Tobacco Documents Online

I I I I I I I I I I I I I I I I I British Yournal of Addiction (1987) 82, 407-415 Craving for Cigarettes ROBERT WEST' & NINA SCHNEIDERz 'Department of Psychology, Royal Holloway & Bedford New College, Universiry of London, Egham, Surrey, U.K. and zDepanment of Psychiatry and Biobehavioral Sciences, UCLA, Los Angeles, California and Psychopharmacology Unit, Veterans Administration Medical Center, Brentwood, California, U.S.A. Summary The urge to smoke is by far the most commonly reported subjective response to cigarette withdrawal. It is worse in smokers with higher nicotine intake and can be reduced by a nicotine substitute. Although it may occur at any time during abstinence, it can be increased by stress and boredom and reduced when the smoker is kept occupied. Craving for cigarettes probably has multiple causation arising partly out of the perceived benefits of smoking and negative consequences of not smoking and partly as a direct result of loss of nicotine. Whatever its root causes it would be expected to involve general psychological processes such as associative learning and attributior T7tis paper draws together the threads of pharmacological and psychological levels of explanation for cigarette craving in an attempt to provide a clearer undentanding of this phenomenon. Introduction Craving is a feature of addictions as diverse as compulsive gambling and alcoholism. It is also potentially the most important feature of cigarette withdrawal.' Its importance lies in the fact that it directly undermines a would-be ex-smoker's will to give up cigarettes for good and thereby reduce his or her risk of ill-bealth and premature death. There are many theories addressing the question of depen- dence on smoking (e.g. Pomerleau') but little attention has been paid specifically to the possible bases of craving for tobacco. A major area of debate with regard to craving is the importance of pharmacological factors (specifi- cally the actions of nicotine) versus what may be termed psychological factors (e.g. the activity of smoking). This paper combines explanations at both pharmacological and psychological levels and argues that craving can best be understood in terms of an interaction of the two. Note that the terms `pharma- cological' and 'psychological' are used here merely as labels for phenomena which are most readily interpretable in such terms. It is not intended to suggest that such phenomena can only be thought of in this way. This paper is divided into three main sections. The first is a brief introduction to the notion of craving as used in the smoking literature and how it has been operationalized for the purposes of study. The next section examines existing evidence on craving. The third section considers possible expla- nations for cigarette craving in the light of the evidence. Repruat requesu to: R West, Department of Psqcbolog}•, Royal What is Clivin for Ci ttes? Holloway & Bedford New Colk~e, Egham Hill, E`ham, SutceY S San 'iVio OEX, ux Wants, desires and urges impel us to actions and 407

I I I 1 U I 1 I I U I I I I I I 408 Robert West & Nina Schneider provide them with directionr.The term 'craving' has been used to refer to strong desires or urges for activities which have become compulsive such as drug abuse or gambling. We use 'craving' in that same sense, as a subjective state at one extreme of a motivational continuum_ Although it is possible to engage in a debate about a more precise definition (e.g. whether it should be construed as a desire, an obsession or a behavioural urge), in practice there is little point-first because we do not have techniques for differentiating between the observable manifestations of the various alter- natives, and second because different definitions may simply reflect differing orientations. Measures of craving vary in scope and complex- ity. One which is commonly used combines several ratings of the 43-item Shiffman-Jarvik smoking withdrawal questionnaire including such questions as 'Would you find a cigarette unpleasant right now?' and 'If you could smoke freely, would you like a cigarette this minute?' Smokers respond on a scale from 'very definitely' to 'very definitely not'.' Others involve simply rating 'craving', 'desire to smoke', 'urge to smoke', 'missing a cigarette' or some such terms on fixed point-rating scales (e.g. Ikard & Tomkias; Gilbert & Pope;' West er al.4). An indirect means of studying craving is to measure smoking behaviour (e.g. Lucchesi er al.'). The assumption is that the more people want to smoke, the more they will do so. However, this may get at something different from ratings of craving. For example, the extent of smoking after a period of abstinence may not reflect the extremity of craving because loss of tolerance to nicotine's effects during abstinence may lead smokers to avoid large aicotine doses when they resume smoking despite strong craving. A complicating factor when using amount of smoking as an index of craving is how to operation- alize amount of smoking. Counting the number of cigarettes smoked is one way. Another is to count the number of puffs or the total puff volume. Yet another is to measure the increase in some bioche- mical matker of smoke intake such as plasma nicotine or expired-air carbon monoxide concentra- tion. Each of these measures may give different results. In the remainder of this paper, 'craving' will be used whenever there is the implication of a strong desire to smoke. Where different operationalizations of craving give different results, the measures used will be considered in more detail. Evidence on cigarette craving Craving for cigarettes is normally just one item in studies on cigarette withdrawal symptoms. In this paper, findings relating to craving will be extracted, and evidence on other withdrawal effects only brought in where this helps the understanding of craving as such. The evidence will be considered in terms of which smokers experience more severe craving, what is the time course and duration of craving following smoking cessation, and what are the pharmacological and non-phatatacological influ- ences on cigarette craving. Who Craves? For obvious reasons, it is extremely difficult to obtain prospective data on craving from representa- tive samples of smokers. This is compounded by a conceptual problem of deciding at what point the desire to smoke (presumably experienced by all smokers) becomes craving. However, one can gain a rough idea of the prevalence of craving in smokers at large from retrospective reports. Russell' asked a sample of 1500 smokers attending their family doctor whether they experienced craving when they had to go without cigarettes. Forty-five per cent reported that they experienced craving either 'fre- quently' or 'always'. Among smokers of more than 15 cigarettes per day the figure was 66%. Prospec- tive studies among volunteers or smokers clinic clients suggest that some 80% experience some craving during abstinence.'-' Several studies have tried to predict which smokers will experience the greatest craving. There are many ways in which one might classify smokers for this purpose. One is according to responses on a `smoking typology' questionnaire such as the Rea- sons for Smoking questionnaire.' High scores on subscales of this questionnaire relating to feelings of addiction have been found to predict greater craving during abstinence.10 Another study has found a high correlation between craving and scores on the 'dependence' scale of the Smoking Motivation Questionnaire (SMQ) developed by Russell et al. from a number of previous typology question- naires. u.,: The SMQ ezamiaes smokers' motives for smok- ing along a number of dimensions. These include smoking for reasons of dependence (because of perceived difficulty going without cigarettes), smoking for stimulant effects, smoking for sedative effects, indulgent smoking (smoking for pleasure), smoking automatically (without thinking about it),

I I I I I I I I I I I i j I I I I smoking for psychosocial image and smoking for something to do with one's hands and/or oral gratification. A second study using this question- naire found a positive association between craving over 2 weeks of abstinence and dependent, stimu- lant and indulgent scores." We should point out that the subjects in this study were smokers clinic clients using nicotine gum and this might have affected the correlations (see West et al.'1). However, it is hard to see how the use of nicotine gum could have created correlations which were not there to start with. Another way of classifying smokers is according to how much they smoke. Dose dependency is characteristic of withdrawal symptoms associated with drugs such as opioids and alcohol so establish- ing this effect for tobacco is important for under- standing tobacco dependence.' We have already mentioned one large scale study which found that retrospective reports of frequent craving were more common among smokers with higher cigarette consumption.' In fact, prospective studies have failed to show any consistent relation between strength of craving during abstinence and usual daily cigarette consumption."M However, smokers with higher levels of nicotine and other biochemical markers of smoke intake, do consistently crave more during abstinence."," The Duration and Time Course of Craving Studies in this area suffer from two major difficul- ties. First, one has the problem of getting hold of representative samples of smokers. Secondly, smok- ers inevitably drop out as duration of abstinence increases. Although there is no way of eliminating these biases it is important to restrict any analysis to those who managed abstinence throughout the period of study to avoid apparent changes in craving resulting from differential dropping out of subjects during the abstinence period. The most often cited study on the time course of withdrawal effects comes from Shiffman & Jarvik.' They reported data from 35 smokers during 2 weeks of abstinence and found a decline in craving during the first week followed by little or no change in the second. However, only 11 of their subjects reported complete abstinence throughout the 2 weeks of the study and there were no objective checks to validate claims of abstinence. A more recent study has looked prospectively at the time course of witbdrawal effects during 4 weeks of abstinence.' Data were collected from 52 smokers Craving for Cigarettes 409 clinic clients who managed complete abstinence during a 4 week treatment programme. These subjects used nicotine gum during their treatment, but they did not vary their consumption of gum during the period of study and it is not unreasonable to suppose that any effect of the gum would be constant throughout. The results of this study suggested that on average craving was at its height within the first 24 hours of withdrawal and declined thereafter. However, whereas amount of time spent craving declined in the first couple of weeks, intensity of craving remained high until the fourth week. It is worth mentioning that even after 4 weeks a substantial minority of subjects were still experi- encing strong craving. There appear to be no prospective data on craving beyond 4 weeks but retrospective reports indicate that smokers may experience what they refer to as craving years after stopping smoking.'s The fact that ex-smokers relapse to smoking months and years after having given up indicates that desire for a cigarette is aot necessarily tied to the acute with- drawal phase. Studies which examine time trends over short periods, such as within a single day, suffer least from problems of subjects dropping out and evi- dence is emerging from a number of studies that craving exhibits diurnal variation.s,' It is generally lowest in the morning, rising to a peak in the evening, then declining again at bed time. Non-pharmacological Influences on Craving for Cigarettes Urges to smoke are often reported as being stronger after meals.' They may also be associated with drinking coffee, and being with other smokers.l' Conversely, craving seems to be less in evidence when smokers are busy at work, or engaging in an enjoyable or physical activity." Stress is a common cause of relapse among would-be quitters" and can also lead to increased smoking.u There is evidence that anaesthetising a smoker's mouth, pharynz and lower-airways reduces the desire for a cigarette." This seems to imply that the sensation of smoke being drawn into the airways is one focus of craving. This appears strange at first sight because when youngsters try their first ciga- rette the irritancy of the smoke in the airways is aversive. Learning to smoke involves developing tolerance to the local irritancy of cigarette smoke leaving the smoker able to inhale it freely. Also, the interpretation of the above finding is made difficult I

I I I I I e I I I I I I i I I i I I I 0 Raberr ti'est & Nina Schneider by the fact that there was no difference in the intensity with which cigarettes were smoked follow- ing airways anaesthetization. Neither was there any difference in the adjudged desirability of the ciga- rettes after they had been smoked. Pharmacological Influences on Csaving for Cigarettes We discuss studies on pharmacological influences on cigarette craving under two headings: those involving nicotine and those where some other drug was used to influence craving. Nicotine and Craving. Chewing gum containing 2 mg of nicotine gives plasma nicotine concentrations of about a third of normal smoking levels• and can improve success rates in smokers ciinics.70 Three recent placebo-controlled studies have looked at ratings of craving during abstinence periods from 24 h to 5 days while using either nicotine or placebo gum.u'-n Ratings of 'craving' were not significantly different in the two conditions in any of the studies. However, the study which lasted 5 days found ratings of 'urge to smoke' and 'missing a cigarette' lower in the nicotine gum groupP It is difficult to explain why 'urge to smoke' and 'missing a cigarette' should have been affected by nicotine gum while 'craving' was not. It may be that the term 'craving' is subject to greater variability of interpretation by subjects and thus greater 'noise'. When asked directly whether nicotine gum re- duces their craving for cigarettes, smokers clinic clients usually report that it does." Clearly one has to consider such reports with caution for two reasons. First, because it may be merely a placebo effect. Secondly, the smokers may be reporting what they think the researcher wants to hear. On the other hand, smokers' reports of how well nicotine gum reduces their craving have been found to correlate with their prior levels of smoke intake." Smoke intake was assessed in this study using expired-air carbon monoxide (CO) concentrations. This mea- sure takes account of the amount of smoke inhaled from each cigarette and provides a better guide to nicotine intake than does daily cigarette consump- tion. Thus the greater the smokers' nicotine intake prior to abstinence the more helpful they found nicotine gum in reducing their craving. Usual daily cigarette consumption did not predict helpfulness of nicotine gum. It is unlikely that smokers would know their levels of smoke intake and that placebo effects could mediate the association between CO and reported reduction in craving. This suggests that nicotine gum did in fact help to reduce craving, and that it did so more among higher-intake smokets. It is worth noting in passing that ex-smokers who have been using 2 mg nicotine gum for over a year experience a range of withdrawal effects including craving when they have to abstain from the gum for a day, or when they switch to a low dose gum.=' The craving is not directed at cigarettes but at the gum, suggesting a transfer from one form of nicotine dependence to another. The view that nicotine replacement can reduce craving for cigarettes is strengthened by evidence that nicotine delivered by means of a transdermal patch may reduce craving during short periods of cigarette abstinence. Rose et a1.21 had subjects abstain from smoking for 90 min. In one condition they wore a transdermal patch designed to deliver nicotine through the skin of the forearm while in a second condition a placebo patch was used. The experiment was carried out double-blind. It appears that with the nicotine patch some reduction in craving for cigarettes was observed. This was despite the fact the subjects were unable to discern when they were getting the nicotine. Studies on the effects of nicotine on smoking behaviour tend to show a reduction in smoking when nicotine is delivered by another route. Luc- chesi et al.' found that an intravenous dose of 2-4 mg per hour led to a sma11 reduction in cigarette smoking frequency compared with a placebo condi- tion despite the fact that subjects were subjectively unaware of when they were getting nicotine. A subsequent study by Kumar et al.t' failed to show a decrease in smoking following intravenous nicotine but this may have been due to strong demand characteristics leading the subjects to smoke more than they would have wished (cigarettes were constantly kept lit in a holder that was suspended directly in front of their faces). Further evidence that nicotine substitution leads to a reduction in smoking comes from the finding that nicotine gum produces a slight reduction in ad lib smoking compared with a placebo." We have examined in preceding paragraphs evidence in which replacing nicotine appears to reduce cigarette craving. It is possible in principle to carry out the reverse operation-reducing nicotine intake while leaving the behaviottral component of smoking largely intact. Several studies have at- tempted to do this although only one has reported systematic data on craving.n In this study, 12

I I I I I I I I I I I I I subjects switched to an ultra-low nicotine cigarette for 10 days while a control group continued to smoke their own brand. Switching to the ultra-low yielding cigarette led to a 60% drop in plasma nicotine concentrations. This was accompanied by modest craving. This craving was less than that observed in a comparable group of subjects who were receiving nicotine from nicotine gum but who were not allowed to smoke;' despite the fact that the nicotine intake in both groups was broadly similar. This may have been because of the continued presence of behavioural components of smoking or the fact that the cigarette delivered their nicotine in bursts associated with each puff. It does suggest, however, that ievel of nicotine dose per se is not the only factor in craving. Other Drug Effects nn Craving. Evidence from studies using drugs apart from nicotine seems at first sight confusing in that both stimulant and sedative drugs may reduce cigarette craving. Several studies have been carried out on the effects of amphetamine and other stimulants on smoking. Early resul'ts were mixed. Some appeared to show a decrease in smoking whereas others showed an increase (see Low et al.p). All these studies suffered from methodological drawbacks, however. In a more recent study there was evidence that both ephedrine and d-aatphetamine reduced amount of smoking among heavy smokers over a week's observation compared with a placebo." Amount of smoking was assessed by measuring the weight of tobacco con- sumed. These drugs also led to reductions in feelings of 'enjoyment' from smoking and feelings of 'addictive craving'. On the other hand there was no effect on ratings of 'urge to smoke'. Again we are faced with different measures of craving producing different results, but on balance it does seem that ephedrine and amphetamine may reduce craving during ad lib smoking. One confounding factor when looking at the effects of drugs such as amphetamine on cigarette smoking is that they have a tendency to increase the rate of any ongoing behaviour. Thus an important stidy which has apparently not so far been at- t*-ntpted would be to look at the effects of ampheta- mine on craving during a period of abstinence. Amphetamine and ephedrine are not the only +timulant drugs to have an effect on smoking. There ualso some evidence that caffeine may reduce lmowng.'0 This may seem surprising given that smokers often light up with a cup of coffee. However, this could be because the circumstances Craving for Cigarettes 411 which provide a cue to drinking coffee also provide a cue for smoking. It is not only stimulant drugs which may reduce craving. Glassman et al." have recently reported that craving for cigarettes can be reduced by the noradrenergic agonist, clonidine. This drug causes drowsiness and has been used to combat withdrawal symptoms from drugs such as heroin. Glassman et al. argue that there may be some common mecha- nism underlying many or all withdrawal states. They also found, although this was not the main purpose of their study, that alprazolam reduced craving for cigarettes. It seems therefore that drugs with sedative as well as stimulant actions may reduce cigarette craving. Glassman er al. argue that clonidine's effect on craving was not mediated by drowsiness because it reduced craving more than alprazolam but did not induce significantly more drowsiness. However, to attach confidence to their view one would need to see drowsiness used as a covariate in any analysis involving craving. Our discussion of drug effects on cigarette craving would be incomplete without mention of some studies which appeared to show drug effects in increasing smoking. Stolerman et al.'~ have reported that giving the central cholinergic blocking agent, mecamylamine, to smokers increased their smoking rate. This has been interpreted as showing that mecamylamine partially blocked the central effects of nicotine, and the smokers smoked more to compensate. Carruthers" has reported in a similar vein that a centrally acting beta-blocker increased cigarette consumption. Unfortunately, there is a major problem with interpreting these results. Suppose the drugs had led to a reduction in smoking rate. This could equally be because they reduced the reinforcing properties of smoking. This has been offered as an explanation for the finding that naloxone reduces smoking." Another drug that tends to have a positive association with smoking is alcohol. There is some evidence that smokers puff more on their cigarette when given alcohol even outside social settings." Possible bsses for Cigarette Craving In this last section we look at possible explanations for cigarette craving in the light of the evidence reviewed. It is a trnism that we desire things which we ezpect'to enjoy or which we expect to relieve discomfort. Thus commonsense suggests that this might be the place to start for an explanation of cigarette craving. Of course the distinction betsveen I

I I I I I I I I I I I I i i I I I I I 412 Robert West & Nina Schneidtr pleasure-seeking and discomfort-avoidance is to some extent artificial as relief from discomfort can itself give rise to pleasure, and the banishing of pleasure can be unpleasant. Nevertheless it is convenient to maintain the distinction. Craving as Pleasure-seeking/13iscomfort-avoidance The first task in applying the pleasure-seeking/dis- comfort-avoidance view to cigarette craving is to establish the existence of benefits contingent on smoking and discomfort contingent on abstinence. It is then necessary to establish a link between these contingencies and craving. Perceived Berufus of Smoking. There is no difficulty in finding things about smoking that smokers report as beneficial Responses on smoking rypology questionnaires indicate that smokers may smoke to combat boredom as well as to relieve stress." For many smokers, cigarettes give them positive satisfaction while they are rela:ing.'= Some smokers also report that smoking gives them something to do with their hands," as indeed may the psychosocial facets of smoking such as its image and the opportunities it affords for establishing or maintaining social bonds." It may also be the case that, after a period of adaptation, the feeling of smoke in the throat and lower airways may give rise to satisfaction." Many of these benefits may be dependent on the actions of nicotine. There is a growing body of evidence that nicotine can have stimulant actions such as smokers appear to seek and also a calming effect (see West & Russell"). Other apparently pleasurable featum of smoking such as the feeling of smoke in the throat may have acquired their reinforcing properties by association with the effects of nicotine. Perceived Disadoaxsages of Abstinence. On the other side of the coin, it is now well recognized that when smokers abstain they experience a range of withdrawal symptoms which can make continued abstinence from smoking seem particularly unat- tractive. Most common are irritability, depression, hunger, restlessness and an inability to concentrate.' Many of these symptoms may arise from loss of nicotine.'J'42 It is conceivable that craving might represent a wish to escape from these symptoms. The link with craving. We have noted that positive associations were found between smoking for stimulant motives and craving as well as between smoking for pleasurable relaxation and craving." Added to this is the fact that both stress and boredom can increase smoking as one would expect if craving were associated with the capacity of smoking to sedate and stimulate.'T It also appears that both stimulant and sedative drugs may be able to reduce cravingr-' suggesting that these may spmehow substitute for smoking. It seems unlikely that craving represents merely a desire for an intoxicated state because alcohol seems if anything to increase smoking," and because most smokers experience little or no subjective drug effect from smoking except for from the first cigarette of the day. That is not to say, however, that smokers could not be seeking some very subtle drugged state specific to tobacco Turning to perceived discomiort contingent on not smoking, craving correlates positively with withdrawal effects such as irritability and rest- lessness'~" suggesting that the two might bear a causal relationship. One could also argue that drugs such as clonidine reduce craving at least in part because of their effects in suppressing other with- drawal symptoms." From this analysis it seettu possible that craving might be understood in terms of a combination of perceived advantages of smoking and undesirable feelings contingent on abstinence. This would also explain why smokers should crave cigarettes years after stopping smoking, particularly in trau- matic situations. They remember what they per- ceive as being the benefits of smoking in these situations. There are, however, certain aspects of the evi- dence which create some difficulties for this view. One problem is that, while craving does appear to be associated with positive functions which smoking is perceived to provide and withdrawal discomfort associated with abstinence, it frequently occurs at times when it is hard to see what function smokin€ could perform. A substantial minority of smoker- report craving all or almost all the time during the first week or more of abstinence.' For these smokers, craving is not limited to the desire to si and relax with a cigarette or to combat stress oi boredom. Neither is it associated with a need tc escape from other withdrawal symptoms because i is more commonly reported than these and i continues after these cease to be a problem.' Unde this view it is also difficult to see how intravenou nicotine can reduce craving without producing an other noticeable effects.'• t\.D Cw

I I I I I I I I I I I 1 i 1 I I I I Associative Learning and Craving for Cigarettes It seems that some elaboration of the simple commonsense explanation of cigarette craving is needed. The occurrence of craving in the absence of clear cues to benefits which might accrue from smoking may be due to the fact that smokers cannot know exactly which cigarettes will serve some function. Although, they might well have expecta- tions about certain cigarettes such as the after-meal cigarette, or the cigarette smoked when under stress, they may not be able to tell in advance whether other cigarettes will provide some pleasure, stimula- tioa, or whatever. In learning-theory terms, smoking would be conceived of as under a partial reinforce- ment schedule, and craving would mediate between the discriminative stimuli for smoking, which could be almost any external or interoceptive cues, and the act of smoking.= The ability of nicotine to reduce craving without providing any noticeable benefits is somewhat more difficult to understand. One would have to assume that the nicotine provided some cue, albeit sublimi- nal, which reduced craving by virtue of its associa- tion with benefits contingent on smoking. ~ At our present state of understanding one can- not rule out a complete explanation of craving in terms of the perceived benefits of smoking as compared with abstinence, allied to the vast and complex set of laws of associative learning. Thus the time course of craving may have much to do with extinction processes. If so, craving in the absence of strong cues that smoking will perform some valued function should be the first to go, while craving associated with powerful dis- criminative stimuli for smoking should be more persistent. Nevertheless, one important feature of craving still poses dif5culties. It is that there seems to be a quantitative mismatch between the putative benefits of smoking and the intensity of craving. There are many things in life that we desire very much, whether they be material possessions, relief from, boredom, or whatever. Yet these feelings of desire are mild compared with the 'gut' craving that many smoken report. This seems to suggest that in some instances craving has a genesis which is independent of the perceived benefits of smoking compared with absti- nence., Thus some part of craving may be a direct response to termination of smoking. It is conceiva- ble that mere termination of a behaviour as repeti- tive as smoking could create an urge to sustain the behaviour.sa It is also conceivable that craving could Craving for Cigarettes 413 arise directly from a physiological imbalance caused by termination of habitual nicotine intake. Craving as a Direct Response to Loss of Nicotine This latter explanation seems preferable in the light of the finding that it is smokers with higher nicotine intake rather than higher daily cigarette consump- tion who experience most severe craving. Also, smokers do not find merely going through the actions of smoking much use in relieving craving. The hypothesis is, therefore, that habitual nico- tine intake could in some smokers lead to compensa- tory changes in physiological functioning so that absence of nicotine provokes a rebound which is interpreted as craving for cigarettes. The two key features of this account are (1) that craving is not contingent either directly or through conditioning on any perceived benefits of smoking or other withdrawal symptoms associated with not smoking, and (2) that craving arises from an attributionaI process which interprets a given feeling as need for a cigarette. The physiological state giving rise to craving may well vary with time of day because of interactions with natural diurnal rhythms. This may be why craving is normally quite low in the morning. The physiological state may also be affected by emo- tions, physical exercise and drugs. It is not clear what this state might be-indeed we only label it as `physiological' on the assumption that it can best be described in terms of neural or hormonal function- ing rather than the more abstract concepts normally used in 'psychological' explanations. The involvement of attributional processes is a crucial feature of this view. It is presumed that the sensations associated nicotine withdrawal come to be thought of as craving for a cigarette because they are abolished by smoking. Thus a clear causal relation is established in the smoker's mind. This leaves open the interesting possibility that the same feelings could be given a different attribution in different circumstances and that other, similar feelings could be construed as cigarette craving. Thus one reason why craving was relatively slight after switching to an ultra-low yielding cigarette may be that sensations which would normally be labelled as craving were given some other interpre- tation. It is interesting to note in this regard that the only other subjective effect of switching was a marked increase in hunger.r Another important consequence of the involve- ment of attributional processes is that as far as the I

I I I I I I I I I I I I I I I I I I A 4 4. htAxD, F. F. & Tomritas, S. (1973) The experience of affect as a determinant of smoking behaviour. a series of validity studies, Journal of Abnormal Prychology, 81, pp. 172-181. 5. GtLeEteT, R. M. & PorE, M. A. (1982) Early effects of quitting smoking, Prychopharmacology, 285, pp. 537-540. 6. WErr, R. J., Jetevts, M., RussEU, M., CnnatrtHERS, M. & FEYEnnaarm, C. (1984) Effect of nicotine replacement on the cigarette withdrawal syndrome, British Journal of Addiction, 79, pp. 215-219. 7. Lvcam, B. R., SxusiEtt, C. R. & EmLzY, G. S. (1967) The role of nicotine as a determinant of cigarette smoking frequency in man with observations of certain cardiovascular effects associated with the tobacco alkaloid, Clinical Pharrlacology and ?kera- peu=, 8, pp. 789-796. 8. RussELL, M. A. H. (1978) Smoking addiction: some implications for cessation, in: J. L. ScitvNt'tz (Ed.) Progress in Smokinr Cetsation (New York, American Cancer Society). 9. WEsi, R., HAE]C, P. & BEUmt, M. (1987) Time course of cigarette withdrawal symptoms during four weeks of treataxnt with nicotine chewing gum, Addictive Behavio+s (in press). 10. Wn.i.uc, D. G. (1979) Different cigarette smoker clusi5cadon factors and subjective state in acute abstinence, Prychopharmacology, 64, pp. 231-235. 11. WEST, R. & RussELt, M. A. H. (1985) Pre- abstinence smoke intake and smoking motivation as predictors of the severity of agarette withdrawal symptoms, Psychopharmacology, 87, pp. 334-336. 12. RvssEU., M., Prro, J. & PATEt,, U. A. (1974) The classification of smoking by factorial structure of motives, Journal of the Royal Stasistical Sociery, Series A, 137, pp. 313-346. 13. WEsr, R., HAjEtt, P. & Brs.cEutt, M. (1986) Which smokers report most relief from aaving when using nicotine gum? Psychophannacolopy, 89, pp. 189-191. 14. HArstntAtl, D. K., Huom, J. R& PICtcENs, R. W. (1987) Blood nicotine, smoke exposure and tobacco withdrawal symptoms, Addictive Behafliours (in press). 15. F1.ETmffx, C. & DoLL, R. (1969) A stuvey of doctors attitudes to smoking, British ,joscrnol of Preventive and Social Medicine, 23, pp. 145-153. 16. Gttrtz, E. R. & JnRvtlc, M. E. (1973) Preliminary study: Forty-eight hours of abstinence from smoking, Paper presented at 81st Annual Convention of the American Psychological Association. 17. LtcSIzNSIm, E., ANToNt)IX7o, D. O. & RAttd- vA'rEF, G. (1978) The resumption of cigarette zmoking: a situational analysis of retrospective re- ports [unpublished manuscript]. 18. ScstncsTEit, S. (1978) Pharmacological and psycbo- logical determinants of smoking, Annals of Internal Medicine, 88, pp. 104-114. 19. Rosg, J. E., ZDaatt, M. C., TasHvrt, D. P., NswcoASa, R. & Ex27.E, A. (1984) Subjective re- sponse to ciguette smoking following airway anaesth- etiution, Addictstx Behaviors, 9, pp. 211-215. 20. JNtvts, M. J., RAw, M., Russ€u, M. A. H. & Frmustxa, C. (1982) Randomised controlled trial Craving for Cigarettes 415 of nicotine chewing gum, British Medical Journal, 285, pp. 537-540. 21. HoGtms, J. R., HAntnur,u, D. K., PtcxExs, R. W., KitAtrN, D., MeuNs, S. & LutumC, A. (1984) Effect of nicotine on the tobacco withdrawal syndrome, Prychopharmacology, 83, pp. 82-87. 22. SCxNEmFx, N. G. & Jettvtx, M. E. (1985) Nicotine gum vs. placebo gum: comparisons of withdrawal symptoms and success rates, in: J. Gttesomstu & S. HALL (Eds) NIDA Monograph 53 (U.S. Department of Health Education & Welfare). 23. W>ssr, R. & RussELL, M. A. H. (1985) Effects of withdrawal from long-term nicotine gum use, Prycko- loaical Medicine, 15, pp. 891-893. 24. RosE, J. E., HBttsxovtc, J. E., TRn.utaG, Y. & JARVrx, M. E. (1985) Transdermal nicotine reduces cigarette craving and nicotine preference, Cliaical Pharmaco!- ogy and T6erapeuacs, 38, pp. 450-454. 25. Ruaua, R., Coom, E. C., LADmt, M. H. & RussELL, M. A. H. (1976) Is nicotine important in tobacco smoking? Clinical Pharmacology and Therapnaics, 21, pp. 520-528. 26. RussELL, M. A. H., WusoN, C., FerFF.AsErrD, C. & CoLE, P. V. (1976) Effect on nicotine chewing gum on smoking behaviour and as an aid to cigarette withdrawal, British Medicai Jmmwl, 2, pp. 391-395. 27. Wssr, R., RtraPS.t, M., Jettvts, M. & Ft:YBtAar;rm, G(1984) Does switching to an ultra-low nicotine cigarette induce nicotine withdrawal effects? Prycho- pharntacology, 84, pp. 120-123. 28. Wrsr, R. (1984) Psychology and pharmacology in cigarette withdrawal, Journal of Prychosomaric Rr- uarch, 28, pp. 379-386. 29. Low, R. B., joNEs, M., CNtTgae, B. & CeDOttFr, R. J. (1984) The effect of d-ampbetamine and ephedrine on smoking attitude and behaviour, Addicrier Beha- viors, 9, pp. 335-345. 30. KouASS1Ct, L T. (1976) Effects of caffeine con- sumption on nicotine consumption, Prychopharnta- cology, 47, pp. 165-168. 31. GussmArt, A. H., JACUON, W. K., W,trstt, B. T. & RoosE, S. P. (1984) Cigarette craving, smoking withdrawal and clonidine, Science, 226, pp. 864-866. 32. STot.mtmArt, I. P., GoLnFAtta, R., Fttu, R. & JNevr, M. E. (1973) Influencing cigarette smoking with nicotine antagonists, Prychopharmacologia, 28, pp. 247-259. 33. CrwttTrdas, M. (1976) Modification of noradrenal- ine-related effects of smoking by beta blockade, Pkysioloaical Medicine, pp. 251-256. 34. KMULAs, A. & RertF, J. (1980) Nalosone reduces cigarette smoking, Life Sciences, 27, pp. 1541-1545. 35. Nts., R., Btrm, R. & BA'rnG, R. (1984) Effects of single doses of alcohol and caffeine on cigarette smoke puffing behaviour, Pharmacology Biochemi.ttry and Behavior, 20, pp. 583-590. 36. WFSr, R. & RussELL, M. A. H. (1985) Nicotine pharmacology and smoking dependence, in: S. R+ER- soN (Ed.) Psychopharnwcology: RaunsAdvmrces and Future Prospects (Oxford University Press). 37. Hurrr, W. A. & MATAttnZao, J. D. (1970) Habit mechanisms in smoking, in: W. A. HuxT (Ed.) Lsarniny Mechanirms in Smokin{ (Chicago, Aldine).