Tobacco Documents Online

I I I I I I I I I I I I I I I I I I I Nicotine pharmacodynamics: some unresolved issues Robert J. West Psychology Department. Aoyal Holloway and Bedford New College, London Universty Egham, Surrey TW20 OEX. UK Abstract. This paper focuses on some issues in the field of nicotine pharmaco- dynamics in which widely held suppositions have outstripped the supporting evidence. It considers how far the view that nicotine acts as a stimulant in low doses and as a sedative in higher doses is supported by the data and concludes that within the range of doses ingested by cigarette smokers, only stimulant actions have been reliably observed. It examines evidence for the view that nicotine improves ability to sustain attention and concludes that a positive effect of nicotine not attributable to relief of a withdrawal deficit has yet to be demonstrated. Finally, it considers the issue of physiological tolerance and argues that ideas concerning a role for chronic tolerance in nicotine dependence have yet co be supported empirically. Despite advances in our understanding of nicotine's c.°fects in recent years there is still much work to be carried out before fundamental issues underlying its addictive potential can be resolved. 1990 The biology of nicotine dependence. Wiley, Chichester (Ciba Foundation Symposium / S2) p 210-224 Now that there is a consensus that nicotine is addictive (US Department of Health and Human Services 1988), it is increasingly important to examine what it is about the actions of nicotine that underlies this propensity. This involves examining dose-response relationships and issues of tolerance-in other words, the pharmacodynamics of nicotine. Rather than attempt a comprehensive review of nicotine pharmacodynamics, this paper will focus on issues about which assertions and beliefs have tended to outstrip the supporting evidence. In some instances, individual studies will be selected for critical examination. The purpose is not to make light of these studies, rather to recommend caution in interpreting results which we have been too ready to accept and use as a basis for generalizations. This paper will consider three major areas where more or less widely held beliefs about the effects of nicotine do not as yet have sound empirical backing. In all three, there are potentially important implications for the understanding 210

I I I I I Nicotine pharmacodynamics 211 of cigarette dependence. These areas are: biphasic actions of nicotine on arousal depending on dose and/or concurrent stressors; effects of nicotine on cognitive performance, and the role of chronic tolerance to nicotine in cigarette dependence. Biphasic effects of nicotine on arousal The simple form of the hypothesis is that low doses of nicotine stimulate and high doses sedate. This biphasic action is proposed as an explanation for smokers' reports that cigarettes can both pep them up and calm them down (Ikard & Tomkins 1973, Ashton & Golding 1989, US Department of Health and Human Services 1988). This view has followed from a range of studies in animal and human subjects examining the effects of nicotine on physiological measures of arousal. The neurochemical basis often proposed to underlie this biphasic action is the action of nicotine at neuromuscuiar junctions; when the nicotine molecule first hits the receptor it helps initiate muscular contraction but this is followed by prolonged paralysis possibly due to blocking of subsequent depolarization (Paton & Savini 1968). With high doses one may obtain a net effect of prolonged receptor blockade. It has not been difficult to show effects of nicotine which may be interpreted as increased activation of the autonomic and central nervous systems or arousal. These include increases in the hormones adrenaline and cortisol as well as desynchronization of the resting electroencephalogram (EEG), increased heart rate, and decreased skin temperature resulting from peripheral vasoconstriction (e.g. Cam & Bassett 1983, Frankenhaeuser et al 1968, Golding 1988, Pickworth et al 1988). Although there has been widespread acceptance of the view that larger nicotine doses have an opposite effect, including statements to this effect in the recent US Surgeon General's report on nicotine addiction (US Department of Health and Human Services 1988), the evidence is far less convincing. One of the studies most often cited is that of Armitage et al (1969). Second- hand reports of this study variously interpret the findings as showing that low doses of nicotine cause desynchronization of the resting EEG whereas larger doses cause increased synchronization, or alternatively that low doses desynchronize and high doses cause mixed synchronization and desynchronization effects. However, in the study concerned, the dose was not varied. In one experiment the cats were given 2 pg/kg every 30 seconds, whereas in the other they were given 4 pg/kg every minute. In neither experiment were the results subjected to statistical analysis, but the 'mixed stimulant-depressant effect' of the 4 pg/kg dosing schedule could more parsimoniously be considered 'no reliable effect'. Recent findings have in any case thrown doubt upon the interpretation of all EEG data from smoking studies. Knott (1989) has noted that increased wakeful alertness is often associated with an increase in alpha power rather than

I I I I I I I I I I I I I I I I I I 212 West a decrease and that when smokers are allowed to smoke their own cigarettes in a manner of their choosing, one observes increased alpha power and decreased theta and delta power. The classic desynchronizazion of the EEG found in earlier studies, which involved decreased alpha and increased beta, may well have been due to the use of unfamiliar cigarettes or unduly high nicotine ingestion. Another series of studies cited as showing stimulant effects of nicotine at low doses and sedative effects at high doses is that by Ashton et ai (1974, 1980). In one of these studies (Ashton et a! 1974), smoking a cigarette was reportedly associated with a change in Contingent Negative Variation (CNV), a build up in electrical potential in anticipation of a stimulus to which a response is required. There was an increase in magnitude in four smokers, a decrease in seven and a`biphasic' action in four. Although statistical analyses were reported to support the view that these effects were genuine, these involved post hoc examination of individual results after removal of the 'sign'. This, together with lack of adequate controls for temporal order effects, means that the findings must be treated with extreme caution. It was briefly reported that similar results were obtained on 11 of the same subjects in a second experiment but details were not given and reference was made to biphasic results within individual subjects, which may equally well be interpreted as lack of reliable results. The doses of nicotine obtained by subjects from their cigarettes were not measured but it was argued that one possible reason for the results was a biphasic dose-response effect of nicotine. This was based on a negative correlation between extraversion and butt nicotine content and a positive correlation between the magnitude of change in CNV and extraversion. The effect was examined more systematically in a later intravenous nicotine study (Ashton et al 1980) in which eight subjects were given a range of doses of nicotine intravenously and dose-response curves were plotted. These were interpreted as showing a biphasic relationship, but they could equally well have been interpreted as showing no relationship. A biphasic relationship would involve intermediate points on the curve being, on average, higher than the end points; this was the case in only three or possibly four of the eight subjects. Averaged data were also presented, adjusted to maximize the size of any dose- response relationship by choosing the maxima and minima for each subject. These appeared to show a biphasic effect but the post hoc selection of data points throws doubt on the statistical validity of the exercise. It is important to note when evaluating these findings that the dose at which the maximum stimulant effect was apparently observed was around 0.05 mg and the dose at which the greatest sedative effect was observed was 0.4 mg. Given that the average intake from a cigarette is around 1 mg, all of the doses used in the CNV studies were on the low side. It is not reported whether any or all of the subjects in the experiment were smokers and if so whether they had abstained prior to the tests. However, even if they were non-smokers, the dose creating the apparent stimulant effect would be little more than a placebo.

I I I I I I I I I I I I I I I I I Nicotine pharmacodynamics 213 More recently, another theory relating to possible biphasic actions of nicotine has been proposed and gained acceptance. This is that a given dose of nicotine will stimulate in certain conditions and sedate in others. In particular, de- arousing effects of nicotine will be found when subjects are stressed. The most commonly cited study supporting this view is that by Golding & Mangan (1982). In that study smokers (degree of prior abstinence not reported) smoked a cigarette either while being subjected to loud bursts of white noise or while relaxing on a bed. It was reported that skin conductance level and skin conductance response showed stimulant effects (increased conductance) of smoking in the 'sensory isolation' condition but sedative effects (decreased conductance) in the stress condition. In fact, real and sham smoking responses did not differ significantly in the stress condition. Alpha power in the EEG was reported as being significantly decreased by real smoking compared with sham smoking under conditions of sensory isolation, reflecting greater arousal, but increased under conditions of stress. However, the difference between the alpha power during real smoking and sham smoking was not significant in the stress condition. Thus the critical test for a sedative effect did not receive statistical support. The Golding & Mangan data, while suggestive of an effect, could be interpreted in terms of a general increase in arousal caused by smoking which becomes obscured when subjects are already aroused. The difficulty in replicating effects in this area is illustrated by a recent study testing a more elaborate theory, namely that the subjective de-arousing effects of smoking are associated with a relative reduction in activation of the right hemisphere compared with the left hemisphere (Gilbert et al 1989). In that study reduction in alpha power was apparently found only in the right hemisphere during stressful episodes in a film. No effect was found on the left hemisphere, which is where Golding & Mangan (1982) had placed their electrodes. It appears that the results of the studies most often cited in support of the biphasic action of nicotine on arousal are at best suggestive. Church (1989) in a comprehensive review of the EEG literature has also cast doubt upon the validity of claims that nicotine can have de-arousing effects on the EEG. If a dose-response relationship is to be postulated as a possible factor underlying smoking motivation and smoking dependence, large-scale, carefully controlled studies must be performed in double-blind trials with multiple doses using both smokers and non-smokers. The analysis of the data should be undertaken using predetermined routines, preferably by computer to avoid bias in interpretation, or if they are examined by eye, this should be done 'blind'. All bands within the EEG spectrum should be analysed and a more sophisticated view of arousal adopted, which would include the possibility of both arousal associated with aversive or novel stimuli, and relaxed concentration. Before leaving the issue of sedative smoking effects, it is worth noting that reference is often made in reviews to animal studies which appear to show that high doses of nicotine have depressant effects. For example, Clarke & Kumar I

I I I I I I I I I I I I I I I I 214 'Nest (1983) showed a dose-dependent reduction in locomotor activity followed by increased activity later in the testing session. Unfortunately, whereas the original authors of these studies are careful to use terms such as 'depressant' to refer to locomotor activity, without implying anything about levels of sympathetic or cortical arousal or psychological sedation, reviews have sometimes taken the word 'depressant' out of context and taken the evidence as support for general sedative effects of nicotine. Yet clearly, locomotor activity can change for any number of reasons, for example it can decrease because of induction of nausea, fright or paralysis. In the case of high nicotine doses, it appears that the effect is characterized by a large and generalized loss of muscle tone. This contrasts with the immobility and rigidity caused by high doses of morphine. Effect of nicotine on cognitive performance The second issue to be addressed in this paper is that of the nature and extent of improvements in cognitive performance attributable to nicotine. It has been argued that such effects could be a major factor in why people smoke and could explain why they find it difficult to give up. Foremost among these effects is an enhanced ability to maintain attention over time. This has apparently received support from a wide range of studies in which smokers and non-smokers have variously been asked to smoke cigarettes or use nicotine tablets. The general finding appears to be that under control conditions there is a decrement over time in performance on tasks requiring continuous sustained attention, and nicotine prevents or reduces this decrement (e.g. Tong et al 1977, Wesnes & Warburton 1978). However, a close examination of the evidence calls into question the conclusions which many now take for granted. There is little doubt that when smokers smoke cigarettes, their ability to sustain attention is better than when they abstain. There is greater doubt about to what extent this is a positive effect rather than alleviation of a withdrawal effect, and also to what extent nicotine is involved. One source often cited is that of Wesnes & Warburton (1978). In their study, subjects had to detect changes in regular movement of a hand around a kind of clock face. Unfortunately, a crucial experiment in which non-smokers were given nicotine tablets failed to show any effect. A subsequent experiment in which nicotine tablets were given to light and heavy smokers revealed a marginally significant effect of the tablets only when one subject had been eliminated on the grounds that he or she was an outlier, i.e. his/her results were different from those of other subjects. This effect of the tablets was for the combined group of heavy and light smokers, but conclusions were drawn as though the result had been observed separately for the light smokers. Wesnes et al (1983) also reported an effect of nicotine tablets on sustained attention in the clock task described above-this effect being the same in heavy smokers, light smokers and non-smokers. Some important features of this report

I I I I I N cot ne pharmacodynamics 215 make it difficult to evaluate the conclusions drawn. The first is that all scores were ~ expressed as decrements from the scores obtained during the first 20 minutes. This leaves open the possibility that baseline scores for the subjects given nicotine were worse than those for the control subjects, so that improvements over time were due to habituation to a stimulus that adversely affected performance. ~ Secondly, if one is to claim specifically that nicotine improves performance in non-smokers, it is necessary to test positively for an effect in this group, not to rely on the absence of a significant difference between effects in groups of ~ heavy smokers, light smokers and non-smokers. Finally, examination of the ' data indicates that the pattern of the dose-response relationship is irregular. For example, the expected dose-dependent effect is found only in the second I of three time periods. In the first there was no difference between responses in the presence or absence of nicotine; in the third results with the 1 mg dose differed from those with the 2 mg and the placebo, which were similar. A potentially important study following from that just discussed was carried ~ out by Wesnes & Warburton (1984). In that study nicotine tablets of varying doses were given to non-smokers. These subjects then had to detect repetitions of digits presented one at a time over a period of 30 minutes. It was reported ~ that there was a dose-dependent alleviation of a reduction in performance over time. Thus there was no decrement in performar.ce when subjects were given 1.5 mg nicotine tablets, but a substantial reduction when subjects were given ~ a placebo. As with the previous study, the fact that the results were expressed as a difference from baseline (the first 10 minutes on task) makes it impossible to know whether the result was due to a nicotine-induced temporary decrement in performance in the early part of the session. Secondly, the authors used 'hit ~ probability' rather than stimulus sensitivity as a measure of accuracy. It is not clear whether this was attributable to a change in response bias rather than a genuine improvement in performance. Even so, the only significant difference ~ was between subjects given 1.5 mg nicotine and the combined results from those given tablets containing 1 mg, 0.5 mg or no nicotire in the 10 minutes following the baseline period. No significant effects were observed on response times. I Despite the extremely weak nature of the findings, the above studies are frequently cited as evidence for the view that nicotine has a positive enhancing effect on sustained attention. It may turn out that such claims are accurate, but at present they go beyond the data reported in the literature. It is worth ~ noting in relation to this that Snyder et al (1989) have reported reliable decrements in performance on a range of information processing tasks during cigarette abstinence, and that at least some of these show a return to pre- ~ abstinence values within 10 days. This suggests that certain performance enhancements of cigarettes may turn out to be alleviation of a withdrawal effect. On the other hand, A est & Hack (in preparation) have found a significant improvement in performance on the Sternberg Memory Search task in smokers ~ who smoked a cigarette without prior abstinence. This would suggest an effect I

I I I I I I I I I I I I I I I I I I I 2!E West which is not subject to acute or chronic tolerance and from which one would expect no withdrawal rebound. West & Jarvis (1986) have also shown that a 2 mg nasal nicotine solution continued to improve the maximal rate of finger tapping in non-smokers with repeated doses over the course of a day, suggesting little acute tolerance to this effect. This demonstrates performance-enhancing effects of nicotine without prior deprivation and therefore can be postulated as a positive effect rather than merely relief of withdrawal. Tolerance and nicotine dependence Tolerance is believed to play a role in drug dependence. One view is that tolerance results in physiological changes so that the drug concerned is eventually taken primarily to stave off a highly dysphoric withdrawal syndrome. A variant on this idea is that there is neuroadaptation leading to disturbance of one or more motivation systems, so that chronic ingestion of the drug accentuates drug seeking and/or desire to use the drug, this effect being independent of other aspects of the withdrawal syndrome. Another possibility is that tolerance acts in a facilitatory manner. Thus it may selectively decrease aversive reactions to moderate or high doses of the drug, leaving users free to enjoy positive effects on which they then come to depend. In addition to the notion of acquired tolerance, there is the concept of constitutional tolerance (or its obverse, sensitivity) to one or more drug effects as a predisposing factor to the development of dependence. This has been proposed with respect to alcohol dependence and might apply to nicotine. Tolerance to certain effects of nicotine has been carefully studied and much is known about it, most notably the influence of nicotine on heart rate. It is now well established that the increase in heart rate caused by nicotine is subject to acute tolerance, and that this does not simply reflect a ceiling effect. If one plots the hysteresis loop of the rise in heart rate during nicotine infusion and the subsequent fall in heart rate post infusion against plasma nicotine concentrations, one finds that heart rate at a particular plasma nicotine concentration on the downward part of the curve is less than on the upward part (Benowitz et a! 1983). Moreover, a subsequent nicotine dose will have reduced efficacy (Porchet et al 1988). Studies have also shown that tolerance to nicotine's effect on heart rate disappears very rapidly during the first 24 hours of abstinence (e.g. West & Russell 1987), indicating that there is little residual chronic tolerance. Further to this, a recent study by West & Hack (in preparation) showed that occasional smokers showed the same profile of heart rate boost when smoking a cigarette on a normal smoking day and after 24 hours abstinence as regular daily smokers. Thus we found evidence of minimal chronic tolerance associated with regular as opposed to occasional nicotine ingestion. Unfortunately, claims that have been made for a relationship between heart rate tolerance and psychological dependence (Hughes & Hatsukami 1986) are

I I I I- I I I I I I I I I I I Nicotine pharmacodynamcs 217 based on a different picture of tolerance. Hughes & Hatsukami found a negative correlation between the heart rate boost per estimated unit of nicotine intake from a cigarette smoked prior to a period of abstinence and the severity of withdrawal discomfort during that period of abstinence. Their interpretation was that withdrawal severity is associated with chronic tolerance developed over a period of months or years of use. However, as just mentioned, tolerance to the heart rate effects of nicotine are primarily acute. In a study by West & Russell (1988), the relationship between heart rate boost from a cigarette smoked after 24 hours abstinence when most or all acute tolerance would have vanished, was positively correlated with craving. This was contrary to what one would expect from a chronic tolerance model, but was consistent with the idea of a relationship between constitutional sensitivity to nicotine predisposing to physical dependence. These correlational data suffer from limitations in assessment of nicotine doses obtained from the cigarettes and would need to be replicated using known doses of intravenous nicotine before firmer conclusions could be drawn. In addition, it would be necessary to take into account the fact that different nicotine effects are differentially subject to tolerance. The peripheral vasoconstriction effect does not appear to be subject to tolerance at all (e.g. Benowitz et al 1983). Possibly more important from the point of view of understanding psychological dependence is tolerance to the subjective effects of nicotine. Very little systematic data have been collected to address the issue of acute and chronic tolerance to nicotine's subjective effects. A study by Jones et al (1978) is widely cited but it was not controlled. West & Russell (1988) showed that a considerable amount of tolerance to the dizziness and nausea induced by nicotine disappears after 24 hours' abstinence, suggesting that it is largely of the acute kind. What is required is a series of studies similar in design to those carried out by Benowitz and his colleagues (1983) on heart rate, but using a battery of subjective and performance measures. From the very patchy data so far obtained, it seems unlikely that a simple chronic tolerance model for causation of withdrawal symptoms such as craving will suffice. Examination of individual differences in development of pharmacokinetic tolerance and pharmacodynamic tolerance to nicotine may provide an understanding of why different smokers seek different levels of nicotine from cigarettes and why, even allowing for this, they are differentially dependent on their cigarettes. , Conclusion It is important to reiterate a point made at the beginning of the paper, that ' in cases where studies have been picked out for critical examination, the purpose Z1Z has not been to belittle these. Indeed, it is because they have dealt with important C~ issues imaginatively that they have been widely cited and influential. The point ~P of this paper has been to suggest a more conservative approach to adopting ~ ~ findings in reviews, even though they may appear to confirm existing beliefs. ~ O ~ W ~ , 00 N~

I I I I I I I I I I I I I I I I I I I 218 West The studies described provide an excellent beginning for more detailed examination of the issues, which will no doubt help our eventual understanding of cigarette dependence. References Armitage AK. Hall GH. Sellers CM 1969 Effects of nicotine on electrocortical activity and acetyicholine release from the cat cerebral cortex. Br J Pharmacol 35:152-160 Ashton H, Golding JF 1989 Smoking: motivation and models. In: Ney T, Gale A (eds) Smoking and human behaviour. Wiiey. Chichester p 21-57 Ashton H, !vtiliman JE. Telford R. Thompson JW 1974 The effects of caffeine, nitrazepam, and cigarette smoking on the contingent negative variation in man. Electroencephalogr Clin Neurophysiol 37:59-71 Ashton H. Marsh VR, Miliman JE, Rawlins MD. Telford R, Thompson JW 1980 Biphasic dose-related responses of the CNV (contingent negative variation) to IV nicotine in man. Br J Ciin Pharmacol 10:579-589 Benowitz NL, Peyton J, Jones R, Rosenberg J 1983 Interindividual variability in the metabolism and cardiovascular effects of nicotine in man. J Pharmacol Exp Ther 221:368-372 Cam GR, Bassett JR 1983 The effect of acute nicotine administration on plasma levels of thyroid hormones and corticosterone in the rat. Pharmacol Biochem Behav 19:559-561 Church RE 1989 Smoking and the human EEG. In: Ney T, Gale A (eds) Smoking and human behaviour. Wiley, Chichester p 115-140 Clarke PBS, Kumar R 1983 The effects of nicotine on locomotor activity in non-tolerant and tolerant rats. Br J Pharmacol 78:329-337 Frankenhaeuser M. Myrstea Waszak M, Neri A, Post B 1968 Dosage and time effects of cigarette smoking. Psychopharmacologia 13:311-319 Gilbert DG, Robinson JH, Chamberlin CL, Spielberger CD 1989 Effects of smoking/ nicotine on anxiety, heart rate and lateralization of EEG during a stressful movie. Psychophysiology 26:311-320 Golding J, Mangan GL 1982 Arousing and de•arousing effects of cigarette smoking under conditions of stress and mild sensory isolation. Psychophysiology 19:449-456 Golding JF 1988 Effects of cigarette smoking on resting EEG, visual evoked potentials and photic driving. Pharmacol Biochem Behav 29:23-32 Hughes JR. Hatsukami D 1986 Signs and symptoms of tobacco withdrawal. Arch Gen Psychiatry 43:289-294 Ikard FF, Tomkins S 1973 The experience of affect as a determinant of smoking: a series of validity studies. J Abnorm Psychol 81:172-181 Jones TR. Farrell TR, Herning RI 1978 Tobacco smoking and nicotine tolerance. In: Krasnegor N (ed) Self administration of abused substances: methods for study. NIDA Monograph 20. US Department of Health and Human Services, p 202-208 Knott V 1989 A neuroelectric approach to the assessment of psychoactivity in comparative substance abuse. Paper presented to international workshop on Comparative Substance Abuse, Florence Paton WDM, Savini EC 1968 The action of nicotine on the motor endplate in the cat. Br J Pharmacol 32:360-380 Pickworth WB, Herning RI, Henningfield JE 1988 Mecamylamine reduces some EEG effects of nicotine chewing gum in humans. Pharmacol Biochern Behav 30:149-153 Porchet HC, Benowitz NL, Sheiner LB 1988 Pharmacodynamic model of tolerance: application to nicotine. J Pharmacol Exp Ther 244:231-236

I I I I I I I I I I Nicotine pharmacocynamics 219 Snyder FR, Davis FC, Henningfield JE 1989 The tobacco withdrawal syndrome: performance decrements assessed on a computerised test battery. Drug Alcohol Depend 23:259-266 Tong JE, Leigh G, Campbell J, Smith D 1977 Tobacco smoking, personality and sex factors in auditory vigilance performance. Br J Psychol 68:365-370 US Department of Health and Human Services 1988 The health consequences of smoking: nicotine addiction. Report of the Surgeon General Wesnes K, Warburton DM 1978 The effects of cigarette smoking and nicotine tablets upon human attention. In: Thornton RE (ed) Smoking behaviour: physiological and psychological influences. Churchill Livingstone, Edinburgh, p 131-147 Wesnes K, Warburton DM 1984 Effect of scopolamine and nicotine on human rapid information processing performance. Psychopharmacology 82:147-ISO Wesnes K, Warburton DM, Matz B 1983 Effects of nicotine on stimulus sensitivity and response bias in a visual vigilance task. Neuropsychobiology 9:41-44 West R, Jarvis M 1986 Effects of nicotine on finger tapping rate in non-smokers. Pharmacol Biochem Behav 25:727-731 West R, Russell MAH 1987 Cardiovascular and subjective effects of smoking before and after 24 h abstinence from cigarettes. Psychopharmacology 92:118-121 West R, Russell MAH 1988 Loss of acute nicotine tolerance and severity of cigarette withdrawal. Psychopharmacology 94:563-565 West R, Hack S. Subjective and heart rate effects of cigarettes in occasional and regular smokers, in preparation DISCUSSION I I I I I I I Pomerleau: You have criticized these studies on methodological grounds. Do you think there is no evidence for a biphasic effect of nicotine on behaviour or are you simply saying that the methodological basis for the studies conducted so far is inadequate and therefore it is still an open issue? West: I think it's an open issue. It is very clear that smokers report that smoking peps them up and calms them down. It is quite possible though that the calming effects represent a relief from withdrawal. A study on school children found an interesting positive correlation between the reported calming effects of their cigarettes and the reported severity of their withdrawal symptoms when they abstained (McNeill et al 1980. Pomerleau: I tend to be sceptical about that as the sole explanation. We observed subjective changes (reduction in anxiety and pain) after administration of nicotine under conditions in which there was no change in nicotine withdrawal symptoms (Pomerleau 1986). The problem may come from the lack of precision in measurement of subjective states. Gray: We need to distinguish between two possibilities: 1) that nicotine improves performance because it alleviates withdrawal symptoms, and 2) that nicotine improves performance when performance is bad, and one reason the performance is bad may be that there are withdrawal symptoms getting in the way. There are two examples from our group where there cannot be withdrawal symptoms, but there are clear beneficial effects of nicotine.