Tobacco Documents Online

DRAFT DRAFT DRAFT DRAFT DRAFT To: John Reynolds Date: May 1, 1990 Subject: current ideas on smoker From: Jim Walker behavior The purpose of this review is to commit to writing what I think we know about the interaction of the smoker with the cigarette. I am concerned primarily with non-pharmacological aspects but I think it is possible that several aspects of how the smoker interacts with the cigarette may have important pharmacological consequences. I. Introduction The philosophy behind this review can be summarized as follows : 1. The factors that will ultimately account for differences in the acceptance or rejection of our cigarettes lie not in the quantity of nicotine per cigarette or per puff but in other aspects of the chemistry of mainstream smoke that translate into sensory effects. That is, while nicotine does have clear 1

4^ sensory effects, it is the other compounds in smoke that(U_ modulate how the smoker delivers nicotine to himself. ~~ ~n.~~~ 2. There are physiological responses of smokers to ~ w s S "C;&A-e Arelf- cigarettes that can be used as indicators of the degree to which the sensory properties of the smoke are "pleasing" to the smoker. These are likely to be at least as sensitive and useful as verbal measures (e.g. ballots) as predictors of smoker acceptance and they have the advantage that they are largely unconscious. This review will focus on three of these: puff profiles, amount of smoke that is "wasted" after the puff is taken, and the characteristics of the first inhalation after the puff. 3. At present the solid evidence that these responses can be used to improve marketplace performance is weak although there are clear logical reasons for this hypothesis. There are straightforward experimental ways that we can determine the degree to which these responses can be used in this way and we should proceed to do these studies. 4. There are likely to be many aspects of the smoker-cigarette .,Yc interaction where it would be valuable to know what the ------------------------------ ~ chemical and neural bases of the responses are. One of the reasons that this would be important information is that it would allow us to intelligently use the extensive psychophysical and physiological literature that deals with the responses of animals to many of the compounds in smoke. ? ~.~~ ~ ~ ~- W,5 q ir c0~54u ,0~t 2

5. In order to use any or all of these three responses to better understand how we can make superior products, we need to learn how much each is primarily a matter of habit and how much it is a direct reflection of immediate chemosensory -'~q),-- stimulation. v~~ r In the segment below ("II. General conclusions about puff profiling, wasted smoke and post-puff inhalations") I consider the three stages of the smoker" s interaction with the cigarette (puff profile pattern, wasted smoke and first post-puff inhalation) and discuss, with each one, the following: A. the degree to which the measurement can be made accurately and unobtrusively B. the degree to which the behavior is a matter of habit versus the degree to which it is controlled on a moment- by-moment by chemosensory stimulation or other variables C. the chemical and neural determinants of the behavior D. the ability of the response, alone or in combination with other pieces of information, to predict consumer acceptance on a short-term or long-term basis Following this discussion of each of these responses, the next segment of the review ("III. Summary of smoker responses") presents a simplified schematic diagram of these three responses on a common time line and summarizes what I think are our key areas of ignorance. In the final section ("IV. ~ Recommendations for future research") I point out what I think m 00 , m 3

are the II. General conclusions about puff profiling, wasted smoke and post-puff inhalations A. Puff profiles 1. Can its measurement be made accurately and unobtrusively? This measurement is one of the most straightforward aspects of smoker behavior to measure. The technology was developed by Norman, Reynoldsand others in BGTC and has been refined over the past 15 years or so and has also been applied by several groups outside of the Biobehavioral division in an attempt to better predict the responses of different groups of smokers to different products. The technology has been shown to be extremely accurate (once calibration of each probe is performed) and it is relatively easy to use since it has been computerized to a great extent. In addition the extremely flexible tubing that is connected to the smoke probe would appear to allow the smoker to behave "normally". additional comments needed here by reynolds, robinson, griffith? (concerning ease of use, accuracy, lack of j' interference in normal smoking topography etc) 2. To what degree is this behavior a matter of habit versus being controlled simply by the immediate pattern 4

ke4 / F . 1 • 1S ' htlt a.v~~~~ ! tJ /j i ' ~fi./ a y"4 t „ (wr2C~*r-z9.'•c ~ ~ G~-4~tt "t t -'f., y}.iI - ,_ 1;- ---_ „ - r ...-,-. - (J slat` w'/s iv V.-r T'~k . c o /Y r , St ~ ~ ~ "t zSv~c .~ GN-cr-~L' ~ r' ` l ~ 1`~C { N&LA*.'6Ca r-L.di4-!'L"L+C CA.C'L,9. r Ct-~-'G J` l1/ ~ , ~, ii l .i'~"t'!'+C" " 119 S~ r A A ! ~S~t' S ~~a f ~s 1/Grll / ' '+...zr rc,t ~~S .~ F4s.t- ~ y ~ ~ {/ G i.~ y I o' _G'..L~f~1` CF r ' , '-f ~ olp--I ~" '/l C~~" ~~ G~G 5/ ~ ~ S c Zq Qr-tat tt. A,7 -t k - - - --Az t a ~ ` r~' ~f3l~: St Gd td i ~L ~" " ' cax.C e~-ol3 / l ~ tw f/Vd/e "i`SG ~!uG k vit/ ,y 7+43 S t«~L 9"'a v"p ~'G~ s Esc °^°~ ~+ u"7 1LL4 '~ Jr-a1sf~': 7~' Q.eat~G 5~,~ P. 11! ' '¢ ~~J'~'G fifGi[oc' fiL t~G ~ r -a ~ ~ ,I v sb r.ii~'w ;S>v. _(;~Fi s r t~wGr S ~ a4-t 7 ~-c~ , ) irp E~~ '~ L f -K

of sensory stimulation? This is an important question because it bears directly on the issue of how this response will be used to understand the sensory appraisal of cigarettes. Such an application would seem to be limited in direct proportion to the degree to which puff profile patterns are a matter of habit. On the other hand, if puff profile patterns are completely labile and are "driven" by the immediate pattern of chemosensory stimulation, puff profile patterns would be maximally useful. The experiments done at R&D to date do not directly address this question since, in most experiments, each smoker is tested for only a short amount of time and with only a extremely limited number of products. This issue breaks down into several sub- questions: a.• Within each product (or even across products), does each smoker exhibit a characteristic or signature puff profile pattern or set of patterns? b. Within each smoker, can puff profile data be used to differentiate between products that differ in terms of mainstream smoke chemistry? If so, do puff profile data do the job when ballot responses do not? c. Across smokers and/or products, are there general trends in terms of puff profile patterns? For example, can we say that later puffs are generally of lower volumes than early ones? If therc are consistent changes over the course of a cigarette, to what degree are these due to changes in mainstream smoke chemistry and to what degree might they be due 5

7~~f~i ~!Gls~2tir ef ~~J-~ '~°~~ .~ G~~~t`- ~..~ AM. Es ° ~ J l ,~ .re+ ~ VUC - ~ ~ ' ~ ~~ ~/ ~ .f~~-,~ ~~-~ ~~ ~ , - f ~~ C! (1 ,~ ,~ ~~ ~`'~ ~lC~! .~ S Ir 3 ~ Y• ~~ ~ , ~ ~ ~ ~. '~' V" 0 v , a ~.. ~ ~.~,~ ~ ~t~ ~ . `~ ~ ~~ ~~ R ~ - ~ s s~c . ~ - ~~ ,. ~~ l f2 ll ! e~ ~+ . l~L~l"!J w i. 7 a . F. . i~ + /i s J' '~} ~ • - - - -w` cv C /7°ZA^X.t. (.-l" . ' ~ { / ' ~, A~. ~ T 44~~ ~, ~ ~ , C3` ~ ~~,e~~~J ~r tr ' ~r,.~ l~ ~r 1t~ ~~,~.e:e , e ~ ~ ~ t ~.as 11 G!°' . x y~ -{"`.~"''i ~` ~# f . &"~ ~'ty rf ~,rq 7~ ~ ~t 7L L ~ .~ss~~ ~ V. ~-r n~v Z4

to changes in oral sensory cues (like aftertaste) ? d. What are the most useful statistical parameters of a puff? Is the entire duration of a puff equally useful or only up to the point where flow rate peaks? Does it make sense to do some sort of multivariate description of each puff and then use this value to correlate with acceptance? 3. what are the chemical and neural determinants of the behavior? There is a great deal of speculation about the components of smoke that may drive puff profile results and there is a small amount of speculation about the neural inputs that are most important. However, very little is actually known about either of these important areas. Most would probably feel that, in decreasing order of importance, the most important inputs are oral trigeminal chemoreceptors, true gustatory receptors, common chemical sense receptors of the nerves at the back of the throat and olfactory and/or trigeminal chemoreceptors of the nose. Of course it is possible that puff profile patterns are affected by the smoker's memory of the sensations from the lung from the previous inhalation, the sight of the smoke or his knowledge of the brand. One of the reason that it would be useful to know the chemical and neural bases of puff profile patterns is that psychophysical or electrophysiological data from animals (such as that now being collected by Roger Jennings) could then be used to predict how smoke chemistry changes would translate into 6

puff profile (and thus acceptance?) changes. The issue of the chemical and neural bases of puff profile patterns also breaks down into several important sub-questions: a. What is the sensitivity of different groups of chemoreceptors in the nasal and oral cavities to the major components of smoke? b. What is the effect on puff profiles of having people smoke with their nose clamped or their eyes closed? c. How early in a puff does chemosensory stimulation begin to affect puff topography? Does this stimulation have only an "volume-limiting" effect? d. Is the sole source of negative pressure in the oral cavity during a puff an initial increase in oral cavity volume or are the lungs involved in the generation of a puff? e. How large doe s the Auf_ _f volume ha ve _ t-G-h~e (or _ho:tiT long does the duration have to be) before chemoreceptors in the back of the throat are stimulated? f. Can we measure evoked electrocortical activity in humans in response to a puff and can this be used to determine the degree to which different groups of receptors are stimulated? 4. What is the mathematical relationship between this behavior and the rating by smokers of the product? 7

The answer to this question depends in large part on the answers to many of the questions above. If, for example, olfactory receptors are not involved, then one would not expect this response to be useful for predicting the flavor charaCteristics of smoke. Similarly, if the smoker uses only the experience at the end of the last puff to provide his sensory rating and there are major changes across There is not experimental support for the notion that there are universal relationships, across the population of smokers, between sensory appraisal and puff patterns. However there are good logical reasons to suppose that the same aspects of the chemistry of the smoke that result in shifts in the ballot ratings also may change puff profile behavior. In a study of oral pH, puffing behavior and sensory ratings (by Walker, Kurtz and Mangan) a combination of puffing behavior and smoke chemistry data was useful in predicting derived (by factor analysis) sensory factors. Studies comparing WINSTON and Marlboro using puffing behavior have indicated that puff profiling may reveal differences between products that would not be seen in the usual sensory evaluation tests. Clearly we•need to understand much more thoroughly the relationship between various puffing parameters and the smoker's appraisal of the cigarette. B. "Wasted smoke" 8