Tobacco Documents Online

Ps}choPharmacolo¢y t 199_l 103::9%-s07 I Psychopharnacology C Springer-Verlag 1992 I I I I I I I I I I I I I I Review The role of nicotine in tobacco use John H. Robinson and Walter S. Pritchard Biobehavioral R&D. R.J. Reynolds Tobacco Company, Bowman Gray Technical Center. Reynolds Boulevard, Winston-Salem. NC 27102. USA Received November 26, 19911 Final version January 14. 1992 Abstract. The 1988 US Surgeon General's Report titled "Nicotine Addiction". is cited frequently in the litera- ture as having established the " fact" that nicotine de- rived from cigarette smoke is addictive in the same sense as "classic" addicting drugs such as heroin and cocaine. This manuscripts critically evaluates key research find- ings used in support of this claim and identifies short- comings in the data that seriously question the logic of labeling nicotine as "addictive"• In addition, the man- uscript argues that the role of nicotine in tobacco use is not like the role of cocaine in coca leaf use as argued by the 1988 Surgeon General's Report, but is, in fact, more like the role of caffeine in coffee drinking as con- cluded in the 1964 US Surgeon General's Report. Key words: Tobacco use - Nicotine - Addiction Historically, tobacco and its primary active pharmaco- logic ingredient, nicotine, have been the focus of more scientific study than perhaps any other product con- sumed by humans. Since the publication by Langley and Dickinson (1889) on the effects of nicotine on peripheral ganglia. scientists have studied its effects on various bio- logical tissues and, in particular, the central nervous sys- tem (CNS). A review of reports published by the US Surgeon General and the National Institute of Drug Abuse (NIDA) indicates that the physiological, pharma- cological, medical, psychological, and sociological as- pects of nicotine and cigarette smoking have been the subject of literally thousands of scientific articles. De- spite this huge literature, however, the question " Why do people smoke?" is still asked and, as behavioral sci- entists, we believe that it is to a great extent still unans- wered. One approach to answering this question is to ask a large numoer of smokers why they smoke. In general, two broad categories of responses have emerged, one Correspondence ro ~ J,H, Robinson related to pharmacological aspects of smoking and the other related to non-pharmacological factors (Spiel- berger 1986). The pharmacological effects produced by cigarette smoking have been attributed, for the most part, to nicotine absorbed by the smoker from inhaled mainstream smoke. We will use the term " smoking/nico- tine" to refer to this determinant of the pharmacological effects. The non-pharmacological aspects of smoking include the "taste" of the cigarette smoke, sensory responses in the throat and upper airway, manipulation of the cigarette itself as well as the smoke it produces, and the social aspects of cigarette smoking. A demonstration of the importance of non-pharmacological factors in smoking motivation can be seen in the findings reported by Rose and colleagues (Rose et al. 1984) that anestheti- zation of the mouth, throat, and upper airway signifi- cantly reduced self-reported desire to smoke in a group of regular smokers. The "nicotine paradox" The responses related to smoking's pharmacological ef- fects can be categorized on two predominant dimen- sions. The peripheral effects of smoking/nicotine are iargely stimulatory (e.g., increased heart rate, most no- ticeable for the initial cigarette of the day; Benowitz 1987). Indeed, a large segment of smokers report that increased mental alertness produced by smoking/nico- tine is an important aspect of their smoking motivation. However, an even larger segment of smokers report that smoking/nicotine helps them to function in the face of environmental stress by having a calming effect on their mood, and that this effect comprises a major aspect of their smoking motivation. For example, Frith (1971) reported that over 80% of 2000 smokers questioned listed "pleasurable relaxation" as an important smokirg motive. These two seemingly contradictory motives (smoking for purposes of mental stimulation and smok- ing for purposes of mental relaxation) form the so-called

I I I I I I I I I I I I I I 1 ~ I 398 "nicotine paradox" (see Gilbert .979). Coupled with the non-pharmacological motives, this paradox provides some insight into the difficulty in answering the seeming- ly simple question "Why do people smoke?". One hypothesis re!ating to the question of why people smoke, which was formalized in thr 1988 Surgeon Gen- eral's Report (SGR; US DHHS 1988), is that people smoke because they are "addicted", either to cigarettes or to nicotine. A study by Eiser (1990) reported that for a sample of British smokers many (but by no means all) endorsed the "addiction" hypothesis by agreeing with statements such as " I'm not going to be able to give up smoking unless someone helps me" and dis- agreeing with statements such as "If I really wanted to, I could give up smoking" t. However, Gallup Poll data indicate that, in the US, smokers often hold disso- nant views, with 61% answering "Yes" to the statement "Do you consider yourself addicted to cigarettes or not?" while 78% of this same sample of 1240 adult smokers answered "Yes" to the question "Do you feel you would be able to quit smoking if you made the decision to do so or not?" (Gallup Poll National Survey Data, July 1990; Public Opinion On-Line Database). Moreover, Spielberger and Jacobs (see Spielberger 1986) factor analyzed responses from 1029 smokers to a large number of smoking motivation questions representing, psychometrically, the "best" questions from a number of previous smoking motivation scales. An addiction- like factor labeled "Automatic/Habitual" accounted for less variance than would be predicted based on the number of questions loading on the factor, whereas a factor labeled "Negative Affect Control" (cf previous paragraph) accounted for more. We believe that a more reasonable hypothesis con- cerning why people smoke, a hypothesis that is consis- tent with the smoking motivation literature, is that smokers use cigarettes primarily as a "tool" or "re- source" that provides them with needed psychological benefits (increased mental alertness; anxiety reduction, coping with stress). This "resource" hypothesis (Wesnes and Warburton 1984a; Warburton 1988a, b; Warburton et al. 1988; see also Pritchard 1991 a) stands as a major alternative to the addiction hypothesis, and we argue that the resource hypothesis passes a "common-sense" test that the addiction hypothesis fails. In addition, the comparison of nicotine to heroin and cocaine (US DHHS 1988) contrasts sharply with the findings of the ` Eiser has proposed that some smokers endorse the "addiction" hypothesis in an attempt to explain their smoking behavior in a way that removes any personal responsibility in making a choice that is viewed by others as unsound or even irrational. After all, if a smoker is "addicted ", he can argue that he has no control over his smoking behavior and can use this helplessness as a "shield" for the criticism he raxives for engaging in this behavior. In addition, Eiser has identified another group of smokers who ascribe to an "illness" definition of addiction, even to the extent of blaming others for their failure to "cure" this "illness" (see Eiser 1990 for discussion). In essence, by expressing the belief that he is addicted, a smoker can use this as an excuse for not making the commitment required to accomplish a permanent behavioral change in a habit that is often viewed by others as "wicked" (see Davies 1990). 1964 SGR (US DHEW 1964) which compares the role of nicotine in tobacco to the role of caffeine in coffee. We will briefly review several lines of research leading to the conclusion that the 1964 SGR was, in many ways, more accurate in classifying nicotine as "habituating" rather than "addicting" and will argue that classifying nicotine as habituating represents a more balanced per- spective than classifying nicotine as addicting. The nicotine addiction hypothesis The 1988 SGR titled The Health Consequences ojSmok- ing - Nicotine Addiction (US DHHS 1988) resolved, for many people, the question of why people smoke. Quite simply, the report stated that people smoke because they are addicted to nicotine. To ensure that this impression would not be missed by anyone, including the popular press, the report highlighted three major conclusions: (1) cigarettes and other forms of tobacco are addicting; (2) nicotine is the substance in tobacco that causes addic- tion; and (3) the fundamental processes'that determine tobacco addiction are similar to those that deterrtune addiction to drugs such as heroin and cocaine. Verification that these conclusions have been widely accepted by the medical and scientific community' re- quires only a quick review of the recent literature, where it is invariably asserted that the "addictiveness" or "abuse potential" of nicotine equals (or even exceeds) that of substances such as cocaine, opiate narcotics (e.g., heroin), barbiturates, benzodiazepines, and alcohol (e.g., Jasinski and Henningfield 1988 ; Goldstein and Kalant 1990; Henningfield et al. 1990, 1991 ; Russell 1990; Sto- lerman 1990a, b). Yet, the conclusion that cigarette smoking is the same as heroin or cocaine addiction seems fundamentally flawed from a common-sense point of view, and is diametrically opposed to the conclusions stated in the 1964 SGR that nicotine was definitely not an addicting drug. How persuasive is the evidence pre- sented in the 1988 SGR supporting the conclusion that smoking/nicotine is the same as addiction to heroin or cocaine? Habituation versus addiction There is currently no universally accepted scientific defi- nition of addiction. While the term "dependence" has been given a somewhat more precise definition (e.g., World Health Organization (WHO) 1981) and on that basis is more acceptable from a scientific standpoint (as acknowledged in the 1988 SGR), the 1988 SGR con- sistently uses the word "addiction" interchangeably with the word "dependence" on the grounds that the former provides "information at a more general level" (p 7). In other words, the word "addiction" is familiar to the average layperson, even if her/his conception of what addiction typically means is based on common mis- perceptions and lacks scientific foundation. To the aver- age layperson, the word "addiction" carries affective I

I I I I I I I I I I I I I I I I I connotations of a strongly negative nature (Davies 1990). a ploy " not lost sight of by the Sur¢eon General's Office" according to Pandina and Huber in their review of the nicotine addiction question (1990. p 55). In addirion. as noted -above. the conclusions drawn in the 1988 SGR are in sharp contrast to those of the 1964 SGR (Smoking and Health - Report of the Adcisorr Committee to the Surgeon General of the Public Health Sercice. US DHEW 1964), When a universally accepted scientific definition of addiction did exist, the 1964 SGR concluded " In medical and scientific terminology the practice (smokin¢] should be labeled habituation to dis- tinguish it clearly from addiction, since the biological effects of tobacco, like coffee and other caffeine-contain- ing beverages, ... are not comparable to those produced by morphine, alcohol, barbiturates, and many other po- tent addicting drugs" (p 350, emphasis in original). The basis for the distinction developed in the 1964 SGR was the established WHO definitions of addiction and habituation as outlined on p 351 of the 1964 SGR. Addiction was defined as "a state of periodic or chronic intoxication produced by the repeated consumption of a drug (natural or synthetic). Its characteristics include: 1) an overpowering desire or need (compulsion) to continue taking the drug and to obtain it by any means; 2) a tendency to increase the dose; 3) a psychic (psy- chological) and generally a physical dependence on the effects of the drug; 4) detrimental effect on the individual and on society." We may also note that this definition probably comes close to what the average layperson has in mind when using the term "addiction". In contrast, habituation was defined as "a condition resulting from the repeated consumption of a drug. Its characteristics include: 1) a desire (but not a compul- sion) to continue taking the drug for the sense of im- proved well-being which it engenders; 2) little or no ten- dency to increase the dose; 3) some degree of psychic dependence on the effect of the drug, but absence of physical dependence and hence of an abstinence syn- drome; 4) detrimental effects, if any, primarily on the individual [rather than society] ". The 1988 SGR acknowledged the dramatic change in position regarding nicotine "addiction" relative to the 1964 SGR. According to the 1988 SGR, the reclassi- (ication of nicotine as addictive rather than habituating is supported by two key arguments: 1) The WHO no longer used the terms "addiction" and " habituation ", dropping this distinction in favor of a single new entity, drug dependence. (This action by the WHO was perhaps motivated by the trivialization of the term "addiction" in the popular media, where it was used to refer to any behavior that people regularly engaged in: sex, watching TV, exercise, video games, eating chocolate, etc.); 2) Reports appearing in the scientific literature since the publication of the 1964 SGR reportedly demon- strated that nicotine shared many features with proto- typic drugs of abuse and had now been shown to fulfill the three primary criteria (as defined by the Surgeon General for the first time in the 1988 SGR) of an addict- 199 ing drug. These criteria were: (a) high(r controlled or compulsioe use: (b) psrchoactive e%fects: and (c) drug- reinforced behat•ior. The 1988 SGR also asserted that the recent scientific literature had demonstrated that nic- otine passed several of the "additionai criteria" lis?ed as "useful" in defining addictiveness. These included use despite harmful effects, relapse following abstinence. recurrent drug crarings, tolerance, phrsical dependence. and euphoric effects. We will examine both of these claims used by the Surgeon General in support of his reclassification ot' nicotine as addicting rather than habi- tuating. While it is true that the WHO dropped the semantic distinction between "habituation" and "addiction", replacing its concept of addiction with the term "depen- dence" (as stated in the 1988 SGR), a key concept from the original WHO definition of addiction was not dropped (as implied by the 1988 SGR) when the term dependence was adopted by the WHO, namely, the con- cept of intoxication. Intoxication A critical attribute stressed in the original WHO defini- tion of addiction was "a state of periodic or chronic intoxication produced by the repated consumption of a drug" (1964 SGR, p 351). Since this phrase was not included in the definition of habituation, it served as a key point of distinction between addicting and habi- tuating drugs. This definition of addiction also implied that someone under the influence of an addicting drug not only had a diminished capacity to decide whether to continue taking the drug, but also suffered impaired cognitive performance. One consequence of the latter was the potential for adverse impact on others (e.g., workplace accidents, impaired driving ability leading to vehicular accidents). In 1978 the WHO reiterated the importance of the concept of intoxication, stating that "psychotropic" substances considered for control by the international community must be capable of producing both a state of "dependence" and "central nervous system stimula- tion or depression, resulting in hallucinations or distur- bances in motor function or thinking or behavior or perception or mood" (WHO, Technical Report Series, No. 618, 1978, p 8). The concept of "intoxication" (neg- ative disturbances in psychological or motor function) is viewed by many as critical in determining if a drug is addictive. In 1984, NIDA presented the First Triennial Report to Congress from the Secretary of Health and Human Services (US DHHS Publication No. ADM 85-1372). The report presented the state of drug abuse and drug abuse research as determined by NIDA. In describing the effects of psychoactive drugs (psychoactivity now being a primary criterion in the new definition of addic- tion), the report states: O ~ ~ O A predictable effect of the use of almost any psychoac- Q tive drug is a distortion of the perception of time, space, Ij and the location of objects within space. A corollary ~ F-~

I I I I i I 1 i I I I I I I I I 400 effect is a dose-related reduction in physical coordina- tion or psychomotor functioning. Normally easy tasks like placing a top on a jar or walking become difficult to perform. The ability to visually locate objects in space. judge their distance and-track them is irrpaired. (pp 19- '0). In short, the NIDA report states that psychoactive drues produce intoxication, and that psychoactivity is a primary criterion for addiction. The concept that ad- dictive drugs produce disturbances in psychological and/ or motor functions was thus forcefully stated. It should be noted however, that the concept of a psychoactive drug producing intoxication does not appear in the 1988 SGR, perhaps because the literature demonstrates that it simply does not apply to nicotine, as discussed below. The same NIDA report (1984) did conclude, however, that compulsive tobacco use ... is a form of drug abuse, or drug addiction, in which nicotine is critical. Specifically, it is evident that the role of nicotine in cigarette smoking is similar to the role of cocaine in coca use, of THC [delta-9-tetrahydrocani- binol] in marijuana smoking, and of ethanol in alcoholic beverage consumption (p 113). The NIDA report comes to this conclusion with re- gard to nicotine, but makes no mention of nicotine pro- ducing psychological or motor disruption. This is in sharp contrast to the other substances to which nicotine was compared, and, in fact, the NIDA report contains extensive documentation regarding disturbances in psy- chomotor, behavioral, and cognitive functions of people and animals under the influence of (or experiencing with- drawal from) these other substances, including THC (marijuana), alcohol, amphetamines, cocaine, narcotics, barbiturates, and hallucinogens. We believe that the omission of any discussion regard- ing the effects of nicotine on psychological function was not simply an oversight by the author of the nicotine chapter of the NIDA report (Burling) or an indication that he disagreed with statements in the report concern- ing the adverse psychological and motor effects of the other substances discussed in the report. This seems evi- dent from the footnote that accompanies the nicotine chapter, which states: A concept central to many discussions of drug abuse is that the substance produces "damage" or "debilita- tion. " This aspect of cigarette smoking will not be ad- dressed here as there are extensive data indicating 1) the actual toxicity of tobacco, and 2) the widespread perception by smokers that their habit is harmful. (p 97, emphasis added). In this NIDA report, Burling has subtly changed the meaning of "damage" or "debilitation" from the psy- chological domain to the health domain in an attempt to classify nicotine as an addicting drug A person who has ingested nicotine in amounts characteristic of normal smoking does not suffer the "distortion of time, space and the location of objects within space" associated with the other substances reviewed in the report. A smoker does not find it difficult to perform "easy tasks like placing a top on a jar or walking" (ibid.), or driving a car. Indeed, nicotine in the majority of circumstances results in improt•ed performance as well as other psycho- logical benefits to the smoker (Wesnes and Warburton 1978. 1983, 1984a, b: Wesnes 1987; Warburton 1988a, b; Pritchard 1991 b; Pritchard et al. 1991). In addition. Hindmarch and colleagues have con- sistently reported distinct differences in the effects of nicotine and "classic" addicting drugs on performance tasks, including "driving" automobile simulators (see Hindmarch et al. 1991 ; Kerr et al. 1991). In these tests, smoking/nicotine clearly resulted in improved mental and motor performance in contrast with "classic" ad- dicting drugs, which resulted in reduced mental and mo- tor performance. The concept of "intoxication", central to the issue of whether or not a drug is addicting, simply does not apply to nicotine. Moreover, smoking/nicotine is clearly compatible with performing everyday tasks and is perceived by the smoker as providing psychological benefits. To reiterate, the author of the nicotine chapter of the 1984 NIDA report, referenced the smoking and health literature and not the cognitive performance liter- ature to illustrate the "debilitating" effects of nicotine. However, others (US DHHS 1988 ; Froggatt and Wald 1989; Roe 1989), have suggested that the smoker's in- creased risk of health-related problems is believed to be due primarily to the ingestion of the "tar" fraction of cigarette smoke, not to the ingestion of nicotine. While some may argue that this is a subtle distinction, and that nicotine and " tar" should not be separate issues, the distinction becomes important when attempting to compare nicotine with "classic" addicting drugs (whose intoxicating effects diminish cognitive performance as well as the capacity of the user to decide whether or not to use that drug). Increased risk of developing cer- tain diseases in smokers has been principally ascribed to the "tar" fraction of cigarette smoke, and should remain a separate issue from that of the "addictiveness" of nicotine, especially given the potential availability of cigarettes yielding nicotine but having minimal biologi- cal activity (Russell 1991). Thus, it is clear that nicotine lacks a feature many consider essential to the definition of an addictive drug. In our opinion, the fact that smoking/nicotine clearly does not result in "intoxication" (psychological debilita- tion) is consistently overlooked in the debate on nicotine "addiction". This obvious difference between nicotine and the "classic" addicting drugs has been ignored by the Surgeon General in changing the classification of nicotine from habituating to addicting. To this end, he cited evidence appearing in the scientific literature of certain cotnmonalities between nicotine and "classic" addicting drugs (cocaine, heroin, alcohol, etc.). The un- derlying rationale for these studies has been termed the "analogy" argument (Warburton 1989). From the per- spective of the Surgeon General, if studies demonstrated that two drugs share a certain number of features in common, then by analogy, those drugs may be consid- ered to share other, perhaps untested, features. I

I I I I I I ~ i I I I I I I I I I I The "analogy argument" The "analogy argument" is seriously flawed from a sci- entific standpoint. As Warburton (1989) pointed out in his critique of the 1988 Sf'rR, no matter how many fea- tures X has in common with Y, it takes only one feature oj X not found in Y to make the analogy fallacious. We have already identified one key concept (intoxica- tion) in the WHO definition of an addictive drug that nicotine does not share with "classic" drugs of abuse. Yet, the 1988 SGR goes to great lengths to establish commonalities between nicotine and other drugs in an attempt to prove that nicotine is addicting. The shortcomings of this line of reasoning should be obvious. Imagine a researcher interested in studying pain. He chooses to use two stimuli, a bowling ball and a cinder block, and begins to identify the physical and behavioral traits these two objects have in common. Both objects weigh about the same. When dropped on a subject's foot, both objects accelerate at the same rate, inflict similar amounts of pain, produce similar physio- logical responses and result in similar verbal reports by the subject. The two objects appear to possess a number of common properties. It would, however, be inaccurate to conclude that the subject could achieve his usual bowling average using the cinder block. The logic of this admittedly facetious example is not too far removed from that used to draw inferences in some of the studies cited in support of the conclusions reached in the 1988 SGR. We will now critically examine some of the key evidence used to support the 1988 SGR's conclusion that "the pharmacologic and behavioral pro- cesses that determine tobacco addiction are similar to those that determine addiction to drugs such as heroin and cocaine" (p 9). Nicotine and the three primary 1988 SGR Criteria for defining addictive drugs The 1988 SGR employed a definition of "addiction" that included three main criteria along with several "ad- ditional criteria". The list of additional criteria included such factors as tolerance, withdrawal/physical depen- dence, drug craving, relapse following abstinence, and stereotypic patterns of ingestion behavior. Warburton (1989) and Collins (1990) have addressed these issues and concluded that while nicotine may, under limited circumstances or to a limited degree, possess or result in some of these, the literature in general does not sup- port the concept that nicotine is equivalent to cocaine, heroin, or other potent addicting drugs in terms of these additional criteria. Since the task of adequately reviewing the literature regarding the "additional" criteria is beyond the scope of the present paper, the reader is referred to Warburton (1989) and Collins (1990) for summaries of these topics. Only the three primary criteria were deemed by the 1988 SGR to be necessary for defining addiction. We will fo- cus our comments mainly on the three primary criteria as specified in the 1988 SGR. We will also examine the s0t "euphoriant" model of nicotine addiction, since this model is often cited as an explanation of the reinforcing psychoactive effect of smoking, nicotine that contributes to the "addictive" nature of smoking (e.g.. Henningfield 1984a; Carmody 1989; Bourne 1991). Criterion 1: highlv controlled or compulsit•e use The concept of highly controlled or compulsive use has been included in many definitions of addiction (Warbur- ton 1985) and fits extremely well with the layperson's idea of drug addiction, i.e.. the addict who is unable to resist uncontrollable drug cravings (Davies 1990), However, as Warburton (1990) has noted, this criterion does not fit particularly well when one considers ciga- rette smoking, where extended periods of abstinence may be readily managed by many smokers. In today's restric- tive environment, many smokers are prevented from smoking on planes, in public places, or at work, without significantly disrupting other aspects of their behavior or performance. Formal restrictions aside, some smokers may smoke only at work because their spouse objects to smoking in the home, and may abstain from smoking in specific social situations if others object. Others smoke on a regular basis throughout the day, and do develop strong habitual behaviors relative to when and where they smoke. Since many smokers perceive smoking as providing benefits in terms of enhanced cognitive performance or stress reduction (Gilbert 1979; Warburton 1988a, b; Wesnes 1987; 1988 SGR), it is not surprising that regular patterns of use develop. By applying the 1988 SGR defi- nition of compulsive use as a criterion for addiction, anything that a person ingests, enjoys, and therefore in- gests again under regular circumstances fulfills this crite- rion for being "addicting". This would apply to such behaviors as the proverbial morning cup of coffee, regu- lar use of aspirin or non-steroidal anti-inflammatory drugs for relief of arthritis pain, or obtaining psychoac- tive effects from sweets or a bedtime glass of milk due to alterations in brain levels of the neurotransrnitter ser- oconin (Fernstrom 1983) '. Criterion 2: psychoactivitv The second primary criterion proposed by the 1988 SGR for an addicting drug, psychoactivity, was characterized by Warburton (1990, p 166) as "trivial" when one con- siders nicotine. While it is true that nicotine is " psy- choactive" which can be defined as producing pharmaco- logical effects in brain tissue that result in changes in nerve cell activity, so are a number of other drugs that j Nicotine gum has been reported not to relieve craving for ciga- rettes (Hughes et al. 1984; Schneider and Jarvik 1985) while adrnin- istration of otner sustances does, e.g., dextrose (West et al. 1990) or the antidepressant doxepin (Murphy et al. 1990). In fact, telling subjects that they are receiving nicotine gum during cessation has been reported to attenutate craving regardless of whether the gum in fact contains nicotine or not (Hughes et al. 1989).

I I I I I I I I I I I I I I I 402 are consumed every day oy millions of people. Theophyl- line in tea, theobromine in cocoa products. and the most widely consumed drug in the world, caffeine (coffee, soft drinks) are all "psychoactive". The fact that a drug is psychoactive (capable of pro- ducing central nervous system effects) cannot establish whether or not it is addicting. The determining factors should be the specific effects produced and the magni- tudes of these effects. The vast majority of published research findings show that the subtle "psychoactive" effects of nicotine are clearly distinct from the mental and rnotoric disturbances produced by the "classic" ad- dicting drugs. In spite of these clear differences, some researchers have reported studies of nicotine-induced "euphoria" in an attempt to demonstrate similarity between nicotine and "classic" addicting drugs on the bases of specific psychoactive effects (Henningfield et al. 1983, 1985; Ja- sinski et al. 1984; see Henningfield 1984b; Jasinski and Henningfield 1988, and Henningfield et al. 1986 for re- views). In these studies, subjects (typically with histories of hard-core drug abuse) were allowed to self-administer (by lever pressing) injections of nicotine or physiological saline, or were given intravenous (IV) injections if nico- tine. Henningfield stated that because of their drug histo- ries, these subjects represent "a very discriminating pop- ulation, like fine wine tasters" (Charlotte Observer, April 24, 1982). An extension of this logic would indicate that manufactures of fine wines and spirits waste huge sums of money for trained taste panelists when they could get better information from derelict "winos". The subjective responses of these chronic, illicit drug- abusing subjects were measured using questionnaires de- signed to measure "liking" or "euphoria". The re- sponses for nicotine were then compared to the ratings of "classic" drugs of abuse. Among the major weak- nesses of these comparisons, it should be noted that the ratings were not necessarily from the same study or the same subjects, and the drugs were not administered via the same procedures. Nevertheless, Henningfield and his colleagues (Henningfield 1984b; Jasinski et al. 1984; Henningfield et al. 1985), concluded that high intrave- nous doses of nicotine (3 mg/kg administered in 10 s) resulted in significant elevations in "euphoria" or "lik- ing-scores". They also claim that the effects of this very large, rapid injection of nicotine are frequently misiden- tified by these "discriminating" drug abusers as being similar to the effects of cocaine or amphetamine. For example, in the Henningfield et al. (1985) study six of these "experienced" addicts misidentified the high dose of nicotine as cocaine and one misidentified nicotine as amphetamine. However, as Clark (1990) correctly notes, the value of these data is highly questionable since four of the seven subjects that misidentified nicotine had no prior experience with stimulants! Since hard-core drug addicts obviously represent a highly deviant group the logic of generalizing the results oi these studies to the population at large can be ques- tioned. We concur with Collins' (1990) observation that the conclusions drawn from these studies are of limited value because of the methodological reasons that have been noted. Consequently, in our opinion. the results of these studies cannot be readily extended to normal smokers without extensively testing subjects who are not drug addicts. To illustrate another example, the results (''liking score" figure) of a single study (Jasinski et al. 1984) have been reported numerous times (Henningsfield 1984a, b, 1987; Burling 1984; Henningfield et al. 1985: Henn- ingfield and Nemeth-Coslett 1988: Henningfield et al. 1987; NIH Publication No. 86-2874, Jasinski and Henn- ingfield 1988; US DHHS 1988) to support establishing as "fact" that nicotine is a euphoriant just like cocaine. We believe that this study and others are seriously flawed and vastly over-interpreted by subsequent authors for the following reasons. The results of several studies taken together (Henn- ingfield et al. 1983; Jasinski et al. 1984; Henningfield et al. 1985) ' indicated that hard-core drug addicts would lever-press to self-administer nicotine but not saline. However, these data can be readily explained without invoking the "nicotine addiction" hypothesis. In a bor- ing or stressful laboratory session, subjects might achieve some benefit from nicotine's capacity to provide mental stimulation or reduce stress (see Gilbert 1979; Gilbert and Welser 1989). Drug addicts typically score high on Zucketinan's Sensation Seeking Scale (Zuckerman 1983, 1987), a measure including.thriil- and adventure-seeking as important components, and have been reported to experience drug-like responses when presented with visu- al cues associated with drug taking apparatus or in re- sponse to placebo drug injections (O'Brien ec al. 1978). Preference for the nicotine lever over the saline lever may merely reflect the subjects' preference for any phar- macological stimulus over no stimulus at all. In addition, heroin addicts have reported physiological responses fol- lowing injections of inactive substances (sterile saline: O'Brien et al. 1978). This may simply be a conditioned response to the stimuli associated with drug taking be- havior. ' Precisely identifying a single study where Henningfield and his colleagues determined that nicotine is a euphoriant similar to co- caine is difficult. The "Liking score" figure that has appeared in the literature numerous times is from Jasinski et al. (1984). How- ever, this paper is inadequate in describing the subjects and proce- dures used to make this determination. For example, the number of subjects in the nicotine group is not reported and no statement about the subjects misidentifying the nicotine injections is made. However, in this report the authors also published a figure (Fig. 3) illustrating cumulative lever pressing records for subject KU self- administering nicotine and saline. This figure was published with the note "Submitted for publication in Pharmacology, Biochemisrry and Behavior " and indeed did appear in Henningfield et al. (1983) where the subjects were reported to have misidentified nicotine as "morphine or cocaine". Since data from subject KU appears in both reports and since data from subjects KU, SK and PE appear in both Henningfield et al. (1983) and Henningfield ct al. (1985) (where subjects were also reported to have misidentified high IV doses of nicotine as cocaine) it is not kuown whether these subjects were tested in all three studies, or if all the data were collected in a single study and repocted separately, or if sub- jects were asked to identify the drug they received in the Jasinski et al. (1984) study and the data were simply not reported.

I I I I I I I I I I I I I I I I I These studies fall short on other methodological grounds as well, If a goal of these studies was to compare the effects of nicotine to cocaine, then cocaine should have been included in the experiment. Cocaine is gener- ally accepted as one of the most addictive drugs known. Results of animal self-administration studies (see below) suggest that it is quite likely that, if given a choice, these drug addicts would have consistently chosen the cocaine lever over the nicotine lever. Apparently the authors were willing to accept the subjects at their word that IV nicotine felt like cocaine. This is despite indications that hard-core drug addicts typically have antisocial per- sonality disorders, including histories of pathological ly- ing in order to cover their drug abuse (Schuckit 1973; Lewis et al. 1983) and that some of the subjects who mis-identified nicotine as cocaine had apparently never used cocaine (Clark 1990). Other important questions can be raised. What kind of informed consent did the subjects sign? Were they led to believe that some of the injections they received during the test might be cocaine? Did the subjects therefore experience some expectancy bias? Did they tell the experimenters what they thought the experimenters wanted to hear in hopes of participat- ing in other sessions where they might in fact receive injections of cocaine?'. Clark (1990) is again correct when he notes that the authors did not conduct the proper test :o support their conclusions. In addition, while the nicotine "liking" graph origi- nally taken from Jasinski et al. (1984) has been published numerous times in the studies listed above, we have never seen this graph published with accompanying (standard) error (of the mean) bars. This leaves un- known the nature of the individual differences that were found between responses to the nicotine injections and the "classic" addicting drugs that were tested. Further- more, it is apparent that not all subjects experience "eu- phoria" following IV injections of nicotine. In this con- text, two important points that are often omitted or ig- nored in subsequent discussions of nicotine "euphoria" are provided by two earlier studies (Henningfield and Goldberg 1983a; Henningfield et al. 1983): 1) nicotine does not necessarily maintain a sustained rate of lever pressing that is higher than placebo (saline), and 2) the subjective responses that follow a rapid (10 s) uerv high (3 mg) IV dose of nicotine typically include respiratory problems, tightness in the chest, and a lightheaded (faint) sensation lasting approximately 15-20 s. Especial- ly interesting is one subject's report that he "would be willing to (pay) seventy-five dollars" not to receive an- other nicotine injection. In addition, the "liking" score for nicotine is significantly different from the placebo "liking" score only at the 3 mg dose of nicotine. This amount of nicotine (administered in a very short period of time, 10 s) is much greater than is typically achieved during smoking. Another important question to be considered when critically evaluating the results of the Henningfield group is the method of nicotine administration. How do the effects of these bolus (rapidly-administered, single ad- ministration) IV injections of a large dose of nicotine (3 mg in 10 s) compare to smoking a 0.5-1.5 mg cigarette 403 over a 5-10 min period? The key datum here is the com- parison of the degree of liking euphoria that people ex- perience when they smoke a cigarette versus the liking euphoria subsequent to an IV injection of nicotine. The "liking" of a smoker who has just enjoyed a cigarette is obviously not the same as the "liking" of a heroin or cocaine addict in response to drug taking (often de- scribed as the ultimate pleasure or in sexual terms; War- burton 1990). In fact, Benowitz and Jacob (1990) have shown that when nicotine is infused at a slower rate (2 µg; kg/min for 30 min), smokers cannot tell the differ- ence between nicotine and saline. The authors of two reports (Henr,:::gfield 1984b: Henningfield and Nemeth-Coslett 1988) attempted to address the important distinction between rapid, bolus administration of nicotine and rates that would better approximate the absorption of nicotine during normal cigarette smoking. These authors presented graphs (again without error bars, also note that the relevant graph in the second reference is published with the wrong legend) illustrating subjects' feelings following ad- ministration of nicotine (IV and cigarette smoking), d- amphetamine, morphine, diazepam, ethanol, pentobar- bital, marijuana, and "simulated gambling". At first glance, the results appear to support the authors' conten- tion that both IV nicotine and tobacco smoking produce similar euphoric feelings of the.same magnitude as those seen with known addicting drugs. It should be noted however, that the graphs are plotted on quite differently scaled y-axes. The Henningfield graphs give the (casual) reader the impression that tobacco smoking and nicotine are similar to known drugs of abuse, but the data clearly do not support the conclusion that nicotine is a euphoriant simi- lar to morphine or d-amphetamine. Cocaine is not di- rectly compared with nicotine or smoking, despite the repeated references to the similarity of nicotine to co- caine. When Warburton (1988) re-plotted the euphoria data as difference scores between drug and placebo using identically-scaled y-axes for each compound tested (Fig. 1), smoking and IV nicotine (again 3 mg in 10 s) had the lowest difference scores, with the difference be- tween cigarette smoking and placebo being the smallest of all (even lower than "simulated gambling "). Subsequent research by Henningfield's group (Henn- ingfieid et al. 1985) has expressly failed to find a differ- ence between cigarette smoking and placebo with regard to self-reported euphoria, even using a very high-yield (2.9 mg nicotine, FTC yield) cigarette. Nicotine gum also fails to produce euphoria (Snyder et al. 1989). In fact, Nemeth-Coslett et al. (1987) reported that 4 mg nicotine gum produced lower "euphoria" scores than placebo. Warburton (1988) presented data on the pleasure-stimu- lation and pleasure-relaxation effects of nicotine and other substances. Nicotine was rated as less stimulating than alcohol, amphetamine, cocaine, heroin, marijuana and sex and approximately equal to caffeine and choco- late. Finally, when McNeill et al. (1987) queried a sample of 104 regular smokers, not one reported that " feeling high" was an effect that they experienced following smoking. The predominant effect (64% of the sample) I

I I I I I I I I I I I I I I I 404 I iei o 3c~Grt~ Q e.o~a o 1, ,ra`1o\ zQac` . was feeling calmer. In our opinion the data do not sup- port the contention that nicotine is a euphoriant and are presented in a misleading fashion to give the impres- sion that nicotine is similar to "classic" addicting drugs when, in fact, it is not. In summary, Henningfield and his colleagues have drawn conclusions without conducting a proper drug discrimination study (see Clark 1990). However, the data from these and other studies, coupled with the determi- nation that nicotine (under limited contingencies) can serve as a reinforcer in animal self-administration para- digms are used to support the contention that nicotine possesses all the properties of "a prototypic drug of abuse" (Henningfield 1984a, p 197). In our opinion the data clearly do not support such a contention, but this conclusion and these data have, nevertheless, been wide- ly publicized. A variation of the "euphoria model" of smoking has been argued by Pomerleau (see Pomerieau and Pomer- leau 1984). The results of this study have been widely interpreted as evidence that nicotine obtained from smok- ing has the capacity to cause the release of the endoge- nous opioid beta-endorphin. This research has also been used to support the notion that smoking is addictive by implying that "euphoria" produced by smoking re- sults from increases in circulating levels of beta-endor- phin (Bourne 1991). The Pomerleau model also proposes that chdnges in the circulating levels of beta-endorphin or other hormones and neuromodulators (e.g., cortisol, adrenocorticotrophic hormone (ACTH)) are responsible for the psychological effects of smoking. However, this model is also inadequate. The neurohumoral changes (especially beta-endor- phin) that were reported by Pomerleau, were seen only in subjects who achieved extremely high plasma concen- trations of nicotine (>60 ng/ml). These unusually high plasma nicotine concentrations were achieved as a result of smoking two very high-yield cigarettes (2.87 mg nico- tine versus 1.1 mg for leading, filtered, "non-light" ciga- rettes) in an extremely short period of time. Research has shown that this smoking paradigm in all likelihood Fig. 1. When Warburton (1988) re-plotted Henningfield's (1984b) "euphocia" data as a differential from baseline (placebo) score, he concluded that "nicotine is. at best, a weak euphonant and ... it is not like mor- phine in opium use," From "The Psycho- pharmacology of Addiction" (1988) Lader MH (ed) by permission of Oxford Universi- ty Press resulted in symptoms of acute nicotine toxicity (primari- ly nausea; see Gilbert and Welser 1989; Gilbert et al. 1992). The cigarettes and smoking paradigm (2 cigarettes 5 min apart) used in this study are unrelated to the real- life smoking patterns of the vast majority of smokers. Furthermore, although often overlooked in subsequent reports of these data, these neurohumoral effects were not observed when the subjects smoked two "low" nico- tine (0.48 mg, FTC yield) cigarettes in succession. What has been assumed to be a specific rewarding effect of nicotine in humans was, in all probability, a non-specific response to the stress of nausea and malaise brought about by toxic levels of nicotine. Pomerleau has stated that " Because the hormonal profile associated with nico- tine-induced nausea resembles that of nausea produced by other manipulations (motion sickness, administration of other drugs), it is likely that the observed hbrmonal pattern at this level of stimulation is characteristic of nausea and not unique to nicotine" (Majchrzak et al. 1987). However, this "qualification" of the original re- port is not generally cited when these data are refer- enced. In sum, the studies purporting to demonstrate that smoking/nicotine is a euphoriant similar to cocaine and heroin are seriously flawed from a number of per- spectives. A critical review of this literature leads to the conclusion that nicotine absorbed from cigarette smoke is simply not a euphoriant in humans. Criterion 3: drug reinforced behavior The few extant human studies of the self-administration of nicotine also indiate that nicotine is not a particularly good reinforcer. Although the primary focus of this re- view is on human studies, we will briefly examine litera- turc regarding nicotine self-administration in animals. NIDA has established such studies as an important test of a drug's "abuse potential" and the 1988 SGR relies heavily on these studies to support the third major crite- rion of an addicting drug. I

I I I I I I I I I I I I I I I I i animal self-administration The 1988 SGR provided a table (pp 183-188) summariz- ing nicotlne self-administration studies. tituch of this ta- ble had appeared verbatim in a report by Henningfield and Goldberg (1983b) in which they concluded that "... nicotine shares many salient ieatures of other drugs of abuse. However. nicotine differs from other drugs in that the range of environmental conditions under which it ;erves as a reinforcer appears to be more restricted" (p 991, emphasis added). Yet. the 1988 SGR concluded that the "... pharmacologic and behavioral processes that determine tobacco addiction are similar to those that determine addiction to drugs such as heroin and cocaine" (p 9, emphasis added). This conclusion is based in large part on nicotine's ability to serve as a primary reinforcer in animal self-administration studies. A review of the studies outlined in the 1988 SGR supports the original conclusion of Henningfield and Goldberg (1983b) that the range of environmental con- ditions under which nicotine reinforce-s behavior appears to be more restricted than "classic" addicting drugs, in particular heroin and cocaine. There are clear-cut dif- ferences between the self-administration behavior of ani- mals for drugs such as cocaine and heroin versus nicotine and caffeine (which has also been shown to be self-ad- ministered relative to saline; Griffiths et al. 1979). For example, animals given unlimited access to cocaine and heroin often ignore food and water, self-administering the drug until death (Bozarth and Wise 1985). Environmental stimuli play a much more important role in the self-administration of nicotine than heroin or cocaine self-stimulation (see Bozarth 1990). With the exception of some reports by Henningfield and Goldberg (cf 1988 SGR), studies that directly compare the rein- forcing efficacy of nicotine and cocaine generally find nicotine to be a much weaker reinforcer than cocaine (see US DHHS 1988. Table 4, pp 183-188). We should also note that self-administration behavior does not prove that the reinforcing stimulus is producing euphoria nor should that stimulus be considered "addic- tive ". Given the proper schedule of reinforcement, mon- keys have been trained to self-administer painful electric shocks to themselves (Morse et al. 1967). In one study (McKearney 1968), three monkeys lever-pressed over 800000 times to receive some 3000 painful electric shocks. One would hardly argue that electric shocks pro- duce euphoria or that they are addicting. Thus, while IV nicotine can serve as a reinforcer in self-administration paradigms (as caffeine can), the cir- cumstances under which nicotine is reinforcing are much more limited than the "classic" addicting drugs such as cocaine, heroin, and amphetamine (see Bozarth 1990). Specifically, even IV nicotine is a much weaker reinforcer (Dworkin et al. 1991) than these other substances and in all likelihood is much closer to caffeine on a hypothet- ical "reinforcement continuum" than it is to cocaine or heroin. It seems clear that the role of nicotine in cigarette smoking is similar to the role of caffeine in coffee or cola drinking as concluded in the 1964 SGR. The pursuit of the goal of "creating" a smoke-free soci- ety by the year 2000 has led the authors at SGR to rely on data that do not often support clusions they have drawn. Summary and conclusion The goal of this paper was to critically evaluate evidence used to support the conclusions expressed in the 1988 SGR that have been popularly reported as proving the claim that nicotine is addicting, just like heroin and co- caine. NLuch of the data offered as evidence for this conclusion simply do not stand up under critical review. If this paper provokes a spirited and open discussion of these data, it will have served an important goal. In our opinion, it is political zeal rather than scientific merit that supports the conclusion of the 1988 SGR that the "pharrrtacologic and behavioral processes that deter- mine tobacco addiction are similar to those that deter- mine addiction to drugs such as heroin and cocaine ". We recognize that nicotine plays an important role in smoking behavior for many people. Of hundreds of veg- etable materials that could readily be put in pipes or rolled in cigars or cigarettes and smoked, people have chosen to smoke nicotine-containing tobaccos. We also recognize that these people enjoy whatever pharmacological and psychological effects they achieve from smoking, and that this enjoyment can positively reinforce the smoking habit. Most importantly, however, common sense tells us that nicotine is not like heroin, cocaine or any other "classic" addicting drug in its phys- iological and behavioral effects. One does not have to be a trained behavioral scientist to come to this conclu-, sion. Simply ask and honestly answer the question as to how many people would board a plane piloted by someone who had just consumed an addicting drug (al- cohol, heroin, cocaine, barbiturates) versus a plane pilot- ed by someone who had just had a cup of coffee and smoked a cigarette. Interestingly, this latter pilot would be classified as a "poly-drug" abuser by NIDA since, like nicotine, caffeine produces "euphoria" (Henn- ingfield 1986) and possesses "all the cardinal features of a prototypic drug of abuse (Griffiths et al. 1986, p 416; see also Griffiths and Woodson 1988; Holtzman 1990). If nicotine is not an addicting drug as the 1988 SGR has painted it to be, the next question that arises is: what motivates the continued use of tobacco products by humans? We believe that Warburton (1990) has devel- oped a balanced, functional theory of nicotine use that recognizes the beneficial psychological effects of nico- tine. This "resource" or "psychological tool" hypothe- sis holds that people smoke cigarettes primarily for pur- poses of enjoyment, performance enhancement and/or anxiety reduction. This theory also passes the common- sense test of why people smoke. They smoke, not because they are addicted to nicotine, but because they achieve some benefits from smoking, enjoy tLese benefits which are totally compatible with everyday tasks and stresses, and choose to continue to enjoy these benefits. While this manuscript cannot "prove" the resource hypothesis, I

I I I I I I I I I I I I I I I I I I I 406 we note that this hypothesis still stands, using the same data. after the addiction hypothesis has fallen. We believe the distinctions are clear and cannot be stated more clearly than what was said in the 1964 SGR: "the practice (smokiflg) should be labeled habituarren to distinguish it clearly from addiction, since the biologi- cal effects of tobacco, like coffee and other caffeine- containing beveraees. ... are not comparable to those produced by morphine, alcohol, barbiturates, and many other potent addicting drugs" (p 350, emphasis in origi- nal). If we lose this common-sense perspective of the role of nicotine in tobacco use, those of us who enjoy the " lift" we receive from that first cup of coffee in the morning or that cola drink in the late afternoon may Fnd that a few years from now a small group of researchers have equated our coffee/cola-drinking be- havior to that of a hard-core crack or heroin addict. Acknowledgemenrs. The authors wish to thank Drs. Charles D. Spielberger and John H. Reynolds, IV for comments on an earlier draft of this manuscript. References Benowitz NL (1987) The human pharmacology of nicotine. In: Cappell HD et al (eds) Research advances in alcohol and drug problems, vol 9. Plenum Press, New York, pp 1-52 Benowitz NL, Jacob P (1990) Intravenous nicotine replacement suppresses nicotine intake from cigarette smoking. J Pharmacoi Exp Ther254:1000-1005 Bourne SJ (1991) Towards a neuro-endocrine explanation of tobac- co addiction. 1 Smoking-Related Disord 2:105-110 Bozarth MA (1990) Drug addiction as a psychobiological process. In: Warburton D (ed) Addiction controversies. Harwood, Lon- don, pp 112-134 Bozarth MA, Wise RA (1985) Toxicity associated with long-term intravenous heroin and cocaine seif-administration in the rat. JAMA 254:81-83 Burling TA (1984) Tobacco. In: Department of Health and Human Services Publication No. (ADM) 85-1372. US Government Printing Office, Washington DC, pp 85-104 Carmody TP (1989) Affect regulation, nicotine addiction, and smoking cessation. 1 Psychoact Drugs 21:331-342 Clark D (1990) Discriminative properties of drugs of abuse. In: Warburton D (ed) Addiction controversies. Harwood, London, pp 185-200 Collins AC (1990) An analysis of the addiction liability of nicotine. In: Erickson CK, Javors M, Morgan WW (eds) Addiction po- tential of abused drugs and drug classes. Haworth Press, New York, pp 83-101 Davies JB (1990) Life stress and the use of illicit drugs, alcohol, and tobacco: Empirical findings, methodological problems and attributions. In: Warburton D (ed) Addiction controversies. Harwood, London, pp 283-295 Dworkin SI, Broadbent J, Guarino R, Robinson JH (1991) Envi- ronmental determinants of the relative reinforcing efficacy of nicotine, Soc Neurosci Abstr 17:425 Eiser R (1990) Social cognition and comparative substance use. In: Warburton D (ed) Addiction controversies. Harwood, Lon- don, pp 271-282 Fernstrom JD (1983) Role of precursor availability in the control of monatrtine biosynthesis in brain. Physiol Rev 63:484-496 Frith CD (19'1) Smoking behaviour and its relation to the smok- er's immediate experience. Br 1 Soc Clin Psychol 10:73-78 Froggatt P. Wald NJ (1989) The role of nicotine in the tar reduction programme. In: Wald NJ, Froggatt P'eds) Nicotine, smoking and the low tar programme. Oxford tlniversity Press, Oxford, pp 229-236 Gilbert DG (1979) Paradoxical tranquilizing and emotion-reducing effects of nicotine. Psychol Bull 86:6-t3-661 Gilbert DG. Welser R (1989) Emotton, anxiety and smoking. in: Ney T, Gale A (eds) Smoking and human behavior. Wiley, New York, pp 171-196 Gilbert DG. Meliska CJ, Williams CL, Jensen RA (1992) Subjec- tive correlates of smoking-induced elevations of plasma beta- endorphin and cortisol. Psychopharmacology 106:275-281 Goldstein A. Kalant H (1990) Drug policy: Striking the right bal- ance. Science 249: 1513-1521 GrifTiths RR. Woodson PP (1988) Reinforcing effects of caffeine in humans. J Pharmacol Exp Ther 246:21-29 Griffiths RR. Brady JV, Bradford LD (1979) Predicting the abuse liability of drugs with animal self-administration procedures: Psychomotor stimulants and hallucinogens. In: Thompson T. Dews PB (eds) Advances in behavioral pharmacology, vol 2. Academic Press, New York, pp 163-208 Griffiths RR, Bigelow GE, Liebson IA (1986) Human coffee drink- ing: reinforcing and physical dependence producing effects of caffeine. J Pharmacol Exp Ther 239:416-425 Henningfield JE (1984a) Behavioral pharmacology of cigarette smoking. In: Thompson T, Dews PB, Barrett 1E (eds) Advances in behavioral pharmacology, vol 4. Academic Press, New York, op131-210 Henningfield JE (1984b) Pharmacological basis.and treatment of cigarette smoking. J Clin Psychiatry 45::-t-.34 Henningfield JE (1986) Nicotine: An old-fashioned addiction. In: Snyder S (ed) The encyclopedia of psychoactive drugs. Chelsea House Publishers, New York Henningfield JE (1987) Nicotine. In: The second triennial report to congress; Department of Health and Human Services Publi- cation No. (ADM)87-1486. US Government Printing Office. Washington DC, pp 93-119 Henningfield JE, Goldberg SR (1983a) Control of behavior by intravenous nicotine injections in human subjects. Pharmacol Biochem Behav 19:1021-1026 Henningfield JE, Goldberg SR (1983b) Nicotine as a reinforcer in human subjects and laboratory animals. Pharmacol Biochem Behav 19:989-992 Henningfield JE, Nemeth-Coslett R (1988) Nicotine dependence: interface between tobacco and tobacco-related disease. Chest 93:37S-55S Henningfield JE, Miyasato K, Jasinski DR (1983) Cigarette smok- ers self-administer intravenous nicotine. Pharmacol Biochem Behav 19:887-890 Henningfield JE, Miyasato K, Jasinski DR (1985) Abuse liability and pharmacodynamic characteristics of intravenous and in- haled nicotine. J Pharmacol Exp Ther 234:1-12 Henningfield JE, Lucas SE, Bigelow GE (1986) Human studies of drugs as reinforcers. In: Goldberg SR, Stolerman IP (eds) Behavioral analysis of drug dependence. Academic Press, New York, pp 69-122 Henningfield JE, London ED, Jaffe JH (1987) Nicotine reward: Studies of abuse liability and physical dependence potential. In: Engel J, Oreland L (eds) Brain reward systems and abuse. Rsven Press, New York, pp 147-164 Henningfield JE, Clayton R, Pollin W (1990) Involvement of tobac- co in alcoholism and illicit drug use. Br 1 Addict 85:279-292 Henningfield JE, Cohen C, Slade JD (1991) Is nicotine more addic- dve than cocaine? Br J Addict 86:565-570 Hindmarch I, Kerr JS, Sherwood N (1991) The comparative psy- chopharmacology of nicotine. In: Adlkofer F, Thunu K (eds) Effects of nicotine on biological systems. Birkhauser, Basel, pp 509-519 Holtzman SG (1990) Caffeine as a model drug of abuse. TIPS 11:355-356 Hughes JR, Hatsukami DK, Pickens RW, Krahn D, Malin S, Luknic A(1984) Effect of nicotine on the tobacco withdrawal syndrome. Psychopharmacology 83:82-87 Hughes JR, Gulliver SB, Amori G, Mire.ault GC, Fenwick JF (1989) Effect of instructions and nicotine on smoking cessation,