laboratory] in Germany suggests that whatever the characteristics of cigarettes as determined by smoking machines, the smoker adjusts his pattern to deliver his own nicotine requirements (about 0.8 mg per cigarette) .

It is recommended that such studies should be considered for application in other countries where B.A.T. has a substantial interest in understanding more about smoking behaviour either for direct commercial or for health reasons [emphasis added]. {1125.01, p. 2}

Later in the report, an additional experiment to measure compensation in the context of cigarette ventilation is recommended: putting small holes in the filter or paper so that air is drawn into the smoke stream, thereby diluting it and reducing smoke delivery.

The effect of ventilation on smoking behaviour should be explored in a McKennell-type test. {1125.01, p. 3}

By the 1970s dilution of mainstream cigarette smoke with room air had become a major technique for reducing machine-measured tar and nicotine deliveries. Termed "ventilation," dilution was accomplished with a variety of techniques, including the use of porous cigarette paper and the drilling of holes around the filter tip. We do not know what a "McKennell-type test" is.

An apparent replication of the German work was completed the following year in Canada. The "Chronology of B&W's Smoking and Health Research" includes the following item for 1975:

Compensation study conducted by Imperial Tobacco Co., a BATCo affiliate, [shows that a smoker] adjusts his smoking habits when smoking cigarettes with low nicotine and TPM [total particulate matter] to duplicate his normal cigarette nicotine intake (Imperial Tobacco Project T-8077) [emphasis added]. {1006.01, p. 27}

Overall cigarette consumption data in the United Kingdom between 1965 and 1973 demonstrated a compensation effect {1190.09}. In 1975 G. F. Todd of the Tobacco Research Council analyzed per capita cigarette consumption for men and women for each year over this nine-year period, during which tar yields were steadily declining. Per capita cigarette consumption rose for both men and women through the period, indicating that smokers were smoking more in partial compensation.

Additional work on compensation was conducted by an outside contractor but reported through the R&D lab at Southampton by J. R. Courtney and A. K. Comer {1208.02}. Completed in 1978, the study analyzed the butts of cigarettes smoked by humans and compared the smoke constituents with those found in machine-smoked cigarettes. The technique permitted an estimation of actual delivery of smoke to the consumer. The study found compensation effects.

By 1980 B&W was beginning to recognize the legal and political dimensions of smoker compensation. In response to a questionnaire circulated by Dr. Alan Heard of BAT R&D about the proposed agenda for the upcoming research conference at Sea Island, Georgia {1132.01}, Dr. R. A. Sanford, vice president of R&D at B&W, indicated that B&W had "definite interest" in the proposed discussion of "compensation/smoker behaviour," but he also noted that this area of research was "dangerous" and questioned, "Is this in our best interests?" {1132.01, p. 3}. Research on smoker compensation was showing that consumers smoked cigarettes in ways that defeated the low-tar designs of the products. This finding undermined the implied benefit of low-tar smoke as less hazardous than normal cigarettes. In any event, the minutes of the conference did not include a discussion of compensation {1177.01}.

The 1983 research conference in Rio de Janeiro covered a variety of topics related to nicotine {1180.07}, including smoker compensation. The minutes of the conference note,

Compensation is now attracting the interest of Government and medical authorities in many parts of the world. This is based on the increasing number of new studies and, in part, by the evidence submitted by the industry to the FTC [Federal Trade Commission] in the Barclay investigation—much of which has already been communicated to Government authorities in Australia, Belgium, Finland, Holland, Switzerland and the UK.

There is now an urgent need to assess whether there are ways in which the industry can either counter the situation or alternatively turn it into a commercial advantage.

A direct consequence of this growing interest in compensation is the possibility that the FTC, and other authorities, may call for a change in the standard smoking machine test procedure for all products. If this were simply to be a modification to the existing standard procedure (increased puff volume, duration or interval) the effect would be to increase delivery levels but it would probably have little effect on League Table rankings [relative tar and nicotine yields]. A more extreme possibility is that an entirely new test procedure could be developed, eg a biological index. ...

Either move would weaken the concept of low tar and would both confuse and concern the smoker . Operating Companies around the Group should, therefore, do everything possible to defend and maintain the present standard test procedure. If, however, the FTC or any other authority takes action to change the procedure the strategy should then be to stretch out any discussions (both with the authorities and later at ISO) until exhaustive studies have established that an alternative procedure is in fact more relevant.

In the meanwhile it is essential that we should increase our own research into how and why people smoke: eg what the smoker needs or gets from the cigarette in terms of nicotine and other sources of satisfaction. Until we have such knowledge we shall not be in a position to judge what would be best for the industry in the longer term.

...

Whatever the outcome of the various public debates on compensation and test procedures, we must aim to use our knowledge to develop products that give improved smoker satisfaction. The concept of 'smoke elasticity' can be expected to play an important role [emphasis added]. {1180.07, pp. 8–9}

The Barclay investigation was spurred by B&W's competitor's crying foul to the FTC about the test results for B&W's Barclay brand of low-tar cigarettes. Competitors charged that the Barclay cigarette was designed to be smoked one way by the machine and another way by the consumer. The resulting investigation focused FTC attention on the compensation problem for the first time and provided the agency with some of the industry's data on the problem. This increased attention did not lead to any change in regulatory approach for many years. (It was not until late 1994 that there was some visible movement on this issue. In

December the National Cancer Institute convened a meeting to examine alternatives to the FTC test method.)

BAT scientists were concerned that a governmental agency might impose a new measure for product labeling which would have an uncertain or at least initially unknown relationship to product strength and nicotine satisfaction. The company's strategy was to jawbone regulators to delay any settlement of the matter until it could be sure that any resulting labeling scheme would not upset its ability to market cigarettes effectively to its various market segments.

BAT scientists, meanwhile, continued to work on compensation. By 1984 the consensus among them, as the proceedings of a joint R&D/marketing conference indicate, was that consumers adjust their smoking behavior to achieve specific nicotine dosing. The immediate signal for gauging nicotine dose is the "impact" of the puff on the throat. The impact, in turn, is directly related to nicotine dose.

[It] is accepted that nicotine is both the driving force and the signal (as impact) for compensation in human smoking behaviour. {1226.01, p. 56}

The 1984 Montreal conference materials also demonstrate that BAT laboratories were evaluating design features of cigarettes that influenced a parameter called "elasticity." Elasticity quantifies the ability of the consumer to affect nicotine dose through compensation (by smoking the cigarette differently). The study appeared to measure the effect of different cigarette ventilation designs on smoke composition under conditions of differing puff volumes {1226.01, p. 58}.

The fact that people smoke in ways that defeat the engineering tricks of low-tar, low-nicotine cigarettes did not become widely appreciated in public health circles or the medical profession generally until the early 1980s, when Neal Benowitz published his studies showing that consumers smoke low-yield cigarettes more vigorously than they do higher-yield ones (25). In fact, the doses of nicotine delivered from all but the lowest-yield cigarettes are remarkably similar, and those from the lowest-yield versions are far higher than would be predicted from strict comparisons of machine-measured yields. Nonetheless, until about 1983 standard primary care medical textbooks advised physicians to recommend low-tar cigarettes for patients who were unable to stop smoking (26). Indeed, the tobacco industry promoted this view (see chapter 9).

Had the results of these internal BAT studies been generally known in the mid-1970s, medical advice in this matter might have changed earlier. The fact that smokers compensate for low-yield cigarettes seriously

undermines the implicit claim that low-tar, low-nicotine cigarettes reduce the risk of harm from smoking. In 1994 Jack Henningfield and his colleagues at the National Institute on Drug Abuse advanced a proposal to change the way the Federal Trade Commission (FTC) requires nicotine delivery to be reported by manufacturers, so that the amounts reported will more closely reflect the actual bioavailability of nicotine to the consumer (27).

A research conference held at Pichlarn, Australia, in 1981 focused some discussion on the regulatory aspects of compensation.

It is felt that the time is close when Government agencies worldwide will take more notice of compensation—and of the scale of the differences, for a given commercial product, between smoking machine numbers and the dose of smoke actually obtained by smokers. This issue may well go beyond the simple technical measurement of deliveries. If for no other reason than defence, we must pay increasing attention as to how our products—especially new products—are smoked by different categories of smokers. {1178.01, pp. 13–14}

The concern was misplaced at two levels. First, governments did not get seriously interested in this problem for another dozen years. Second, dealing with the compensation problem in a conscientious manner requires more than taking a merely defensive, public relations-oriented posture.