Tobacco Documents Online

National I nstitute on Drug r1 ^ Abuse 11) )'-~ 1 t e earch~ ~J A MONOGRAPH SERIES 11 S. OEPApTMFNT [1F HFAI TH Fnll(`&TFnN emn wGi tApG . DIihM Hsakh Cenilrn , ew,hm n..- ew.., TIlVLN 0152357

Cigarette Smoking as a Dependence Process Editor: Norman A. Krasnegor, Ph.D NIDA Research Monograph 23 January 1979 DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Alcohol, Drug Abuse, and Mental Health Administration National Institute on Drug Abuse Division of Research 5600 Fishers Lane Rockville, Maryland 20857 TIIVVIN 0152358

The NIDA Research Monograpn series is preparea ay me vwIaIon u1 RtnVUra1 r u1 the National Institute on Drug Abuse. Its primary objective is to provide critical re- views of research problem areas and techniques, the content of state-of-the-art conferences, integrative research reviews and significant original research. Its dual publication emphasis is rapid and targeted dissemination to the scientific and professional community. Editorial Advisory Board Avram Goldstein, M.D. Addiction Research Foundation Palo Alto, California JeromeJaffe, M.D. College of Physicians and Surgeons Columbia University, New York Reese T. Jones, M.D. Langley Porter Neuropsychiatric Institute University of California San Francisco, California William McGlothlin, Ph.D. Department of Psychology. UCLA Los Angeles, California Jack Mendelson, M.D. Alcohol and Drug Abuse Research Center Harvard Medical School McLean Hospital Belmont, Massachusetts Helen Nowlis, Ph.D. Office of Drug Education, DHEW Washington, D.C. Lee Robins, Ph.D. Washington University School of Medicine St. Louis, Missouri NIDA Research Monograph series Karst J. Besteman ACTING DIRECTOR, NIDA William Pollin, M.D. DIRECTOR, DIVISION OF RESEARCH, NIDA Robert C. Petersen, Ph.D. EDITOR-IN-CHIEF Eleanor W. Waldrop MANAGING EDITOR Parklawn Building, 5600 Fishers Lane. Rockville, Maryland 20857 ~~' ,'~IMN 015~~~

I Cigarette Smoking as a Dependence Process TIMN 0152360,

I ACKNOWLEDGMENT The meeting on which this monograph is based, held June 19, 1978, was sponsored by the National Institute on Drug Abuse and the Conmittee on Substance Abuse and Habitual Behavior of the National Research Council. The National Institute on Drug Abuse has obtained permission from the copyright holders to reproduce certain previously published material as noted in the text. Further reproduction of this material is prohibited without specific permission of the copyright holders. With these exceptions, the con- tents of this monograph are in the public domain and may be used and reprinted without special permission. Citation as to source is appreciated. Library of Congress catalog card number 79-60046 DIO;W publication number (AIM) 79-800 Printed 1979 NIDA Research D'fonographs are indexed in the Index Medicus. They are selectively included in the coverage of BioScienees inforntation Service, ChemicaZ Abstracts, PsychoZogicaZ Abstracts, and Psycho- pharmacoZogy Abstracts. iv . - T'jMN 0152361

Foreword The 1979 Smoki.ng and Health report and its press coverage are giv- ing new proms.nence toTeTiealth risks of smoking. At the same time, the tobacco industry advocates "personal choice" and no restrictions on smoking, and manufacturers are producing more cig- arettes of lowered tar/nicotine content. Most importantly, millions of individuals--from preteens to adults--are making personal de- cisions about smoking: to begin to smoke; to change to cigarettes delivering less tar and nicotine (but no less carbon monoxide); to stop smoking; and, for most of those'who stop, to "light up" again after a period of abstinence. Once a person starts to smoke, future choices are less freely made, because smoking is addictive. The nature of the dependency is not well understood, and new knowledge is vitally needed. Hence the de- pendence process'is the focal point of this volume. The findings of researchers who are now seeking answers to questions about smoking should someday make it possible to influence more in- dividual smoking decisions and to lessen the high economic and social costs of smoking. This will come through increased knowledge of such areas as: social and psychological factors in the initiation of smoking; just what makes cigarette smoking "rewarding"--and also dependence-producing; the effects of nicotine and the many other sol- id and gaseous compounds in cigarette smoke; the sites of action of nicotine in the brain and of possible compounds which may alter or block its effects; the nature of smoking withdrawal and abstinence syndromes. v TIMN 0152362

This volume on Ci aSmoking as a Dependence Process presents the results of some o t e researcTi ai is now eg6-innning to answer such questions and to raise still further questions for which answers are needed. The National Institute on Drug Abuse offers this monograph with thanks to all those who are carrying forward this work and with an invitation to other qualified investigators to join them. As part of NIDA's fulfillment of its charge to carry out and support research on tobacco-smoking behavior, this book takes its place with Research Monograph No. 17, Research on Smok- npon ~in Behavior (1977), and sections on smoking witFin~ monograp~o roa er subject of substance abuse: No. 20, Self-Adm.i.nistration of Abused Substances (1978), and No. 25, Behaviora a ysss an 'I7eatment o stance Abuse (forthcoming)'-- We hope that these pub lications as well as the research reported are a significant contribution to the continuing effort to reduce the toll of disabilities and deaths related to cigarette smoking. William Pollin, M.D. Director Division of Research National Institute on Drug Abuse vi TIM.N 0152363

Preface Cigarette smoking is a dangerous habit. It can lead to a variety of ailments and serious disorders, from impaired breathing to de- bilitating and fatal illness, such as heart disease and cancer. Cigarette smoking is "slow-motion suicide," in the words of Joseph A. Califano, Jr., Secretary of Health, Education, and Welfare. Such facts are widely known and accepted by most--including many who smoke. Yet, some 54 million Americans smoked 615 billion ciga- rettes last year. The immediate question that comes to mind is "why"? Surveys find 90 percent of smokers saying they have tried to quit or probably would if they had an effective way to do so. In the past 15 years, 30 million smokers have quit the habit, almost all of them on their own. Again, the question is "why"? If this habit were merely an irritating or unpleasant practice, it might be interesting to explore as another curious facet of human behavior. But the smoking habit is a case of widespread self- injury with.enormous health consequences that cost the Nation an estimated $27 billion a year, not to mention the personal toll in sickness and death. We face a major public health problem created by millions of individual decisions to risk well-being for a puff of smoke. Why? There are many theories to explain such behavior but only pieces of possible answers. Because of the magnitude of the problem, those concerned with protecting the, public health are compelled to promote the prevention of smoking in the first place and to find ways to help others break the habit. To achieve those ends, we must understand the basis of the smoking habit and its mechanism. The papers collected in this monograph are steps toward that under- standing. vii i TIMN 0152364

i The authors of these papers are among the first to admit that they do not have final answers, that questions far outnumber answers ~ in this field. But good questions lead to solutions, and this effort represents a move toward that goal. The health consequences of smoking have been studied intensively, as reflected by more than 30,000 articles on the subject and the research summarized in the 1979 Report of the Surgeon General on ~Sm~ok~i~n ~and FIealth. But, until recently, little serious research F~a.s~~een con u3 ctea into the reasons why the smoking habit begins and how it maintains its hold. Simply put, that is the subject of these scientific papers: To define the habit that leads indi- viduals to act counter to common sense, and to discover why. It is appropriate that the National Institute on Drug Abuse assume leadership in the exploration of these questions because smoking is surely the most widespread example of drug dependency. The drug, nicotine, for example, is but one of the 2,000 components of cigarette smoke and a pack-a-day smoker of cigarettes self- administers more than 50,000 puffs--or doses--a year. That is in- deed drug dependency. The smoking habit is an excellent subject for research into the biological, behavioral, and psychosocial aspects of the dependency process. This kind of research can help us learn why more than 4,000 young- sters take up the habit daily; why women, smoking in greater num- bers than ever, find it more difficult than men to quit; why some individuals can break the habit "cold turkey" and others find it impossible to quit. There are many other questions that need an- swers: Is the smoking habit a "true" drug addiction, or is it a learned social or behavior pattern? Are there nicotine receptor sites in the central nervous system that can be blocked by medi- cation to break the habit? Why, in other words, do people persist in the habit despite know- ledge of the health consequences? And how can they be helped? Such questions are explored.in these papers in the hope that the theories they present, the clues they offer, and the ideas they prompt will lead to the answers we need. John M. Pinney Director Office on Smoking and Health "T1MN 0152365

Contents Foreword WiZZiam PoZZin . . . . . . • • . • . . . . . v Preface John M. Pinney . . . . . . . . . . . . . . vii Chapter 1 Introduction Norman A. Krasnegor . . . . . . . . . . . . . 1, Part I. PSYCHOSOCIAL FACTORS Chapter 2 Smoking as an Addictive Disorder Jerome H. Jaffe and Maureen KanzZer ........ 4 Chapter 3 Psychological Analysis of Establishment and ' Maintenance of the Smoking Habit Danie Z Horn . . . . . . . . . . . . . . . . 24 Chapter 4 Cigarette Smoking as a Precursor of Illicit Drug Use John A. 0'DonneZZ . • • • • • • . • • • • • • 30 Chapter 5 Patterns of Tobacco Use in the United States Dorothy E. Green • • . • • • • • • • • . . 44 Part II. BEHAVIORAL FACTORS (hapter 6 Nicotine as a Discriminative Stimulus to Behavior: Its Characterization and Relevance to Smoking Behavior John A. Rosecrans • . . . . . • • • • • • • . 58 Chapter 7 Nicotine Self-Administration in Rats H. M. Sanson, C. A. Ivester, and B. R. Morton ••• Chapter 8 The Effects of d-Amphetamine, Meprobamate, and Lobeline on the Cigarette Smoking Behavior of Normal HLunan Subjects C. R. Schuster, B. R. Lucchesi, and G. S. EbaZey .. ix • .70 TIMN 0152366

r„" Part II. BEHAVIORAL FACTORS (continued) Chapter 9 Tobacco Dependence: Is Nicotine Rewarding or Aversive? M. A. H. RusseZ2 . . . . . . . . . . . . . . 100 Chapter 10 Regulation, Withdrawal, and Nicotine Addiction StanZey Schachter . . . . . . . . . . . . . .123 Part III. PSYCHOBIOLOGICAL FACTORS Qiapter 11 Acute and Chronic Effects of Nicotine in Rats and Evidence for a Noncholinergic Site of Action L. G. Abood, K. Lorvy, and H. Booth ... .....136 Chapter 12 Tolerance to the Effects of Tobacco Murray E. Jarvik . . . . . . . . . . . . . .150 Chapter 13 The Tobacco Withdrawal Syndrome SauZ M. 9hi ft3nan . . . . . . . . . . . . . .158 Part IV. IND?LICATIONS AND DIRECTIONS FOR FUIURE RESEARCH Chapter 14 Implications and Directions for Future Research Norman A. Krasnegor . . . . . . . . . . . . .186 Participants in Symposium on Cigarette Smoking as a Dependence Process . . . . . . . . . . . . . . . . . .190 List of Monographs . . . . . . . . . . . . . . .191 x TIMN 0152367

Chapter 1 i ntroduction Norman A. Krasnegor, Ph.D Cigarette smoking is of interest to the National Institute on Drug Abuse both because of the public health problems associated with this form of substance abuse and our view that this behavior repre- sents a prototypic dependence process. The scientific data which link cigarette smoking with risks to health have been well and amply documented in the first Surgeon General's Report on Smoking and Health (USDHEW 1964) and the recently updated version of that docu- ment (USDHEW 1979). Despite this linkage, relatively little scientific research has been conducted to describe and analyze the cigarette smoking habit itself or the factors which are responsible for its initiation, development, maintenance, and cessation. Health risks associated with tobacco use are predicated upon the necessary existence of a chronic, habit- ual pattern of cigarette smoking. Scientific data which characterize the smoking habit are essential, therefore, because they can provide an understanding of the dependence process and guide the develop- ment and testing of efficacious treatment strategies. This monograph is based upon a meeting held at the National Academy of Sciences in June 1978, sponsored by NIDA and the Committee on Substance Abuse and Habitual Behavior of the National Research Council. The intent of the meeti.ng was to review current knowledge concerning the psychosocial, behavioral, and psychobiological factors which characterize the dependence process associated with cigarette smoking and make cessation of it difficult. This volume, which includes papers presented at the symposium, is designed to provide an overview for the scientific commtmity on the smoking habit and an agenda to guide future research in this area. The monograph is divided into four sections. In the first, psycho- social factors relating to the dependence process associated with cigarette smoking are explored. A stinulating discussion of how to characterize the habit is presented by Drs. Jerome Jaffe and Maureen Kanzler. Patterns and trends in tobacco use in the United States are detailed in Dr. Dorothy Green's chapter. Cigarette smoking as a precursor 1 ; TIMN 0152368

of illicit drug abuse, based on a sample of young men, is discussed by social scientist Dr. John A. 0'Ibnnell. The dean of American researchers in the area of cigarette smoking, Dr. Ifaniel Horn, pro- vides a perspective on the psychological factors involved in the establishment and maintenance of the habit. The second section, on behavioral factors, is devoted to discussions of theoretical and empirical data on the role played by nicotine in the dependence process. Dr. Rosecrans develops arguments concerning the properties of nicotine as a discriminative stimulus. The well- known English researcher, Dr. Michael A. H. Russell, presents his perspective on the dependence liability of nicotine and the smoking dependence process. Dr. Schuster and his coworkers report findings on the effects of nicotine on smoking behavior, while Dr. Schachter discusses his social psychological experiments designed to determine the relationship of nicotine to withdrawal and addiction. The paper by Dr. Hanson and his colleagues provides convincing empirical evi- dence that nicotine is a reinforcer. This information is of special interest because it demonstrates an experimental model of nicotine self-administration and because it provides a method for studying pharmacological and behavioral variables associated with the rein- forcing efficacy of the drug. The third section is devoted to psychobiological phenomena associated with the smoking process. The paper by Drs. Abood and Lowy provides evidence suggesting the existence of a central noncholinergic recep- tor that is specific for nicotine, an exciting field of investigation. Further, the techniques described offer a methodological approach to the study of ways to centrally block the reinforcing effects.of nicotine. The papers by Drs. Jarvik and Shiffman discuss their observations respectively on the development of tolerance to cigarette smoking and the withdrawal symptoms associated with cessation of smoking. 'Ihi.s latter work is of particular importance because abstinence symptoms have been correlated strongly with the relapse to smoking after cessation. The final section, by Dr. Krasnegor, is a brief agenda for future re- search on•smoking. It is hoped that this listing of research needs will be used by members of the scientific commoity as a focus in planning and carrying out their research and as a guide in requesting extramural funding support from the National Institute on Drug Abuse. REFERENCES U.S. Department of Health, Education, and Welfare, Public Health Service. Smok'ng_ and Health, Report of the Advisory Committee to the Surgeon'~enerato t e blic Health Service, P.H.S. Publica- tion No. 1103, U.S. Government Printing Office. 1964. U.S. Department of Health, Education, and Welfare, Public Health Service. ~Smoking and Health, A Report of the Surgeon General, P.H.S. Pub ic~"-ation-W.79-5-0 066, U.S. Government Printing Office. 1979. 2

Partl Psychosocial Factors I ; TIMN 0152370

Chapter 2 Smoking as an Addictive Disorder Jerome H. Jaffe, M.D., and Maureen Kanzler, Ph.D. Bishop Bartolnme de las Casas, observing the use of "tabacos" by Spanish settlers In the New World, wrote that Vhen reproached for such a disgusting habit, [they] replied that they found it inpos- sible to give it up. I csnnot understand what enjoyment or advan- tage they derive from it" (de las Casas, in Corti 1932, pp,42-43). 7bday, appmxfmately 450 years after de las Casas racorded those observations, we are still considering the same two guestions in regard to cigarette snaking: hhy don't people give it up? 14nd what advantage or enjoyment do they derive fram it? Some wnrk has been done in the interval and some of the researchers who have contributied greatly to our kmwledge are participants in this sym- posium. Because of their •aork we can now fornulate same reasonable hypotheses about the origins of the "enjoynent or advantage" people derive from the smoke of tobaooo leaves, and we even have a aon- siderable body of experience about helping people give up the habit. Die las Casas would be happy to kmw that giving it up is not irpos- sible, although for some tobacco users.giving it up is difficult and relapse is ooamnn. The title of this monograph refers to smoking as a "dependence process." Pesemblance between tobaooo use and`oonsmption of other substances that produce dependence has been debated throughout his- tory. In 1604, James I, in his O~unterblaste to 7bbaoco ((brti 1932) appeaxed to view bobacao process o habitu- ated as quite anaLoc.pu.s to the process by which a drinker of alca- hol becane a drudcasd. Three huidred years later, Sir Humphrey Bolleston, whose conmittee reooinmandations in 1926 set the tone for the British response to opiate dependence, was asked whether tobacao smdcing was not properly viewed as an addiction. In his reply, Sir Humphiey differed fran James I: This question turns on the meaning attached to the ward "addiction", and nay therefore be a verbal problem. 7he Ministry of Health's Departmsntal Cbnmi.ttee on Nlorphine and Heroin Addietirn (1926) defined an addict as a "person 4 -,TIMN 0152371

who, not requiring the continued use of a drug for the re- lief of the symplrnis of organic disease, has acqaired, as a result of repeated administration, an overpowering de- sire for its continuance, and in wham withdrawal of the drug leads to definite symptans of inental or physical dis- tress or disorder." That smking pzmduces a craving for moxo when an attempt is made to give it up... is undoubted, but it can seldom be accurately described as overpowering, and the effects of its withdrawal, tlrough there may be definite restlessness and instability, cannot be compared with the p!iysical distress caused by withdrawal in mor- phine addicts. Zb regard tobacco as a drug of addiction may be all very well in a humrous sense, but it is hard- ly accurate (Iaolleston 1926). In at least one sense Rolleston was correct: this issue is a senen- tic one. And when senantic problems arise, there is always a pos- sibility that the argunents about whether tobacco snoking is prop- erly grouped with other forms of r»nmedieal drug use will divert energy fran more pragmatic questions. We do not have to asoertain whether all aspects of tobacco use resemble other drug-using be- haviors in all of their particulars. The problems posed by aloo- hol, opiate and eocaine use differ from each other in a number of significant ways. Zhe essential question is to what degree oocloeptualizing tobacco use as one of the ar3dictive disorders is of help in directing us toward appropriate means to deal arith pxoblems that tobacco use causes. Not all dependence on drugs results in problens for society and/or the individual. Chffeine consunQtion is viewed by many as appro- priately classed with other fornis of dependence, and caffeine de- penc3enoe can be found in the International Classificat3on of Dis- eases (ICD 8). There is a caffeine withdrawal syndmne and caffeine can be abused to the point where it causes problems and disrupts behavior (Gilbert 1976). But so lorg as the price of coffee re- veins within reasonable bounds, scientists and policymakers alike will think aboutwaaffeine primnrily as samething that adds inmeas- urably to the beverage served at the coffee break and without which it is difficult to start the day. Fbr the most part, no grant applications are submitted to develop preventive trsatments for caffeine dependenees there are no debates in the hails of cbngress about taxing it; and coffee drinkers are not forced to sit in the rear of airplanes. Perhaps saoe day caffeine may beeaos of concern to behavioral scientists, but, for the present, the personal and social costs of this dependence appear to be relatively low. Th- bacco dependenee, on the other hand, has enornpus cost to the indi- vidual who develops sroking-related diseases, and these diseases in turn affect the eoonanic wel.l-being of society. There are several significant areas in which tobacco use resenbles other drug use, as well as a few areas in which it diverges. 5 ,TIMN 0152372 '

F3ien we begin to examine the ways in which tobacco use resembles other drug-using behaviors, we find ourselves asking the identical qusstions: Wiat factors - biological, psychological, sociological and pharmacological - determine whether there will be experimenta- ticn with the drug, a progression to casual or recreatiociall use, or on to intensive (or excessive) use? Wat factors aze associ- ated with (or cause) cartpulsive (addictive or dependent) use, and which factors are associated with relapse after abstinence has been achieved? Although the factors are interactive, we asstme with tobacco, as with other drug-using behaviors, that certain factors could act primarily at one stage while others might expst their effects at other stages. The objective of this brief overview is not to attempt to stmmt.~r- ize all the factors involved in tobacco-using behavior, but to point out a few notable similarities and differences between to- bacco use and the drug-using behaviors that are nnt+re cammnly viewed as "addictions" and to speculate on what the future may hold. SCME INfFm.S'tIIv- PSYmIAGICAL PARAIdEIS FzATID TO INITIAL USE As with most other forms of nonmedical drug use, the initial ex- perimentation and regular use of tobacco begin in youth. in the present climate, which is considerably less approving of.cigarette use than it once was, the behavior often is seen nnre canronl.y among the less well adjusted (Snith 1970) and less scholastically successful (Barland and Rudolph 1975; Simon and Prinavera 1976), and especially atnong those who have friends who smoke (Iarsai and Silvette 1975). Although there is great overlap between the psychological charac- teristics of snakers and ncnsmokers, in study after study, ciga- rette saakers on average tend to be nore extroverted (9mith 1970), more intolerant of rules, nore adventurescme and risk-taking and, in some studies, nnre angry (T4oms 1973) than appropriately matched nonmnokers (for additional references see Larsai and Silvette 1971, 1975). Mile it can be argued that some of these differences may be a result of smdcing, they are observed even anmong young people just beginning to smoke (Smith 1969) and they seem to persist when the smker becaaes abstinent (2lsonns 1973). Eysenck (1973) has postulated that the smoker is an extrovert who is usually at less than his or her optimal level of arousal and therefore uses nicotine to raise the level of arousal. Sudi a view leads to a "normalizing" hypothesis to accamt for the maintenance of the habit in at least same smokers. Yet many of these sane personality characteristics seem to be associated with experimentation with other drugs, e.g., LSD, opiates and alcohol (Haeburg, Kra.eeer and Jahnke 1975; Jaffe, 1977) which are not acambnly viewed as inducing arousal, an obser- vation which is difficult to reconcile with Eysenck's hypothesis. Most of those who begin to smoke cigarettes believe that they will some day give thesn up (Lieberman 1969). Very few cigarette smok.ers at present start out to become dependent. We must infer fram their behavior that gradually the capacity to choose is eroded and, while 6 TIMN 0152373

the user may want to believe s/he can stop at any time, the behav- ior indicates that this is not the case. 7he attitudes and beliefs about the likelihood of becaning dependent are not very different among those who begin to use opiates and alcohol. With tobacco usars, as with users of other drugs, there are nimer- ous theories that attempt to account for the transition fran experi- mentatioai to continued use. With most other drugs obsQrvess are willing to attribute the ongoing behavior, at least in part, to the effects of the drug itself; but even with the opiates and alcotrol, researchers recognize that, in certain social settings, the act of using the drug (rather than its pharmacological effects) may con- tinue to provide some of the reinforcenent. So it is with ciga- rettes - it is a matter of degree. Russell, Peto and Patel (1974) investigated nm4tives for smoking in two groups - one camposed of "normal" smokers, the other an ad- dicted group of heavy smokers attending a witthdxaeaall clinic. A factor analysis of their responses to a questionnaire separated a "pharmacological addiction" dimension fram the sensorinotor, indul- gent,and psychosocial factors. The sensorim»ter and indulgent fac- tors appeared to be related to the individual's ability to experi- ence pleasure or,its enhancement by smoking and to the act of man- ipulating the cigarette. 7he psychosocial.,factor reflected associ- atian of cigarette smoking with a,desired public image and with ease in social situations. In this, as in other studies using factor analysis (e.g., Ikard, Green and IHorn 1969; Mc Kennell 1970) an addictive dimension repeatedly emezges, but is always acoompanied by nonpharmacological factors. Iiowever, the mez+e presmzce of non- pharmacological factors in the maintenance of smoking does not serve to distinguish cigarette smoking fram other drug-aaing behaviors. Social reinforcers and symbolic aspects of the drug-taking behavior are also postulated to play a major role in the developnsnt of a variety of drug-using behaviors and, indeed, of deviant behaviors in general (Jessor and Jessor 1977). Again, the objective here is to point to parallels rather than to survey the literature. PHARNACO?MCAL FAC'1C42S IN OCtTPIIVI]ID USE For many years researchers'hanre+ assused that, afte= smoking has been initiated through psyehosocial factors, the behaniar beatmes habitual because the pharmacological effects of nicotine are rein- forcing (for references eee Ejrup 1965; Larson and Silvette 1971 and 1975= Jarvik 1973). Russell (1971 and 1976) has emphasized that a ama7.l "bolus" of nicotine reaches the brain within seconds after a puff froam a tobacco cigarette is i.nhaled. If nicotine is a reinforcer, then the hiaidreds of puffs inhaled each day should produce a well-established puff-inhalatien habit. There appears to be support for the viedv that it is, indeed, nicotine which is the major reinforcing eonponeet in cigarette smoking (although it may not be the only reinforcer). tdien nicotine and tar content are varied independently, it is the nicotine content that is eorre- lated with ratings of strength and satisfactian (Goldfarb et al. 1976). Men provided with lear or rnn-nieotine cigarettes, I i 7

1. most smokers oanglain bitterly or refuse to cositinne smoking them (sm Jarvik 1973). Nevertheless, reliable laboratory evidence that nicotine is a reinforcer of drug-taking behavior has been more dif- ficult to develop than cxitparable evidence for drugs like morphine, anQtietamine or cocaine. In contrast to the latter drugs, which anima].s will self-aaninister over a wide range of doses, an4nall self-administration of nicotine has been more difficult to induce (however, see Hanson, this volume). When it does oocur, it appears to be a less powerful reinforcer of behavior than drugs suah as cocaine and amphetatnine, at least as judged by the nurber of lever presses that the animal will aeke for a single dose of nicotine (Yanagita 1976). Nicotine has both peripheral and central effects. The pgeri.pheral effects, such as inhibition of stansch eontractions, aeceleration of heart rate, release of epinephrine fian the adrenal gland, and effects mediated by periphera]l release of noradrenaline do not seee to be of major inQortance in reinforcing sndcing. Most of these can be blocked without altering the psychological effects in man (Carruthers 1976). The central effects are obniously more relevant. But which ones? Nicotine appears to produce nmultiple effects -- and in this respect the problem of identifying the site of the re- inforcing effects of nicotine is not uilike the problem of deter- mining which of the nultiple effects produced by the opioids or alcdhol are responsible for their reinforcing properties. In man, nicotine produces an alerting pattern in the EHG and behavioral arousal (Danino 1973; Larson and Silvette 1975). It also stinulates release of a nunber of hormanal snbstanoes from the CNS (Husain et al. 1975; Winternitz and Quillen 1977; CCryer et al. 1976). Ani- nel studies indicate that nicotine releases norepinephrine and dopa- mine frcm brain tissue (See Goodman 1974; Larson and Silvette 1975; Russell 1976). Depending on the dose, it mny increase or decrease the release of acetylcholine (Axmitage, Hall, and Sellers 1969; Russell 1976). It may also affect brain,levels of serotonin. How- ever, with nicotine, as with other drugs, these effects on neuro- transtnitters do not tell us.how nicotine reinforces sndcing behavior. BIOIAGICAL FAGTCI2S IN CCNTINJID ANID DF.PENDfNr USE Many people drink alcahols a relatively sroall proportion beocme de- pendent.-'Not all of those who use opiates beoane dependent. With the latter drugs, those who becaone dependent tend to cane fran dis- turbed families where alcaholima and, often, a history of sociopathy or other psychiatric illness is prani.nent (See Jaffe 1977). <1ne at first suspects that it is the stress of growing up in such a family that leads to the later tendency to overuse aleohol or illicit drugs. However, the search for the basis of vulnerability to dependence on drugs has taken saae surprising turns over the last decade. For males, a genetically transmitted biological vulnerability to alco- holism, that may be independent of disturbed family background, ap- pears to be fairly well established (Scliuckit, Goodwin, and Winokur 1972; Goodwin et al. 1973s Geockwin et al. 1974). less defined and only suggested by the discovery of opioid receptors and endogenous opioids is a possible vulnerability to opiate addiction. Even be- fore this discovery, sone opiate users maintaineci that they used TIMN 0152375

opiates not to get high but to feel rbrnal - not to alleviate with- drawal symptoans but to experience a state of noxma].ity that the ncnusPx enjoys without the benefit of exogenous substances. The existence of a syndrane characterized by relative inactivity in an endogenous opioid sysiten such that exogenous substances could at least theoretically act to "narnwlize" the user's feeling state has not yet been dennnstrated. Aithough the existence of a biolog- ical vulnerability to tabaceo dependence is also speculative, the possibility of such a biologicall substrate carmot be excluded. 4hil.e we are speculating, we might note that ane of the effects produced by cigarette smoking is a sharp rise in oortisol (Winteznitz and Quillen 1977; Cryer et al. 1976). Wb can assune that this ef- fect is mediated by release of ACTEI. Since it seems that a nale- cule of s-encbrphin is released each time a molecule of AClS is xe- leaseri (Giillemin et a1.1977) ve cannot rule out the possibility that cigarette smoking and release of S-endorphin are related. We might also note that aleoholics and opiate users are not anly more likely to be smdcers but they smdce maeh nnre heavily (Dreher and Fraser 1968) and often find it easier to give up opiates or alcohol than cigarettes. Only the arrogance of ignorance accepts as proven that which can anly be an hypothesis at present - that each and every alcoholic, opiate user, and tobaceo smaker iaaild be able to functiaa better without the substance in questian. Oertainly we believe that there are mi].lians of crarent tobnceo users who will ftnction better if they stop smdcing, but that is nat the same as assuning that all will be able to give it up without cost in te.am of psycho- logical functioning. S,Ta need to know more about the where and the how of tobaeco's ef- fect an the brain and the rest of the central nervous systan. Ub also need to lQnow whether there are people who actually function better when smdcfng. 7he studies that have been carried out on heavy aedcers acutely deprived of nicotine do not anawer this question- PHYSICAL DbRMFiJCE ATD WITFDRAYM 7he existence of physicall dependence is an inference nede fraa the observatien of a stereotyped arithdraw®1 syndrome which oocurs when, a dix+anically adainishered drug is discontinued. 7he withdrawaaX syadraae following smoking cessation is not as well stuaied as other fot~as of withdrawal, but few who are )Qna+rle3geable doubt that it exists. It differs in time course and character fran that follow- ing alcohol or opiate deprivation. The onset of smoking withdrawal sympta: may occur within hours of the last cigarette or may be delayed for days. The synptans may last frcm days to moaths. Like the other withdrawal syndromes, thez+e are associated physiological changes, e.g., decreased heart rate, !:l1G slowing. In addition to craving for tobacoo, other spmptems have been reported following the cessation of snddng, such as restlessneas, dullness, sleep disturbances, gastrointestinal diatutbances, drowsiness, heedache, amnesia, and inpni.1 t of eoncentration, juagmsnt, and pspohamtor pcrfaYnoance (for references see Gdzilfor+d 1966; iarsco and Silvette 1975; Russell 1971; Jaffe and Jarvik 1978; Shiffnan and Jarvik 1976). 9 , TIMN 0152376

As little as 90 minutes of deprivation may increase irritability (Schedhter and hand 1974). However, there are areas of smolcing withdrawal that are virtually unexplorefl, including the factors that determine the severity of the syndrooe, the time eourse of its detrelepnent and decay, the degree to which it can be oonditiened to internal and external (e.n- viroonental) stimil.i, and the relation of the type and severity of the withdrawal phenanena to subject craving, to suocessful cessa- tion and to relapse after cessation. With other drugs, the problem of tolerance is usually discussed in relation to physical dependence. It is apparent that tolerance to nicotine can develep quickly and that tolerance levels can be in- flueace8.. 7here are now hundreds of thousands of aidtpss who every day purchase cigarettes delivering only 0.1 mg of nicotine. Many smolaers who at first find such cigarettes to be totally unsatis- fying can becane axwstaned to tten. Bnt vae know very little about the natare of tolerance to nicotine and the factors regulating its developtient. TInwICN If smdcing leads to physieal dependence and if the withdrawal syn- dmme is an aversive state, then one might expect to find that smdc- ers would try to avoid withdrawal by aeintaining their tobacco (or nicotine) intake. Indeed, it has been proposed that hsavy smcke.rs don't get any significant positive effects fran nicotine, but snaice primarily to avoid wittxlrawal (Schachter 1978). This general prop- osition has provided the basis for a nunber of experiamts that ask the related questions: Do senicers smdce tobacco prinerily tA' get nicotine? (i.e., Is nicotine the reinforcer in tobacco smdc- ing?) Do ffinokers adjust (titrate) their levell of amddag to nmin- tain a given level of nicotine in the body? Tl:e experimsnts havn incluaed: (1) administration of nicotine (orally, sabcutan®ously, and intravenously) eouples3with observations of the number of cigarettes smaked (or nunter of pu.ffs taken) and/or degree of satisfaction (see Jarvfk 1973; Ih,issell 1976; Kuanr et al. 1977; Schuster, this wl- ume); (2) changing the nicotine content of the tobacco and measuring the smking patterns (Jarvik 1973) and sonetiires the pl.asm levels of niootine (Russell et al. 1975); and (3) administering various drugs that alter the pharnaoologieal effects of nieotine (Jarvik 1973) or its disposition (Schachter 1978) and then ohserving stmking behavior and the psychological effects of smaking. The evidence fran most of these studies strongly supports the view that within limits heavy sndcers do attezrpt to regulate their plasma nicotine levels by ad- justing the rate and mount of tobacco smoked. lieavy smataers appear to be more consistent in reducing intake of high nicotine cigarettes to avoid unusually high plasma levels of nicotine than in increas- ing the number of low nicotine cigarettes to avoid unusually low levels (Russell et al. 1975). There are several possible explana- tions for this general finding: (1) Very high plasina nicotine lev- els may be aversive for all smohers,whil.e sharp drops in plasma levels may be aversive for only a subgroup of srokers. Alternatively, 10

the task of trying to maintain plasma levels of nicotine by ssmking a very law nicotine cigarette may itself be aversive because of the characteristics of the cigarette, the accumu].at.ion of carbon mon- oxide, or because the absence of the reinforcer (nicotine) is it- self a frustrating situation. Most titration and nicotine manipulation experiments also demon., strate that in confirmed cigarette smokers factors other than nico- tine cai1tribute to the maintenance of behavior, at least over the short run. For example, although they oomplain bitterly, partici- pants in experinents will continue to smoJce cigarettes with little or no nicotine for a period of a week or aiore (Jarvik 1973). Such participants are of course free to stop smoking at any time. In one experiment, an in'travenous dose of nicotine equivalent to the oontent of a single cigarette had no effect at all on the latency to puff on a real cigarette or on the amoamt smaked (ICanar et al. 1977). That short run puffing behavior does not carrespond per- fectly with nicotine intake or plasm levels does not necessarily mean that tobaeco simking is a distinct form of substanee-using behavior. Perfect titration of plaaea levels of drug is not found in the classic addictive disorders. Alcdholics titrate alcohol levels inperfeat7.y. Depending on envixrnmesital circaastanees and schedules of reinfarcement, such individuals may tolerate major decreases in plasina levels. Similarly, some opiate-dependent sob- jects may continue to work for intravenous opiates when maintained on high cross-toleranoe-inducing doses of methadone (Martin et al. 1973). Many, given adequate n+otivation, will voluntarily accept withdrawel fran chronically administered opiates. In additian, addicts may eontinue to inject theeselv+ss and experience opiate- like effects while en blocking doses of narcotic antagonists (O'Brien 1976). it is lilasly that with nicotine, as with other drugs, the act of drug use itself acquires secondary reinforcing prop- erties; and also, despite the developnent of tolerance, the acute effects of each dose of nicotine may continue to produce effects that are distinct fran relief of withdrawal. we should not expect a perfect oorrelation between nicotine levels and smalcing behavior any more than we expect such correlations amaig the "classic" ad- dictive disiorde.rs. (E$SATION AND RFTAASE Most stmkers, like most drug users, haae.a remnrkable capacity for denial; they appear to beliene that the bad effects caused by drugs happen to other people. But for many smkers there caoes a time when si elenent of self-presernation or concern about cost dic- tates discantinuing tobaeeo use. At such ttmss an attengt to stop is often suocessful; at least in the short run - just as it is for users of other drugs. Hat.the relapse rate is high, just as it seems to be for other addictive substaeoes. Hamt, Barnett, and Branch (1971) plotted relapse rates for heroin users, cigarette smdaers,ard alcaholics. The plotted curves for percent of relapse over a one year pesiod were virtually identical. In considering these curves we need to keep two pointa in mind. They represent data taken frm individuals who sought "treatment." Evidence is 11 TIMN 015237g

In mount3ng that for both alcohol and opiate use the nunber of indivi- duals who change their use patterns without foreal help may exceed, by far, the number who seek formal treatmeat (Cahalan, Cisin, and Crossley 1969). It is estimated that, at present, thirty million nonsmoking Americans were onee regular cigarette smolaers. Most of these stopped without formnll treatment. We cb not have a great deal of information on the relapse rate for this large population. It is entirely possible that ttrose who seek forme7l tx+eatment mny con- stitute a selected group with niore severe feQies of dependence or a greater propensity to relapse. The "treatment" of tobacco dependence, like that of other dependence disorders, has been approached from many perspectives, none of which is entirely satisfactory and some of which appear to be relatively ineffectual. This area has recently been reviewed (Bernstein and Mc Alister 1976; Schwartz 1977). As with heroin and alcohol, the reasons for relapse in tobacco de- pendence appear to be nultiple. Indeed, relapse is prabably over- deteYmirei and is the result of an interaction annrxg several facr tors - the personality of the individual, degree of ocncern about health# the fmctional utility of 9noking, susceptibility to the phaxie;cological effects of tobaox, the degree to which withdrawal phenomenon& beccme linked to enviramental stimiti, and exposure to envisarmental situations which in scme way induce smoJcing. Fhat is not at all clear is the relative eontributioa of these various factors. 9QMg: POPPNrIALLY II410[tTANr DIFFE[ffiJCES BMW= ZCBAE70U USE ADD dl'HM FtRi+1S CP DRi7G DEP'EMFIJCE Despite the parallels which may be drawn between tobacco use and the use of other dependence-producing substances, there are some important diffe.rences. Perhaps the most iinportant is that tobacco does not in8noe the acute behavforal toxicity that is seen with alcohol, opiates, amphetamines, cocaine,and hall.uainogens. The adverse effects of tobeeoo are exclusively remote and sten fran chronic use rather than from occasional indulgence. There are no reports of acute fatal tobaeaoovercbsage as a result of smdcing or induced aberrant behavior. The tobacco user presents little danger to other people (albeit there my be some annoyance from the smoloe and sflae smahers cause fires) and his/her productivity is not low- ere<i until sroking persists for nany years and leads to impai.neent of health. fiurthernare, the likelihood that regular users of to- bacco will suffer a medical illness appears to be directly related to total dose of certain tobacco constituents over tine (Gori 1976). This situation offers hope that there aay be some level and sane patterns of regular use which are relatively free of hazard. ufii].e this is theoretieally true of other drugs, we have thus far been unsuooessful in finding ways to control their acute toxicity or their effects on behavior. 12 + TIRN 0152379

-~.......,,; , In the case of other drug-using behaviors, we believe we have iden- tified the pharm®c.~ological reinforcer. We believe that most drinkers of gin and tonic want the gin aiore than the fizz or the quinine, that people who drink pategoric want the morphine and not the cam- phor. In contrast, we are not entirely certain that nicotine is the only reinforcer of tobacco-usfng behavior. Although Risse.ll (1976) has stated his eanviction that nicotine is the reinforcing canponent of cigarette stnaking, there have been inconsistencies amoag the experiments performed over the past twenty years that re- main to be resolved. Certainly, the acute effects of nicotine in the noasmoacing hunan volunteer have not been as thoroughly explored as have the effects of opioids, alcdhol,or anphetaadnes in non3e- pendent subject groups. For a substance that exerts such a remark- able hold over so many people, nicotine appears to be less than our srost dramatic reinforcer in animals. 2here is no "mystique" about the nicotine "high" that in any way resembles that associated with the use of alcohol, opiates, cocaine, or marihuana. It is curious that it is easier to find descriptions of the acute'effects of the latter substanoes,which are used by only a sma]1 fraction of the population,than eomparable descriptions of the subjective effects of tobacco, which is used by millions. . ... Campared to other drugs, the cost of a daily ration of tabaeco is low. Zb a large degree this is because the cost of other drugs, such as alcohol, nerihuana,and most synthetic substances, has been sharply increased as a result of deliberate social policies - either through taxation or by legal prohibition that adds the costs of maintaining an illicit distribution system to the costs of the drug itself. In teans of understanding the differences between tobacco and other substances, we need to recognize that at present it re- m4ins an3g the relatively inexpensive canrodities in most indus- trialized countries, and for most people it is mmng the easiest to obtain. Since there is evidenoe to suggest that the eonsueption of cigarettes, like other eamodities (Pebo 1974, 1976; Setmt 1978) and other drugs (Rophaae, Sdaniift, and De Lint 1975), is responsive to changes in price, a number of countries are considering taxation schedules that aou].d redisce overall coasunption. In 1976, a bill was introduced i.nto ths ifiited States Senate that would have raised the taxes-on various cigarettes in propartion to the content of tar and nicotine (tlar,t 1976). The social acceptability of tobacco use and depeadenoe is, at pmes- ent, in a class by itself. In most developed eaaitries, moderate use of alcahol is acoepbed and approved. Public oonstanptien of such beverages is part of the fabric of society. Nevertheless, it is oonsidered dishonorable to be seen as an excessive user or to be de- pendent on alcohol, and (despite the pr+ominenee of a number of former alcoholics) most people would rather not advertise their difficulties in keeping alooholl use at nmoderate levels. Most people dependent on tobaooo, on the other hand, do not behave as if the continu®d use represents either a personal inadequacy or a behavior that onght to be kept out of the public eye. For a long time,regular snokfng, including heavy smoking, was so 13 TIMN 0152380

commn that it was almost "noanal: ' Under such circumtstances, it is understandable that textbooks and diagnostic schenes casrerned with psychiatric, behavioral and medical disorders nmad,e no mentien of tobacoo dependence (see Jaffe 1977). Curiously, the eighth edi- tion of the Internatieeal Classification of Diseases (ICa (k8) listed caffeine de . not pendenee. In the forttcoming Im # 9, tobacco dependence will be inclir3ed. Sim- ilarly, 7he American Psychiatric Associatian's Dia stic and Sta- tistical Manual, second edition (DSR-+II) inclndes g ne sn g as a disorder but not tobacco use. In the forthooming revisfan, DSZK- III, a proposal has been mede to include, among the substance use . disorders (i.e., alcohelistn, opiate dependence, etc.), a category designated 'Robaeca Use Disorder." Since DS4-III is, at this time, still in draft form, the criteria for 7.bbaceo Use Disorder cannot be considered final. It is aorthwhile emphasizing, however, that just as the inclusion of alcoholisn (or alcohol dependence) as a disorder does not ieply that every use of alcohol is a disorder, the inclusion of the category Tobacco Use Disorder does not imply that every use of tobacco is an indication of a psychiatric abnor- mnlity. Indeed, in the curreat version, even the presence of phys- icall dependence on tobacco (as evidenced by withdrawal symptomatology, uupon cessation) is not a sufficient criterion for classifying the behavior as 2bbaceo Use Disorder, in an otherwise healthy individual. Meatien is mecie of this possible change. in D3d-III primarily to in- dicate that the climnte in which tobaceo use occurs is changing rapidly and this c3ia-ge may change the behavior itself (Jaffe 1977). A IDoK TD THE FITl[JRE If our analysis of similarities and differences in`relatiai to other addictive disorders is accurate, what changes should we expect in the near future and what kinds of questions merit our eoncern? First, there will be no total victory. Despite our concern about the adverse health consequences of tobaooo use in general, and about cigarette sedcitig in pnrticuIar, a campaign against tobacco use and tobacco users is not likely to lead to an isyoa~ditianal surrender - certainly not in the near future, and probably not within the fore- seeable future. If history teaches us anything about the nanmsdi.ca]l use of drugs, it is that, within the limits of the means which free societies will tolerate, tAe- naynstlical use of drugs that prodvice relaxation or pleasure can be redueed but not eliminated. This view does not imply that 3ncteased'societal concern and efforts to dis- courage tobacco use will be without effect. It does inQly that in all probability ten years fran oow there will still be millicns of people smoking tobacco every day. Second, we can predict that the characteristics of the populaticn that smoke will change. History tells us that when there was no clear cut socia3l disappzoval of opiate use (as in Britain and the United States in the mid-19th century) opivn and nprphine were in- cluded in patent medicines and could be purchased at the local gro- cery steares opiates were eonst.med openly and dependence on opiates was widespread. UKn social reform and legislatieo focussed on the dangers of opiate use, the social climate changed. "Respectablew 14 TIMN 0152381

elements within society disavowed both medical and nonmedical use. Eventually opiate use, even in a medical context, became tinged with a hint of immorality, and namedical use of opiates became more and more identified with sociopathic elements (See Musto 1973). 3cme trends suggest that the changing attitudes toward smaking are now beginning to produce changes in the population of srtokers. Not long ago cigarette smoking was a behavior that was typical of men in the upper socioeconomi.c grrnpst now, men from lower socio- eccrmni.c groups and wrxrnn from higher socioeconanic groups may be overrepresented (Task Force 1976). Zhese deamgraphic changes imply that scme of those who continue to smoke at present may have a greater need to do so, or more erotional liabilities. The heav- iest smokers and those with the most emotional difficulties are the least likely to successfully give up smoking (Dieher and Fraser 1968). The heavy smokers of toirorrow may be even more dif- ficult to help than those of today. Third, we can expect cigarette manufacturers to eontinue to caipete in lowering tar levels in cigarettes. over the past decade, both nicotine and tar levels have been reduced by all major cigarette eonpanies. The mean tar level of the average cigarette is substan- tially lower'than it was only five years earlier (See Wynder 1976). This trend is likely to centinue until all the cigarette manufac- turers have at least one entry to campete with the very low tar/nico- tine brands that deliver only 0.1 - 0.2 ng nicotine and 1 - 2 mg tar per cigarette. If the degree of dependenee is a function of the daily intake of nientine, this trend toward lower nicotine de- livery aay tend to offset the effects of the self-selection process induced by the changing social climate. The average smoker of to- morrow may have more emotianal need for nicotine but may be less pharmacologically dependent. Flowever, mx:h depends on the degree to which tolerance develops to the effects of nicotine and to which effects. Just as we have found no single "best approach" to alcoholisn or opiate dependence, we will find no single best approach to preven- tion and treatment of tobaeco dependence. Yet it is possible that studies looking for an optimall approach will be undertaken; it is equally likely that the populations using cigarettes and the pat- terns of use will have changed by the time the work is published. If the piedictien that lewer nicotine levels in cigarettes of the future will lead to lower levels of nicotine dependence is correct, the major problem will continne to be mativating stoak:ers to try to stop rather than finding new and dramatic ways of helping those who do seek help to sucoeed. Cn the other hand, efforts to under- stand the reasons for relapse will take on increasing importanee. Third party payment of costs of snoking tneatment pr+ograms will be eonsidered, discussed and argued - but our crystal ball is too murky to see the likely outoane of the axgiment. Since, at present, medical insurance is expected to cover the costs of treatment of the health pxoblems caused by stmki.ng, it would seem to be eoonanic 15 TIMN 0152382 -

ccnmon sense to cover costs of effective smoking cessation treat- ments. But two realities of the present health system in the U.S. make snch a change unlikely. First, with national attention riveted on spiralling costs of health care, adding the potential costs of treating fifty million smokers to the total health care bill will not be popular. Second, it is quite apparent that most smakers can stop withont formal help. 4dat criteria can be used to allocate limited treatment resources to those who really require them? vftt criteria might be used to define a qualified provider of treatment? And what procedures could be used to deter the developnent of a smoking treatment "ripbff" in which marginal practiti,oners sutmit bills for the treatment of hundreds of smokers - over and over and over again? We nust assume that unless there develops a totally unforeseen Puri- tanical view of tabacco use, we will continue to be concerned with the adverse health effects of tobacoo rather than the use of tobacco ~.rse. This being the case, the issue of less hazasdous smoking w31.1~ecome more important as time goes on. Gori (1976) has re- viewed the studies which showed that health hazards of smoking were dose-related over time. He has compnbed critical values for intake of tar, nicotine, and other cigarette ecmponents below which there did not appear to be a statistically significant increased risk of morbidity. More recently, Oori and his cowor)oers have ooaclnded that some cigarettes now an the market are "less hazardous" than others. The nuances of terminology are worthy of note. Currently, it is considered "bad form" in the mec3ical eamnity to speak of "safer" cigarettes since that would imply that existing cigarettes have saae degree of safety. "Iess hazardous" oi the other hand implies that all cigarettes are hazardous, but some mny be less eo. Despite his sensitivity to snch sementic nuances, Gori's views have generated considerable contraversy in the medical and smoking eessa- tien camnnities. Cigarette manufacturers are noir offering cigarettes with markedly reduced tar and nicotine. In general, there is a positive eorrela- tion between tar and nicotine content and amrnnt of carbon necxadde delivered when the cigarettes are smaked on a mediine (Philip Nbrris Co. personal ocimtmicatioz). Howver, then+e is eonsiderable vari- ability in OD delivery among brands with similar tar delivery. In any event, OD delivery of individual cigarette brands has not yet been made poblic information and may not be available for another year. Since cardiovascular disease may be me=e related to carbMT- h%Mlobin (CbH9b) levels than to nicotine and tar intake (see Arronow 1976), it is 3mportant to know what happens to the CoHgb of the smdcer as tar and nicotine aze reduced. Vogt (1977) reported that OD in expired air in his subjeets exfiibited a dose response relationship with the nunber of cigarettes/day his subjeets said they smoked. Vogt, working with men who volunteered far the Mul- tiple Risk Factor Internention Trial pr+ogrffin, fonnd no correlation between the acrount of tar and nicotine delivered by the cigarette and the expired CO level; bat he did not indicate the brands his subjects smdced or the pmportion of sabjects smoking the very low tar/nicotine brands. 3h our own work, we have also fouaul that 16 TIMN 0152383

carbon monoxide in expired air correlates substantially and posi- tively with the number of cigarettes smoaed per day (r = .52) in a group of 34 female s+m)Cers. The correlation with CO is increased (r = .62) if the total weight of the tobacco smlaed is used in the calculations (Jaffe and Kanzler,impublished data). Schachter (1978), reviewing his own work as well as studies of other workers on ciga- rette smoking, finds much support for the view that heavy smokers tend to "titrate" their body levels of nicotine. Such a view leads us to be concerned that those who attempt to reduce tar and nico- tine intake by changing brands may merely smoke more cigarettes or inhale those they smoke more deeply, thereby developing higher car- boxyhemoglabin levels as the price for reduced tar and nicotine intake. t F In a study in which we induced twelve femal.e smkers to switch to less hazardous brands, we did not find CO in expired breath to in- crease as the snnlaers switched to brands with lownr tar and nicotine. Hawever, in air group of twelve fema].e smokers, there was no appre- ciable increase in nunber of cigarettes smaked per day. It is pos- sible that the monitoring of Co may have served to inhjbit what might have been an increase in ni.arber of cigarettes smoked tmder other circvnstanees (Jaffe et al. 1978). There are obviously sane parallels between this approach to "less hazardous" cigarette smoking and the use of methadone in the treat- ment of heroin use or the "responsible drinking" approach to alco- hol problems. Both views assuoe that the drug use in question will continue and that the most pragmatic approach is to find Wnys to live with the drug that cause the least harm to the individual and society. But, before we can beoane advocates of "less hazardous" smoking we need to laiow more about the "tradeoffs." How much are the risks reduced, and are the benefits, if any, also reduced? To what extent does the body adjust to decreased nicotine intake so that lower levels produce effects similar to those once produced by higher levels? And of equa7l practical significance, what are the best ways to induce smohers to switch to less hazardous brands? The "less hazardous use°.approach to srnking-related problems can generate hostile reactions fran sane anti-smdcing groups, reactions that are analogous to those that temperanee and "total abstinence for alcoholics" advocates often exhibit toward the proponents of "responsible drinking" and social drinking for selected ex-alcdholics. There are also obvious parallels to the bitter schism that has di- vided those concerned about opiate abuse into pro- and anti-methadone maintenance camps. Many of the criticism of the "less hazardous" smoking approach reflect realistic ooncerns (just as do sane of the reactiens to mettmdom maintenance and responsible drinking). For esample, what levell of risk would be acceptable? And for which kinds of disability - cancer, enphysema, heart disease? Eqnally important is the coneern that even discussion of the possi- bility of a less hazardous cigarette will iandermine the effort to deter smoking among yotaig people who, as with their use of other cTrugs, generally deny the possibility of becaning dependent or of sustaining +mpai*+++e*+t of health. we might expect that the effort 17 Tj1'1N 015 2384

to develop inforaaticn on the nature of tobacco dependence and the risks of altered patterns of tobaxo use may have a nunber of valu- able byproducts for ttie understanding of hunan bi,ology and behav- ior. Certainly, there is still much to be learned about where and how tobacco produces its reinforcing effects. Can we be certain that the same sites and receptors involved in these effects are responsible for its adverse effects on health? Is there a possi- bility of developing agents anre selective than nicotine? Might such agents offer people who enjoy the effects of tobacco (or find that tobacco helps them regulate their internal state in sclae use- ful way) same of the sa¢ne effects without the risks? in the pres- ent climate, it is difficult to forget that the discovery of en- dorphins was a byproduct of research that was motivated by concerns about opiate dependence. With tobacoo, as with caffeine, dependence per se carries no social stigma. This eonttasts sharply with our present attituc7es tmaard dependence on opiates and alcohol, which elicit such grave concern that it is often difficult to persvade physicians to provide dying patients with enough opiates for relief of pain. It is possible that the increasing awareness of tobacco as a substance that induees dependence will not in itself alter or have a significant effect on the use patterns of the public at larqe; and that dependence will eontinue to be of inQortance only when people wish to change their behavior but find that change is exceedingly difficult or requires special help. We can expect the rediscovery of tobacco as a dependenee-inducing s::bstance to sharpen public recognition that in proportion to its iaQact on health we have grossly uedersupported research on tobacco dependence. Over the past decade we have spent hundreds of millions of dollars on the study of diseases known to be directly caused by the prolonged use of large aamunts of tobacyo. But the few scientists who felt that it was important to examine tobaooo dependence itself rather than the diseases it in- duces were viewed nnre with tolerance than with respect. 7his monograph is a sign that this attitude is changing and that the future will see more interest in the nature of tobacco dependence within the scientific eommmity. If additional funding follows the increasing interest, we would do weil to begin nav to set priorities and frame the imgortant<questions. REFERENCES Amnow, W.S. Carbon annoxide and cardiovascular disease. In: Wynder, E.L.; Hoffmazns, D.; and Gori, G.B., eds. Modi£~x the Risk for the Smker. Vol. I, DHEW Pcb.No.(NtH) 76-1221, 1976, pp. 321-328. Armitage, A.K.; Hail, G.H.;and Sellers, C.M. Effects of nicotine on electroeortica7l activity and acetylcholine release fran cat eere- bial cortex. Br J Pharmacol, 35:152-160, 1969. Bernstein, D.A., and Mc Alister, A. ztse nodification of smoking be- havior: Progress and prablems. Adc7ictive Behav, 1:89-102, 1976. 18 TIMN 0152385

gorland, B.R.,and Rudolph, J.P. Relative effects of low socio- ectinomic status, paz+ental smoking and poor scholastic perfoxmanoe on smoking among high school students. Soc Sci Med, 9:27-30, 1975. Cahalan, D.; Cisin, I.H.;and Crossley,•H.M. American Drinkin Prac- tioes: A National Study of Drinking Behaviors and Att t es. NeW Bnmsw , N.J.: Rutgers Center o co Studs.es, 1969. Carruthers, M. Modification of the noradrenaline related effects of smoking by Beta-blockade. Pspatal Med, 6:251-256, 1976. Corti, C. A History of Smoking. New York: Harcourt Brace, 1932. Cryer, P.E.; Hapnnml, M.W.; Santiago, J.V.; and Shah, S.D. Norepi- nephrine and epinephrine release and adrenergic mediation of snoking- associated heirodynamic and metabolic events. New England J Med, 295(11):573-577, 1976. Danino, E.F. eology of nicotine and tobaeco smok- ing. In: Diaui, W.L., Jr., ed. Sreoking Behavior: Motives and In- centives. Washingtan, D.C.: Winston and Sons, 1573, pp. T-=. Dreher, K.F., and Fraser, J.G. Siroking habits of aleoUolic out- patients. Int J Addictions, 3:65-80, 1968. Ejrup, B. The r+ole of nicotine in seiaking pleasure, niootinisin, and trreatment. In: Van Euler, V.S., ed. Toba.cco Alkaloids and Re- lated Ccmnotnds. (bcford: Pergam, 1965. pp. -3 . Eysenck, H.J. Pereanality and the mtintenanoe of the s:ro)Ciag habit. In: Dum, W.L., Jr., ed. gnokLxg Behavior: Motivies and Is:oentivies. Washington, D.C.: Winston and , 1973. pp. - . Gilbert, R.M. Caffeine as a drug of abuse. In: Gibbons, R.J.; Israel, Y.; Kalant, H.•. Popham, R.E.; SclmLtdt, W.; and Snart, R.G., eds. Research Advar:oes in Aloohol and Drug Pxrobleens. Vbl. 3. New Yorkt Wiley, 1976. pp. . Galdfazb, T.; Gritz, E.R.; Jarvik, M.E.; and Stolernnn, I.P. Re- _ actions to cigarettes as a fuwtian of nicotine and "tar." Clin Phannacol Ther, 19:767-772, 1976. Goodean, F.R. Effects of nicotine on distribution and release of 14C-Nozapinephri.ne aad 14C-DOpxnine in rat brain striatun and hypo- thalmffi slices. Neui~ophezma~cal, 13:1025-1032, 1974. Goodwin, D.; Schulsinger, F.; Heamsnsesi, L.; Guze, S.B.; and Winokur, G. Alcahol problems in adoptees raised apart fxcm aloo- holic biological parents. Arch Gen Psydiiat, 28:238-243, 1973. Goodwin, D.; Schulsinger, F.; Moller, N.; Hezri~asen, L.; Winokur, G.; and Guze, S.B. Drinking psoblems in adopted and nan adopted sons of alodholics. Arch Gen Psychiat, 31:164-169, 1974. 19 I TIMN 0152386

Gori, G.B. Low risk cigarettes: A prescription. Scieaae, 194: 1243-1246, 1976. Guilford, J.S. Factors Related to 3nccessful Abstinence from Srtek- i.n,~c. American Institute for Re , P , 196 . Guillenul, R.; Vargo, T.; Rossier, J.; Minick, S.; Ling, N.; Rivier, C.; Vale, W.; and Bloan, F. 6-endorphin and adrenooortioo- tropin are secreted conoomitantly by the pituitary gland. Science, 197:1367, 1977. Hatfiisg, B.A.; Kraeme.r, H.C.; and Jahnke, W. A hierarchy of drug use in adolesoence: Behavioral and attitudinal correlates of sub- stantial drug use. Psner J Psychiat, 132:1155-1163, 1975. Hart, G. Tax reform amendments of 1976-H.R.10612. Congressioea]. Record, 122(99):Washingtm, D.C., June 24, 1976. Hunt, W.A.; Barnett, L.W.; and Branch, L.G. Relapse rates in addia- tion programs. J Clin Psychol, 27:455-456, 1971. Husain, M.K.; Frantz, A.G.; Ciaroc:hi, F.; and Robinson, A.G. Nico- tine-stinulated release of neurophysin and vasopressin in htmens. J'C1in Endocrin Metab, 41:1113-1117, 1975. Ikard, F.F.; Green, D.E.; and Horn, D. A scale to differentiate between types of swking as related to the m®nageoent of affect. Int J Addictions, 4:649-659, 1969. Jaffe, J.H. Factors in the etiology of drug use and drug depend- enee - two nodelsa Opiate use and tobacco use. In: Schecter, A. e3. Rehabilitation and Treatment AsFects of Drug Dependence. Cleveland: cRC Press, 1977. ~pp. 23= . Jaffe, J.H. Tobacco use as a mental disorder: The rediscovery of a madiaal problem. In: Jaxvik, M.E.; Cullen, J.W.; Gritz, E.R.; Vogt, T.M.; and NBst, L.J., eds. Research on Behavior, National Institute an Drug Abuse Research Mor:ogra1.DHEW pub. No. (ADM)78-581. Washington, D.C.: Superintendent of Dociments, U.S. Govierrmest Printing Office, 1977. pp. 202-217. Jaffe, J.H., and Jarvik, M.E. Tnbaoco use and tobacco use disorder. In: Lipton, M.A., Di Mascio, A., and Ki].lam, K.F., eds. Psycho- pharmaoology: A Generation of Piogress. NeW York: Raven s, . pp• -16. Jaffe, J.H., and Kanzler, M. Unpublished data. Jaffe, J.H.; Kanzler, M.; Cohen, M.; and Kaplan, T. Inducing low tar/nicotine cigarette smoking in waren. Br J Addict, 73:271-281. 1978. Jarvik, M,E. Further observations on nicotine as the reinforcing age"It in s:roking. In: Dunn, W.L., Jr., ed. Smoking Behavior: Motiv~es and Incentives. Washington, D.C.: Winston~'J3, pp. 33-49. 20 TIMN 0152387

I Jessor, R., and Jessor, S.L. Problem Behavior and P ial Devel t: A Long itudinal S o You . Nea Yor ; Press, 1977. Kumar, R.; Cooke, E.C.; Lader, M.H.; Russell, M.A.H. Is nicotine ,imortant in tobaaco stmmki.ng? Clin Pharwwol Ther, 21:520-529, 1977. Larson, P.S., and Silvette, H. Zbbacco• E: erimental and Clinical Studies, Supplement 2. Baltimore ~Wi ceas,1971. Larson, P.S., and Silvette, H. Tebaxo: Euperitrental and Clinical Studies, Supplement 3. Baltimoze: W ams an WiMeRns, 1975. '-" Liebernm Research, Inc. The Teenager Looks at Cigarette Smatcing. Amarican Cancer Society, 1969. Martin, W.R.; Jasinski, D.R.; Haertzen, C.A.z Kay, D.C.; Jones, B.E.; Mansky, P.A.; and Carpenter, R.W. Methac7a;e - A reevaluation. Arch Gen Psychiat, 28:286-295, 1973. Mc Kennell, A.C. Smoking motivation factors. Br J Clin Psychol, 9:8-22, 1970. Musto, D.F. The American Disease. New Haven: Yale University Press, 1973., O'Brien, C.P. EScperimental analysis of conditioning fackors in hussn narcotic addiction. Phaanacel Rey, 27:533-543, 1976. PetD, J. Price and eoesunptian of cigarettes: A case for inter- ventiea? Br J Pre Soc Med, 28(4):241-245, 1974. Peto, J. Taxes, staoking, and health. (Latter) Lancet 1(7954): 301, 1976. Philip Morris Co. (Personal canamication) 1978. Pophain, R.E.; SdYnidt, W.; and De Lint, J. The prevetstica of alco- holism: Epidaniological studies of the•effects of ganerimant eon- trol measures. Br J Addict, 70:125-144, 1975. Rollesbon, H. M®dical aspects of tabaoco. Lancet, 1:961-965, 1926. Russell, M.A.H. Cigarette smdcing: Natural history of a dependenoe disorder. Br J Med Psychol, 44:1-16, 1971. Russell, M.A.H. Smoking and nicotine dependence. In: Gibbints, R.J. Israel, Y., Kalant, H., PophaRn, R.I., Schmidt, W., and SnaLt, R.G., eds. Researgh Advances in Aledhol and Drug Problems, Vol. 3. New York: jc-hn RI1ey~ 1976, pp. 1-. Russell, M.A.H.; Peto, J.; and Patel, U.A. The classification of smoking by factorial structure of motives. J R Statist Soc, 137(3): 313-333, 1974. 21 TIIVIN 0152388

Russell, M.A.H.; Wilson, C.; Patel, U.A.; Feyerabexi, C.; and Cole, P.V. Plasna nicoti.ne levels after smoking cigarettes with high, nmedi.ua and loer nicotine yields. Br Med J, 2:414-416, 1975. Schachter, S. Ptiarnacological and psychological determinants of smoking. Azm Iat Med, 88(1):104-114, 1978. Schechter, M.D., and Rand, M.J. Effect of acute deprivation of smoking on aggression and hostility. Psydooarniacologia, 35: 19-28, 1974. Schuckit, M., Goodvin, D., armd Winokur, G. A study of alcohplim in half siblings. Aaer J Psychiat, 128:1132-1136, 1972. Schwartz, J.L. smoking cures: Ways to kick an unhealthy habit. In: Jarvik, M.E.; Cullen, J.W.; Gritz, E.R.; Vogt, T.M.; and West, L.J. Research on Smdcin Behavior. Natienal Institute on Drug Abuse Re~Monograpn 17. Dom Pub. No. (ADM)78-581. Wash- ingtoa, D.C.: Superintendent of Doetments, U.S., COvexnnent Print- ing Office, 1977. pp. 308-336. Setnn, M. Price and aggregate demand for cigarettes (author's ab- stract). smoking and Health Ball. USDHEW Pub. No. (CDC) 78-8702: 175, 1978. Shiffinen, S.M., atxi Jarvik, M.E. Snaking withdrawel syaptams in two weeks of abstinence. PsychoQharmscoloqia, 50:35-39, 1976. Simon, W.E., and Primaviera, L.H. The personality of the cigarette smeker: Sane enpirical data. Int J Addictions 11:81-94, 1976. Smith, G.M. Relations between personality and staddng behavior in preac3u].t subjects. J c7onsult Clin Psychol, 33:710-715, 1969. Smith, G.M. Personality arad srtoking: A review of the empirical literature. In: Hunt, W.A., ed. Learning Mechanistns in Szroic3x~g. Chicago: Aldine Pub Co., 1970. pp.4. Task Fozce on Tdaacco and Cancer, Report to the Board of Directors, Amsrican Cancer Society, 1976. 17:onns, C.B. T!e relationship of andcing and habits of nesva:e tensicn. In: Dunn, W.L., Jr., ed. emoking 8ehavior: Motiyes and inosatives. c+Tashingtan, D.C.: Winatan arX1 S.ns 397'3.- pp. T5T-'ITO. Vogt, T.M. Smokii:g bek:avioral factors as predietors of risks. In: Jarvik, M.E.; Cullen, J.W.; Gritz, E.R.; Vogt, T.M.; and VABSt, L.J. Research on gmking Behavior. National Institute on Drug Abuse ReseaMonogrla.-17. DFMK Pub. No. (ALM)78-581. Washington, D.C.: Superinteix3ent of Dacianents, U.S. Goaea~t Printing Office, 1977. pp. 98-111. Winternitz, W.W., and Quillea, D. Acute ha2nrooal response to ciga- rette smoking. J Clin Phazmacol., 17:389-397, 1977. 22 TIN.IN 0152389

Wyrn3er, E.L.j Hoffinen, D.; Gori, G.B., eds. Modifying the Risk for the 9mker. Pz+oceedings of the 3rd mbrid Conference on 9roking and Health, Vol. 1. DHEW Pob. No. (NIH) 76-1121, 1976. 564 pp. Yanagita, T. Brief review on tbe use of self-administration tech- niques for predicting drug dependence potential. In: 1lnnpson, T., and Unna, K.R., eds. Predicting Depenc'lenoe Liability of S+'ie.,lant and Depress~aatD s. Baltina vers ty P Press, 3T6. -" pp• _ ~2. AUTHORS Jerome H. Jaffe, M.D. Professor of Psychiatry College of Physicians & Surgeoas of C7binnbia University 722 West 168th Street New York, New York 10032 Maureen Kanzler, Ph.D. Assistant Professor of Clinical Psychology College of Physicians & Surgeons of Ciolumbia University 722 West 168th Stxieet New York, New York 10032 23

Chapter 3 Psychological Analysis of Establishment and Maintenance of the Smoking Habit Daniel Horn, Ph.D. I began studying smoking habits about 30 years ago, primarily b,e- cause of an interest in the epidemiology of cancer, and continued partly to help me decide whether or not I should continue to smoke. Hventually I became persuaded that I would be better off if I .stopped smoking cigarettes, and I accomplished this quite easily by continuing to smoke only a pipe. Because of the publicity asso- ciated with the publication of our American Cancer Society studies on the effect of smoking on health (Hammond and Horn 1954, 1958), I found myself the target of innumerable personal accounts of how people had given up smoking. Not only would friends and associates do this, but strangers would approach me on the commuter trains into Manhattan with detailed accounts of how they had accomplished the feat. Partly because it seemed a waste to ignore this fund of information, I began to move into a more scientific concern about the smoking habit, its development, and its alteration. During these many years I have developed certain prejudices about statements on cigarette smoking. First, I deplore attempts by people who apply their personal standards to others' behavior to characterize it as "illogical" or "abnormal." A form_of behavior that was regularly engaged in at some time by approximately three- fourths of the males in the United States born around the time of World War I, as I was, hardly seems appropriately characterized as abnormal. It was not until the early 1950's that scientific infor- mation became available which persuaded a majority of us that we would be better off if we did not smoke. Although more than half of us did eventually quit smoking, many of us found that the behav- ior had become so integral a part of our way of functioning that it was difficult to quit. For some, quitting demanded such an invest- ment of effort that it seemed better to postpone trying to a later, more propitious time, but in the meantime to "hedge one's bet" by seeking a filtered cigarette. My prejudices also encompass those who feel that only the rare indi- vidual successfully quits smoking. Roughly a million adults a year have successfully quit smoking in the past few years. During the 24 TIMN 0152391

,r „ late 1960's, with the support of a bombardment of antismoking announcements on television, the number of persons quitting reached as high as 3 or 4 million a year (Horn 1970, 1979). Since the long term success rate during that period was between one-fourth and one- third of those who were trying to quit, we have had as many as 10 million persons trying to quit in a single year. Also, I deplore those who characterize quitting smoking as a tor- tured, almost impossible process. For many people, it is easy; for most it is somewhere between easy and difficult; and only for a minority is it really difficult. Finally, the demands of the smoking public have pushed the industry into developing and promoting cigarettes that produce lower and lower emissions of harmful ingredients. These have now been demon- strated to be of lesser degrees of harm than those being used for- merly, though they are far from being demonstrated as innocuous. Average tar and nicotine levels, which have been dropping since the early 1950's, have dropped another 10 percent in the past 2 years alone and are now down to 16.6 mg of tar and 1.09 mg of nicotine (Horn, unpublished data), compared to the 42 mg of tar and 3 mg of nicotine in the cigarettes before the 1950's. In addition to those who quit smoking and those who change to cigar- ettes delivering less tar and nicotine, increasing numbers of indi- viduals do not become smokers. Until about 2S years ago, cigarette smoking was a habit that was growing rapidly in each successive cohort of males born in the United States. It began to take hold even more rapidly among successive generations of young women. This phenomenon has slowed and turned around during the past 25 years. The taking up of smoking by young men is now substantially lower than it used to be, and the growth of the habit among young women has probably now peaked, at a level appreciably below the former peak among men (Horn et al. 1959; U.S.DHE4V 1972, 1976). MKII3G AS A LEARNING PROCESS I have explored the initiation and establishment of thesmoking habit as a learning process partly, no doubt, because of the assumption • that what can be learned can be unlearned. This has been basic to . the descriptive model of personal choice health behavior about which I have written, sumnarizing much-of our resear on smoking (Horn 1976). We have investigated the factors that characterize this class of behavior with the objective of finding out how to maximize its benefits while minimizing its harmful effects. Initiation of the Behavior 1 The initiation of personal choice health behavior is usually explora- tory in nature. 'iypically, it takes place in rather young people, sometimes in young children. It is largely dependent on: first, the availability of opportunity to engage in the behavior; second, having a fairly high degree of curiosity about the effects of the behavior; and third, in finding it a way of expressing either con- formity to the behavior of others (such as parents, older siblings 25

I or age-equals) or rebellion against what are seen as unreasonable proscriptions against the behavior. Gharacteristically, the greater the availability or the opportunity for expressing the behavior, the less is the necessity for a strong commitment either to conformity with the behavior of others or to rebellion against proscriptions by others (Horn et al. 1959). We know that smoking, for example, is much more common in children of parents who are themselves regular smokers, and this is partly be- cause of the ready availability of cigarettes to children when there is already a smoker in the household and partly because use of cigarettes by older members of the family sets an example of accept- able behavior and stimulates curiosity about what makes the cigar- rette so attractive (Horn et al. 1959). When smoking first begins to be widespread in a culture, it tends to be taken up with increasing frequency by successive waves of young people. Nevertheless, substantial numbers of older people may turn to it, especially if it can serve as a convenient substi- tute for previously well-established behavior, as was the case with cigarettes replacing cigar and pipe smoking in many male populations between 1910 and 1950. We are seeing a similar phenomenon now in Third World countries where the promotion of cigarette smoking is increasing rapidly (Eckholm 1978). Establishment of the Behavior The establishment of personal choice health behavior can be influ- enced by at least three groups of factors. In the case of cigarette smoking developing as a continuing habit in adolescents, these are, first, the costs/benefits evaluation of the behavior; second, common stereotypes that characterize perception of behavior; and third, psychological factors characterizing both personal structure and personality integration factors, particularly as they reflect the relationship of the individual and his needs to society and its demands. The costs that go into the evaluation of the costs/benefit balance may include the harmful effects on health (bot=p ys~.caI~an menta ), economic cost, and the harmful effects on society, such as economic or as a form of pollution. The benefits may include positive effects on health (both physical and mental), economic advantages, social utility (especially in terms of the facilitation of personal inter- actions), psychological utility (both in terms of the increase of positive effect or the reduction of negative effect) and benefits to society. There may also be quite separate evaluations of a costs/benefits analysis for the individual and a costs/benefits analysis for society, since for some individuals, one of these may be more per- suasive than the other in producing attempts to change behavior. For political leaders, the costs/benefits for society usually take precedence over the costs/benefits analysis for the individual. 26 TIMN 0152393

When the behavior which appears to be most logical on the basis of a costs/benefits balance and the actual behavior of the individual are quite different, another set of factors consisting of rational- izations or some other set of beliefs may come into play to reduce the dissonance between the "logical" behavior and the actual be- havior. In the case of cigarette smoking, the costs for the past twenty or so years have largely reflected the increasing evidence of the harmful effects of smoking on health, but in addition to that, such concerns as esthetic values, the contribution of smoking to various forms of pollution, and economic values related to the financial cost either to the individual or to society, have come into play. The benefits are wide, ranging from the facilitation of social interaction, which is perceived or appears as one of the most valu- able benefits of taking up smoking, to the reduction of tension and to the enhancement of states of pleasure which tend to be a later development in the appreciation of benefits. Such common remarks as "I can always give up smoking before it hurts me," or "I don't really smoke enough for it to do any damage," or "The kind of cigar- ette I smoke (or the way I smoke it) is not very likely to hurt anyone" are characteristic of beliefs of the individual who per- ceives potential costs as higher than benefits and yet who continues to smoke. Per tual ster~e~o~ty ~estend to develop as a kind of mythology about at smoking is like, what smokers are like, and why people smoke. These tend to be superficial and frequently inaccurate systems of beliefs and are likely to be derived either from the exaggerations of advertising on the one hand or the exaggerations of counter- advertising by antismoking groups on the other. In general, the greater the role played by superficial and inaccurate beliefs about the behavior, whether positive or negative, the more difficult it is to develop a sound decision-making process on the part of the individual or society as a whole. A variety of patterns of s cholo ical forces may enter into the determination of personal choice ea t TFie -vior, in particular, those that reflect the conflict engendered in individuals by the demands of society and his own inner demands. In the case of smoking behavior and health, we have identified two such factors (Milne and Colman 1973) one depends on the strength of the conflict perceived by the individual between the satisfaction of his own needs and the demands imposed on him by society or by its authority figures. The second factor is a reflection of the urgency to the individual of maintaining control over his own behavior and over his own destiny as opposed to being subject either to the control of others or to the vagaries of chance as represented by "good luck" or by "bad luck." Management of Affect Although the factors that contribute to the establishment of the habit probably continue for a time to be important in the mainten- 27 4IMN 0152394

ance of smoking behavior, we have found the model developed by Tomkins (1966) and extended by us (Horn and Waingrow 1966; Ikard, Green, and Horn 1969) to describe the psychological utility of smoking in terms of its affect management capabilities describes well the use of smoking even in as young a population as college students. The use of cigarettes to help in the management of affect, by augmenting positive affective states and reducing negative affect- ive states, is probably developed quickly; and as the number of cigar- ettes used per day increases usually up to the ages of 35 or 40, the individual may shift from using cigarettes for one purpose to using them for multiple purposes. This approach to characterizing smoking behavior has been of some use in developing educational programs for altering smoking behavior, as a logical application of this knowledge carries with it implications for interfering with the establishment and maintenance of the behav- ior. Our critical need at this point is to identify the common pat- terns that predict establishment of the smoking habit and to find ways of interfering with that process. FOOTNOTE 1. With minor changes, the sections of this paper on "Initiation of the Behavior" and "Establishment of the Behavior" were included in an article by Dr. Horn entitled "A model for the study of personal choice health behaviour," in The International Journal of Fiealth Fducation 19(2):89-98, 1976. Dr. Horn was t en Drirector o t e Nat~Clearinghouse for Smoking and Health, Bureau of Health Fducation, Center for Disease Control, Department of Health, Educa- tion, and Welfare. During 1975 and early 1976 he served as a Special Consultant to the World Health Organization. REFERENCES Eckholm, E. ~Cutti~n Tobacco's Toll. Worldwatch Paper 18. Washing- ton, D.C.: oW rhdwat Institute, 1978. 40 pp. Hammond, B.C., and Horn, D. The relationship between human smoking habits and death rates. JAMA, 155(15):1316-1328, 1954. Hamnond, B.C., and Horn, D. Smoking and death rates--report on forty-four months of follow-up of 187,783 men. JAMA, 166(10):1159- 1172, 1958. Horn, D. What's happening to smoking behavior? In: National Con- oio ference on Smokinsz and Health: A Surranary of Proceedina's.'_M£~i nteragency Counci on mo ing an a t, 1970. pp. 15-20. Horn, D. A model for the study of personal choice health behaviour. International Journal of Health Fducation, 19(2):89-98, 1976. 28 ,T'IMN 0152395

I Horn, D. How much real progress have we made in the fight against smoking? American hmg Association Bulletin, 65(1):6-9, 1979. Horn, D.; Courts, F.A.; Taylor, R.M.; and Solomon, E.S. Cigarette smoking among high school students. American Journal of Pubiic Health, 49:1497, 1959. Horn, D., and Waingrow, S. Some dimensions of a model for smoking behavioi~t change. American Journal of Public Health, 56:21, 1966. Ikard, F.F.; Green, D.E.; and Horn, D. A scale to differentiate between types of smoking as related to the management of affect. International Journal of Addictions, 4:646, 1969. Milne, A.M., and Colman, J.G. Development of a teenager's self- testing kit (cigarette smoking). Final Report. Education and Public Affairs, Washington, D. C., 1973. Tomkins, S.S. Psychological model for smoking behavior. American Journal of Public Health, 56:12 (Part II), 17-20, 1966. U. S. It-IE<N. National Clearinghouse for Smoking and Health. Teena e Smokin --National• Patterns: 1968 1970. U.S. Department of a t, ucation, an e are Pu ication .(HSM)72-7508, 1972. U.S.IHEW. National Clearinghouse for Smoking and Health. ~Teenage_ 9nokin --National Patterns: 1972 1974. U.S. Department of ea t , Education, an We are Pub i.ca.tion No. (NIH)76-931976. AUTHOR Daniel Horn, Ph. D. R. D. 1, Box 182 Frenchtown, New Jersey 08825 29 ; TIMN 0152396

Chapter 4 Cigarette Smoking as a Precursor of Illicit Drug Use John A. O'Donnell, Ph.D. This paper will focus on one question: can the variable of ciga- rette use be seen as a predictor, and possibly a cause, of other forms of drug use? Alcohol use and marihuana use will be•employed ' as competitive predictors and causes. Many surveys establish that there is a statistical association between cigarette use, or tobacco use, and use of other drugs. An association has been found for junior high and senior high school students of both sexes (Block 1975; Blum et al. 1970; Josephson 1974; Roth 1972; Shapiro 1975; Spevack and Pihl 1976; Whitehead, Smart and Laforest 1972; Wolfson et al. 1972). It has been reported at the college level (Goode 1972; Groves 1974; Johnson 1973; Steffenhagen, McAree and Nixon 1972), for U.S. military personnel in Germany (Prendergast, Preble and Tennant 1973), and for medical students (Lipp et al. 1972). only one study (Svhitehead 1974) reports no association between tobacco and most other drugs, and this was based on a methodological error; when the correct analysis is done the usual associations emerge. None of these studies paid adequate attention to the time order in which drugs were used. Even investigators who conceptualized ciga- rette use as`the predictor of other drug use,Iike Goode (1972) and Johnson (1973), made the error of measuring current cigarette use, and use of heroin and other drugs at any time. One study has quasi-longitudinal data. Johnston (1973) asked about use during the high school years, and separately about use in the year after high school graduation. Those who started use of any drug in that year showed a higher percentage of cigarette users in the high school years than did the continuing nonusers. This is highly suggestive, but not conclusive, because the new users of each drug were outnumbered by those who had used it in high school. Conceivably, though not probably, that early use had preceded cigarette use, which would reverse the relationship for the entire group of users. 30 ; TIMN 0152397

1 In a true longitudinal study, Smith and Fogg (1978) measured cigarette use in 1969, when their 651 Boston students were in the 7th or 8th grade. This use predicted their classification five years later into one of three categories--nonusers, early users, and late users (of marihuana and hashish). Measures of at- titudes towards smoking and smokers were also predictive. Finally, Kandel and her associates (Kandel 1975; Kandel, Kessler and Margulies 1978) surveyed 5500 students representative of New York State high school students. Questionnaires were completed in the fall of 1971, the spring of 1972, and by almost 1000 of the seniors again in late 1972. Much of the time Kandel's variable TABLE 1 Time Order of Use of Drugs in the NationaZ SampZe (Percentages). Number Who Have Used Percenta e in which Each Pair Drug t er Time of Drugs (n = 2510) in CAPS First Drug First Order Unknown ALCOHOL and Cigarettes 1722 59 24 17 Marihuana 1377 93 5 2 Psychedelics 548 98 2 ~ Stimulantsl 578 98 2 * Sedativesl 407 97 2 ~ Heroin 147 97 3 0 Opiatesl 491 94 6 * Cocaine 351 99 1 * CIGARETTES and Marihuana 1077 73 14 13 Psychedelics 451 82 10 9 Stimulantsl 488 81 9 9 Sedativesl 339 84 10 6 Heroin 134 89 _. 6 5 Opiatesl 409 82 11 7 Cocaine 291 90 5 5 MARIHUANA and Psychedelics 546 80 14 6 Stimulantsl 562 73 21 7 Sedativesl 394 84 12 5 Heroin 146 90 7 3 Opiatesl 449 77 20 3 Cocaine 350 96 1 3 1Quasi-medical use excluded h Less than half of one percent 31 TIMN 0152398

is "legal drugs," meaning cigarettes or hard liquor or beer or wine, and sometimes it is "cigarettes or hard liquor," so that it is not possible to separate out the independent effects of ciga- rettes. Some data make it clear that cigarette use alone, at time 1, predicted use of marihuana and other drug use at time 2, beyond the effects of hard liquor and beer or wine. But the large majority were using cigarettes and one or more kinds of alcohol at time 1, and there seem to be no data on the time order of use existing at time 1. Kandel did regard her data as sufficient to establish that drug use involves four stages, and that each is almost a necessary condition for the next. The first involves beer and/or wine, the second hard liquor and cigarettes. Marihuana use is the third stage, and other illicit drugs the fourth. Let us now examine the question with data collected in late 1974 and early 1975. The primary sample was a nationwide probability sample of 2510 men, then 20 to 30 years old inclusive, selected from Selective Service records. It may be regarded as repre- sentative of young men in the United States, except for Alaska and Hawaii (O'Donnell et al. 1976). A secondary sample of 294, of the same ages, was drawn from high drug use areas in New York City. It should not be regarded as representative of any popu- lation, but it is a sample of the general population, not a sample from a prison or treatment agency. 'Table 1 presents the temporal relationships of alcohol, cigarette and marihuana use with each other and with other drugs. It should be noted that no drug was invariably the first. Still, the preponderance is clear, and any of the three could be an efficient predictor of the use of other drugs. The first row of table 1 shows that alcohol normally preceded cigarettes in this sample; if the indeterminate cases are distrib- uted like the known cases, alcohol was first in 71 percent of the cases where both were used. Cigarettes, in turn, normally pre- ceded marihuana. It should be added that the interview inquired about-regular cigarette`u'se. The word "regular"-ghould always be understood, though I will often omit it. Earlier studies (O'Donnell and Clayton 1978; O'Donnell and Clayton, forthcoming) have suggested the hypothesis of a causal chain, begin- ning with alcohol use, leading to marihuanaa use, leading in turn to the use of other drugs. Cocaine is used as an example of "other drugs" in table 2. Lifetime use of marihuana is shown as an ordinal variable across the top, ranging from no use, through increasingly heavier degrees of use; the total row at the bottom shows that the percentage of those who used cocaine increases as marihuana use in- creases. This zero-order relationship, seen in the bottom row, is repeated in all other rows essentially unchanged, with extent of alcohol use controlled. 32 TIMN 0152399

;he zero-order relationship of cocaine use with alcohol use may be seen in the last column; again, cocaine use increases linearly with alcohol use, though not as strongly as it did with marihuana use. But this zero-order relationship disappears in each of the other columns, when marihuana use is controlled. The findings are pre- cisely what one would expect if marihuana use comes after alcohol use but before cocaine use in the causal chain. The marihuana- cocaine relationship is not reduced by control on a prior variable, alcohol, while the alcohol-cocaine relationship disappears when con- trolled on an intervening variable, marihuana use. Despite this evidence for alcohol's causal status, we have re- garded marihuana use as the crucial variable. Part of the reason is an historical accident; most of the debate in the literature has been in terrtfs of the marihuana-heroin relationship. But part relates to practical considerations. Almost all men used alcohol, and the users tended to bunch at the high use end of the scale, as can be seen from the column of n's in table 2. The marihuana variable offered a better•distribution for prediction purposes. Finally, while t}oth alcohol and marihuana were associated with other drugs, the zero-order relationship was stronger for marihuana. •Table 2 shows that the percentage of cocaine use ranged from under 1 to 58 percent as marihuana use increased from none to heavy, while the corresponding range for alcohol was only from zero to 22. . This, of course, is to be expected if almost all of the effect of alcohol on cocaine use is indirect, through marihuana, with little or no direct effect. One practical use for predictors would be in prevention of drug use. The analysis thus far suggests that one way to prevent or reduce cocaine use would be to reduce the marihuana and alcohol use that predict it. But to fall in the "heavy" categories of alcohol and marihuana use might take years,of use, and the use of cocaine (or other drug) could have started before the prediction is firm. For prevention purposes, what is needed are predictors which can be easily measured at an early point in time. Age at first use of marihuana seems to be such a predictor. Table 3 shows, for both samples, that the earlier marihuana use began, the more likely was the use of heroin and of other opiates. Similar tables, not included here, show that it is an equally powerful predictor of the use of other drugs, of illicit drug sales, and of criminal behavior. To sum up the above and much other work, marihuana use normally, though not universally, precedes other kinds of illicit drug use, and has strong statistical associations with them. We have tested for spuriousness on the variables we have available, have seen no indications that these associations are spurious, and we believe that intervening variables between marihuana use and, for example, heroin use have been identified. These statements are sufficient •for the assertion of a causal connection, by one conception of cau- sality widely used in survey research (O'Donnell and Clayton 1978). 33 TIMN 0152400

TABLE 2 Use of Cocaine, by Extent of Lifetime Use of AZcohoZ and Marihuana (Percentages) Extent of Extent of Lifetime Use of Marihuana Lifetime Use of Alcohol na (2510) None (1128) Experimenta (423) ig t (231) Ma rate (227) Heavy (501) Total None (76) 0 b 0 0 _ - 0 Experimental (93) 0 0 0 0 (1 of 2) 1 Light (491) 0 0 4 37 53 7 Moderate (318) 0 2 0 24 40 8 Heavy (599) *c 1 2 16 54 14 Heaviest (933) 0 1 6 20 63 22 Total (2510) * 1 4 22 58 aParentheses indicate numbers~of individuals; figures without parentheses are percentages. bA dash indicates there were no cases in the cell. cAn asterisk means less than one-half of one percent.

TABLE 3 Use of Heroin and Use of Opiates, IncZuding Heroin, by Age at First Use of Marihuana, in TWo ScmrpZes A. National Sample Age at First Use of Marihuana Percent Used Heroin Percent Used Opiates 16 or less (288) 25 68 17-18 (329) 11 45 19-20 (346) 8 42 21 or more (419) 2 30 Never used (1128) * 16 Total (2510) B. 6 New York City Sample 32 14 or less (41) 73 80 15-16 (61) 36 62 17-18 (53) 30 51 19 or more (65) 9 35 Never used (74) 1 7 Total (294) 26 43 * One user among.,1128 men. The statistical associations and time order hold for alcohol too. We have not directly examined the possibility of spurious relation- ships for alcohol, but it seems most unlikely that they could be completely spurious, because of the apparent causal path through marihuana. The question now becomes: do cigarettes too have some causal impact? The question can be clarified by considering the hypothetical causal model in figure 1, where the four variables are arranged from left to right in the time order established by table 1. The problem would be to assign values to the path coefficients of the 35

! X i FIGURE 1 5 Marihuana 6 ~ Illicit T ~ Drugs z Y HypotheticaZ CausaZ ModeZ arrows shown, especially the values of arrows 4 and 5; does ciga- rette use affect marihuana use, and does it affect illicit drug use directly, in addition to whatever effect it has through marihuana use? Problems arise because of the unnumbered arrows, which represent all other ir.fluences which impinge on the endogenous variables. There must be many variables which at one point influence a person's alcohol behavior, later his starting to use marihuana, and so on-- factors such as unconventionality, rebelliousness, and curiousity. If these had been measured and could be added to the model, the model would be complex, but the solution of its linear structural equations would be straightforward. These measures are not available, however, so any attempt to use the model as it stands would confound the effects of these variables with those of-the variables included " in the model, producing incorrect values for the coefficients. A second problem arises with the arrow from alcohol to cigarettes. Are we justified in determining time order, and using a single- headed arrow, when alcohol is known to be first only 59 percent of the time? It would appear safer to take cigarette use as an exogenous variable too, and replace the arrow between it and alcohol with a symbol indicating that we are taking the correlation between the two as given and unexplained. But this in turn would mean that we can never isolate the effect of cigarettes; part of its apparent effect would be due to its correlation with alcohol use. The converse is also true; the effect of alcohol use would be partly-confounded with that of cigarette use. For both practical and theoretical reasons, therefore, we cannot expect an exact measurement of their relative influences. 36 ~TIMN 0152403

{Ve can get some hints, however, from simpler approaches. Two smoking variables are available. One, which might be labeled "Heaviest Smoking", refers to the number of cigarettes smoked daily during the heaviest period of smoking. It is related to marihuana use in table 4. There is a low degree of association--a gamma of .218. Table 5 uses the second smoking variable, age at first regular use of cigarettes. The expected negative association with marihuana is found, but it is weak--a gamma of only -.107 when based on the age differences among cigarette users, increasing to -.248 when the nonusers also contribute to the association. By both measures, then, it appears cigarette use may be seen as a predictor, and possibly a cause, of marihuana use, but hardly as an efficient predictor or an important cause. To examine its relationship with other drug use, an index of non- medical drug use was constructed. This index summarizes in one number each man's lifetime use of sedatives, stimulants, psyche- delics, heroin, other opiates and cocaine. The index weights each category of use, for each drug class, by the frequency of that category in the sample, so that the drugs used by fewer men, and the higher frequencies of use, receive higher scores. The range of scores in the'sample is from 420 to 992, but 70 percent of the sample score below 500, because few men used drugs beyond marihuana. The relationship of cigarettes to this index was examined in two ways. First, the index was divided into five categories and cross- tabulated against the "Heaviest Smoking" variable, within cate- gories of lifetime extent of alcohol and marihuana use. Using Kendall's tau c and the .05 level, the pattern of findings was clear. Smoking continued to be significantly, though not strongly, associated with the index when controlled on alcohol use, suggesting that alcohol is a prior variable. When marihuana use was controlled, the association between smoking and the index was not significant for one category, the men who did not use marihuana, but was significant in each of the other four categories where the extent of marihuana use varied. This is precisely the opposite of what happened when the alcohol association with the index was controlled on marihuana use. There, the association remained significant among the non marihuana users, but disappeared in the four categories of users. The implications may be summed up in a few statements. 1. Use of the six drug classes with which we are concerned here was rare and minimal among men who did not use marihuana. 2. The effect of alcohol on nonmedical drug use operates almost entirely through its effect on marihuana use, with little or no direct effect. 37 I TIMN 0152404

TABLE 4 Lifetime Extent of Marihuana Use, by Number of Cigarettes Used DaiZy During Period of Heaviest Smoking (Percentages) Number of Ci arettes Used Dail Lifetime no. reg. Less - - 4 1- 25-34 35-44 45+ Marihuana Use (n)1 cig.' use (766) than 1 (25) (116) (229) (491) (308) (365) (208) Never 60 48 54 41 38 39 34 33 w 00 Less than 10 times ' 16 16 16 14 17 17 20 18 10-99 times 10 16 13 14 14 12 21 20 100-999 times 9 8 14 16 17 20 12 12 1000 or more times 5 12 3 15 14 13 12 17 100 100 100 100 100 101 99 100 gamma = .218 12 cases missing; unknown on nunber of cigarettes used daily H

TABLE 5 Lifetime Extent of Marihuana Use, by Age at First Regular Use of Cigarettes (Percentages) Lifetime 14 or 15-16 17-18 19+ Tobacco but no Never Used Marihuana earlier reg. cig. Tobacco Use (n)1 (518) (452) (425) (348) (467) (299) Never (1128) 35 36 42 42 52 73 ~ Less than 10 times (423) 17 17 19 17 20 9 10-99 times (356) 15 16 16 17 13 6 100-999 times (338) 15 18 14 14 10 8 1000 or more times (264) 19 13 10 9 5 4 101 100 101 99 100 100 gamma = -.107 (on 4 age columns) = -.248 (on all 6 columns) 10ne case missing; age at first use of cigarettes unknown J

3. The effect of cigarette use on nonmedical drug use is direct, as well as indirect through marihuana use. But the effect is small. While statistically significant, the value of tau-c is usually below .10 with the control on marihuana, and no higher than .20 without that control. The second approach uses the second of the two smoking variables employed earlier, age at first regular use of cigarettes. Corres- ponding to this are age at first use of alcohol and age at first use of marihuana. These were not grouped into categories, but the full range of ages was used, as was the full range of index scores. The correlations and some partial correlations are presented in table 6, using computer labels like "CigAge" for "Age at first regular use of cigarettes," to save space. TABLE 6 CorreZations and PartiaZ CorreZations Among Age VariabZes and Drug Use Index A. Correlation Matrix (Pearson rl) CigAge MarAge Drug Use AlcAge .324 .314 -.281 (1718) (1375) (2427) CigAge .242 -.112 (1074) (1741) MarAge -.419 (1379) 1A11 correlations are significant at the .001 level. B. Partial Correlations Variables CigAge-Drug Use Control on AlcAge Partial r -.023 (n) (1717) Signif. NS CigAge-Drug Use MarAge -.012 (1073) NS CigAge-Drug Use AlcAge & MarAge .037 (1072) NS AlcAge-Drug Use CigAge -.261 (1717) .001 AlcAge-Drug Use MarAge -.174 (1374) .001 AlcAge-Drug Use CigAge $ MarAge -.177 (1072) .001 MarAge-Drug Use CigAge -.407 (1073) .001 MarAge-Drug Use AlcAge -.363 (1374) .001 MarAge-Drug Use CigAge & AlcAge -.364 (1072) .001 40 TIMN01524__ 07

As might be expected, the three age variables are positively cor- related--men who began to use alcohol or cigarettes or marihuana early tended to use the others at early ages too. Also as expected, each of the age variables is negatively correlated with the drug use index--the earlier a man started to use one of these three, the more likely he was to become involved in nonmedical drug use. The marihuana correlation with drug use is -.419. This leaves most of the variance unexplained, but .4 is a quite respectable correlation in social science research, and would be enough to suggest what we already know from table 3, that age at first use of marihuana may be a useful predictor of other drug use. Further, as may be seen in the lowest panel, the correlation is hardly reduced at all by controls on the other age variables. The correlation of Alcohol Age with Drug use is lower, at -.281, and is appreciably reduced by a control on Marihuana Age, though still significant. It might have some independent and additive predictive power. Finally, and of most relevance here, the correlation between Cigarette Age and Drug Use is only -.112. It is statistically significant, largely because it is based on 1700 cases, but its substantive importance could not be great. And even that little importance completely disappears when the relationship is con- trolled on either Alcohol Age or Marihuana Age. DISCUSSION It is the writer's impression, though not yet confirmed by any data found in the•older literature, that up to a decade or so ago it was rare for anyone without prior smoking experience to experi- ment with marihuana. That inhibitory factor would have produced an association between smoking and drug use, easily seen as a causal connection, and it would have been plausible to predict that a decrease in smoking would lead to a decrease in drug use. But that inhibition, if indeed it did exist, seems to be dis- appearing. In late 1974, when the sample of young men was inter- viewed, of 299 men who had never used tobacco in any form, 27 percent had used marihuana. Of 467 who had used tobacco, but not cigarettes, 48 percent had used marihuana, while of the 1743 who had used cigarettes regularly, 62 percent used it. The differences suggest that some inhibition may still persist, but 27 and 48 are not negligible percentages. By-a few years ago there were confirmed marihuana users, and users of other drugs, who quite seriously stated they would never use tobacco because tobacco has health hazards. The association of cigarette use with drug use therefore may be partially a causal connection. It appears likely that much of it may be spurious, due to the fact that both are connected with alcohol use, and almost certainly with personality and social 41 TIMN 0152408 ~

variables that have been studied with respect to single substances, but not with respect to several simultaneously. Whatever the nature of the association, it is small, and would not suggest that cigarette use would be a useful predictor of later drug use. REFERENCES Block, J.R. Behavioral and demographic correlates of drug use among students in grades 7-12. In: Lettieri, D.J., ed. Pre- dicting Adolescent I Abuse. Rockville, Md.: National Institute on Drug us5. pp. 26-276. Blum, R.H.,and Associates. Students and Drugs II: ~C~olleg-e and High School Observations. San Francisco: Jossey-BassO 3§-9--pp . Goode, E. Cigarette smoking and drug use on a college campus. Intl J Addictions, 7(l):133-140, 1972. Groves, W.E. Patterns of college student drug use and lifestyles. In: Josephson, E. and Carroll, E.E., eds. Dru Use: Ep~id~_emio- _e~y,,ff74. ~lo ~ica~l and Sociological Approaches. New Yo Jofin Wi_ pp . Johnson, B.D. Marihuana Users and Drua Subcultures. New York: John Wiley, 1973. 290 pp. r. Johnston, L. Drugs and American Youth. Ann Arbor, Mich.: Institute for Social Research, 1973. 273 pp. Josephson, E. Trends in adolescent marihuana use. In: Josephson, E. and Carroll, E.E., eds. ~Dru Use: E idemiolo ical and Sociological Approaches. Newlorc:I- Jo Wiley, 1 4. pp. 177- 0~5-- Kandel, D. Stages in adolescent involvement in drug use. Science 190(4217):912-914, 1975. ' Kandel, D.; Kessler, R.C.; and Margulies, R.Z. Antecedents of adolescent initiation into stages of drug use: a developmental analysis. J Youth and Adolescence 7(1):13-40, 1978. Lipp, M.; Tinklenberg, J.; Benson, S.; Melges, F.; Taintor, Z.; and Peterson, M. Medical student use of marijuana, alcohol and cigarettes: a study of four schools. Intl J Addictions 7(1): 141-152, 1972. O'Donnell, J.A., and Clayton, R.R. The stepping-stone hypothesis-- marihuana, heroin and causality. Paper presented at Meeting of the American Sociological Association, San Francisco, September, 1978. To be published in Addictive Diseases. 42 TIMN 0152409

0'Donnell, J.A., and Clayton, R.R. Determinants of early marihuana use. In Friedman, A.S., Beschner, G, and Pittell, S. Youth Drru~ Abuse: Problems, Issues, and Treatment. Lexington, Mass.: ~.-exington 0o s, forthcoming. 0'Donnell, J.A.; Voss, H.L.; Clayton, R.R.; Slatin, G.T. and Room, R.G.W. Young Men and ~Dru s: A Natiwide~Su~. National Institute on D~rug Abuse ~ear-ch_~lonograpn ~. ,,nnr~ ruo. No. (ADM)76-311. Washington, D.C.: Superintendent of Documents, U.S. Government Printing Office, 1976. ` Prendergast, T.J.; Preble, M.R.; and Tennant, F.S. Drug use and its relation to alcohol and cigarette consumption in the military community of West Germany. Intl J Addictions 8(5):741-754, 1973. Roth, R. Student drug abuse in southeastern Michigan and profiles of the abusers. In: Einstein, S., and Allen, S., eds. Student Drug Suryeys. Farmingdale, N.Y.: Baywood, 1972. pp. 5S-66~- Shapiro, R.D. Alcohol, tobacco and illicit drug use among adoles- cents. Intl J Addictions 10(3):387-390, 1975. Smith, G.M. and Fogg, C.P. Psychological predictors of early use, late use and nonuse of marihuana among teenage students. In: Kandel, D.B. ~Lo_n it~udina__l Research on Drug Use. New York: Wiley, - 1978. pp. 101_1I6. Spevack, M. and Pihl, R.O. Nonmedical drug use by high school students: a three-year survey study. Intl J Addictions 11(5): 755-792, 1976. - Steffenhagen, R.A.; McAree, C.P.; and Nixon, H.L. Drug use among college females: socio-demographic and social psychological correlates. Intl J Addictions 7(2):285-303, 1972. Whitehead, P.C. Multidrug use: supplementary perspectives. Intl J Addictions 9(2):185-204, 1974. Whitehead, P.C.; Smart, R.G.; and Laforest, L. Multiple drug use among marihuana smokers in eastern Canada. Intl J Addictions 7(1):179-190, 1972. - Wolfson, E.A.; Lavenhar, M.A.; Blum, R.; Quinones, M.A.', Einstein, S.; and Louria, D.B. Survey of drug abuse in six New Jersey high schools: I. Methodology and general findings. In: Einstein, S., and Allen, S., eds. Student Drug Surveys. Farmingdale, N.Y.: Baywood, 1972. pp. 9-32. AUTHOR John A. O'Donnell, Ph.D. Department of Sociology University of Kentucky Lexington, Kentucky 40506 43 TIMN 0152410

r Chapter 5 k Patterns of Tobacco Use in the United States Dorothy E. Green, Ph.D. From the time of World War I until the mid-1960's, when the health hazards of cigarette smoking became conmon knowledge, the number of smokers grew at increasingly rapid rates. After the publication of the Report of the Advisory Committee to the Surgeon General in 1964 (USDHSW 1964), both government and private organizations undertook to inform the public of the health consequences of smoking and thus to halt the increase in numbers of smokers. This paper describes the patterns of tobacco use since this effort began. Shortly after the Surgeon General's report was issued, the National Clearinghouse for Smoking and Health (NCSH) was established in the Public Health Service. Since 1964, several surveys of tobacco use patterns among adults, teenagers, and health professionals have been conducted by private organizations for the NCSH. The data presented below are drawn from these surveys. DHtifOGRAPHIC DATA ON ADULT CIGAREITE SMOKII2S Surveys of adult tobacco use were made in 1964, 1966, 1970, and 1975 ((JSDHEW 1969; USDHENI 1973; USDHEW 1976a). In 1975, sample size was approximately 12,500; in each of tfie three earlier surveys, it was about 5,000. Additionally, except in 1964, followup studies were made of persons previously surveyed. Age and Sex At the time of the Surgeon General's report, a little over half the men (52.4 percent) and almost one-third of the women (32.5 percent) were cigarette smokers. (Data from the two surveys of national probability samples of adults age 21 and over conducted in the fall of 1964 and the spring of 1966 have been combined.) By 1975, these percentages had dropped to 39.3 percent of the men and 28.9 percent of the women. In that ten-year period, the greatest decreases in smoking among men occurred during the first five years, when male sTMking dropped from 52.4 percent to 42.2 percent. Decreases occurred in every 44 ,TIMN 0152411

age group, except that between 1970 and 1975 there was a slight increase in the proportion of men age 65 and over who smoke. The decrease in female smoking between 1964 and 1975 was not so large, but occurred in every age group except those age 55 and over. There was a small increase in the youngest age group--age 21-24--from 1970 to 1975, but a substantial decline as compared with the 1960's. The increases in the proportion of smokers in the older age groups do not, of course, reflect the taking up of smoking by senior citizens, but simply the aging of a population which included large numbers of smokers. The percentages are much smaller than would have been predicted from earlier data on this cohort. The slight increase in young women who smoke follows the trend in teenage girls' smoking, discussed below. Marital Status In 1975, the majority of the respondents (78 percent of the men and 71 percent of the women) replied that they were married. Since married persons comprise such a large part of the total population, their smoking rates parallel those of the population as a whole. In the married group there is a slight drop since 1970 in the proportion of current smokers: from 40 percent to 38 percent for males, and from 32 percent to 28 percent for females. About the same percentage of males who have never been married and males who are married are current smokers. Single females have a slightly higher rate than those who are married (31 percent and 28 percent respectively). Both males and females show a substan- tially lower smoking rate than in 1964-66 or in 1970 (61 percent of single males and 33 percent of single females were current smokers in 1964-66; comparable data for 1970 were 56 percent and 36 percent. The highest smoking rates are among those who are divorced or separated. While only one-third of the respondents who are married and living with_ spouse are smokers, 60 percent of the men and 50 percent of the women who are divorced or separated are smokers. This represents a substantial decrease from the 1970 data showing that 76 percent of the men who were divorced or separated were smokers; but at the same time, it represents an increase from the 44 percent rate for women reported in that year. This is compar- able to findings in the earlier surveys. Men who are widowed have a slightly lower smoking rate (36 percent) than those who are married (38 percent). Widowed women, however, have a much lower rate (19 percent) of cigarette smoking than do married women (28 percent). This probably reflects the lower smoking rate among older women. The proportion of smokers among both males and females in this marital category (widowed) is little different from that observed in 1964-66 or 1970. 45 C3152412

'i To summarize the marital status category, the data show that men and women who are divorced or separated are more likely to be cigarette smokers than are persons who are married, widowed, or single. In all marital categories, a greater proportion of men than of women are smokers. Educational Level The smoking rates are lowest for those who never went to high school--37 percent for men and 18 percent for women. Because educational level is related to income, there may be a correlation between income and the cost of cigarettes for persons in this education group. At all other educational levels there is a decided inverse rela- tionship between the amount of education and smoking behavior, with a smaller proportion of smokers at higher educational levels. Among males, those who attended high school but not college showed 46 percent current smokers, those with some college, 36 percent, and those who graduated from college, 28 percent. Comparable proportions for women are 32 percent, 32 percent, and 21 percent. An examination of quit rates shows that half of the college grad- uates who ever smoked cigarettes are now former smokers. Occupation Among males, white collar workers (including sales personnel) are much less likely to be current smokers (36 percent) than are those in all other occupations (47 percent). This finding is consistent with the relationship between educational level and smoking behavior. Two-fifths of the women in the sample reported that they are employed outside the home. Of these employed women, one-third are current smokers, while only 27 percent of those who classify them- selves as housewives are smokers. This finding tallied with pre- vious evidence of greater prevalence of smoking among working women. Among the employed women, white collar workers are somewhat more likely to smoke than are those in other occupations (34 percent and 32 percent, respectively)--a relationship opposite to that found among men. Income Family income is related to smoking rates with both males and females, but in different ways. Men in relatively affluent fam- ilies are less likely to smoke, while women in this group are more likely to be current smokers. Only 35 percent of the men who reported an annual family income of $20,000 or more are cigarette smokers, while 46 percent of the men in the $7,500-$10,000 range are smokers. Among women, however, there is an increase in smoking from 24 percent for those in families earning under $3,000 to 34 percent for those with incomes of $20,000 or more. 46 TIMN 0152413

s~ _ . . ~ ,.....~.~. ~ . _ . r ; RHMEDIAL ACTION--SEITING LIb4ITS There has been a growing feeling that nonsmkers have a right to be allowed to breathe air free from contaminants in cigarette smoke. A dramatic change in attitude has taken place over the years toward agreement with the statement, "The smoking of cigar- ettes should be allowed in fewer places than it is now." Only 52 percent agreed with this statement in 1964 and 1966, but 57 percent agreed in 1970, and 70 percent in 1975. Between 1970 and 1975, this increase took place among current smokers (from 42 percent to 51 percent), among former smokers (from 61 percent to 77 percent), and among those who have never smoked (68 percent to 82 percent). This means that more than half the present smokers would like to see smoking allowed in fewer places than it is now, despite the fact that there are more and more restrictions on places where people are allowed to smoke. Since so many people would like to'see smoking allowed in fewer places, it is not surprising to find that nearly two-thirds (63 percent) of the respondents say that it is annoying to be near a person who is smoking cigarettes. Ten years ago less than half of the total number of respondents (46 percent) agreed with this, and now even more than a third of the smokers (35 percent) do so. A ban on cigarette advertising was advocated by a little over one- third (36 percent) of the 1964 respondents. By 1970, this had increased to 60 percent. Since then, television and radio commer- cials advertising cigarettes have been banned. Because these two media were the most widely used for advertising cigarettes, it might be presumed that such a ban would satisfy the public. This is not so. In 1975, 56 percent believed that cigarette advertising should be stopped completely. This view is subscribed to by 51 percent of the men and 60 percent of the women. Two out of five smokers would like to see cigarette advertising stopped, perhaps because some smokers who are trying to quit find that the adver- tisements make it more difficult. A second possibility is that, while they continue to smoke, they are loath to see the younger generation take it up. Because of the growing feeling that smoking should be regulated more stringently, questions were added to the 1975 survey to explore the matter of regulation more fully. First, the question, "Do you favor stronger Federal regulations concerning cigarette smoking?" elicited 52 percent "yes" responses from men and 60 percent from women. More than 40 percent of current smokers even favor such regulations. Another area explored was whether management has the right to prohibit smoking in its place of business. Only 16 percent of the. respondents felt that management should not have the right, while 78 percent thought they should have the ri t; the other 6 percent were undecided. fiighty percent of the women and 75 percent of the men thought that management should have the right to prohibit 47

smoking. Of those who felt this way, 28 percent believed that they should be able to do this only when smoking is a safety hazard, while 72 percent said they should have this privilege whether it is a safety hazard or not. Respondents were asked if they found it annoying to be near a person smoking cigarettes. Those who said "yes" were asked whether they found cigarette smoking more annoying than several other things that might be bothersome. The majority of these respondents expressed the belief that cigarette smoking is more annoying than someone's smoking a pipe, cracking knuckles, chewing gum, and humming or whistling. At the same time, a greater number of people said that smoking cigars-is more unpleasant than cigarettes, and about the same number found cigarette smoking more annoying than drinking alcoholic beverages as found drinking alcoholic beverages more irritating than smoking. TEENAGE SbfOKING The widespread use of cigarettes among teenagers, with its attendant health hazards, has prompted four national surveys of teenage smok- ing in the United States. The first was completed in January 1968 and the last in January 1974, with intermediate surveys in 1970 and 1972 (USDHB9 1972; USDI-IEW 1976b). About 2500 young people (12 through 18 years old) were interviewed in each survey. Trends in Prevalence of Teenage Smoking During the six years covered by the surveys, the proportion of boys who smoke at least once a week changed very little, except for an increase in 1970. This increase was most marked in the 17- and 18-year olds. In 1972, this proportion had dropped back to the 1968 level and remained there in 1974. Among girls, however, there has been a gradual increase in the proportion of smokers in every age group., This has had the effect of practically eliminating the difference in smoking behavior of the two sexes. In 1968, the proportion of girls reporting that they smoked cigarettes regularly was only about one-half that of boys. By 1970, this ratio had risen to almost two-thirds, and in 1972 to about 85 percent. The difference between the smoking behavior of boys and that of girls had disappeared by January 1974. Discussion Overall, the proportion of teenagers who smoke cigarettes regularly, including those who smoke at least once a week, has increased from 12 percent in 1968 to 16 percent in 1974. This increase has taken place almost exclusively among girls. Traditionally, cigarette smoking was a custom indulged in by men and boys and was not usual among women and girls. About the time of World War II, smoking among women began to increase, but it never quite reached the level of male smoking. In the late 1960's, men began to quit smoking, so that the proportions of inen and women smokers approached each other more closely. We have already mentioned that the smoking behavior 48

4 of boys did not cha~ige appreciably during the late 1960's and early 1970's, but smoking among girls gradually increased. In fact, it rose to the point where, in 1974, boys and girls were smoking in the same proportions. While the reasons for the disappearance of differences between boys' and girls' smoking behavior are not immediately apparent, it is likely that they are the same as the reasons that account for the disappearance of many other differences between boys and girls, in dress, hair styles, etc. In order to understand the phenomenon of teenage smoking, it is important to recognize the differences between those who take up smoking and those who do not. Since adolescence is a period of exploration, it is natural for teenagers to experiment with smoking, just as they do with other "adult" behavior. But which experimenters become smokers and which do not? By the time a boy or girl is 18 years old, only about one in three is a smoker, although most have tried it. How do we explain the differences in behavior? One way is to look at the differences between teenage smokers and teenage nonsmokers. Until adolescence, there is no question that the family exerts the greatest of all influences upon children; but, as they grow up, outside influences become stronger. How much does the family situation affect the smoking practices of the teenager? Several characteristics of the family were examined, and all were found to be related to smoking practices. First, teenagers in intact families (that is, those in which both parents were present) were less likely to smoke than'were those in homes where there was not both a mother and a father. About one teenager in five lives in a home without both parents, some because one parent is dead or the parents divorced, and some, particularly' the older teenagers, because they have left the family home and have set up their own living arrangements. They have adopted a lifestyle that is somewhat precocious; perhaps smoking is one characteristic of this lifestyle. Another family characteristic that is related to teenage smoking is socioeconomic status, as measured by education of the parents. Those with better-educated parents are less likely to smoke than are those with less well- educated parents. This is consistent with studies of adults which find that the higher the educational level a person has attained, the more likely he is to have quit smoking, if he ever did so. In spite of the notion that adolescents, in rebellion against their parents, reject the customs and beliefs of their elders, they do not turn away from their parents' smoking practices. Parents who smoke are likely to have children who'smoke. In fact, teenagers with two parents who use cigarettes are more than twice as likely to smoke as are those with no parent who indulges in this habit. Teenagers emulate older brothers and sisters, too. A boy or girl with an older sibling who smokes is extremely likely to be a smoker as well. If a teenager has both a parent and an older sibling who smoke, the likelihood of becoming a smoker is amplified. In fact, he or she is four times as likely to smoke as is one who has no smoking example in the immediate family. Smoking appears to be one of those customs which families as a whole either adopt 49 T jMN 415 _ 2416

or do not adopt. Just as in some families a coffee pot is always on the back of the stove, in some homes cigarettes are readily available for family members to help themselves. Teenagers not only have families who smoke; they also have friends who smoke. Almost nine out of ten smokers acknowledge that at least one of their four best friends smokes on a regular basis, while only one in three of the nonsmokers claims a smoker among his or her four best friends. At the other extreme, one in five nonsmokers claims that none of his or her four best friends has even experimented with smoking, while only one in a hundred of the smokers makes this claim. When we discuss parents' and older siblings' smoking patterns, we can talk about their influence on teenagers, since presumably these older family members set the stage for them. But when we talk about friends' smoking, there is no way to guess who influenced whom. Did Tom's friends exert pressure to get him to smoke? Did Tom urge his friends to smoke? Did he select friends because they smoked, or did they select him because he smoked? Or do they share the same kind of lifestyle, congenial to all in this group, that includes cigar- ette smoking? Tom cannot tell us; no one can. It is likely, however, that Tom and his friends share a life pattern that includes smoking. One aspect of the teenage lifestyle is the practice of working out- side the home, either full time or part time. Teenagers who work at some time during the year are twice as likely to be smokers as are those who do not work. Perhaps the nonworking minority is somewhat more protected, and less likely to have achieved much independence. Those who work often participate with adults in a work situation, and therefore are more likely to experiment with smoking and other adult behaviors. Another aspect of the teenage lifestyle is reflected in educational and vocational aspirations. The college-bound teenager probably has priorities that differ from those of peers who plan to terminate _formal academic education with high school. -The high school student- in a college preparatory course is less likely to smoke than is one in any other course. This is consistent with the finding that off- spring of parents who went to college are less likely to smoke. Thus the theory that smoking is related to socioeconomic status is supported. HEALTH PROFESSIONALS: SMOKING AND HEALTH People in the health professions are the authorities on matters pertaining to health. They are looked to for advice and are expected to be exemplars in preventive medicine. This is certainly true of cigarette smoking. For example, a recent survey of adults showed that more than three out of four feel that doctors should set a good example by not smoking. Thus, it is essential to know how these important exemplars see their roles, how they perceive smoking in relation to diseases, and how they act with regard to cigarette smoking both personally and professionally. 50 ; TIMN 0152417

~ 1 an NCSH survey (unpublished) approximately 2,500 individuals representing each of four groups of health professionals were studied in 1975: physicians, dentists, pharmacists, and nurses. The same groups had been studied in the period between the spring of 1967 and the summer of 1969.1 One purpose of the current study was to assess the changes that have taken place during the inter- veni.ng years. Findings In three of the four groups the proportion of smokers has decreased markedly. Only among nurses has the proportion of smokers stayed at about the same rate as in 1969. The proportion of physicians who smoked dropped from 29.6 percent in 1967 to 21.0 percent in 1975. During the same 8-year period, smoking among dentists decreased from 34.3 percent to 23.3 percent. In 1968, 34.5 percent of pharmacists smoked, compared with 27.5 percent in 1975. Nurses were smoking at about the same rate in both surveys--37.3 percent in 1969 and 38.9 percent in 1975. pmong physicians, there was little or no difference in smoking behavior by age group: 21 percent of those under 40 are current smokers, and 21 percent of those 40 and over are smokers. Both these proportions represent substantial decreases from the 1967 survey, when 33 percent of those under 40 and 29 percent of those 40 and over were smokers. The same pattern holds for dentists: in 1975, 24 percent in both the younger and older age groups.were current smokers. Both were down from 1967, when 36 percent of those under 40 and 33 percent of those 40 and over were current smokers. Among pharmacists, the decrease in proportion of smokers was more noticeable in the younger age group than in the older. The percent of pharmacists under 40 who were current smokers decreased from 36 percent in 1968 to 25 percent in 1975. The cor- responding percentages for those 40 and over were 35 percent in 1968 and 29 percent in 1975. As the other health professionals do, nurses under 40 show a decrease from 1969 to 1975--from 39 percent to 34 percent. However, 35 percent of the nurses 40 and over were smokers in 1969, and 42 percent were smokers in 1975. This increase reflects the fact that women 40 and over in 1975 were in the 33- to 40-year age group in 1969 and smoked at a much higher rate than the older nurses who have since left the profession. In comparing health professionals with the adult population in general, it is appropriate to compare three of the groups with males and one group with females. Physicians in the sample included 92.9 percent males; dentists, 99.2 percent; and pharmacists, 90.0 percent. Nurses included 97.5 percent females. A 1975 national probability sanple of adults showed that 39 percent of the males and 29 percent of the females were current smokers. The proportion of smokers in each of the three predominantly male health professional groups was much lower than that of males in the general population, but nurses were smoking at a higher rate than women in general. S1 ; TIMN 0152418

Three out of every five health professionals surveyed had a history of smoking, but the majority in all groups except nursing had quit smoking. The highest quit rate was found among physicians--64 percent of all those who had ever smoked were former smokers at the time of the 1975 survey. This was closely followed by 61'per- cent of the dentists and 55 percent of the pharmacists who had once smoked but had quit. These proportions are higher than the quit rate of 43 percent in the male population'who had ever smoked. Nurses had a quit rate of 36 percent--slightly above the quit rate of 34 percent of smokers in the female population. Physicians have been the leaders in giving up smoking; 55 percent of former smokers have not smoked in 10 or more years. This com- pares with 46 percent of former smokers in the male population. Dentists were next in quitting, with 50 percent of former smokers having quit 10 or more years ago. Pharmacists began quitting at about the same time as other men in the population; 47 percent of former smokers among them quit 10 or more years ago. Women did not start to quit smoking in large numbers as early as men. Of the former smokers among the female population in 1975, only 36 percent had not smoked for 10 or more years. However, like other health professionals, nurses lead their general popula- tion group in quitting smoking. Among their former smokers, 43 percent of nurses have not smoked for 10 or more years. Smokgrs in the predominantly male health professions smoked fewer cigarettes per day than males in the general population, while there was little difference between nurses and their female coun- terparts in the general population. As well as smoking fewer cigarettes, physicians, dentists, and pharmacists also tended to smoke cigarettes lower in tar and nicotine than did males in the general population. While one in five in these health professions smoked cigarettes with a "tar" level of 15 mg or below, only one male in eight in the general population smoked these low "tar" cigarettes. Similarly, a larger proportion of nurses smoked low- tar cigarettes than did smokers in the general female population. b4ost health professionals who are still smoking cigarettes started smoking before the relationship between smoking and disease was well known. A majority of them have made at least one attempt to stop, at rates similar to those among the general population of smokers: physicians 63 percent, dentists 64 percent, pharmacists 61 percent, nurses 55 percent, adult males 64 percent, and adult females 60 percent. In summary, changes in the smoking habits of these groups of health professionals have been towards the reduction of smoking and, for the most part, have exceeded and predated changes among the general population. Nurses were formerly among the group of heaviest smokers in the female population. Now that there is an increasing rate of quitting among older nurses and a reduced rate of taking up smoking by younger nurses, this is beginning to change. 52 1IMN 0152419

Relationship between Cigarette Smoking and Selected Diseases people in the health professions generally see cigarette smoking as either a major cause or a contributing cause of diseases of the lung. In fact, 9 out of 10 physicians and dentists say that smoking is a major or contributing cause of lung cancer, chronic bronchitis, and pulmonary emphysema. Between 80 percent and 90 percent of pharmacists and nurses agree. In every case, this represents an increase from the population holding these opinions in earlier surveys. Even greater increases were observed in the proportion seeing smoking as a cause of heart disease. The proportion of physicians increased from 71 percent to 78 percent. Corresponding proportions for dentists were 60 percent to 75 percent; for pharmacists, 46 percent to 63 percent; for nurses, 60 percent to 74 percent. Perception of Responsibility as a Health Professional People in the health professions see cigarette smoking as a serious health hazard. How do they see themselves as serving to reduce the incidence of disease and death caused wholly or partly by snnking? First, they believe it is their responsibility to set a good example by not smoking cigarettes. Ninety-one percent of physi- cians, 88 percent of dentists, 87 percent of nurses, and 73 percent of pharmacists agree they have this responsibility. The majority also agree it is their responsibility to convince people to stop smoking. Three-quarters of physicians and nurses, 61 percent of dentists, and 51 percent of pharmacists subscribe to this. Many in the health professions also believe they should be more active than they have been in speaking to lay groups about cigarette smoking. In fact, 82 percent of physicians, 74 percent of nurses, and 68 percent of dentists and pharmacists agree they should be more active. Since health professionals profess a feeling of responsibility toward their patients' (or customers') smoking, smokers in these ,groups have changed their behavior accordingly. Health profession- als are much less likely to smoke in the presence of patients or customers than they were in the earlier surveys. In all four pro- fessions, more of them report that they never smoke in front of a patient (or customer) (physicians, 39 to-7percent; dentists, 50 to 65 percent; pharmacists, 22 to 41 percent; nurses, 60 to 89 per- cent). Although a slightly smaller proportion of nurses, compared with doctors and dentists, feel that they should set a good example, they include the largest proportion who say they do not smoke in front of patients. Perhaps they face more restrictions of where they are allowed to smoke at their places of work. Indeed, 80 percent of the nurses who are presently employed say that smoking is restricted to certain times or places, or prohibited entirely, at the place where they work. • 53 ' TININ 152420

Just as health professionals feel increasingly that they should be doing something about smoking, they are less pessimistic than they were earlier about their ability to effect change. For example, in 1967 three-fourths of the physicians agreed that "there is no method around today to really help a smoker who wants to quit but can't do it on his own." By 1975, fewer than half of them agreed with this statement. The proportion of dentists agreeing with it dropped from 58 percent to 33 percent. It is interesting to note that nurses have been most optimistic about the possibility of helping people to quit. USE OF OTHER FOR'AS OF TOBACCO Since the mid-1960's, the proportion of pipe smokers among males has decreased from about one in five to about one in eight. Among women, it has remained at less than one in a hundred. Cigar smoking has also decreased, from 28 percent in the 1964-1966 surveys to 18 percent in 1975 in males. Women have not usually smoked cigars; fewer than 1 percent reported smoking cigars in any of the surveys. The use of snuff has remained at about 3 percent for males and less than 2 percent for females over the time of the surveys. Chewing tobacco is used, and has been used, by about 6 percent of the males and 1 percent of the females (USDHEW 1976a). These forms of tobacco are used by such a small percent of the population that little attention has been paid to trying to decrease the prevalence of their use. • •SUhMARY Overall, then, the picture of decrease in cigarette smoking is an encouraging one, with the exception of the increased rate of smoking among teenage girls. However, this change probably reflects the tenor of the times and an alteration in the concept of what has traditionally been considered "feminine" behavior. FOOTNOTE 1. .The 1975 survey was conducted for NCSH by Chilton Research Service; the 1967 and 1969 surveys were conducted by National Opinion Research Center. REFERENCES U.S. Dept. of Health, Education, and Welfare, Public Health Service. Smoking and Health, Re ort of the Adviso Committee to the Sur eon Generalof tI e-Mb ic ea t rvice. Pub. 103. Was ington, D.C.: Superinten ent o Doctnnents, U.S. Govt. Printing Office, 1964. 387 pp. U.S. Dept. of Health, Education, and Welfare, Public Health Service. Use of Tobacco--Practices, Attitudes, Knowled es and Beliefs U.S. Fall 1 an Spring 1966. National Clearing ous.e for Smo ing and Health, 1969. 54 TIMN 0152421

U.S. Dept. of Health, Education, and Welfare, Public Health Service. Teen-A e Smokin : National Patterns of Ci arette Smoking, Ages 12 t_rou __ 18, in 1 an 1970. DHEW . EW 72-7508 , Washington, U.S. Dept. of Health, Education, and Welfare, Public Health Service. Adult Use of Tobacco--1970. National Clearinghouse for Smoking an e. t. Pub. No. (HSM) 73-8727, 1973. U.S. Dept. of Health, Education, and Welfare, Public Health Service. Adult Use of Tobacco--1975. Center for Disease Control and National Cancer Institute, a. U. S. Dept. of Health, Education, and Welfare, Public Health Service. Teen-A e Smokin : National Patterns of Ci arette Smoking, Ages t rou 18, in an Pub. . TF1IRj-"7"6=93 ,v'asTiungton, . . , 1~3'7bb_. AUTHOR Dorothy E. Green, Ph.D. Consulting Research Psychologist 3509 N. Dickerson Street Arlington, Virginia 22207 r 55 ,TIMN 0152422

Part II Behavioral Factors TIMN 0152423

Chapter 6 Nicotine as a Discriminative Stimulus to Behavior: Its Characterization and Relevance to Smoking Behavior John A. Rosecrans, Ph.D. INTRODUCI'ION One of the major objectives of research being conducted in our laboratory over the last decade has been to determine the neuro- chemical and pharmacological mechanisms by which nicotine pro- duces its behavioral effects in experimental animals. The goal of this research has been to assist in identifying which of nicotine's effects may contribute to its use in tobacco by humans. The generalization of our findings to tobacco use, however, is contingent upon our ability to develop behavioral procedures which are specific to nicotine and analogous to nicotine's effects in humans. Much progress has been made in this regard, and it is felt that we have been successful in developing an animal model of nicotine effects which will provide important information relevant to drug dependency problems. The experimental model developed in this laboratory relies upon the ability of nicotine to exert stimulus control of behavior (Rosecrans and Chance 1977; 1978). In this approach an animal or human subject is required to discriminate between drug and non- drug states in order to-receive a positive reinforcement. For example, food-deprived rats are trained to press one of two levers (for liquid reinforcement) in an operant chamber after they are administered nicotine (200-400 ug/kg s.c.), while the opposite lever is reinforced following saline administration on a different day. After repeated daily exposure to this procedure, the animal soon learns to detect the effects of nicotine. Thus, the drug is acting as a discriminative stimulus. Human subjects can be studied in an analogous way. In studies just initiated in our laboratory, human subjects can obtain a financial reward for pressing the correct lever in a similar two-lever drug discrimination task. In this situation each subject must detect differences in nicotine level when allowed to smoke cigarettes differing in nicotine content (0.3 vs. 1.3 mg nicotine per cigarette). 58 ITIMN 0152424

Tar content, taste, and other parameters have been equalized to pre- vent the smoker from using other cues to solve the problem. As with our animal studies, human subjects will be allowed access to the discrimination task following random exposure in each session to either the high or low nicotine cigarette. This latter experiment is extremely important, as it will allow us to determine whether nicotine can be studied in human subjects using procedures developed from our animal experiments. USES OF TI-IE MODEL It should be understood that we are not studying the effects of nicotine on behavior. Rather, these procedures are used to measure drug detection. This is an important distinction, since drugs from different pharmacological classes may be shown to induce the same degree of behavioral disruption in animals, but may not be similar in terms of quality as a discriminative stimulus (DS). Put in anoth- er way, a rat may exhibit equivalent behavioral disruption by spe- cific doses of both d-amphetamine and morphine, but it will be able to discriminate between them because of the qualitatively different DS properties. These procedures, therefore, provide a very specific model which can be used to study the mechanism of drug action within a given pharmacological class. The discriminative stimulus model has provided us with a unique tool by which to study nicotine. To determine mechanism of action or central cite of action, several neuropharmacological manipulations can be utilized in animals trained to detect nicotine, to attenuate or mimic the nicotine DS. In the experiments to be summarized in this paper, rats are first trained to discriminate between doses of nicotine and saline. Qice this is achieved, animals can be challenged with suspected antagonists or given experimental or known compounds to determine if such compounds produce effects similar to those of nicotine. From such studies one can determine: 1) the receptor mechanisms involved, or 2) the structure-activity relationships in- volved in the nicotine-receptor interaction. In addition, such studies can be conducted centrally via chronically implanted cannulae at various sites to study the locus of action of the nicotine DS with- in the brain. Neurochemical interactions have been studied by evaluating the ability of nicotine to produce its DS effects in rats chronically depleted of specific biogenic amines. At present, we are also exploring ways of studying ongoing brain area biogenic amine function under nicotine and saline conditions, via the use of radio- isotopes and centrally placed push-pull cannula. Specific Behavioral Techniques Used in Animal Studies At approximately 10 weeks of age, rats were trained to press first one lever, then the other, in a two-lever operant chamber. Discrimi- nation training began with four preliminary training sessions of 15 minutes duration in which each correct bar press was reinforced (Chance et al. 1977). Subsequent sessions started with a 2.5 minute period during which no responses were reinforced; a variable interval of 15 seconds (VI - 15 sec) schedule was in effect for the remaining 59 TIMN 0152425

12.5 minutes. Every session was preceded by 10 minutes with sub- cutaneous injection of either nicotine or saline. During the four preliminary training sessions, nicotine and saline injections were alternated daily; thereafter, 2 days of one treatment were followed by 2 days of the other. By means of this double alternation schedule of drug administration, each treatment was preceded equally often by a session with the same and the opposite treatment. One lever was reinforced after the injection of nicotine, and the opposite lever was reinforced following saline. For half of the subjects, the right lever was rewarded after nicotine and the left lever was re- warded after saline. These conditions were reversed for the remain- ing animals. After 40 discrimination training sessions, responding was relatively stable. The same animals continued to receive either 200 or 400 pg/kg of nicotine and saline according to a double alternation sequence. However, test sessions 2.5 minutes in duration were inter- posed during which no responses were reinforced. fin odd number of training sessions, usually one or three, separated two successive test sessions. All experimental manipulations were accomplished during test sessions. That is, drug antagonism or drug generaliza- tion studies were conducted during these test sessions. During test sessions, rats made 85-95 percent of their responses on the nicotine- correct lever following a training of nicotine, while response rates following a dose of saline were only 10-15 percent on the same lever. Thus, these animals exhibited an ability to discriminate between nicotine and saline of 70-85 percent. NICOTINE AS A DISCRIMINATIVE STIMULUS: MEQiANISM AND SITE OF ACTION The DS approach has been very useful in studying the mechanism of action of nicotine, especially since this drug appears to elicit behavioral changes within other tasks which are contingent on baseline arousal levels prior to drug exposure (Rosecrans 1971a; 1971b, Thus, nicotine produced stimulant effects when the baseline arousal levels were low, while reverse effects were observed if arousal levels were high. These variable effects, while important to the overall behavioral pharmacology of nicotine, cause difficulty in attempts to correlate behavior with central mechanisms or neuro- chemical events. The advantage of DS procedures is that the subjects are not responding to a specific effect of the drug on behavior, but to the presence of the drug within the CNS. Furthermore, the speci- ficity of and sensitivity to the nicotine cue did not change in relation to time-duration or dose-response relationships across different schedules of reinforcement which elicit high (FR 10 sec) or low (DRL 10 sec) rates of responding ((hance et al. 1977). Furthermore, to learn a DS task, an experimental subject must de- velop some degree of behavioral tolerance to a specific training dose of nicotine to perform the operant schedule. Domino (1967) was one of the first to investigate nicotine's CNS mechanisms in a systematic way and clearly showed that nicotine acts at specific non-muscarinic receptor sites. He and coworkers, 60 TIMN 0152426

by studying various neurophysiological parameters in cats, showed that nicotine's arousal effects could be antagonized by mecamylamine, a centrally acting nicotinic blocking drug, but not by the muscarinic blocker, atropine. In addition, they were unable to block the nico- tine effect by the quarternary nicotinic peripheral blocker, hexa- methonium, indicating nicotine's central site of action. These workers also showed that arecoline (a muscarinic agonist) had the exact opposite profile (atropine, but not mecamylamine, blocked arecoline's arousal effect), suggesting that each drug was acting on separate N-cholinergic and M-cholinergic receptors. This research was extended using the DS paradigm (Schechter and Rosecrans 1971; Hirschhorn and Rosecrans 1974) and an identical profile for the ac- tion of nicotine was observed (see table 1). TABLE I Mechanism of the Nicotine (200-400 ug/kg s.c.) DS: Drug Antagonism Studies Results vs. ` Drug Antagonist Receptor Blocked Rot ne ic~ Mecamylami•ne (s.c.) N-cholinergic complete 1 mg/kg; 30 min prior central acting antagonism Hexamethonium (s.c.) N-cholinergic 1 mg/kg; 10 min prior peripheral acting no effect Atropine (s.c.) M-cholinergic 0.5-20 mg/kg; 30•min prior central acting no effect Dibenamine (i.p.) 10-20 mg/kg; 30 min prior a-adrenergic no effect Propranolol (i.p.) 1-4 mg/kg; 30 min prior g-adrenergic no effect a-Methyl-paratyrosine (i.p.) catecholamine sc.ae 30-200 mg/kg; 2 hr prior synthesis I blockade p-Chlorophenylalanine (p.o.) 5-hydroxytryptamine 300 mg/kg; 72 hr prior synthesis no effect Thus, it appears that the behavioral effects of nicotine are contin- gent upon stimulation of a specific N-cholinergic receptor. The possibilities that other animal systems may be involved and that there may be a dopamine (DA) neuron link have also been considered (Rosecrans, (hance, and Schechter 1976). Research conducted in this laboratory indicates that rats trained to discriminate nicotine (400 ug/kg, s.c.) vs. saline will generalize to nicotine injected into the hippocampus; but if the rats are dopamine-depleted (65 percent), no generalization to the trained s.c. nicotine cue occurs following injection of nicotine into the hippocampus. Additional support that L1A is involved with nicotine's pharmacological effects 61 r i TIMN 0152427 ,

:ames from human research in which smokers appear to have a lower incidence of Parkinson's disease than nonsmokers (Kessler 1973). The theoretical model proposed suggests that nicotine is somehow metabolized to nicotinic acid, which will alter the amount of DOPA available to DA neurons. Support for this theoretical model has also been obtained at the Medical College of Virginia in studies where nicotinic acid, given to rats orally, facilitated the accumu- lation of DA following i.p. doses of DOPA (Black 1977). The hippo- campus may thus be a primary site of nicotinic action and DA may also be involved in the action of nicotine (figure 1). DA systems, however, appear to affect cholinergic systems quite indirectly, via either the septum or N. Accumbens. Schechter and Rosecrans (1972a) also studied arecoline in this para- digm and showed that atropine, but not mecamylamine, antagonized its DS effects. Diethylatropine, a peripherally acting muscarinic blocker, was unable to antagonize the DS, indicating arecoline produced its effect via a central M-cholinergic stimulation. Schechter and Rosecrans (1972b) also showed that rats could be trained to discrimi- nate nicotine from arecoline, which suggests that there are separate central N- and M-cholinergic receptors. Other research (Rosecrans and Chance 1977) has indicated that peri- pheral doses of nicotine (100-400 ug/kg, s..c.) will generalize to nicotine administered via the intraventricular (ivt) route of doses ranging from 16-32 ug. The mechanism of ivt generalization also appears to be a function of N-cholinergic stimulation. In addition, generalization studies involving nicotine metabolites (cotinine) have indicated that nicotine is not producing its effects via some active metabolite. Support for this has also been obtained by cor- relating behavior to brain area nicotine levels (Rosecrans and Chance 1977). Time-duration and dose-response experiments indicated a close relationship between behavior and physiological drug level. This research delineates how nicotine produces stimulus control of behav- ior in rats. Parametric studies have provided information describing the duration and mechanism of action of nicotine (Rosecrans and Chance 1977). SPECIFICITY OF TfE NICOTINE DS Four years ago a research program designed to search out known chemi- cal compounds which might produce DS effects similar to th9se of nicotine was initiated in this laboratory. We began by studying d-amphetamine under a variety of training conditions and observed this stimulant to be quite unlike nicotine (Chance et al. 1977). In addition, other types of stimulants such as caffeine and magnesium pemoline were found to be devoid of nicotine-like effects. The responses elicited indicated that the rats studied perceived these compounds as unlike either nicotine or saline. Response rates on the nicotine-current lever never averaged 60 percent following doses of d-amphetamine, indicating that a CNS effect was perceived but was qualitatively different from nicotine. In addition, drugs such as LSD or morphine elicited only saline-like responding. 62 ; TIMN 0152428

, FIGURE 1 Hippocampus /------------------\ i N-CAolinerqic yChnc ~ JNEI ~rv~ Y-~./ 1 Neostrnotum . A6,/ i . A10, I '~A6 i Locus Midbrnin Interpeduncular Substontio Cerulaus Retlcutar Nucleus Niqra FormaNon CATECHOLAMINE - CHOLINERGIC INTERRELATIONSHIPS A description bf the interrelations between catechol- amine and neuron connections. Abbreviations are: DA ffi dopamine; ACh = acetylcholine; NE a norepinephrine. FIGURE 2 N` 'N ~ N \ ~ ~! CH3 N Nicotine 3- pyridylmethylpyrrolidine (3 PMP) Nornicotine Cotinine Structures of nicotine analogs and metabolites. Respective ED-50 generalization doses with nicotine were as follows: Nicotine = 52.4 ug/kg; 3PMP - 263.4 ug/kg; nornicotine se 960.5 ug/kg; cotinine did not generalize with nicotine. 63 :TIMN 0152429

To complete this study we also evaluated a series of nicotine and pyridine derivatives. Nicotine metabolites such as cotinine were inactive, but nornicotine did produce a partial generalization. Interestingly, lobeline, a compound considered to act like nicotine, was also without effect (Schechter and Rosecrans 1972c). The con- sensus of studies indicated that only one compound, 3_NP (figure 2), which is a chemical isomer of nicotine, had a significant nicotine- like effect (Rosecrans et al. 1978). In fact, Kallman et al. (1978) have shown that 3 PIT appears to act like nicotine centrally. That is, it appears to be producing its stimulus effects by stimulating N-cholinergic receptors. Finally, it should be added that the ability to exert stimulus con- trol of behavior is not a property only of nicotine. Most psychoac- tive drugs exert DS control of behavior. In many instances there is a good correlation between stimulus strength, drug abuse, and thera- peutic potential within a specific psychopharmacological class of compounds (Overton 1978). The intriguing aspect of nicotine in relation to other psychoactive drug classes is its apparent speci- ficity. It should be kept in mind, however, that we have been limited in our research to the numbers of chemical compounds avail- able for study. The picture could change as other chemicals are synthesized. RELEVANCE TO SD40KING BEHAVIOR To understand better how our research efforts may correlate to real world events, a model describing the sequence of events leading to drug dependency in man has been developed (figure 3). We have broken the sequence into three stages, which helps separate out all the con- tingencies involved in the development of drug dependency. The T6 properties of a specific agent appear to us to fit into stages II and III. For a drug to be abused, it first must be recognized by the individual (Discriminative Stimulus). Thus, there must be a state change, which is contingent upon the detection level of the individual. This, plus the quality of the state change, will have a direct bearing on the probability that the drug will be used again (Reinforcement Potenti- al). For example, cocaine produces strong qualitative state changes in animals and is rapidly self-administered, suggesting it to be reinforcing. ISD, on the other hand, also produces a profound state change, but the probability of continued self-administration is low, suggesting it may be somewhat aversive. Nicotine produces a potent D6 effect in animals, but some will argue that it does not produce the same rate of continued self-administration observed with other reinforcing drugs. At least, many investigators have had difficulty demonstrating self-administration of nicotine by'animals. Thus, what is it about nicotine which leads to its being used initially ? This is an issue we have discussed much in our laboratory, and as yet we have not developed an adequate hypothesis, at least in terms of a behavioral or neurochemical mechanism. We are satisfied at this time that the nicotine DS model does provide us with a good animal analog to human drug effects and that much of the information we obtain appears quite relevant to how this drug produces its effects'in man. 64 TIMN 0152430

In the process associated with nicotine use, the drug may not always be the primary reinforcer for smoking behavior. Instead, other psy- chological and sociological variables associated with the act of smoking may also be primary reinforcers, and nicotine becomes a secondary reinforcer once the habitual use stage occurs (figure 3). At this stage the discriminative stimulus properties of nicotine become important. The repeated pairing of task performance with the nicotine may provide the necessary signal for maintenance of the smoking behavior. This phenomenon, sometimes called state-dependent learning, is also a prominent property of other psychoactive drugs (Weingartner 1978; Overton 1978), and there is no reason to believe that it is not also the case with nicotine. Our hypothesis, then, suggests that the maintenance of smoking behavior (or the dependency stage) may be in part related to the state-dependent or discriminative stimulus properties of nicotine. Experimental evidence for such a hypothesis is lacking in man, but we do have some support for such a concept in animal subjects (Rosecrans and Chance 1977). In one study, rats were trained in a one-trial learning paradigm to avoid a shock under saline or nico- tine conditions. After the initial training sessions, rats were reexposed to the apparatus 24 hours later. The data indicated that rats trained in the nicotine state avoided the shock less often after a repeated exposure unless nicotine was readministered prior to the test session. Thus, nicotine provided the stimulus which enabled the rat to avoid being shocked. According to our hypothesis a smoker would need to maintain a constant level of nicotine to adequately perform certain behaviors. This point is speculative, though, and should be tested. However, the specific nature of the nicotine DS suggests that this property of nicotine is one of the many factors important for the maintenance of smoking behavior. DIRECTIONS FOR FUf[JRE RESEARCH There are three research areas that need to be expanded: 1) Sites and mechanisms of nicotine action There is much to be learned about the way in which nicotine pro- duces its effects. Until now, little effort has been given to pursuing how and where nicotine exerts its many pharmacological effects. In this respect, areas to be further explored include: a) brain sites of nicotine action; b) neurochemical mechanisms, specifically DA-ACh interactions (figure 1); and c) studies investigating the mechanismq of behavioral tolerance, physical dependence, and state-dependent learning. This last area is of special importance as it relates to dependence in human subjects. 2) Human research The hypothesis developed in this paper suggests that nicotine could be acting as a reinforcer of smoking behavior via its DS or state- 65 TIMN 0152431

FIGURE 3 Stage I Acute Pharmacological Experience Recognition of a State Change (discriminative stimulus) + Quality of L the State Change (stimulation, sedation, etc.) ~ Nature of the REINFORCEMENT QUALITY ~ Stage II Development of Habitual Behavior Drug will serve as a REINFORCER IF: Good Good ~ Subjective or Sociological ( ( ~ Experienc:e Reward I * ! Psychological Needs of the User ! ~ ~ Habitual Behavior Need to continue drug taking 1 k behavior / ~ ~~~~.... - - - - - - - - - - ..- , / Drug Dependecy will ensue to Stage I11 / maintain specific reward systems I n dence - Drug Depe 1 ~ Physiological Behavioral Sociologicai ~ Dependence Dependence Deoendence + / ( Toferance State Possible loss ~ ~ Withdrawal Dependency f of Social ~ ~ Symptoms Tolerance Contingent I Behavior 1 ~------_--_--__---__-_~- ~ An overview of the sequence of events leading to drug dependence, as viewed by this laboratory. 66 A TIMN 0152432

dependent properties. This question needs to be studied in depth in human subjects. The necessary methodologies have been developed, and what is needed now is some fundamental research into the ques- tion. 3) Organic molecular mechanisms of nicotine : t eTi rapeutic imp ications - The fact that millions of humans smoke tobacco suggests that many people obtain some benefit from this behavior. Furthermore, if this behavior can be associated with nicotine, then one could suggest that nicotine could have potential as a therapeutic agent. The history of the pharmacology of marihuana is relevant to this point. At present, much research is being devoted to studying the possible therapeutic potential of the tetrahydrocannabinols and synthetic derivatives, even though many consider this drug to be potentially dangerous. Thus, one wonders if some time should be devoted to studying nicotine analogs with the same point of view. Such research would serve two purposes. First, a nicotine-like compound could conceivably be developed which has psychotherapeutic potential. Second, a research program designed to synthesize nicotine analogs would also assist us in determining the structure and configuration of the nicotine receptor. ACKIVOWLEDGMENTS This research has been supported by grants from the A. M. A. Education Research Foundation and the r-ouncil for Tobacco Research. In addition the author wishes to express his thanks to Drs. M. D. Schechter, I. D. Hirschhorn, W. T. Chance and M. J. Ka11man for their research efforts over the years. REFERENCES Black, M.J. Alteration of L-Dopa effect on brain dopa and dopa- mine levels with nicotinic acid and N-Methyl nicotinamide. Rich- mond, Va.: Medical College of Virginia, 1977. (Ph.D. dissertation; unpublished.) Ghance, W.T.; Murfin, D.; Krynock, G.M.; and Rosecrans, J.A. A description of the nicotine stimulus and tests of its generali- zation to amphetamine. Psychopharmacol, 55:19=26, 1977. Domino, E.F. Electroencephalographic and behavioral arousal effects of small doses of nicotine: A neuropsychological study.. Ann N.Y. Acad Sci, 142:216-244, 1967. 67 TIMN 0152433

Hirschhorn, I.D.,and Rosecrans, J.A. Studies on the time course and the effect of cholinergic and adrenergic receptor blockers on the stimulus effect of nicotine. Psychopharmacol, 40:109-120, 1974. Kallman, M.J.; Spencer, R.M.; Chance, W.T.;and Rosecrans, J.A. A comparison of nicotine and structurally related compounds as discriminative stimuli, 1978 (in press). Kessler, I.J. Parkinson's disease. Perspectives on epidemiology and pathogenesis. In: Perspectives on Parkinson's disease. New York: Academic Press, Inc., 1973. pp. 88-105. Overton, D.A. Major theories of state dependent learning. In: Ho, B.T.; Richards, D.W., III-,and Chute, D.L., ed. Dgru •Discrim- ination and State Dependent Learning. New York: Academic Press, Inc., 1978. pp. 283-318. Rosecrans, J.A. Effects of nicotine on behavioral arousal and brain 5-hydroxytryptamine function in female rats selected for differences in activity. European J Pharmacol, 14:29-37, 1971a. Rosecrans, J.A. Effects of nicotine on brain area 5-hydroxy- tryptamine function in male and female rats separated for dif- ferences of activity. European J Pharmacol, 16:123-137, 1971b. Rosecrans, J.A.,and Chance, W.T. Cholinergic and non cholinergic aspects of the discriminative stimulus properties of nicotine. In: Lal, Harbans, ed. Discriminative Stimulus Properties of Drugs. New York: Plenum ress, 1977. pp. 155-185. Rosecrans, J.A.,and Chance, W.T. The discriminative stimulus properties of N- and M-cholinergic receptor stimulants. In: Ho, B.T., Richards, D.W., III;and Chute, D.L., ed. D~rug Discrimin- ation and State De endent Learning. New York: Academic Press, Inc. , IS'I8. pp, t19- 0. Rosecrans, J.A.; Chance, W.T.; and Schechter, M.D. The discrimina- tive stimulus properties of nicotine, d-amphetamine, and morphine in dopamine-depleted rats. Psychopharm Consn, 2:349-356, 1976. Rosecrans, J.A.; Spencer, R.M.; Krynock, G.M.; and Chance, W.T. Discriminative stimulus properties of nicotine and nicotine related compounds. In: Battig, K., ed. Behavioral Effects of Nicotine. Basel, Switzerland: S. Karger; I'=. pp. 70- 82. Schechter, M.D., and Rosecrans. J.A. CNS effect of nicotine as the discriminative stimulus for the rat in a T-maze. Life Sci, 10: 821-832, 1971. , 68 TIMN 0152434

Schechter, M.D.,and Rosecrans, J.A. Nicotine as a discriminative cue in rats: Inability of related drugs to produce a nicotine- like cueing effect. Psychopharmacol, 25:374-387, 1972a. Schechter, M.D.,and Rosecrans, J.A. Effect of inecamylamine on discrimination between nicotine- and arecoline-produced cues. European J Pharmacol, 17:179-182, 1972b. Schechter, M.D.,and Rosecrans, J.A. Atropine antagonism of arecoline cued behavior in the rat. Life Sci, 11:517-523, 1972c. Schechter, M.D.,and Rosecrans, J.A. Nicotine as a discriminative stimulus in rats depleted of norepinephrine or 5-hydroxytryptamine. Psychopharmacol, 24:417-429, 1972d. Weingartner, H. Human state dependent learning. In: Ho, B.T.; Richards, D.W., III;and Chute, D.L., ed. Drug Discrimination and State Dependent Learning. New York: Academic Press, Inc., 1978. pp. 361-382. AUTHOR John A. Rosecrans, Ph.D. Professor of Pharmacology Virginia Commonwealth University M.C.V. Station Box 726 Richmond, Virginia 23298 69 TIMN 0152435

Chapter 7 Nicotine Self-Administration in Rats H. M. Hanson, Ph.D., C. A. Ivester, and B. R. Morton INTRODUCTION Contrasted with the voluminous and still burgeoning literature devoted to laboratory study.of the self-administration of mor- phine, cocaine, amphetamine and the barbiturates, studies of the self-administration of nicotine have been few in number. The entire literature available to us is less than a dozen papers; Clark (1969), Lang et al. (1977), and Hanson, Ivester and Morton (1977) reported data collected with rats, Deneau and Inoki (1967) and Yanagita (1973) with primates. The reasons for this apparent neglect of a compound self- administered by a large percentage of the human population are uncertain. Techniques for such studies have long been available, and the feasibility of the study of the self-administration of nicotine was demonstrated by Deneau and Inoki (1967) 10 years ago. It is likely that the less dramatic effects demonstrable in the laboratory with nicotine compared to the effects possible with morphine or cocaine may have overshadowed the potential importance of such studies and have resulted in investigators directing their efforts toward the study of-substances yielding more striking and voluminous data.. The studies of nicotine self-administration to be reported in this paper were undertaken based on the assumption that nicotine is the substance in tobacco smoke that leads to man's use of tobacco. It was further assumed that such studies were relevant to understanding smoking in humans, and as a model could lead to a successful pharmacological "treatment" for smoking. METHODS Because of their small size, ease of maintenance,and the possi- bility of assembling a large colony for study, the rat was chosen as an experimental animal. Four hundred to 500-gram male albino 70 ,' TIMN 0152436

rats (Sprague-Dawley-derived, Buckshire Corporation) were im- planted with Weeks-Davis type jugular cannulae using ether anesthesia. After recovery from the anesthetic, the rats were fitted with Weeks-type "saddle and leash" restraint devices to protect the catheters and pladed in individual operant con- ditioning cages. For two days following surgery the rats were infused with nicotine solution or saline every 30 minutes; on the third day operant conditioning levers were placed in the chambers and all further infusions were dependent on lever depressions. Thirteen-second injections of 0.1 ml of solution were delivered following every lever depression (FR 1). Responses emitted during an infusion were ignored. Food and water were available at all times; the room housing the rats was maintained at 21-22°C; lights were on in the room 12 hours daily from 6:00 a.m. to 6:00 p.m. Patency of individual catheters was checked at weekly intervals or following the completion of a particular study. All data collected with a particular rat for a study were discarded if testing indicated that the catheter had failed. A group of 42 identical test cages were used for the studies to be reported. Rats were randomly assigned to test chambers and to studies. Where noted, more than a single compound was tested in individual rats. A total of 276 rats were used to collect the data. RESULTS Self-Administration of Nicotine; Dose-Response Relationship Groups of 9-10 rats (N=59) were implanted with jugular cannulae and placed in the test chambers with the operant levers removed. Every 30 minutes for two days the rats were given automatic injections of saline, or 3.75, 7.5, 15, 30, or 60 ug/kg of nicotine. At the end of the 48-hour period the automatic injections were stopped and the operant conditioning levers were placed in the chambers; for the next 10 days each lever press resulted in an injection of the same dose of nicotine or saline as was given in the°pretreatment period. The data collected are shown in figure 1 as mean square root responses for each of the groups. The horizontal line indicates the mean number of responses emitted by•the saline control group over the 10-day test period. The vertical lines are standard errors. A linear function fitted to the data by the method of least squares is also shown. The value for 60 ug/kg of nicotine was omitted from this estimation. Doses greater than 60 ug/kg resulted in even lower response rates, suggesting that for naive rats•doses of 30-60 itg/kg are maximally reinforcing; it can be noted, however, that doses in excess of 100 pg/kg, which resulted in convulsions following each injection, were self-administered at rates significantly higher than saline controls. 71

FIGURE 1 6 14 -1 7 ~ i SALINE 3.75 7.5 15 30 60 , _. µg/kg NICOTINE Dose response reZationship for seZf-administration of nicotine. Data points represent average number of infusions of various nico- tine soZutions over a 10-day period. The horixontaZ Zine is the saZine controZ group mean; the verticaZ Zines show standard errors. It was concluded based on these data that rats will self- administer nicotine at rates significantly higher than control levels over a relatively wide range of doses, and that it is possible to demonstrate a relationship between dose level and rate of self-administration. 72 TIMN 0152438

Effects of Pretreatment with Nicotine on Nicotine Self-Admini- stration Groups of 10-14 rats were implanted with jugular catheters and placed in individual cages as in the preceding experiment. The rats were assigned randomly to two groups receiving automatic injections of either saline or nicotine (60 Vg/kg) every. 30 minutes. After two days the rats were divided into four groups and assigned to one of four treatment conditions, self- administering nicotine or saline after pretreatment with either saline or nicotine. FIGURE 2 60 0 1 Nicotine-Nicotine N=13 Saflne-Nicotine N=10 ~ Saline-Saline Ns10 ' _ Nicotine-Saline N=14 _J Six Hour Periods Cumulative response curves of seZf-acr'nmistration of nicotine (60 ug/kg) or satine over a 5-day period foZZowing pretreatment with injections every 30 minutes of either nicotine (60 .ug/kg) or saZine for two days. The curves are identified by the treatments given during the two phases of the study. The mean vaZues on the finaZ day of the study are significantZy different: The data, collected over a 5-day period of self-administration, are shown in figure 2 as cumulative curves. Only the data collected during the last three days of the study were analyzed. The means of the groups were significantly different from each other. The shape of the curves indicated that both groups self-administering nicotine were increasing their rates of response while the other two groups self-administering saline were decreasing their average response rates over the 5-day period. 73 tiTIMN 0152439 .

It was concluded that a significant increase in response rates as a result of nicotine self-administration compared with saline is demonstrable in as short a period as 5 days. Pretreatment with automatic nicotine injections was found to significantly increase the rate at which nicotine is self-administered, compared with animals pretreated with saline. Pretreatment with automatic nicotine injections was found to significantly depress the rate of saline self-administration: the decreased rates might be considered in some way analogous to withdrawal symptoms seen following exposure to opiates. Further study of this effect is indicated. Based on these data, in order to facilitate training, all rats except for those used in special studies were subjected to a 2-day period of automatic injections of nicotine before being allowed to self-administer the compound. Course of Nicotine Self-Administration The data examined so far were gathered over very short periods of time (up to 10 days) following first exposure to lever_controlled injections of nicotixie. If longer periods of exposure are considered, the trend toward increasing daily dosages noticed in figure 2 becomes even more pronounced. Data collected with 4 rats self-administering nicotine (60 ug/kg) over a 30-day period are shown in figure 3 as cumulative curves. The rats, which were selected to represent a range of response rates, showed an overall acceleration in response rates. It is possible this tendency to increase daily dose levels is due to the development of tolerance,.either to the reinforcing actions of nicotine (requiring higher dosages) or alternatively, to the limiting side effects produced by nicotine (allowing higher dose levels). This tendency to increase the rate of administration has been noted in all rats studied to date. Correlated with the increase in total nicotine intake is a slow but consistent increase in the amount of nicotine necessary to induce a convulsion. __ That this phenomenon is not restricted to short exposures is illustrated in figure 4. In a manner similar to that in figure 3, cumulative curves, representing responding over a 5-month period, are shown for 4 additional rats. The increments in this case equal 5 days, which "washes out" minor day-to-day fluctuations. The cumulative curves of these animals also show accelerating rates of response. It should be noted that all 4 of these rats were subjected to a variety of pharmacological treatments, which changed response rates on a daily basis but did not seem to have affected the overall course of responding. 74

FIGURE 3 1300 1200-~ I 100-I I000 -~ 900-I w 800 f!) 2 0 vai 700 W ~ W ? 600 F a 400-~ 300-I 200-~ I00-I CurmcZative eurves of self-administration of nicotine for 4 rats over a 30-day period. The totaZ monber of responses for each ani- maZ is indicated under its respective identification nwnber. 75 ;TIMN 0152441

I FIGURE 4 I I I I I I I I I I i I 1 1 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 0ays G'umuZative curves of seZf-administration of nicotine for 4 rate over a period of five months. The numbers are for animaZ identi- ... fication. During the last 5-day period shown in the figure, the rat with the greatest nicotine intake (Number 15-64) averaged 44 responses a day, equalling 2.64 mg/kg of nicotine, a high but by no means unusual rate of self-administration. Extinction of Lever Pressing Reinforced by Nicotine Reports of systematically collected extinction data following reinforcement with drug infusions are practically nonexistent in the self-administration literature. The data that are available are almost entirely devoted to the study of 'tcross-self-admini- stration" of selected compounds and are usually single animal studies. rSee, however, Yokel and Pickens (1976)]. We felt that 76 TIMN 0152442

in order to unequivocally categorize nicotine as a reinforcing substance, extinction data was necessary. Following our standard procedure, 12 rats were implanted with cannulae, placed in individual cages, pretreated for 2 days with nicotine infusions (60 u g/kg) every 30 minutes and on the third day given operant conditioning levers. All further nicotine infusions were dependent on lever depressions. After 12 days of self-administration, saline was substituted for the nicotine solution, and lever pressing was recorded for 24 more days (extinc- tion). FIGURE 5 4-7 3-I 2-i Z a w Z f- DAYS Extinction of Zever pressing foZZor,ting reinforcement by nicotine infusions. Data shown are means and standard errors of the data transformed to square roots. Extinetion was begun on the traeZ,fBh day of the study, as indicated by the verticaZ arrow. The data collected are shown in figure 5 expressed as the means and standard errors of the square roots of the values. The group of rats showed the usual increase in response rate for the first 12 days of the study. The mean rate of response on Day 12 was approximately 50 percent higher than on Day 1, a statistically significant difference (paired values t-test). After withdrawal of nicotine the mean rates continued to increase on Days 13-15, 77 N=12 0 0 0 EXTI NCTION- 10 15 20 25 30 35 40 ;TIIVIN 0152443

the difference in mean rates reaching significance compared with Day 12 by Day 15. The rates thereafter decreased day by day, reaching the lowest point on Day 22 followed by a slight overall increase through Day 36. The rate of responding in extinction was significantly different from the mean rate on Day 12 by Day 18 and thereafter. Nicotine self-administration appears to generate extinction data comparable to that following other types of reinforcement. The prolonged period of increased rates of response at the beginning of extinction is unusual but is somewhat similar to the effects we have seen in extinction following reinforcement with morphine. Considering the fact that the animals were reinforced on a FR 1 schedule and were exposed to this schedule and to extinction 24 hours a day, the course of extinction was dramatically pro- longed, which again is very similar to extinction following morphine (unpublished data). The Effects of Selected Compounds on Nicotine Self-Administration The effects of 9 compounds were studied in groups of rats self- administering nicotine (60 ug/kg). Compounds with a range of pharmacological actions were selected, all of which have received intensive study in a variety of behavioral and neurophysiological experiments. Groups of 10 or more rats self-administering either nicotine solution or saline were selected from the colony for a particular study. Individual rats were used repeatedly; however, no rat was used more than once for the study of a particular compound. No rat was selected for study that had not been in the colony for at least 3 weeks; nicotine-self-administering animals were not selected for study unless their average daily response rate fell outside the 95 percent confidence limits of an historical saline- self-administering reference group and showed clear evidence of circadian rhythm in responding (greater number of responses during 12-hour dark period). A11, compounds were administered orally twice daily for 3 days (9:00 a.m. and 9:00 p.m.) after a 3-day period of twice daily dosing with water at the same times. Data were also collected for 3 days following drug administration as an additional con- trol. The patency of the catheters was tested on Day 10 of the study. I The data collected are shown in figures 6-14 as the means and standard errors of the square roots of the number of injections for consecutive 6-hour periods over the 3-day period. The arrows below the abscissa indicate the times of dosing. The doses selected for study were large enough in all cases to show pharma- cological activity in some appropriate test system. All dosages of the compounds were calculated as the weight of the base compound and administered dissolved or suspended in water. 78

FIGURE 6 5 ( I Jllz W 2 NICOTINE N=10 i 1 l T T 1 T j I i iSALINE ii_• "jS(c X X N=10 T~XX~ x X xx ~,X~ i ~ ijT X•\% X ~X,~ XX x~•X •X•X / •z z z'X X`x`X~/ z • T 5 MECAMYLAMINE 4 mg/kg p.o. b.i.d. T 10 _ 25 - t 15 t t 20 t t t 6 HOUR PERIOD S _T_ - 30 35 Effects of 4 mg/kg mecarmyZamine b.i.d. on nicotine and saZine setf- administration. The verticaZ Zines indicate standard error of the mean. MeeamyZcnttine was adninistered at the times indicated by the arrorvs; dosing with water was done at comparabZe time periods for the preceding 3 dabs. Mecamylamine and Pentolinium: The effects of 4 mg/kg p.o. of the ganglionic blocker, mecamylamine, are shown in figure 6. The difference between the two curves was significant at all tim~ intervals, the mean response rate of the nicotine group being generally 7 times that of the saline group. Although it is more pronounced in the later figures, a clear circadian rhythm is in evidence for the data collected preceding treatment. Following the first dose of mecamylamine, a dramatic and highly significant increase (approximately 4-fold) in nicotine self-administration occurred. The rats appeared active and slightly tremulous within 5 minutes of dosing. Repeated lever responses were emitted during the next 15-20 minutes followed by quiescence and assumption of the usual sleep posture. The total effect lasted no more than 45 minutes. No overt signs of high nicotine dosage were seen. The remaining doses'did not produce similar effects but resulted in a slow decrease in response rates. The average number of nicotine injections for the 24-hour period immediately following the 3-day mecamylamine treatment period was signifi- cantly lower than the average number of responses emitted the day preceding treatment. A similar, but nonsignificant, decrease was also seen for the control group. The last 3 days of the study. 6 + : 79 TIMN 0152445

resulted in a slow daily increase in response rates. Mean responses on the last day of the study are not significantly different from the mean responses on the last pretreatment day, suggesting the original control levels had been recovered. This effect of mecamylamine, considering the known nicotine- blocking activity of this compound rstone, Meckelnburg, and Torchiana (1958)], was not unexpected. The dramatic increase in self-administration following mecamylamine appears very similar to the effect seen following the administration of naloxone to rats self-administering morphine (unpublished data), and perhaps in similar fashion could be understood as a direct receptor blockade. The marked lack of effect following the remaining five doses in the series is puzzling and no ready explanation is presently available. FIGURE 7 5 c 0 ~4-~ w a ~ x ~ PENTOLINIUM 10 mg/kg p.o. b.i.d. Pi 3.4 - r! A /l I -A _wNICOTINE N 2 ~ 1-- '~ 2-~ W I z wI-~ ~ • 11 + I Y ~ T T ~TT x- x~SNL10E zX ~T T xTT x.X.x.X\~ X T/ X~\~-z~x~z\T x-X\ x-x,x \X\ ~ 5 10 15 t t 20 t t 25 30 35 6 MOUR PERIODS Effects of pentoZinizan, 10 mg/kg p.o. b.i.d. See figure 6 for details of the study. For comparison with mecamylamine, the peripheral ganglionic blocker pentolinium was tested in groups of rats self-admini- stering nicotine or saline. The data collected are shown in figure 7. The distinctive pattern of responding controlled by the light- dark cycle is clearer with these groups of rats. The separation in response rates between the two groups is clear and was statis- tically significant throughout the study. In sharp contrast to the effects produced by mecamylamine, pentolinium had no measur- able effect on either nicotine or saline self-administration. 80 'TIMN 0152446

4 These data lend much weight to the conclusion that the effects seen with mecamylamine were due to the central nervous system activity of the compound and, by extrapolation, that the locus of the reinforcing activity of nicotine is also in the central nervous system. FIGURE 8 AMPHETAMINE SULFATE 4 mg/kg p.o. b.i.d. 6 HOUR PERIODS Effects of amphetamine, 4 mg/kg p.o. b.i.d. See figure 6 for detaiZa. Amphetamine and Caffeine: Figures 8 and 9 show the effects of the central nervous system stimulants, amphetamine and caffeine, on nicotine and saline self-administration. The doses tested, amphetamine, 4 mg/kg p.o. and caffeine, 20 mg/kg p.o., produced increased activity and, in the case of amphetamine,'stereotypy within 15 minutes following the first dose and throughout the testing period. The rate of responding was significantly in- creased for the animals self-administering saline during the drug test period for both amphetamine and caffeine. No significant change in nicotine self-administration was•seen following dosing with caffeine at any interval; however, amphetamine produced a statistically significant increase in responding during the usually inactive periods when the room lights were on (i.e., the 13th, l4th, 17th, 18th, etc., 6-hour recording periods) compared with similar data collected before and after dosing. Response rates were not significantly changed during the "dark periods." 81

FIGURE 9 6 HOUR PERIODS Effects of caffeine, 20 mg/kg p.o. b.i.d. See figure 6 for detaits. These effects would appear to be due to general increases in activity and not to any interaction with the pharmacological agents studied. It is interesting to note that the rats did not increase responding (increase nicotine intake) above the control level during the normal periods of high responding (dark periods), even though the immediate effects of dosing as well as the long-term effects would have been present. It is possible that limiting side effects of nicotine may have controlled the actual maximal amount tolerated, and the rats were demonstrating some- thing akin to titration of the dose of nicotine. Ethanol, Chlordiazepoxide, Phenobarbital and Diphenylhydantoin: The data collected with ethanol, chlordiazepoxide, phenobarbital and diphenylhydantoin are shown in figures 10, 11, 12 and 13. These four compounds, although differing widely in their pharma- cological activities and tested at doses which produced observ- able sedation and ataxia, unexpectedly increased nicotine self- administration following the 3-day dosing period, compared with predosing control levels. This increased responding was not seen in the saline self-administering control groups. In the case of chlordiazepoxide and diphenylhydantoin, the increase in rates occurred soon after dosing was initiated,and in the case of diphenylhydantoin had reached levels some three times higher than the predosing control levels by the last day of the study. 82 TINLN 0152448

NICOTINE N=10 SALINE N=10 6 HOUR PERIODS Effects of ethanoZ, 905 mg/kg p.o. b.i.d. See figure 6 for detaiZs. An explanation for this unexpected effect is not readily avail- able. The fact that the saline self-administering group did not show a comparable increase suggests that simple stimulation was not a factor. Since the response rates remained high or con- tinued to increase after dosing had stopped, some fundamental change may have occurred, perhaps a change in the metabolic rate of nicotine, or possibly an increase in tolerance to the limiting side effects of nicotine. Some support is offered for this first possibility by the known ability of these compounds to modify enzymatic systems. Assuming that convulsions and the preconvul- sant state are some of the,,limiting side effects of nicotine, dosing with anticonvulsants, such as phenobarbital and dilantin, could possibly unmask the reinforcing properties of high doses of nicotine by preventing these side effects and thereby "allowing" higher dosages of nicotine to be self-administered, leading to the more rapid development of tolerance. From the data availablejnone of these four compounds could be considered to have a specific effect on nicotine self-admini- stration. 83 ; TIMN 0152449

FIGURE 11 CHLORDIAZEPOXIDE 10 mg/kg p.o. b.i.d. 6 HOUR PERIODS Effects of chZordiaxepoxide, 10 mg/kg p.'o. b.i.d. See figure 6 for details.

FTGURE 12 PHENOBARBITAL 30 mg/kg p.o. b.i.d. 00 Ul 6 HOUR PERIODS . Effects,of phenobarbital, 30 mg/kg p.o. b.i.d. See figure 6 for detaiZs. -

FIGURE 13 00 0% 5 DIPHEHYLHYDANTOIW 300 mq/kq p.o. b.i.d. 5 10 25 30 35 ~ 15 20 6 HOUR PERIODS Effects of diphenylhydantoin, 300 mg/kg p.o. b.i.d. See figure 6 for detaiZs. I N

FIGURE 14 5-1 CHLOR'PROMAZINE 2 mg/kg p.o. b.i.d. 5 10 15 20 25 30 35 6 HOUR PERIODS Effects of chZorpromazine, 2 mg/kg p.o. b.i.d. See figure 6 for detazls._

Chlorpromazine: The data collected with chlorpromazine are shown in figure 14. The dose used, 2 mg/kg p.o., produced ataxia and sedation which became more pronounced with succeeding doses. In spite of the severity of the side effects produced by the dosing regimen, there were no significant changes in the overall rates of nicotine self-administration other than disruption of the normal circadian rhythm. The saline self-administration control group,surprisingly,did not show a similar effect. The response rates for this group were, of course, very low. Normal patterns of responding were recovered 2 days after cessation of dosing; at approximately the same time sedation and ataxia were no longer noted. DISCUSSION While clearly possessing reinforcing properties, nicotine can only be classified as relatively weak in this regard when it is compared with morphine, amphetamine, or cocaine (Pickens et al. 1978). Of particular interest in the study of nicotine self-administration, and a factor which differentiates it from other compounds, is the remarkably slow onset of self-administration in the rat, comple- mented by the tendency gradually to increase the daily dosage over extended periods of exposure. This slow increase in the daily rate of nicotine self-administration could be explained as due to the development of tolerance either to the reinforcing properties or to the limiting side-effects of nicotine. Con- current administration with an appropriate dose of the peri- pherally-acting ganglionic blocker, pentolinium, might yield data of interest in this regard since at least some of nicotine's side effects, but not its actions on the central nervous system, would be blocked by this treatment. „ The difficulties experienced by Lang et al. (1977) in demonstra- ting the reinforcing properties of nicotine are easily understood considering they studied acquisition of lever pressing for only 90 hours. It would seem likely that longer periods of exposure would have yielded data comparable to those presented here. The surprising resistance to extinction generated following a short period of reinforcement by nicotine suggests that "feed- back" from this reinforcement system is poor; we would expect that partial schedules of reinforcement would be poorly supported by nicotine. To date we know of no studies of partial reinforce- ment schedules using nicotine infusions as a reinforcer. It would appear, however, that any attempt to demonstrate the control of lever pressing by partial schedules of reinforcement would be most likely to succeed if done after a significant daily intake of nicotine had been established. The implications of this unusual resistance to extinction are obvious in any comparison with human smoking and may be of use in understanding the prob- lems of withdrawal from smoking. 88 ? TIMN 0152454

7;5T7~~7S The effects on nicotine self-administration of standard compounds studied clearly demonstrate the difficulties in achieving pharmaco- logical manipulation of nicotine self-administration. Mecamylamine, which is a specific blocker of nicotine's actions, produced a dramatic but short-lived increase in nicotine self-administra- tion. Several.explanations are possible; the most likely would appear to be a direct blockade of nicotine at the receptor site in the central nervous sytem, the appropriate comparison being with the interaction of naloxone and morphine. If this is true, an effort to identify specific nicotine receptor sites would be of value. Still to be accommodated in such an explanation is the unusual resistance to extinction observed (the rats were self- administering saline in extinction, the perfect analog of "blockade" of nicotine) following a short period of nicotine reinforcement. Also to be explained is the apparent ineffective- ness of additional doses of mecamylamine in either increasing or decreasing nicotine self-administration. The constancy of the rates of nicotine self-administration following heroic doses of a stimulant, such as amphetamine, or depressants, such as chlorpromazine or phenobarbital suggests that in spite of all its apparent fragility as a reinforcer, in these studies of nicotine we were dealing with a major behavioral control which was not easily disrupted. The implications for the study of smoking are again obvious. RECOMMENDATIONS FOR FURTHER RESEARCH 1. Systematic studies of self-administration of nicotine should be extended to other species, such as the monkey, even though the rat may well remain the test animal of choice. 2. Self-administration of the main metabolites of nicotine should be studied, since it is possible the "reinforcing" properties of nicotine are shared'with those of a.metabolic byproduct. 3. To better bring nicotine into the mainstream of self-adminis- tration studies, partial schedules of reinforcement could be studied with either rats or primates. 4. While technically somewhat difficult, the study of "sensitive" schedule-controlled behaviors, reinforced by food and water concurrently with nicotine self-administration, might allow the study of the nicotine abstinence syndrome in laboratory animals. 5. Comparative studies of several representative compounds, using resistance to extinction as a measure of reinforcing "potency," might allow the ordering of compounds in terms of abuse po.- tentie.l. Techniques other than substitution or repeated extinction would probably be most fruitful. 89 TIMN 0152455

K*1 REFERENCES Clark, M.S.G. Self-administered nicotine solutions preferred to placebo by the rat. Brit J Pharmacol, 35:376, 1969. Deneau, G.A., and Inoki, R. Nicotine self-administration in monkeys. Ann NY Acad Sci, 142:277-279, 1967. Hanson, H.M.; Ivester, C.A.; and Morton, B.R. The effects of selected compounds on the self-administration of nicotine in rats. Fed Proc, 36:1040, 1977. Lang, W.J.; Latiff, A.A.; McQueen, A.; and Singer, G. Self- administration of nicotine with and without a food delivery schedule. Pharmacol Biochem Behav, 7:65-70, 1977. Pickens, R.; Meisch, R.A.; and Thompson, T. Drug self-adminis- tration: An analysis of the reinforcing effects of drugs. Vol. 12. In: Iverson, L.L.; Iverson, S.B.; and Snyder, S.H.; eds. Handbook of PharmacologJr. New York: Plenum Press, 1978. pp. 1-37. Stone, C.A.; Meckelnburg, K.L.; and Torchiana, M.L. Antagonism of nicotine-induced convulsions by ganglionic blocking agents. Arch Int Pharmcodyn Ther, 117:k19-434, 1958. Yanagita, T. An experimental framework for evaluation of de- pendence liability of various types of drugs in monkeys. Bull Narcotics, 25:57-64, 1973. Yokel, R.A., and Pickens, R. Extinction responding following amphetamine self-administration: Determination of reinforcement magnitude. Physiol Psychol, 4:39-42, 1976. AIITHORS H. M. Hanson, Ph.D. C. A. Ivester B. R. Morton Merck Institute for Therapeutic Research West Point, Pennsylvania 19486 90 ; TIMN 0152456

Chapter 8 The Effects of d-Amphetamine, Meprobamate, and Lobeline on the Cigarette Smoking Behavior of Normal Human Subjects C. R. Schuster, Ph.D., B. R. Lucchesi, M.D., Ph.D., and G. S. Emley, M.S. In the mid 1960's at the University of Michigan my colleagues, B.R. Lucchesi and G.S. Emley, and I conducted a series of experi- ments designed to investigate the effects of various drugs on the cigarette smoking behavior of normal volunteer subjects. One of our major interests was the role of nicotine in cigarette smoking behavior. The subjects used in this experiment were male and female volun- teers ranging in age from 21 to 30 years. Each volunteer was given an interview and a complete physical and psychiatric examination before being accepted as a subject. They were told that the purpose of the experiment was to determine the effects of psychoactive drugs on several types of performance tasks and to see how these behavioral changes correlated with changes in various physiological systems. At the end of the experiment the subjects were completely informed of the details of the experiment and the drugs they had been given. ~ Each subject was confined to a soundproof room for six hours per day for a minimum of 12 experimental days. Heart rate, EKG and blood pressure were continuously monitored. Smoking was permitted ad libitum and a standard breakfast and lunch were provided. Behavi.oral procedures including time estimation, hand steadiness and a vigilance or monitoring task were scheduled at various times throughout the six hour session. All experimental days had identical testing schedules. The number of cigarettes smoked during the six hour period was recorded and the residual portion of each cigarette was weighed at the end of each session. The subjects were instructed to report any subjective effects attribut- able to the drug. In the first experiment we investigated the effects of intravenously infused nicotine bitartrate. The results of this experiment have been presented in detail elsewhere (Lucchesi, Schuster and Emley 1967). In summary, this study demonstrated that doses of 2-4 mg/ hour of intravenously infused nicotine produced a small but 91 7

significant decline in cigarette smoking frequency and the amount of each cigarette consumed, as estimated by weighing the butts at the end of the day. Of great importance is the fact that at this dosage of nicotine the subjects did not report any unpleasant side effects and were unable to discriminate the drug days from saline control days. Further, the nicotine infusion produced no changes in any of the behaviors generated by the performance tasks,suggest- ing that its effect on cigarette smoking behavior was specific. We interpreted our data as suggesting that cigarette smoking was a form of drug-seeking behavior and that the drug was nicotine. T1iat the infusion of nicotine produced only small decrements in cigarette smoking behavior could readily be attributable to the fact that cigarette smoking was under strong stimulus control and that the act itself had through association with nicotine become a powerful conditioned reinforcer. Although recent experimental evidence has seriously questioned this role of nicotine (Ktunar et al. 1977),it is still my opinion that cigarette smoking is a form of drug-seeking behavior and that nicotine is the drug being sought. EXPERIMENTS WITH d_-AMPI-iETAMINE, MEPROBAMATE, AND LOBELINE In subsequent experiments, which have not been reported previously, we studied the effects of orally administered lobeline, d-ampheta- mine and meprobamate on cigarette smoking behavior. These drugs were selected because at the time these experiments were conducted they were being used alone or in combination by clinical investi- gators attempting to develop a program of pharmacotherapy for aiding smokers to reduce or cease their cigarette use. Lobeline is an a],kaloid which is the principal constituent of Lobelia,which is obtained from the herb, Lobelia inflata. Lobeline shares many of the actions of nicotine and cross tolerance has been demonstrated. Because of the pharmacologic similarity, lobeline has been suggested as a substitute for nicotine in people attempting to decrease their tobacco use (Dorsey 1936). Various commercial preparations containing lobeline have been marketed as smoking deterrents. Meprobamate and amphetamine were suggested as,_pharmacotherapeutic aids in smoking programs to alleviate the anxiety (meprobamate) or sleepiness (amphetamine) associated with the cessation of cigarette smoking. The experiments to be reported here differ from the clinical work in that we used naive subjects who were unaware of the drug being administered. They did not come to us for aid in reducing their cigarette consumption and were not aware that we were studying their cigarette smoking behavior. In contrast, in all of the clinical investigations the subjects were selected because of a desire to stop smoking. The method used in this study was identical to that used in the first.study except that the drugs were given orally rather than intravenously. All drugs were encapsulated in identical gelatin 92 ,- ! TIM,N 0152458 , _

capsules. The subjects took 2 capsules with milk just before breakfast. The d-amphetamine was given in dosages of 5.0, 7.5, and in some cases 10 mg. The meprobamate was given in dosages of 400 and 800 mg. The drug sessions were always randomized with placebo sessions. Following the testing of these drugs, the subjects were given 3 placebo capsules daily which they took with each meal for 3 consecutive days. This was followed by a 5 day chronic lobeline regimen. The lobeline capsules (2 mg/ tablet) which were made from commercially available lobeline sulfate (with antacids) were crushed, mixed with lactose,and encapsulated in a gelatin capsule. The subjects took 3 lobeline capsules per day (one with each meal) for 5 consecutive days and were tested in the experimental procedure on days 1, 3, and 5 of the chronic regimen. This was the dosage regimen suggested by the manufacturer for the use of this product. Six subjects were used in this series of studies. One subject failed to complete the lobeline portion of the study,reducing the number of subjects to five during the final stage of the experiment. Figure 1 (bottom section) shows the number of cigarettes smoked during the placebo control and d-amphetamine sessions for each subject. The placebo values are means of at least 6 sessions. The d-amphetamine data represents individual sessions. This illustrates that for subjects 2, 3, 4 and 6 there was a significant increase in cigarette smoking frequency during the d-amphetamine sessions. The upper graph shows the corresponding residual weights of the cigarettes. It can be seen that the amount of each cigar- ette consumed was not consistently affected by d-amphetamine. On the other hand, there was no decrease in the amount of each cigarette consumed,at least as revealed by the weight of the cigarette butts. Figure 2 shows the corresponding data for meprobamate sessions and placebo sessions. As can be observed, there was no significant change attributable to meprobamate at the dosages of 400 and 800 mg in cigarette smoking frequency or.the residual weights of cigarettes consumed. Figure 3 shows the effects of chronic lobeline on cigarette smoking. Lobeline did not produce a decline in the number of cigarettes smoked or the amount consumed in any of the 5 subjects tested. In fact, subject 1 actually showed a sizable increase in the number of cigarettes smoked as well as the amount consumed on day 5 of the chronic lobeline regimen. From these data it can be concluded that only d-amphetamine had any significant effect upon cigarette smoking Trequency under the present experimental conditions. At the dosage levels employed, d-amphetamine produced both physiological and behavioral effects. 93 TIMN 0152459

FIGURE 1 ~. t N ti. .500 .400 .300 20 16 12 8 6 4 ~ z 0 ff Placebo O d-Amphetamine I Standard Error a f 2 3 4 5 6 Subject The effects of various doses of d_-amphetamine on the cigarette smoking behavior of six subjects. The top paneZ shows the average weight of the unsmoked portion of the cigarette and the Zmver paneZ shows the number of cigarettes smoked. The data for the pZacebo condition represent the average of six sessions and the verticaZ brackets indicate the standard error. The data for the drug sessions represent the singZe determination at each dose ZeveZ. 94 TIMN 0152460

FIGURE 2 600 c rn 3 E .500 _ a ~ o r .400 Ul ~ d ~ .300 22 m 18 . ~ Placebo Meprobamate I Standard Error I 3 4 5 6 Subjects The effects of various doses of ineprobamate on the cigarette smoking behavior of six subjects. The top paneZ shows the average weight of the unsmoked portion of the cigarette and the Zmver paneZ shows the nwnber of cigarettes smoked. The data for the pZacebo condition represent the average of six sessions and the verticaZ brackets indicate the standard error. The data for the drug sessions rep- resent the s'ingZe determination at each dose ZeveZ. 95 cTINi~ 0152461

FIGURE 3 22 r a 18 d a Q Placebo p Lobeline I Standard Error N L u_ 2 L= 4 Subjects W 40 The effects of repeated administration of ZobeZine on the cigarette smoking behavior of five subjects. __Ae top paneZ shows the weight of the unsmoked portion of the cigarette and the Zewer paneZ shows the number of cigarettes smoked. PZacebo sessions represent an average of six sessions and the ZobeZine data are for sessions conducted on days 1, 3, and 5 of the chronic dosing regimen. 96

Figure 4 shows the mean heart rate under placebo conditions and for the highest dosage of d-amphetamine (7.5 or 10 mg). It is clear that on days when suEjects received these doses of d- amphetamine,heart rate was consistently elevated over the 6 hour session. In addition systolic blood pressure showed a small but consistent increase over the 6-hour session. These effects probably represent a combination of the sympathomimetic actions FIGURE 4 C 110 0 Is 100 .M 90 C 0 70 a: - Placebo ---- Amphetamine (highest dosage) ~ Standard Error 50 I I I ` I I L Resting I 2 3 4 5 Hour Intervals The effects of d-cnnphetrnnine on heart rate over the six-hour session averageU for the six sub,jects. The pZacebo data represent an average of six sessions for each subject. The drug data are derived from a singZe session in which the highest dose of d-ampheta- mine mas administered. - 97

of d-amphetamine and the effects of increased nicotine intace, since the subjects were smoking more frequently. It is of importance to note that d-amphetamine also produced increased rates of responding on the vigilance task. This suggests that the actions of d-amphetamine on smoking frequency were not specific. In contrast, neither lobeline nor meprobamate produced any changes in heart rate, blood pressure,or behavior generated by the perfor- mance tasks. Further, with the exception of drowsiness reported by one subject at the 800 mg dose of meprobamate, subjects did not report any subjective effects associated with the administration of any of the drugs. If this study had been conducted more recently, a number of additional measures could have been obtained. Clearly, blood nicotine levels would provide important additional data. For example, it is conceivable that the drugs employed in this study might have altered puff frequency and/or duration. Further, the depth of inhalation of smoke might have been changed by these drugs. Measures of blood nicotine levels would be the ideal way to resolve these questions. Unfortunately, at the time these experiments were carried out, the methodology for the measurement of nicotine levels in blood was not available. CONCLUSIONS Despite the limitations of this study, I believe we can still reach some important conclusions. First, lobeline sulfate at the recommended dosages does not produce any effect upon cigarette smoking behavior under the same conditions where nicotine does. It thus seems unlikely that lobeline has comparable pharmacologic actions allowing it to substitute for nicotine. The data on d-amphetamine as well are of interest for theoretical reasons. It-has been suggested that the mechanism underlying the reinforcing action of nicotine is its ability to produce EEG arousal through its actions on the ascending reticular activating system (Jarvik 1970). If this is the case, one would assume that d-amphetamine should substitute for nicotine since it as well produces EEG arousal through stimulation of the ARAS. The fact that amphetamines cause an increase in cigarette smoking frequenc clearly does not support this hypothesis. The fact that meprobamate failed to alter cigarette smoking fre- quency does not necessarily indicate that anxiety is not a factor in the control of cigarette smoking frequency. The efficacy of meprobamate as an anti-anxiety agent is highly questionable. Clearly, studies with more efficacious drugs (e.g., the benzodi- azepines) are necessary. One final observation concerns the stability of cigarette smoking. frequency observed in the subjects of these experiments under tht placebo conditions. The small size of the standard errors bracke 98 TIIV11v 0152464

I ing the placebo averages indicates that there was little variance in both the measures of cigarette smoking frequency and the amount of each cigarette consumed. Observation of the times when subjects smoked indicated that this occurred with high probability during "break" times from the performance tasks. It would be of great interest to determine in future research whether smoking frequency could be altered by manipulation of the timing of these "break" periods. This would help to clarify the relative importance of pharmacologic and environmental control over cigarette smoking patterning and frequency. ACI{1'YOWLEDG`dEN'T This study was supported by a grant from the American Nifedical Association Research and Education Foundation, Committee for Research on Tobacco and Health. REFERENCES Dorsey, J.L. Control of the tobacco habit. Ann Int Med, 10: 628-631, 1936. Jarvik, M.E. The role of nicotine in the smoking habit. In: Junt, W.A., ed. Learnin Mechanisms in Smoking. Chicago: Aldine, 1970. pp. 155-190. Kunar, R.; Cooke, E.C.; Lader, M.H.;and Russell, M.A.H. Is nicotine important in tobacco smoking? Clin Pharmacol Ther, 21 (5):520-529, 1977. Lucchesi, B.R.; Schuster, C.R.;and Emley, G.S. The role of nicotine as a determinant of cigarette smoking frequency in man with observation on certain cardiovascular effects associated with the tobacco alkaloid. Clin Pharmacol Ther, 8:789-796, 1967. AUTHORS Charles R. Schuster, Ph.D. The University of Chicago Pritzker School of Medicine Departments of Psychiatry and Pharmacological and Physiological Sciences 950 East 59th Street Chicago, Illinois 60637 B.R. Lucchesi, M.D., Ph.D. The Uriiversity of Michigan Department of Pharmacology Ann Arbor, Michigan 48104 G.S. Emley, M.S. Foundation for Behavioral Research Augusta, Michigan 49012 99 TIMN 0152465

Chapter 9 Tobacco Dependence: Is Nicotine Rewarding or Aversive? M. A. H. Russell, M.B., MRCP, MRCPsych To understand tobacco smoking is.like doing a large and very diffi- cult jigsaw puzzle. I shall focus on the role of nicotine in tobac- co use, partly because it may be a kind of straight edge to the puz- zle which we can usefully begin to piece together without getting lost among all the pieces that go in the middle. But the role of nicotine interests me for another very practical reason. If it is an important factor, and if most people do smoke chiefly to dose themselves with nicotine, the present low-tar, low-nicotine approach to safer cigarettes is obviously wrong, and it would be more logical to develop low-tar, medium-nicotine (or even high-nicotine) ciga- rettes (Russell 1976a). We know that tobacco smoking is highly dependence-producing. We know that nicotine has numerous pharmacological effects, peripherally and in the brain, which are potentially reinforcing. We know that it has effects on behavior and performance which are potentially rein- forcing. We know that it induces tolerance in animals and that they become dependent on nicotine injections to maintain performance. We know that most smokers absorb nicotine in amounts sufficient to produce these rewarding effects, that they acquire tolerance to some of its actions, and that-they suffer from phys'ical as well as psycho- logical withdrawal effects when they stop smoking. We also know that the rapid absorption of nicotine through the lungs produces intra- venous-like high-nicotine boli in the blood which reach the brain within a few seconds after each inhaled puff. The possibility that this produces a rapid puff-by-puff pharmacological reinforcement in the reward centers in the brain some 70,000 times a year would go a long way toward explaining the high dependence-producing potency of cigarette smoking (Russell 1976b). If we could prove that nicotine is what smokers seek, we could be confident that the puzzle was virtually completed. Unfortunately this is not the case and we cannot escape the nagging fact that powerful addictive syndromes occur where pharmacological factors clearly play no part. Qie does not have to look far for examples such as gambling, nail-biting, and the desire for sweet tastes 100

or high-fat and high-cholesterol foods. With tobacco smoking there are numerous nonpharmacological components which could make it equally dependence-producing. Nicotine titration studies have attempted to prove that people smoke for nicotine. But, as I shall go on to show, these data do not disprove (indeed they can be better fitted to support) the hypothesis that people smoke and inhale for nonpharmacologic effects such as taste, aroma, sensorimotor ritual, and the local irritation by nicotine and other components of tobacco to sensory receptors in the lungs and respiratory tracts, and that the pharmacological effects of nicotine are aversive and tend to inhibit inhalation rather than reinforce it. Definition of "Dependence" and "Addiction" If we agree that tobacco smoking is for many a form of dependence, what does this mean? There are some who restrict the term "addiction" to compulsive syndromes which are maintained largely by the need to relieve or avoid physical withdrawal effects. The issue of "phys- ical" versus "psychological" dependence is a somewhat false dichotomy, as the two are so interwoven. I use the terms "dependence" and "ad- diction" interchangeably to refer to a state in which the urge or need for something is so strong that the individual suffers or has great difficulty in going without it, and in extreme cases cannot stop doing it, or using it, when it is available. How high a degree of dependence is required before the condition is labelled a "de- pendence disorder" or "addiction" is somewhat arbitrary. I would like to emphasize one point. It is the strength of the urge or need which is important, not whether it is predominantly pharma- cologically or nonpharmacologically determined. Strong social or psychological rewards can produce a higher degree of dependence than weak pharmacological ones. To what extent, then, does tobacco smoking depend on pharmacological as opposed to social or other nonpharmacological reinforcement? And to what extent do smokers smoke for positive rewards (pharma- cological, nonpharmacological or acquired by conditioning) as op- posed to seeking relief from, or avoidance of, unpleasant withdrawal effects (pharmacological, nonpharmacological or conditioned)? In other words, how much is smoking maintained by positive versus negative reinforcement? Expressed Nbtives of Smokers Many smokers find it difficult to explain why they smoke. Some will say they smoke because they find it pleasurable, that it helps them to relax, or simply that they are addicted. There are obvious pitfalls to accepting such statements at face value. When people find themselves doing something frequently and to some extent against their better judgment, they may tend to attribute this to pleasure, addiction, or some such "rational" explanation. They are, however, unable to tell us what makes the behavior pleasurable, re- laxing, or addictive. Nevertheless, the statements of smokers'them- 101 TIMN 0152467 ,,

selves are an essential starting point in determining the motives underlying the dependence. A number of people, such as Horn in the U.S.A. (Ikard, Green and Horn 1969) and McKennell (1973) in Britain, have used factor analysis to make systematic studies of the responses of smokers to questions on motives for smoking. We have done a similar study (Russell, Peto, and Patel 1974) using a 34-item questionnaire combining various as- pects of the earlier work of Horn and McKennell. We obtained six factors representing various motives for smoking: psychosocial, in- dulgent, sensorimotor, stimulation, addictive, and automatic. Un- like previous studies, we did not obtain a sedative smoking factor. Items designed to form a sedative factor loaded instead on the addic- tive and stimulation factors. It is not intended here to discuss the factor structure, but it may be that the negative affect and agitation which smoking apparently sedates are generated by cigarette withdrawal, and that in such cases "sedative" smoking is withdrawal relief smoking rather than smoking for sedation of negative affect due to other causes. The most striking finding of our study was the clear-cut separation of the factors and their items into two distinct clusters, which we interpreted as representing pharmacological and nonpharmacological mo- tives. This is shown in figure 1 where the items have been plotted according to their loadings on the first two unrotated factors, which respectively accounted for 18 percent and 8 percent of the variance. The "pharmacological" cluster was comprised of the addictive, auto- matic, stimulation, and sedative smoking items. The psychosocial, sensorimotor, and indulgent factor items were all in the "nonpharma- cological" cluster. Some validation is provided by the correlation of the "pharmacological" group of factors with cigarette consump- tion; it was these.factors that also differentiated a criterion group of addicted heavy smokers attending a withdrawal clinic from the main sample of normal smokers (figure 2). The questionnaire has since been modified and further validation will be sought by examining the relation of factor scores to blood nicotine and COHb levels and to clinical outcome. Pharmacological studies have'shown that in smoking doses nicotine is predominantly a stimulant drug but that it also has some sedative actions (Russell 1976b . The self-report data of smokers suggested similar effects, and t is led us to label the two main item clusters as "pharmacological" and "nonpharmacologica.l." The addictive fac- tor items were concerned with craving and the relief of withdrawal symptoms, or, in other words, with negative reinforcement smoking. The high correlation between the positive effects of stimulation and the negative addictive factor items suggests that once sufficient nicotine is taken in to provide stimulation, withdrawal effects are likely to occur. This does not appear to ha pen to the same extent with indulgent smoking (smoking for pleasure). How much the pleas- urable aspects depend on nicotine is obviously an important question. As a final caution, it should be emphasized that stimulation, seda- tion, and withdrawal relief smoking do not necessarily indicate 102 TIMN 0152468

4, . . , m 7,..y,4L- ._ r - _ •5 r FIGURE 1 Loadings of 34 items on the first two unrotated factors IL 26P• •211 281 •4 11 P• •3 5Sm. •2 ' •8P .2Sm •25 •41 .18P •151 •34P 3Sm• * 24Adi14 ~30P .1 •27Sm . 13 Sm •29Sd L -•1 •1 •2 •3 •4 5 •6 •7 8 -1 20St• •17Ad•9Ad 23St• ~6Sd 19A. •31Ad •12St •16St -2 •1 •10A -3 •32A -4 ~ . 7St . 33 St -5 L St: Stimulation, I: Indulgent, P: Psychosocial, Sm: Sensorimotor, Ad =Addictive, A=Automatic, Sd :Sedative. Loadings of 34 questionnaire items on the first two unrotated fac- tors. The upper Zeft cZuster is comprised of "nonpharmacoZogicaZ" motives for smoking from the psychosociaZ, sensorimotor, and induZ- gent smoking factors. The "pharmacoZogicaZ" items representing stim- ulation, sedative, addictive, and automatic smoking are aZZ in the Zoroer right cZuster. (Reprinted with permission from RusseZZ, M.A.H.; Peto, J.; and PateZ, U.A. The cZassification of smoking by factoriaZ structure of motives. JournaZ o~ the R~ Stati~sticaZ Socie Series A, 137:313-333, 1974. ©1974, RoyaZ StatzsticaZ Soczety.). 103 `TIM.N 0152469

3•0 FIGURE 2 Mean factor scores by sex for'Mairi' &'Smoking Clinic sampies n FEMALES Smoking ' ciinlc i MALES sample i FEMALES) . ~ ~'A~ / LMain ; / rsample . ' . / * MALES / . , y_ / / 0 1 1 I 1 I I I I FACTOR 1 FACTOR 2 FACTOR 3 FACTOR 4 FACTOR 5 FACTOR 6"Sedative' Stimulation Indulgent Psychosocial Sensorimotor Addictive Automatic Mean factor,scores by sex for "Main" sample of normal smokers (N=175) and "Smoking Clinic" sample (N 103). There is no difference between the samples in scores on the indulgent, psychosociaZ and sensorimotor factors, -but the differences on the other factors are highly significant (p<.001). (Reprinted with permission from Russell, Peto, and PateZ, 1974. Q 1974, Royal Statistical Society. See citation, figure 1.)

pharmacological mediation. It is possible that a similar factor structure with the same two main clusters of items could have been obtained by applying a similar questionnaire to regular gamblers. Nevertheless, the evidence so far available from this kind of ap- proach suggests that smoking depends on a mixture of pharmacological and nonpharmacological motives and that pharmacological motives dominate in the case of addicted heavy smokers.. Onset and Maintenance of Smoking Daniel Horn (this volume) has discussed the factors involved in the establishment and maintenance of the smoking habit -two quite dis- tinct processes. He has perhaps paid insufficient attention to one outstanding British study by Bynner (1969, 1970) which greatly clari- fies the process of recruitment to smoking in schoolboys. It is also worth emphasizing the finding of McKennell and Thomas (1967) that of those teenagers who smoke more than a single cigarette, only 15 percent avoid going on to become regular dependent smokers. Why this escalation should be almost inevitable may be explained in the following way: The first few cigarettes are usually unpleasant, but skill is quickly acquired to limit the intake of smoke to a com- fortable level. Tolerance to the unpleasant side effects of nico- tine soon develops, thus lowering the threshold for further attempts. If the psychosocial rewards are sufficiently strong to cause the act to be repeated in the face of the side effects and physical dis- comfort, there is little chance that it will not continue, as the side effects rapidly disappear. The factors controlling the onset of smoking are summarized in figure 3, and those determining its maintenance are shown in figure 4. In the early stages, smoking is intermittent and is usually confined to social situations. A few remain occasional social smokers for many years. But, in the majority, consumption gradually rises and ceases to be confined to social situations. After a few years smoking occurs with great regularity. Inhalation also increases gradually in most smokers until nicotine is absorbed in sufficient quantities to exert nuneerous potentially rewarding pharmacological effects. It is not known, however, whether the establishment of inhalation coincides with or determines the development of a regu- lar smoking pattern. Many smokers then progress to a further stage of smoking for a predominantly negative reason: avoidance or relief of the effects of withdrawal. What proportion of smokers reach this stage, how long it takes, and to what extent the withdrawal effects are mediated by pharmacological as opposed to nonpharmacological factors are unknown. Noninhaled Smoking Some cigarette smokers do not inhale and consequently absorb little nicotine. Systematic comparison of noninhalers and inhalers could throw much light on the importance of pharmacological vs. nonpharma- cological factors, but to my knowledge this has not been done. Apart from novices and occasional social smokers, some heavy (two-pack-a- day) smokers do not inhale; this is confirmed by low plasma nicotine 105 - ~-- ::.I TIMN 0152471

FIGURE 3 Parental attitude School attitude Health risks Sensory discomfort Nicotine effects Non-smoker Availability of cigarettes Curiosity Rebelliousness To appear tough Anticipation of adulthood Social confidence Parental example Older sibs smoking Friends smoking Smoker The main psbchosociat factors determining the onset of smoking. On the right are the positive reinforcers or incentives to smoke. An- ticipation of aduZthood inctudes an impatience to be grown up and a tendency to participate in the activities of oZder teenagers, such as drinking, taking an interest in the opposite sex, going to dances and coffee bars, and staying out Zate. On the ZefB are the factors that discourage smoking, two of these-the sensory discomfort of the first few cigarettes and the unpteasant side effects of nico- tine-soon disappear as the act is repeated, thereby Zowering the threshoZd for further attempts. (Reprinted with permission from RusseZZ, M.A.H. The smoking habit and its cZassification. The Practitioner, 212:791-800, 1974. © 1974, The Practitioner. 106 ~TIMN 0152472

FIGURE 4 Health E xpense Social pressure Mastery Aesthetic Example Stop smoking Psychosocial rewards Sensorimotor rewards Pleasure Sedation Stimulation Withdrawal relief/avoidance + # Smoker inki llC0ntuedsm0n9I Factors controZZing the maintenance and discontinuance of smoking. For most peopZe beZow middZe age, the factors motivating conttinued smoking are stronger than the motives to stop. The "Mastery" motive is the wish to show wiZZ power and gain controZ over the habit. "Aesthetic" refers to the feeZing that smoking is dirty and messy, whiZe "ExampZe" concerns those who wish to stop to set a good exam- pZe to chiZdren or other impressionabZe groups. (Reprinted with permission from RusseZZ, M.A.H. The smoking habit ar2d its eZassi- fieation. The ~Prac~titioner, 212:791-800, 1974. 01974, The Practitzo er.l 107 TIMN 0152473

and COHb levels (figure 5). Such smokers may, however, be highly dependent in terms of craving and difficulty in giving up smoking. We do not know how many cigarette smokers are noninhalers, nor do we know whether they have lower ratings of dependence (with or without control for daily consumption). To study this would require measures of COHb or expired air CO, for smokers themselves do not reliably know the extent to which they inhale. FIGURE 5 Time (min) PZasma nicotine concentrations in an inhaZing smoker and a noninhaZ- er during and after smoking one cigarette. The inhaZer smoked onZy one pack a day and had ZittZe difficuZtb•giving up smoking. The noninhaler smoked more than three packs a day and suffered intense craving when he tried unsuccessfuZZb to quit. (Reprinted from Fegerabend, C.; Levitt, T.; and RusseZZ, M.A.H. A rapid estimation of nicotine in bioZogicaZ fLuids. JousmaZ of Pharmacy and Pharma- co2o 27: 433-436, 1975, with permiss n. Q 1975, JourngZ ~ P-;iar~~ac_r ~ac_r~ and PharmacoZoAy. ) 108 TIM 015 414

Though less rapid than absorption through the lungs, it is generally held that with snuffing, tobacco-chewing, and noninhaled pipe and cigar smoking,absorption of nicotine through the nose or mouth is sufficient to provide a pharmacologically rewarding effect. The role of pH in determining the rate of absorption is crucial (Russell 1976b). Nicotine-containing chewing gum with a buffer to keep the pH in the mouth at about 8.5 produces plasma nicotine levels com- parable to inhaled cigarette smoking, but absorption is much slower (Russell, Feyerabend, Cole 1976; Russell et al. 1977). The few snuff users we have tested have had high plasma nicotine levels. But what about noninhaled pipe and cigar smoking? Although the pH of air-cured tobacco smoke is alkaline (about pH 8.5), is the buffering capacity of the smoke sufficient to bring the pH of saliva in the mouth up to this level? I am not aware of any studies which show this. There is, however, one recent English study which suggests that nicotine absorption from noninhaled cigar smoking may be minimal (Turner, Sillett, McNicol 1977). TABLE I Blood Nicotine and COHb Levels from Cigar Smoking in Primary and Secondary Pipe and Cigar Smokers* Before 20 min 40 min 60 min 120 mi•n NtCOTINE(nglml) 10 (N=5) • 3.3 4.0 4.3 5.1 4.5 20 (N-5) 12.8 30.7 45.6 36.2 24.7 COHb ($) 10 (N-S) 0.8 1.0 1.0 0.9 1.0 2(Ns5) ~ 2.9 6.6 8:4 9.6 8.1 The cigar was discarded at 60 minutes. Abstracted from Turner, J.A.M.; SiZZett, R.W.; and McNichob, M.W. Effect of carboxyhaemogZobin and plasma nicotine concentrations in primary pipe and cigar smokers and ex-cigarette smokers. British Medical Journa2 2:1387-1389, 1977. Q 1977, British Medical Jour- naZ. Repro~ by permission. *Primary pipe and cigar smokers are those who have never been regu- Zar cigarette smokers. Secondary pipe and cigar smokers are ex-cig- arette smokers who have saitched to a pipe or cigars, which they then tend to inhale (CastZeden and CoZe, 1973; Cowie, SiZZett and BaZZ, 19?3). 109 .-a

Turner and his colleagues measured plasma nicotine and COIib levels of smokers before, during, and after the smoking of a large cigar (12.4 cm, 6.2 g). The results in table 1 show clearly that the noninhaling primary pipe and cigar smokers absorbed virtually no nicotine. The authors claim that these smokers were nevertheless addicted in terms of craving and suffering when they could not smoke. The authors believed that their study seriously challenges the role of nicotine in tobacco dependence. They did not, however, make systematic ratings of dependence. True primary pipe and cigar smokers are an atypical minority of the smoking population. The fact that ex-cigarette smokers who switch to a pipe or cigars often continue inhaling is an indication that, once experienced, the in- halation of tobacco smoke is difficult to forgo. Though further study is obviously necessary, it would seem that strong dependence may occur in cigarette smokers, pipe smokers, and cigar smokers who do not inhale and who consequently absorb little nicotine. This suggests that tobacco dependence can be mediated by nonpharmacological factors such as taste, smell, sen- sorimotor ritual, and the like. Indeed, nicotine itself may con- tribute to the flavor, sharpness, irritancy, etc., and may be rewarding for these effects apart from its pharmacological actions. Nicotine Intake from Cigarette Smoking It comes as no surprise to find wide individual variation among smokers in blood nicotine levels just after a cigarette. Values range from below 10 ng/ml to more than 50 ng/ml, with an average for heavy smokers of around 35 ng/ml. Neither is it unexpected that the individual smoker tends to obtain a fairly consistent peak nicotine level after each cigarette, whether it is smoked in the morning or afternoon, from one day or week to the next. Of the majority of smokers, who do inhale, we are nowhere near the stage of knowing whether they smoke for some positive effect of the blood nicotine peaks, whether they smoke to avoid the pharmacological effects of falling below a certain blood nicotine trough, or indeed whether the nicotine intake and its pharmacological effects are merely incidental to inhalation which is determined by nonpharmaco- logical factors. It would seem from preliminary pharmacokinetic findings (Russell and Feyerabend, 1978) that the predominant blood profile for inhalers who smoke one cigarette per hour or less is one of repeated high nicotine peaks (figure 6), whereas the accumulation of nicotine in the body of those who smoke at least one cigarette every 30 minutes would tend to show smaller peaks relative to the overall level (figure 7). They might more likely, therefore, be motivated by the need for "trough maintenance." Very tentatively I suggest that if nicotine has a pharmacologically reinforcing role, trough maintenance may be the main motive for the addicted smoker, while opti.mal peak effects may be more important to indulgent smokers, wiio smoke less heavily. The peak blood nicotine levels of the two types, however, do not appear to differ greatly, and peak blood nicotine does not correlate with cigarette consumption, r = .02 (Russell et al. 1975 and 1977). 110 _ -___ ~1524~6 TpA1~ -

FIGURE 6 50 E 0 a IO~ 0 I • O4 10 II 12 13 14 IS 16 Time (hours) Cigarette smoked t } t f t t t PZasma nicotine concentrations in an inhaZing smoker whiZe smoking at a rate of one cigarette per hour for seven hours. AZthough there is some aecumuZation of nicotine, the predominant profiZe is one of prominent bZood nicotine peaks foZZozaing each cigarette. This iZ- Zustrates the characteristic pattern of the "peak seeker" type of smoker mentioned in the text. (Figure reproduced with permission from RusseZZ, M.A.H.; Feyerabend, C.; and CoZe, P.V. PZasma nico- tine ZeveZs after cigarette smoking and chewing nicotine gum. British MedicaZ JournaZ, 1:1043-1046, 1976. 6)1976, Fritish Medi- ca~JournaZ.1 111 T jMN 0152477

1 60 -1 50 1 0 u c 30 -I 20-I s   : .  2   1 • ° cigarette smoked) FIGURE 7 L ~ ~ i ~ 3.  . :   ~   :  ' Time (h) PZasma nicotine concentrations in an inhaZing smoker whiZe smoking at a rate of three cigarettes per hour for seven hours. BZood sam- pZes were taken just before and two minutes after each cigarette. The bZood nicotine buiZds up to a "steady state" with peaks which are smaZZ reZative to the overaZZ ZeveZ. This iZZustrates the characteristic pattern of the "trough-maintainer" type of smoker '" suggested in the text. (Reprinted from RusseZZ, M.A.H., and Peyerabend, C. Cigarette smoking: a dependence on high nicotine boZi. Drug MetaboZism Reviews, 8:29-57, 1978, by courtesy of MarceZ Dekker, Inc., d 1978.) 112 TIMN 0152478

T r It is tempting to postulate a three-stage process, with smokers be- ginning to smoke for psychosocial reasons, the majority then learn- ing to inhale and progressing to smoking for the effects of the sharp blood nicotine peaks that follow each cigarette (peak-seekers), and some finally going on to a third stage of negative reinforcement smoking to avoid the withdrawal effects of dropping below a certain blood nicotine level (trough-maintainers). As mentioned, however, the issue is far more complex: Noninhalers may also progress to a second stage of regular indulgent smoking for positive nonpharmacological rewards, as well as to a third stage of negative withdrawal relief smoking mediated by nonpharma- cological factors. Further complexities arise from interactions between pharmacological and nonpharmacological components, due to conditioning and other processes. Plausible as all this may seem, with many resemblances to other drug addictions, it has yet to be established that smokers do indeed smoke for the pharmacological effects of nicotine rather than for nonpharmacological reasons. Use of Low-Tar, Iow-Nicotine Cigarettes What can be learned from studies of the use of low-nicotine ciga- rettes ? Among English cigarette smokers, only 10 percent of men and 18 percent of women regularly smoke low-nicotine brands. Most low-nicotine brands smoked in Britain have standard deliveries of 0.6 to 0.8 mg nicotine; yet it is still possible to maintain high blood nicotine levels, of 40 ng/ml or more, by smoking these brands. It is only when the nicotine yield is reduced to 0.4 mg or less that this becomes difficult, and relatively very few people smoke ciga- rettes of this type. The lack of popularity of low-nicotine cigarettes may seem to provide evidence of a need for nicotine. Similarly, the tendency when smoking them to smoke more cigarettes or take larger and more frequent puffs and inhale more deeply has been interpreted as reflecting a need to titrate nicotine. There is a serious objection to this interpretation, however. Iow-nicotine cigarettes also have low yields of other consituents, such as tar, which contribute to the taste and satisfaction of the smoke. Indeed, they are often also difficult to light and to smoke. Regulation of Nicotine Intake Self-regulation of nicotine intake, or nicotine titration, is probably the most useful approach for studying the role of nico- tine in smoking. It has been comprehensively discussed by Schachter (this volume). As he points out, there are numerous studies showing that smokers modify their smoking patterns to compensate for a reduction of the tar and nicotine yields of their cigarettes. But, as mentioned above, owing to the high correlation (~>0.9) of tar and nicotine yields, we cannot know 113 TI~~ 0152479

I from such studies whether the smokers are regulating their intake of nicotine rather than tar or some other factor. Qily one study, by Goldfarb et al. (1976), has used cigarettes in which tar and nicotine yields were independently varied. The num- ber of cigarettes smoked was unaffected by tar yield but varied inversely with nicotine yield, although insufficiently to maintain urinary nicotine excretion constant. While most studies have shown that compensation on low nicotine cigarettes is often only partial (Russell 1976b, Sutton et al. 1978), we have found that smokers lower their intake very accurately when switched to high nicotine cigarettes. Blood nicotine levels on cigarettes yielding 3.2 mg nicotine averaged 29 ng/ml, compared to 30 ng/ml on cigarettes yield- ing 1.3 mg nicotine (Russell et al. 1975). It seems, therefore, that that self-regulation downwards to avoid an excessive intake may be more sensitive and complete than compensation upwards to avoid a reduction in nicotine and/or tar intake. There have been very few direct studies of nicotine titration, i.e., those which avoid confoundment with tar intake and other factors. They have, however, with one exception, been most convincing in their support for nicotine titration. For example, Stolerman et al. (1973) showed that central cholinergic blockade with mecamylamine caused smokers to increase the number of cigarettes smoked by 26.5 percent and the number of puffs by 25.3 percent compared to placebo. In the study by Iucchesi et al. (1967), intravenous nico- tine infusion reduced the number of cigarettes smoked by an average of 29.8 percent compared to saline. Furthermore, 22 mg of IV nico- tine bitartrate (about 7.3 mg base) caused a mean decrease of about 2.5 cigarettes during the 6-hour infusion period. The nicotine de- livered by 2.5 cigarettes in the mid-1960's, though not stated by the authors, would have been about 7 mg. Titration of similar accuracy was demonstrated during Schachter's manipulation of urinary nicotine excretion via its pH (Schachter, Kozlowski and Silverstein 1977). It would take too long to go fully into our own failure to replicate Iucchesi's findings (Kumar et al. 1977). This could have had something to do with the extreme artificiality of our ex- perimental conditions. Although our subjects continued to puff at baseline levels during and just after the IV nicotine, it is possible that they titrated by inhaling less. We did not measure blood nico- tine levels, so we do not know whether or not this was so. The sensitivity of the nicotine titration demonstrated in these more direct studies is in keeping with the view that self-regulation of nicotine intake downwards is more sensitive and complete than com- pensation upwards. One other direct approach has been to study the effect on smoking behavior of nicotine-containing chewing gum. Trm studies have shown a modest inhibitory effect (Kozlowski, Jarvik and Gritz 1975; Russell et al. 1976). In our study (table 2), a dose of 20 mg nicotine per day taken in the gum reduced cigarette consump- tion by 17 percent (p <.05) and CnHb by 41 percent (p <.01) compared to placebo gum. About 90 percent of the nicotine in the gum is re- leased, but much of this is no doubt swallowed, absorbed in the gut, 114 I ' TIMN 0152480

TABLE 2 Effect of Nicotine Chewing Gwn on Plasma Nicotine Levels During ad Zibitwn Smoking (means of 41 subjects) No gum Placebo gum Nicotine gum Cigarettes per day 33.3 23.0 20.9 M ($) 8.5 7.2 6.3 Plasma nicotine (ng/ml) 30.1 24.7 27.4 Note: One piece of gum containing 2 mg nicotine was chewed hourly for 30 min to a'total of ten pieces per day. The specific effect of nicotine (as opposed to the effect of placebo gum) accounted for 17% of the reduction in daily cigarette consumption (23.0 20.9) =(33.3-20.9) x 100 = 17. By similar calculation it accounted for 41% of the reduction in COHb and hence degree of inhalation. Adapted from RusseZZ, WiZson, Feyerabend, and CoZe. Effect of nicotine chewing-gum smoking behaviour and as an aid to ciga- rette withdrarvaZ. British MedicaZ JournaZ, 2:391-393, 1976. Qc 1976, British Me iec~ a~ JournaZ. Repr- z'nted by permission.) and metabolized in the liver before reaching the systemic circulation. It is not known how much of the 20 mg daily dose would have been absorbed through the mouth. Nevertheless, despite the extra nico- tine from the gum, the average plasma level of 27.4 ng/mg (table 2) was not significantly different from the base-line level of 30.1 ng/ml when smoking normally without gun,,showing once again fairly accurate downward titration. On the placebo gLmi, the small decrease in plas= ma nicotine (p G 05) suggests that a drop in plasma nicotine may be better tolerated than an excess. Adaptation and Compensation Theoretically, a smoker may adjust to a dilution of nicotine (and tar) content of smoke either by adapting to a lower dose or by com- pensating to maintain intake, smoking more cigarettes or increasing the intensity of puffing and inhalation. The degree of compensation may range from nil, to partial, to complete. The reduction of nico- tine (and tar) intake to which adaptation is required will obviously vary accordingly. Adaptation to a lower dose will only be complete when any negative affect or discomfort has subsided. The time courses of adaptation and of compensatory changes in smoking pattern 115 TIMN 0152481

may differ, and they may also interact. For example, the degree of compensation may be reduced as adaptation to a lower intake reduces the urge for a higher intake. Increase in the intensity of puffing may generate discomfort of a different kind (from that generated by the reduced intake), to which adaptation of a different kind may or may not occur. For example, to increase puff volume from 40 ml to 60 ml to maintain smoke intake per puff may involve little effort, but an increase from 60 ml to 80 ml may make the smoker aware of some awkwardness or discomfort. There are further complexities. For example, a smoker may be able to take a larger puff of diluted smoke, sufficient to maintain nicotine (and tar) dosage per puff, but the dilution may reduce impact and flavor below an acceptable level. Response to such interactions will vary, depending on the degree to which the smoker requires to maintain nicotine (and tar) intake on a per puff basis or on a per cigarette or per unit of time basis. Furthermore, the unit of time may refer to the time taken to smoke one cigarette or to longer periods in which several ciga- rettes are smoked. Most studies of smokers' responses to changing to low-nicotine ciga- rettes have been focused on demonstrating a significant compensatory change in a group of smokers. They have been less concerned with individual differences and have not attempted to measure the degree of compensation. We have recently examined the extent to which smokers compensate for the dilution of smoke produced by ventilated cigarette holders (Sutton et al. 1978). There were no changes in the number of cigarettes smoked. However, measures of plasma nico- tine and CIHb showed, on average, partial compensation of about 40 percent on Holder 2 (60 percent dilution), but on Holder 1 (20 per- cent dilution) there was no significant compensation (see figure 8). Individual differences were marked, ranging from full compensation to none at all. About half the subjects were clear-cut noncompensa- tors. A similar 50/50 differentiation of individual smokers into compensators and noncompensators has been observed by others (Cherry and Forbes 1972; Freedman and Fletcher 1976; Forbes et al. 1976). A further finding, shown in figure 8, was that the degree of compen- sation on fblder 2 did not change between 2 days and 7 days, but adaptation to the lower nicotine intake did improve over this time. Withdrawal symptoms subsided and subjective satisfaction on Holder 2 increased. How much this difference in the short term trends of these two processes would be reflected in the long term is obviously an important question. The adaptation acquired over this short period was transient. Plasma nicotine and COHb had returned to former levels five days after the holders were abandoned. The only long term study, by Freedman and Fletcher (1976), suggested that adapta- tion acquired over several months could be maintained and that smokers who had adapted to a lower tar and nicotine intake at mouth level (based on butt analysis) did not increase it again when they reverted to smoking stronger cigarettes. These findings highlight the fact that compensation is, on average, only partial and that some smokers compensate while others do not. This contrasts strikingly with the very accurate downward titration 116 TIMN 0152482

FIGURE 8 100 25 Compensation at 2 days Compensation at 7 days Nicotine COHb Nicotine COHb . El 100 r . I 75 50 25 I 0 1 2 0 1 2 0 1 2 0 1 2 0=No holder 1= Holder 1 2 = Holder 2 Amount of compensation when using ventiZated cigarette hoZders, based on the means of 18 subjects. The observed and "expected" bZood nicotine and COHb ZeveZs on the hoZders are expressed as percentages of the mean no-hoZder ZeveZs. The "expected" ZeveZs are based on the diZution factors of the two hoZders, about 20 percent for HoZder 1 and 60 percent for HoZder 2. The shaded areas represent,.the difference between the observed and "expected" ZeveZs, which is a measure of the amount of compensation. The Zined area (for COHb on HoZder 1 at 2 days) indicates a negative 0-E difference, i.e., observed ZeveZs were Zorver than "expected" ZeveZs. Thus, though there was significant compensation on HoZder 2 on aZZ measures, there was cZearZy none on HoZder 1. Significance of 0-E differences: *p<.02, **p<.01, ***p<.001. (From Sutton, S.R.; Feyerabend, C.; CoZe, P.V.; and RusseZZ, M.A.H. Adjustment of smokers by ventiZated cigarette hoZders. CZinicaZ PharmacoZo and Thera eutics, 24:395-405, 1978. 0 1928, T~ie C.V. Mos Reprznte m2t permission.) 117 ITIMN 0152483

of nicotine in the experiments cited previously and may be partly explained by the fact that it requires some effort by the smoker to increase smoke intake but no effort to reduce it. Although satisfaction is reduced when the smoke is diluted, this is not neces- sarily due entirely to failure to maintain nicotine intake. Other factors such as the effort of compensation, loss of taste and impact of the smoke in the mouth, throat, and lungs may contribute. It would appear on balance that smokers tolerate a decrease in nicotine intake better than an increase. The Role of Nicotine Our uncertainty about the role of nicotine in tobacco dependence is reflected in the title of this conference where tobacco and nicotine are linked, ambiguously, as "tobacco/nicotine." As I see it, the stroke between them proclaims loudly what we don't know, and it is on this issue that I have tried to focus. In terms of unequivocal evidence that people smoke for nicotine, even the few direct studies which purport to show that people smoke to obtain a pharmacological effect from nicotine and that titration indicates a need for nicotine, could be interpreted as showing nico- tine to be pharmacologically aversive. Thus, when Lucchesi adminis- tered nicotine intravenously, smoking was inhibited (Lucchesi, Schuster, and Hnley 1976); but when Stolerman blocked the pharmaco- logical effects of nictotine, his subjects could smoke more (Stoler- man et al. 1973). Similarly, Schachter's subjects could smoke more when nicotine excretion was increased (Schachter, Kozlowski, and Silverstein 1977). In other words, these studies could be seen as showing that excessive nicotine is aversive; they do not show that smaller amounts are pharmacologically rewarding. Can we be sure that people are not dependent on inhalation for nonpharmacological reasons such as taste, smell, ritual, sensory stimulation to the lungs and respiratory tract- to which nicotine might contribute? Are people perhaps addicted to inhalation for nonpharmacological rewards but inhibited from indulging more because they find excessive nico- tine pharmacologically aversive? Simply because nicotine has so many pharmacological effects in smoking doses, it does not necessarily follow that these effects are reinforcing rather than aversive. Is there, indeed, any evidence that nicotine is rewarding? Evidence that nicotine can be pharmacologically rewarding is scanty. Firstly, throughout history people have never shown a strong incli- nation to inhale smoke, however aromatic or flavorful, which does not contain a psychoactive drug. This historical evidence is com- pelling but only circumstantial. on the other hand, there is a surprising lack of evidence beyond the anecdotal level that animals will learn to self-inject nicotine as avidly as they do other ad- dictive drugs. Until attending this conference I was aware of only two incomplete reports (Clark 1969; Deneau and Inoki 1967). The careful studies of Hanson and his colleagues (this volume) are a major contribution to this area. A preliminary reservation is that they may not have shown conclusively that their rats were indeed 118 s TIMN 0152484

seeking nicotine as opposed to increasing the rate of lever-pressing due to its stimulant action. The study design did not include a free choice situation, which could have gone a long way to settle this question. Apart from the historical evidence and animal self-administration studies, there is the well-known pioneer study by Johnston (1942). He reported that when smokers were given hypodermic injections of nicotine they "almost invariably thought the sensation pleasant." He also reported that the craving following withdrawal of cigarettes was relieved by injection of nicotine. Has Johnston, all those years ago, done what all of us have failed to do, namely shown that nicotine is both a positive and negative reinforcer for human smokers? Unfortunately, he has not. His 'studies were uncontrolled and nonblind. However, his approach to the question has been more direct and relevant than any of our efforts. We still lack a direct study in humans to show that nicotine is pharmacologically rein- forcing. Whether or not it is rewarding in optimal doses, it cer- tainly seems to become aversive when these are exceeded, and the implications for safer cigarettes remain the same. Less tar and W will be taken in by smokers if their cigarettes combine a low tar and low CO delivery with a medium to high, rather than low, nicotine delivery. Sunanary and Conclusions The role of nicotine in tobacco dependence is discussed. Review of the data available in the literature raises many questions but pi•ovides few answers beyond the following conclusions. ( 1) Pharmacological reinforcement is not an essential feature of addictive behavior. ( 2) There are many nonpharmacological factors involved in tobacco smoking, and these appear to be sufficient to generate strong depen- dence in smokers who do not inhale. ( 3) The low acceptability of low-nicotine cigarettes is nat neces- sarily due to the low nicotine. Nonpharmacological factors are also involved. ( 4) Smokers who inhale seem to tolerate'a decrease in nicotine in- take better than an increase. ( 5) Simply because nicotine has many pharmacological effects in smoking doses, it does not follow that these effects are reinforcing rather than aversive. ( 6) Evidence is scanty that animals will self-inject nicotine as avidly as they do other addictive drugs. ( 7) Apart from circumstantial historical evidence that people have never shown a strong inclination to inhale smoke that does not con- 119 TIMN 0152485

tain a psychoactive drug, there is no direct experimental study which shows that nicotine is pharmacologically rewarding or re- inforcing in humans. I ( 8) Whether or not nicotine is pharmacologically rewarding in op- timal doses, it seems to become aversive when these doses are ex- ceeded. ( 9) The hypothesis that people smoke and inhale for nonpharmaco- logical rewards, including the taste and irritancy of nicotine it- self, but are inhibited from smoking more because they find exces- sive nicotine pharmacologically aversive, has not yet been dis- proved. (10) The implications for safer cigarettes remain the same no mat- ter whether nicotine is rewarding or aversive. The safer cigarette should have a low tar, low CO, but medium to high rather than low nicotine yield. ACKNOWLEDGMENTS I thank the Medical Research Council and Department of Health and Social Security for financial support. I am also grateful for helpful comments from my colleagues, J. Richard Eiser, Martin Raw, and Stephen Sutton. REFERENCES Bynner, J.M. The Young Smoker. Government Social Survey. London: fII,iSO, 1969. Bynner, J.M. Behavioural research into children's smoking: Some irrplications for anti-smoking strategy. Royal Society of Health Journal, 90:159, 1970. Castleden, C.M., and Cole, P.V. Inhalation of tobacco smoke by pipe and cigar smokers. Lancet, 2:21-22, 1973. Cherry, W.H., and Forbes, W.F. Canadian studies aimed toward a less harmful cigarette. J Natn Cancer Inst , 48:1765-1773, 1972. Clark, INi.S.G. Self-administered nicotine solutions preferred to placebo by the rat. Brit J Pharmacol, 35:367P, 1969.. Cowie, J.; Sillett, R.W.; and Ball, K.P. Carbon monoxide absorption by cigarette smokers who change to smoking cigars. Lancet, 1: 1033-1035, 1973. Deneau, G.A., and Inoki, R. Nicotine self-administration in mon- keys. Ann N.Y. Acad Sci, 142:277-279, 1967. 120 TIMN 0152486

Feyerabend, C.; Levitt, T.; and Russell, M.A.H. A rapid gas-liquid chromatographic estimation of nicotine in biological fluids. J Pharm Pharmacol, 27:434-436, 1975. Forbes, W.F.; Robinson, J.C.; Hanley, J.A.; and Colburn, H.N. Studies on the nicotine exposure of individual smokers. 1. Changes in mouth-level exposure to nicotine on switching to lower nicotine cigarettes. Int J Addict, 11:933-950, 1976. Freedman, S., and Fletcher, C.M. Changes of smoking habits and cough in men smoking cigarettes with 30% NV tobacco substitute. Brit Med J, 1:1427-1430, 1976. Goldfarb, T.L.; Gritz, E.R.; Jarvik, M.E.; and Stolerman, I.P. Reactions to cigarettes as a function of nicotine and tar. Clin Pharmacol Ther, 19:767-772, 1976. Ikard, F.F.; Green, D.E.; and Horn, D. A scale to differentiate between types of smoking as related to the management of affect. Int J Addict, 4:649, 1969. Johnston, L.M. Tobacco smoking and nicotine. Lancet, 2:742, 1942. Kozlowski, L.T.; Jarvik, M.E,; and Gritz, E.R. Nicotine regulation and cigarette smoking. Clin Pharmacol Ther, 17:93-97, 1975. Kumar, R.; Cooke, E.C.; Lader, M.H.; and Russell, M.A.H. Is nico- tine important in tobacco smoking? Clin Pharmacol Ther, 21:520- 529, 1977. ! Lucchesi, B.R.; Schuster, C.R.; and Bnley, G.S. The role of nico- tine as a determinant of cigarette smoking frequency in man with observations of certain cardiovascular effects associated with the tobacco alkaloid. Clin Pharmacol Ther, 8:789-796, 1967. f McKennell, A.C. A comparison of two smoking typologies. Research Paper No 12. London: Tobacco Research Council, 1973. i McKennell, A.C., and Thomas, R.K. Adults' and Adolescents' Smokin 1 Habits and Attitudes. Government Soc~i T"Survey. on on: , 1967. Russell, M.A.H. The smoking habit and its classification. The Practitioner, 212:791-800, 1974. ^ Russell, M.A.H.; Peto, J.; and Patel, U.A. The classification of smoking by factorial structure of motives. JRoy Statist Soc.Pi, r 137:313-333, 1974. ~ + r Russell, M.A.H.; Wilson, C.; Patel, U.A.; Cole, P.V.; and Feyerabend, C. Plasma nicotine levels after smoking cigarettes with high, medium and low nicotine yields. Brit AIed J, 2:414-416, 1975. Russell, M.A.H. Low-tar medium-nicotine cigarettes: A new approach to safer smoking. Brit Med J, 1:1430-1433, 1976a. 121 :`;TIMN 0152487 , _

Russell, M.A.H. Tobacco smoking and nicotine dependence. In: Gibbins, R.J., et al., eds. Research Advances in Alcohol and _D__r_u_g Problems. Vol. III. New York: W~.'3ey an ns,79~.~4~ Russell, M.A.H.; Feyerabend, C.; and Cole, P.V. Plasma nicotine levels after cigarette smoking and chewing nicotine gum. Brit Mad J, 1:1043-1046, 1976. Ttussell, M.A.H.; Wilson, C.; Feyerabend, C.; and Cole, P.V. Effect of nicotine chewing-gum on smoking behaviour and as an aid to ciga- rette withdrawal. Brit Med J, 2:391-393, 1976. Russell, Pd.A.H.;•Sutton, S.R.; Feyerabend, C.; Cole, P.V.; and Saloojee, Y. Nicotine chewing-gum as a substitute for smoking. Brit Med J, 1:1060-1063, 1977. Russell, M.A.H., and Feyerabend, C. Cigarette smoking: a depen- dence on high nicotine boli. Drut? Metabolism Reviews, 8:29-57, 1978. Schachter, S.; Kozlowski, L.T.; and Silverstein, B. Effects of urinary pH on cigarette smoking. J F..~ Psychol (Gen), 106:13-19, 1977. - Stolerman, I.P.; Goldfarb, T.; Fink, R.; and Jarvik, M.E. Influen- cing cigarette smoking with nicotine antagonists. Psychopharmaco- logia, 28:247-259, 1973. Sutton, S.R.; Feyerabend, C.; Cole, P.V.; and Russell, M.A.H. Ad- justment of smokers to dilution of tobacco smoke by ventilated cigarette holders. Clin Pharmacol Ther, 24:395-405, 1978. Turner, J.A.M.; Sillett, R.W.; and McNicol, M.W. Effect of cigar smoking on carboxyhaemoglobin and plasma.nicotine concentrations in primary pipe and cigar smokers and ex-cigarette smokers. Brit Med J, 2:1387-1389, 1977. r - AUTHOR Michael A.H. Russell, M.B., MRCP, MRCPsych Addiction Research Unit The Institute of Psychiatry The Maudsley Hospital Denmark Hill London, S.E.5 England 122

Chapter 10 Regulation, Withdrawal, and Nicotine Addiction Stanley Schachter, Ph.D. Most of us who do research on smoking have at some time championed the hypothesis that cigarette smoking, with nicotine as the active agent, is an addiction. Sometimes, however, it is difficult to figure out why that conviction is so strong. The data supporting the proposition are not particularly good; in fact, looked at with a ruthless eye, they are rather flimsy. I intend to review and evaluate the data relevant to the addiction hypothesis and to at- tempt to understand just why they are so weak. Most studies of the addiction hypothesis have focused on the regulation of nicotine intake. I know of twelve studies in which subjects were supplied with high- or low-nicotine cigarettes and a record kept of how much they smoked. Two of these (Finnegan, Larson, and Haag 1945; Goldfarb, Jarvick, and Glick 1970) found no differences--subjects smoked no more low- than high-nicotine cigarettes. ane study (Ashton and Watson 1970) found exquisitely precise regulation, and nine (Frith 1971; Goldfarb et al. 1976; Herman 1974; Jarvik et al. 1970; Johnston 1942; Kozlowski et al. 1975; Kumar et al. 1977; Russell et al. 1973; and Schachter 1977b) did find that subjects smoked more low- than high-nicotine cigar- ettes. It was not precise regulation, by any means; but in all cases subjects did smoke somewhat more of the low- than of the high-nicotine cigarettes. So far so good: in 10 of 12 studies the amount smoked varied inversely with the nicotine content of the cigarette. Though these studies indicate some degree of regulation, they do not as yet establish just what is being regulated; for, as we know, the nicotine and tar contents of cigarettes covary. When I last correlated the nicotine and tar contents of commercially available cigarettes, the correlation coefficient was +.91. We note then that two of these twelve studies attempted in some fashion to manipulate tar and nicotine content independently. Both studies (Goldfarb et al. 1976; Schachter 1977b) found the same result-- smokers track nicotine content, not tar content. There have been three studies which preloaded subjects either with nicotine chewing gum or nicotine capsules (Brantmark et al. 1973; 123

Jarvik et al. 1970; Kozlowski et al. 1975). All report the same effect: Subjects on nicotine gum or capsules smoke less than subjects on placebo. The effects in all cases are quite small but statistically significant. In addition, Stolerman et al. (1973) have demonstrated that administration of a nicotine antag- onist increases cigarette smoking. Finally, there have been two studies using the IV infusion tech- nique. Lucchesi, Schuster, and Pmley (1967) find good evidence for regulation, and the study by Kumar et al. does not. In other contexts (Schachter 1977a), I have spelled out why I think this is so. It looks as if a reasonable case can be made for the propos- ition that smoking is addicting and something of a case for the proposition that nicotine or one of its metabolites is the addicting agent. Why then do I speak of the data supporting the proposition as "flimsy"? For two reasons: First, when we speak of regulation, we tend to think in terms of fairly precise set-point models. Presumably there is a "nicostat" sensing nicotine and regulating intake so as to keep nicotine at some constant level. Yet with a single exception (Ashton and Watson 1970), the data indicate extraordinarily crude and imprecise regulation. A typical instance of just how crude can be seen in Table 1, which presents the data of a study I did some years ago (Schachter 1977b) on a group of long time, very heavy smokers. In alternate we7cs, subjects smoked high- or low-nicotine cigarettes. The low-nicotine cigarettes contained 0.3 mg of nicotine per cigarette; the high-nicotine cig- arettes contained 1.3 mg of nicotine--more than a fourfold difference. Though all of the subjects did smoke somewhat more low- than high- nicotine cigarettes, on the average they smoked only 25 percent more low-nicotine cigarettes--hardly the exquisite titration that set- point models suggest. The second reason for discomfort with the addiction hypothesis is that exceptions to an addiction model are so common. Though it is known (Isaac and Rand 1972) that plasma-level nicotine is zero on awakening in the morning, there are heavy smokers who will not light up their first cigarette of the day till afternoon. There are smokers who smoke only at parties or while they are working, and otherwise not at all. Some orthodox Jewish smokers, forbidden to smoke on the Sabbath, appear able to do so without a whimper. And so on. Just how to cope with such blatant exceptions is problematic. Perhaps it is necessary to invent typologies (e.g., blcKennell 1973; Russell 1974; Tomkins 1968) to accommodate the distressing apparent variety of smokers, but I find this an unsatisfying scientific strategem and admission of defeat. As a working hypothesis, I propose instead that we consider vir- tually all long-time smokers as addicted and attempt to understand the exceptions to maintaining a constant nicotine level in terms of'such notions as self-control, concern with health, restraints, and so on. Certainly all smokers are aware of the dangers and 124 TIMN 0152490

TABLE I The Effects of Nicotine Content on Cigarette Smoking Cigarettes Smoked per Day Low (0.3 mg) High (1.3 mg) % Increase High to Low Subjects Nicotine Nicotine Cigarettes JA 31.25 21.50 + 45.3 SS 55.00 40.50 + 35.8 RR 42.50 30.75 + 38.2 RS 22.75 20.00 + 13.8 DR 70.75 58.75 + 20.4 RA 30.25 26.25 + 15.2 JE 48.00 44.25 + 8.5 Mean 42.93 34.57 + 25.3 expense of smoking. To the extent that such concerns are prominent, the smoker probably inhibits smoking by such devices as imposing an upper limit on daily consumption, scheduling smoking, restricting smoking to particular occasions, and so on. All of these are devices which are designed to lower consumption and which would tend to mask such behavioral manifestations of addiction as tracking nicotine con- tent. If this is correct, we should also expect other less obvious manifestations of addiction to be present. Of these I would suggest that the key will be the withdrawal syndrome. Obviously, anyone can give up smoking, limit daily intake, or restrict smoking to particu- lar times or occasions if he is willing and able to put up with the withdrawal syndrome. If it is correct that virtually all longtime smokers are addicted, it should be anticipated that smokers who do not smoke in the morning will be grurpier and more irritable at that time of day than in the afternoon; that heavy smokers who switch to very low-nicotine brands will be chronically more volatile than heavy smokers who don't switch, and so on. In order to test this sort of expectation, Perlick (1977)*designed an experiment in which she measured irritability in two groups of smokers and a control group of nonsmokers. Unrestrained, heavy smokers were compared with a matched group of mostly former heavy smokers who, though still smoking, were deliberately cutting down. In fact, most of this second group were smoking less than half of their former regular intake. 125 iTIlMN 0152491

We have, then, two groups of experimental subjects: one smoking to heart's content and the other deliberately restricting intake by any or all of a variety of devices such as switching to low- nicotine brands, buying only one pack a day, smoking only on the hour, and so on. Within the context of a study of aircraft noise, these subjects individually watched a television drama and rated how annoying they found each of a series of simulated overflights. The labora- tory had been designed by Eugene Galanter to look like a living room in Queens, near Kennedy Airport. Every 30 seconds or so a horrendous racket pervaded the room as the stereo audio system reproduced the noise of a jet taking off or landing. There were three experimental conditions: In one, the subjects were permitted ad lib smoking of high-nicotine (1.3 mg) cigarettes; in a second, ad lib smoking of low-nicotine (0.3 mg) cigarettes; finally, in a third condition, they were prevented from smoking. Before getting to the results, let us examine the actual amount of smoking by these two groups of smokers during the experiment. Do they behave during the experiment as on questionnaires they say they do in real life? They do. In those conditions in which they were permitted to smoke, restrained smokers took an average of 20.6 puffs (3.3 cigarettes) during the two-hour experimental session. This is just about half the amount smoked by unrestrained smokers, who puffed an average of 39.5 times (5.3 cigarettes). Let us examine next the impact of the nicotine manipulation on the heavy, unrestrained smokers. The logic of the general argument demands the demonstration that irritability is a consequence of nicotine shortage. The data are presented in Figure 1. Along the ordinate is plotted the average annoyance rating. It can be seen that when permitted ad lib smoking of high-nicotine cigarettes, unrestrained smokers are just about as irritated by this noxious series of noises as are the control nonsmokers. However, when pre- vented from smoking or permitted to smoke only low-nicotine cigar- ettes, their irritability increases markedly. It does appear that, for the unrestrained smoker, irritability is a clear consequence of depleted nicotine. Incidentally, Perlick found precisely the same pattern for eating behavior, which is also presumed to be one of the key nicotine withdrawal symptoms. At one point during the experiment, subjects had the opportunity to nibble freely at the contents of a large jar of jelly beans. Subjects in the no- or low-nicotine conditions ate twice as many jelly beans as did high-nicotine or nonsmoker subjects. Let us turn•next to the group of highly restrained smokers. If it is correct that the price of successful self-control is a chronic state of withdrawal, we should anticipate high irritability in all conditions, for even in the high-nicotine condition these former heavy smokers are getting considerably less nicotine than was their wont. The results are presented in Figure 2, where it can be seen that in all conditions these highly restrained smokers are irritable. Further,-7erlick showed that these restrained smokers eat more when 126 I STIMN 0152492

FIGURE 1 (D z Q ~ 0 LOW HIGH NO (0.3mg) (1.3mq) SMOKING NIC NIC 0 NO SMOKING NICOTINE MANIPULATION The effects of nicotine on the irritabiZity of heavy smokers and nonsmokers. 127 TIMN 0152493

FIGURE 2 300 250 200 50 0 LOW HGH NO (0.3mg1 (L3mgl SMOKING NIC NiC 0 NO SMOKING NICOTiNE MANIPULATION 0 LOW HIGH NO (0.3mg) U.3mg) SMOKING NIC NtC The effects of nicotine deprivation on the irritability of heavy smokers, nonsmokers, and restrained smokers. 5

given free access to jelly beans and also do not do as well at a proofreading task requiring concentration. Restrained smokers, then, appear to be chronically more irascible, to nibble more, and to have poorer concentration than unrestrained smokers. It is possible to control and restrict smoking, but at a price, and the price appears to be a chronic state of withdrawal.l It does appear that one of the exceptions to a purely addictive view of smoking is in reality no exception. I suspect that this will be the case with most of these exceptions, and that by taking account of withdrawal we can understand those studies (Finnegan, Larson, and Haag 1945; Goldfarb, Jarvik, and Glick 1970) that fail to show nicotine regula- tion. I would certainly suggest that any future studies of nicotine regulation make systematic provision for the measurement of with- drawal. I would also suggest that of the numerous possible indices of withdrawal, weight change may be by far the simplest and most reliable. Given what we know of the effects on weight of giving up smoking and what we know of the effects of nicotine deprivation on eating, it's not an unreasonable guess that those subjects who fail to regulate nicotine will be the most likely to increase weight. I would like to conclude that by jointly considering withdrawal and regulation we can cope with the numerous apparent exceptions to an addictive model of cigarette smoking and also dispose of the hateful necessity to invent typologies to account for the apparent diversity of types of smokers. Though I do believe that this conclusion is valid for most smokers and smoking phenomena, I am not yet ready to declare it universal, for there does seem to be at least one "type" (dreadful word) of smoker for whom none of this seems to make sense. This is the sort of person who has smoked for most of his or her life and has never smoked more than 10 or 15 cigarettes a day. This person inhales, smokes medium- to high-nicotine cigarettes, but by his own self-description gives no indication that he is addicted. He may not light up his first cigarette o£ the day until cocktail time and can go for days without smoking and not think about it, until he goes to a party or some such event. Further, he insists that he is in no way trying to cut down or limit his smoking. Despite extensive and long-time exposure to nicotine, we seem to have a non- addicted and nonaddictable animal. For years I have tried to prove that among his other attributes this "type" of smoker is a liar, but so far with no success. In a first attempt to study regulation in this group, I was able to identify fo,ur such people and ran them simultaneously with the group already described in the study in which subjects smoked either 0.3 mg or 1.3 mg nicotine cigarettes in alternate weeks (Schachter, 1977b). These subjects, as a group, gave no sign of regulation, for two of them smoked more of the high-nicotine and two smoked more of the low- nicotine cigarettes. Since my theme has been, "If they don't regulate, look for withdrawal," let us next examine this "type" of smoker for indications of witht drawal. Perlick was able to locate 15 such smokers, who incidentally were roughly matched with her group of restrained former heavy smokers. Both groups had smoked for about 8.5 years and both groups currently 129 0152495

averaged about 11.5 cigarettes a day. The irritability data for this group are presented on the right hand side of Figure 3. They stand in fascinating contrast to all of the other groups of subjects, for their irritability rating has absolutely no relationship to the nicotine manipulation. In all three conditions, their annoyance rating is quite low and on a par with the two groups that are not suffering withdrawal--nonsmokers and heavy smokers in the high nicotine conditions. One obvious problem in interpreting these data is that two hours without smoking may be small deprivation to someone who smokes only 10-15 cigarettes a day. It may be that with a longer period of deprivation even these subjects would manifest withdrawal. However, several sub-analyses suggest that this would not be the case. First, even for such subjects, one might expect that annoyance ratings made at the end of the experimental period would be higher than those made at the beginning. This is certainly the case for the heavy and the restrained smokers in the nicotine deprivation condition. Both groups are markedly more annoyed at the end of the experiment than at the beginning. There is no such effect at all for this group of apparently nonaddicted smokers; they are just as tranquil at the end of the experiment as at the beginning. Secondly, thanks to a questionnaire, Perlick knew when the subjects had smoked their last cigarette before coming to the experiment. If it is simply a matter of time, one would expect that those who had smoked their last cigarette long before coming to the experiment would be more irritable than those who had smoked shortly before the experiment. Not so. For this group of subjects, there is no relationship between time of last pre-experiment cigarette and the mean level of annoyance during the experiment. There is, then, experimental evidence suggesting that some smokers are not addicted and apparently are nonaddictable. Some of these subjects have smoked for 30 years and never smoked more than 10 or so cigarettes a day. I think there are few such people--perhaps 5 to 10 percent of the smoking population--but I harp on them because I suspect that they are an extraordinary scientific resource. If we can understand these "freaks," we may begin to make steps in discovering what addiction is about. FOOTNOTE 1. One alternative interpretation of these data must be considered. It is conceivable that naturally irascible people are more likely to restrain their smoking. If so, these results could be attributable to self-selection rather than withdrawal. Acutely aware of this possibility, Perlick (1977) compared these groups on numbers of personality and demographic variables and found no differences between the two groups. 130 TITV~N 0152496

300 (5 250 z 6 O Q ¢ ~ ~ 200 U) W w U Z o 50 ~ O 4 Z :2 100 Q v 50 FIGURE 3 A B C D HEAVY SMOKERS NON - SMOKERS RESTRAINED SMOKERS LIGHT UNRESTRAINED SMOKERS O LOW HIGH NO (0.3mg) (1.3mg) SMOKING NIC NIC 0 NO SMOKING NICOTINE MANIPULATION O LOW HIGH NO (U,3mg) (1.3mg) SMOKfNG NIC NIC O LOW HIGH NO (0.3mg) C1.3mg) SMOKING NiC NIC The effects of nicotine deprivation on the irritability of heavy smokers, nonsmokers, restrained smokers, and light unrestrained smokers.

REFERENCES Ashton, H., and Watson, D.W. Puffing frequency and nicotine intake in cigarette smokers. Brit Med J, 3:679-681, 1970. Brantmark, B.; Ohlin, P.; and Westling, H. Nicotine-containing chewing gum as an anti-smoking aid. Psychopharmacologia, 31:191-200, 1973. Finnegan, J.K.; Larson, P.S.; and Haag, H.B. The role of nicotine in the cigarette habit. Science, 102:94-96, 1945. Frith, C.C. The effect of varying the nicotine content of cigar- ettes on human smoking behaviour. Psychopharmacologia, 19:188-192, 1971. , Goldfarb, T.L.; Jarvik, M.E.; and Glick, S.D. Cigarette nicotine content as a determinant of human smoking behavior. Psychopharma- cologia, 17:89-93, 1970. Goldfarb, T.L.; Gritz, E.R.; Jarvik, M.E.; and Stolerman, I.P. Reactions to cigarettes as a function of nicotine and "tar". Clin Pharmacol Ther, 19:767-772, 1976. Herman, C.P. External and internal cues as determinants of the smoking behavior of light and heavy smokers. J Pers Soc Psychol, 30:664-672, 1974. ~ Isaac, P.F., and Rand, M.J. Cigarette smoking and plasma levels of nicotine. Nature, 236-308, 1972. Jarvik, M.E.; Glick, S.D.; and Nakamura, R.K. Inhibition of cigarette smoking by orally administered nicotine. Clin Pharmacol Ther, 11:574-576, 1970. Johnston, L.M. Tobacco smoking and nicotine. Lancet, 2:742, 1942. Kozlowski, L.T.; Jarvik, M.E.; and Gritz, E.R. Nicotine regulation and cigarette smoking. Clin Pharmacol Ther, 17:93-97, 1975. Kumar, R.; Cooke, E.C.; Lader, M.H.; and Russell, M.A.H. Is nicotine important in tobacco-smoking? Clin Pharmacol Ther, 21:520-529, 1977. Lucchesi, B.R.; Schuster, C.R.; and Fmley, G.S. The role of nicotine as a determinant of cigarette smoking in man with observa- tions of certain cardiovascular effects associated with the tobacco alkaloid. Clin Pharmacol Ther, 8:789-796, 1967. McKennell, A.C. A comparison of two smoking typologies (Research Paper 12). London: Tobacco Research Council, 1973. 132 TIMN 0152498

Perlick, D. The withdrawal syndrome: nicotine addiction and the effects of stopping smoking in heavy and light smokers. (Unpub- lished doctoral dissertation). New York: Columbia University, 1977. Russell, M.A.H.; Wilson, C.; Patel, U.A.; Cole, P.V.; and Feyerabend, C. Comparison of effect on tobacco consumption and carbon monoxide absorption of changing to high and low nicotine cigarettes. Brit Med J, 4:512-516, 1973. Russell, M.A.H. Realistic goals for smoking and health. Lancet, 254-258, 1974. Schachter, S. Nicotine as addicting agent. Proceedin zs othe Conference on Comnonalities in Substance Abuse and Habitual Be avior. Was ington, D.C.: tiona Aca emy o Sciences, 1977a. Schachter, S. Nicotine regulation in heavy and light smokers. J Exp Psychol: General, 106:5-12, 1977b. Stolerman, I.P.; Goldfarb, T.L.; Fink, R'.; and Jarvik, M.E. Influencing cigarette smoking with nicotine antagonists. Psychopharmacologia, 28:247-259, 1973. Tomkins, S. A modified model of smoking behavior. In: Borgatta, E.F., and Evans, R.R., eds., Smoking, Health and Behavior. Chicago: Aldine, 1968. AUTHOR Stanley Schachter, Ph.D. Department of Psychology 1400 Schennerhorn Columbia University New York, New York 10027 133 + T'IlMN 0152499

Part111 Psychobiological Factors ~; TIMN 0152500

Chapter 11 Acute and Chronic Effects of Nicotine in Rats and Evidence for a Noncholinergic Site of Action L. G. Abood, Ph.D., K. Lowy, M.D., and H. Booth Since the discovery that acetylcholine is a neurotransmitter in the autonomic nervous system and that its action consists of a muscarinic and nicotinic component, the action of nicotine has been attributed mainly to effects on nicotinic cholinergic synapses (Domino 1965; Larson and Silvette 1975). Although the action of nicotine in the autonomic nervous system or in certa3.n• brain areas can be understood in terms of either its agonistic or antagonistic action at nicotinic cholinergic synapses, a close perusal of the literature reveals that its neuropharmacological •effects are considerably more complex and cannot be entirely explained by this mechanism. Recently we have observed that when the natural form of nicotine, which is the (-)-isomer, is injec- ted directly into a rat's lateral or third ventricle, there occurs a characteristic prostration-imnobilization syndrome lasting for .from 0.5 - 2 min (Abood et al. 1978). The syndrome can be pre- vented or antagonized by the intraventricular administration of a ntunber of newly synthesized nicotine or piperidine derivatives, but not by a variety of antinicotinic or rn unerous psychotropic agents. These observations,along with those involving electro- physiological and receptor binding measurements,led to the conclusion that the prostration syndrome may not be mediated by cholinergic mechanisms. Additional pharmacological and electrophysiological studies on the acute effects of nicotine as well as observations on the chronic effects following continuous intraventricular infusion of nicotine into the lateral ventricles of freely moving rats are described in this report. MATERIAIS AND NMHODS All studies were performed on male Sprague-Dawley rats weighing between 150-200 g. The anesthetized rat was introduced into a Kopf stereotactic instrument, and a stainless steel cannula (#220 DK 1 rat cannula, obtained from David Kopf) was inserted stereotactically,aiming at either the lateral or third ventricle. 136 ~~1VI1~ 015~5~~

It was attached to the skull by means of acrylic cement and four small set screws. A bipolar Formvar coated nichrome electrode of 0.2 mm thickness and 3/4 mn vertical tip separation was in- serted with leads into the region of the dorsal hippocarrpus. Electrical Recordings From Conscious Rats Electrical recordings were made ffom f!reely moving conscious rats by way of a light shielded flexible cable leading into'a Grass P-15-B preanplifier set for an amplification of 1000 and a filter band between 3-30 cycles/sec. The amplifier output was connected to an FM tape recorder and a Grass dynagraph. Analysis of elec- trical activity was accomplished by two LINC-8 programs operating on the output from the tape recorder. Spectral analysis was done by means of a fast Fourier transform program modified for use by LINC-8. Another program was used to plot histograms of amplitude distribution recorded fran representative sections of the tape recordings. 3H-Nicotine Binding to Glass and Tissue In an effort to develop a receptor binding assay for'nicotine, an extensive investigation was undertaken utilizing various neural membrane preparations, synaptosomes, brain homogenates, and brain slices; however, none of these proved to be satisfactory.' The ligands used in this exploratory study were 225I-a-bungaratoxin, 1''C-d-tubocurarine, and 3H-nicotine. The techniques for measur- ing binding to cell-free preparations included equilibrium dialy- sis, centrifugation, gel filtration, and filtration through glass fiber filters. Although some stereospecific or specific nicotine binding could be demonstrated with fresh rat brain slices, the most satisfactory data were obtained by. measuring 3H-nicotine binding directly to Whatmsn GB/F glass fiber filters in the absence of any tissue preparations. The procedure for measuring 3H-nicotine binding to glass fiber filters is described elsewhere (Abood et al. 1978). Briefly, it consists of filtering a solution (0.05M Tris, pH 7.5) of 10 8M 3H-nicotine (0.2 curies/rmiole) in the absence or presence of 10 5M (-) or (+)-nicotine or nicotine analogues and then fil- tering in vacuo through Whatman GF/B glass fiber filters (2.1 cm). After washing the filters with 10 ml 0.05M Tris, pH 7.5, the filter is transferred to vials and radioactivity measured by liquid scintillation. Although some success was obtained in measuring specific nicotine binding to fresh rat brain slices, no specific or stereospecific binding was demonstrable to synaptosomes, neural membranes, or brain homogenates, employing all known techniques for measuring binding (see "Results"). Chronic Intraventricular Infusion of Nicotine Chronic microinfusion of nicotine into the lateral ventricles of rats was accomplished by means of Alzet osmotic minipumps Model 137 I

1701 (Wei and Loh 1976). The minipunp was inserted subcutaneously and was connected by means of a fine catheter to a 24 gauge stain- less steel cannula stereotactically implanted into the lateral ventricles. The reservoir of the minipump contained 170 u1 of 10 mg/mi solution of nicotine HC1, and the delivery rate was about 1 ul/hr (i.e.,10 pg nicotine/hr). In order to determine the reliability of the minipumps,they were tested for their re- maining contents 1-2 weeks later either by measuring optical absorbance at 260 nM or by the use of 3H-nicotine and measuring residual radioactivity. RESULTS Psychopharmacological Effects of Nicotine Given Intraventricularly Within 1-10 sec following the intraventricular administration of 2-10 pg of (-)-nicotine HC1 the rats became prostrate and immobile, manifesting occasional seizures and tremors together with various autonomic changes (tachycardia, hyperpnea, and urination-defe- cation). The dose-response relationship was monotonic in this range. In order to produce a comparable response with (+)- nicotine, the required dose was at least 100 times greater than for (-)-nicotine. A variety of agents were tested for their ability to prevent the prostration incmbilization syndrome of (-)-nicotine,including a series of synthetic derivatives of nicotine and piperidine (table 1). When 10 ug of either the benzyl or 4-azido-2-nitrophenyl derivatives of either piperidine or nicotine was given intraventricularly 2 min prior to 4 ug of (-)-nicotine,the prostration syndrome could be prevented. A wide variety of psychotropic neurotransmitters, cholinergic agents (e.g., naloxone, d-tubocurarine anticholinergics, and physostigmine),and d-tubocurarine were ineffective. Azapetine, an a-adrenergic blocker having a structural relationship to nicotine, possessed moderate antagonist activity. Neither chlorpromazine nor diazepam had any effect in antagonizing nico- tine or on binding. 3H-Nicotine Binding to Glass Fiber Filters In general,a good correlation was observed between the behavioral antagonism of a given agent and its ability to compete with 3H- nicotine for binding to glass fiber filters (table 1). The most effective antagonists-such as N-ANPP, N-benzyl and N-ONPS deri- vatives of nicotine,and piperidine-were also the most effective agents in blocking nicotine binding. The (+)-isomer of nicotine, which possessed about 1 percent of the efficacy of (-)-nicotine in producing the prostration syndrome, was 10 percent as effective as the natural nicotine in blocking binding. Although some stereo- specific and specific binding of nicotine could be demonstrated with fresh brain slices, the magnitude of the binding was small and somewhat variable; therefore, such data were not included. No specific binding to neural membranes, synaptosomes, or homogenates prepared from rat brain was demonstrable. 138

TABLE 1 Antqgonism of various agents to prostration syndrome in rats and to .iH-nicotine binding to glass filters. The dose of (-)-nicotine was 4 ug, which was given 1 min after the test agent. All drugs were given intraventricularly at doses of 10 jig, except for (+)-n.icotine and piperidine, which were given at 100 ug. At least 8 rats were used for each drug tested behaviorally, while the standard error in the binding studies is w'thin 6 percent of the mean. Binding is expressed as moles x l01~/glass fiber filter. N-ANPP = 4-azide-2-nitrophenyl; N-ONPS = 2-nitrophenylsulfenyl. Agent % Behavioral H-nicotine bindin Antagonism Moles x 10 Glass % corrpetition Glass Control - 12.0 (-)-nicotine 0 3.6 70 + -nicotine 33 11.0 8 -benzy nicotine 92 7.0 42 - - cotine 83 6.5 45 -benzyl piperidine 75 6.0 50 - piperidine 72 6.0 50 N- S-piperidine 35 9.0 25 N-ONPS-nicotine 40 10.0 17 piperidine 40 11.5 4 butaclamol slight 12.0 0 decame honium 0 0.5 19 -quinuc 1 benzilate slight 11.8 2 oxotremorine 0 9.9 17 naloxone 0 12.0 0 a-lobeline 0 11.5 4 azapetine moderate 10.0 •17 chlorpranazine slight 11.5 4 diazepam 0 11.7 2 139 --- xMN 0~525p4 T

Electrical Recordings of Dorsal Hippocampus After (-) and (+)- Nicotine and Henzyl Nicotine Spontaneous electrical activity recorded from the dorsal hippo- campus of unanesthetized, freely moving rats before and after the administration of 4 ug of (-)-nicotine bitartrate revealed marked changes in both the frequency and amplitude of the electrical pattern (figure 1 a-d; figure 2 a-d). Frequency analyses of the electrical record are presented as oscillog•raphic displays from fast Fourier transform (figure 1 a-d). The control record indi- cated a frequency range of 6-8 cycles/sec with a minor corrponent in the 12-14 cycles/sec range (figure 1 a). Within 1 min after the administration of 4 ug of (-)-nicotine, when the rat was conpletely prostrate and imnobile, the 5-8 cycle/sec activity greatly dimisdshed (figure 1 b). When 10 ug of N benzyl nicotine was given intraventricularly about 30 sec after the record in figure 1 b, the electrical pattern within 30 sec reverted to one resembling the control (figure lc). For comparison, a frequency histogram is presented of a record from an animal given 200 jig of (+)-nicotine bitartrate, a dose which produced only minimal prostration, muscle weakness, and inactivity (figure ld). A slight shift to lower frequencies was notedywith a peak value at about 5 cycles/sec. A computer analysis was performed on the ampli- tudes of the electrical recordings, and the results are.presented in the form of histograms (lower tracings, figure 2 a-d). During the control period the anplitudes varied over a large range •between 200-400 uV. After (-)-nicotine,when the anim37l was pros- trate, the higher attplitudes at 400 uV disappeared, while the major corrponent was at 200 uV. The flattening of the electrical recording after (-)-nicotine is evident in the upper'tracing (figure 2b). After 10 ug of benzyl nicotine, the electrical record and the amplitude distribution resembledd the control pat- tern; however, the 300-400 uV an-plitude components did not reach the control level (figure 2c). After 200 ug of (+)-nicotine, the histogram and electrical pattern varied only slightly from the control records, the major difference being an increase in the 200 uV corrponent (figure 2d). Electrical Activity of Hippocanpus During Nicotine Infusion The electrical activity was recorded from the dorsal hippocanpus of freely roving rats throughout the period of the intraventri- cular infusion of nicotine at a rate of 10 ug/hr (figure 3 a-d). The electrical pattern of the hippocampus is displayed in the upper oscillographic tracings of figure 3a and 3b; and in the tracing below is displayed a corresponding anplitude histogram of the upper record. The amplitude range before the infusion began had a broad spectrum ranging up to 400 uV (figure 3a); but after 48 hours of infusion the amplitude range narrowed consider- ably, with a maxirrnun of 100 uV (figure 3b). After 6 days of in- fusion the amplitude histogram and electrical record resembled the control record (figure 3c), and two days following the removal of the miim3.puTrp the records were unchanged (figure 3d). A spec- 140 ` TIMN 4152505

tral analysis of the frequency eaploying fast Fourier transform revealed after 16 hours of nicotine a broad spectrum up to 30/ sec with a m@jor component in 5-7/sec range (figure 4a). After 48 hours of nicotine infusion the higher frequency corrponents alrnost vanished and only the 5-7/sec corrponent remained (figure 4b). After 6 days of infusing nicotine, the frequency spectrum returned to the pattern seen in the control (figure 4c); and two days after removal of the minipurrp the pattern remained unchanged (figure 4d). FIGURE 1 Spectral frequency analyses of electrical recordings of rat dorsal hippocanpus after (-)- or (+)-nicotine and N-benzyl nico- tine. a = control, b= 1 min after 4 ug (-)-nicotine, c= same rat given 200 pg (+)-nicotine 2 days later (control record identi- ca].), d = a rat given 10 ug N-benzyl nicotine 30 sec after (-)- nicotine. The records are typical of at least 5 individual rats for each drug. See text. Each dot on the abscissa represents a scale.of 5 cycles/sec. Chronic Nicotine Administration and Tolerance In an effort to determine whether tolerance developed with repeated intraventricular administration of nicotine, rats were exposed to various dosage regimes of nicotine (table 2). When rats received 10 ug of nicotine for 3 consecutive days, they all responded (prostration inmobilization syndrome) on day 1, while ;TIMN 0152506

, FIGURE 2. ; ;. /. t^ v i `, .: .. ~ : i:, . 4~ ~. ~J' ;O~ ~L + e ++.r.~', c: ~! Electrical record and amplitude histograms corresponding to record$ described in figure 1 a-d. The bar represents a scale of 10-4 V on the abscissa. TABLE 2 Effects of various dosage regime on behavioral response of rats to nicotine. From 5-10 rats were represented in each schedule. Nicotine was administered into left lateral ventricles. Dosage Schedule Incidence of Prostration S drome 10 ug nicotine daily for 3 All of rats responded day 1, consecutive days 50 percent on day 2, 20 per- cent on day 3, 90 percent response after 3 days recovery. Tolerce . 10 pg nicotine every other day st rats ful respondedor for 10 days each day drug was given. No tolerance. 2 pg nicotine day 1, 4 pg day esponse on day to 10 pg 2 10 u d . nicotine. minipunq~ (10 ug/hr Normal response to 10 ug for 10 c~ays. nicotine on day after infusion 142 ended. TIMN 0152507

FIGURE 3 H ~ W Amplitude histogram of hippocampal electrical activity following chronic intra- ventricular infusion of nicotine. Same rat and legend as in figure y a-d. j

FIGURE 4 Frequency analysis of hippocanpal electrical activity following chronic intra- ventricular infusion of nicotine. a = control, b = 48 hrs of nicotine infusion, c= 6 days of nicotine, d = 2 days after infusion was discontinued.

50 percent responded on day 2 and only 20 percent on day 3. If 10 ug of nicotine was given every other day for 10 days, most of the rats responded each day the drug was given. After 10 ug of nicotine was given for 5 consecutive days, none of the rats re- sponded to 10 ug of nicotine on day 6; however, most of the ani- mals responded on the second day after discontinuing the nicotine. If rats were given 2 ug of nicotine on the first day, 4 ug on the second day, and 10 ug of nicotine on the third day, they all responded to 10 ug of nicotine on the fourth day. Effect of Nicotine on Brain Levels of Met-Enkephalin Regional brain levels of inet-enkephalin were measured in rats sacrificed 1-2 minutes following the intravetntricular injection of 10 ug of nicotine. The animals were still prostrate at the time of sacrifice. No significant differences were observed between the saline- and nicotine-injected rats in any of the three brain regions (table 3). The determinations were made by Dr. S. Chanda, utilizing the technique of radioimmune assay. TABLE 3 Effect of nicotine on enkepha.lin of rat brain regions. Enkephalin = TYR-GLY-GLY-PHE-MET in pmoles/g wet wt. All rats received 10 ug nicotine intraventricularly 1-2 minutes before sacrifice. Results are average of 6 rats. Nicotine Saline elencephalon 128 ± 39 97 ± 29 1 Diencephalon 149 f 55 152 ± 36 iKesencephalon 223 ± 48 243 ± 34 DISCUSSION The major conclusion from this study is that the action of nico- tine in producing the prostration imnobilization syndrome does not appear to be mediated by cholinergic mechsnisms in the brain. Among the reasons for this conclusion are the following: 1) the syndrome cannot be simulated by a number of cholinergic agents in doses up to 100 ug; included among such agents were acetylcholine, oxotremorine, pilocarpine, physostigtn3ne, and a-lobeline; 2) a variety of agents known to block the nicotinic cholinergic receptor were ineffective in either preventing or reversing the behavioral effect of nicotine; and 3) the nicotine and piperidine analogues which were effective in blocking the behavioral effect of nicotine did not interfere with the binding 145 I TIMN 0152510

of 125I-a-bungarotoxin or 3H-d-tubocurarine to neural membranes (Abood et al. 1978). In an effort to determine the possible relationship of nicotine to known neurotransmitters,a wide variety of substances, in addition to those described, were injected intraventricularly (in doses up to 100 ug) and found to have no similar effect. These included noradrenalin, dopamine, serotonin, glutamic acid, Y-aminobutyric acid, met-enkephalin, and morphine, as well as a variety of antagonists to known neurotransmitters. It would appear, therefore, that this behavioral effect of nicotine is quite unique and distinct, while being largely unrelated to any known neurotransmitter system. The brain sites associated with the prostration syndrome exhibit a high degree of stereospecificity, the natural (-)-isomer of nicotine being at least 100 times as effective as the (+)-isomer. Since the optical purity of the stereoisomers can be determined with only about 99 percent accuracy, it is conceivable that the (+)-isomer is conipletely inactive. Domino (1965) compared the two isomers of nicotine on established conditioned avoidance and seizure threshold in rats as well as hypertension in dogs and concluded that the (-)-isomer was only 8 times as active as the (+)-isomer. Pfeiffer (personal communication) reported a similar difference in man with the two isomers. Although the (+)-nicotine used by both investigators was not as optically pure as the material used in the present study, the difference between the results cannot be accounted for by the difference in purity. A more plausible' explanation is that the ventricular sites associated with the prostration syndrome are distinct from those involved in the behavioral and peripheral parameters measured by Domino (1965). Another conclusion from this study is that tolerance in rats to the behavioral effect of nicotine develops after repeated in- jections, but the responsiveness returns to the initial level within 2 days following the last injection. Even after the rats received a continuous infusion of nicotine for 7 days, they still responded 1 day after stopping infusion to 10 ug of nicotine injected into the lateral ventricles. With continuous infusion of nicotine into the lateral ventricles,a marked change occurred in electrical pattern of the dorsal hippocampus 24 hours after infusion began, but the electrical activity returned to normal by the end of the third day. Despite the fact that the electrical pattern changed so markedly, there were no marked behavioral effects during 7 days of infusing nicotine. FUPURE OBJEGTIVES AND DIRECTICNS The present study raises a number of important questions con- cerning both the acute and chronic effects of nicotine on be- havior. The most fundamental question pertains to the nature of the prostration-ininobilization syndrome produced by nicotine administered intraventricularly to rats and its possible rela- tionship to the behavioral effects of tobacco in man. It is recognized that the doses used to produce this effect in rats are comparatively large, so that any comparison with smoking would be 146 TININ 0152511 ~

difficult and somewhat conjectural. Some similarity has been ob- served in EEG changes associated with smoking in man (Murphee et al. 1967; Philips 1971) and after intravenous administration of small doses of nicotine in cats (Domino and Yamamoto 1965). Nicotine in both cats and man produces a desynchrony in the EEG, which has been interpreted as both an "arousal" (Domino and Yamamoto 1965; Philips 1971) and an opposite tranquilizing effect (Murphee et al. 1967). In our studies with intraventricularly administered nicotine9both the electrical recordings and behavioral responses tend to suggest that the preponderant effect is a depressant one; however, at lower doses (1 i,g) the rats appear more active and 1-~yperexcitable. Precise and detailed psychophysical measurements are needed in order to distinguish the two kinds of effepts in animals as well as man. Another question concerns the nature and localization of the sites for nicotine's action in the brain. If, as the present study seems to suggest, the sites are noncholinergic and may be unrelated to other known neurotransmitters, the question then arises as to chemical nature of the endogenous substrate for this particular site. As to the location of the sites, they would appear to be somewhere within or near the lining of the lateral or third ventricles. To date it has not been possible to produce the prostration syndrome by injecting nicotine directly into brain structures 3mnediate].y adjacent to the ventricles or into innumerable deeper brain structures. Since a large number of neurohumoral substances, including bio- genic amines and neuropeptides, are present in the hypothalamus (Hokfelt et al. 1978),and since synaptic terminals for known neurotransmitters have been detected in the linings of the lower third ventricles, the hypothalamus has been considered as a• possible site for nicotine. Although the injection of nicotine into the ventricular region in the vicinity of the hypothalamus will elicit the prostration syndrome, the administration of nico- tine into various nuclei of the hypothalamus has failed to produce any behavioral response. In view of the fact that a number of neuropeptides, such as S-endorphin, enkephalins, substance P, and angiotensin II are associated with the hypothalamus, the possibility exists that nicotine may promote their release and that the behav- ioral effects are due directly to the peptides. Nicotine has been shown to promote the release of catecholamines from the hypothalamus (Hall and Turner 1972) and to inhibit the release of serotonin (Goodman and Weiss 1973), but neither neurotransmitter mimics the action of nicotine. The brain is also known to contain amines which possess nicotinic cholinergic actions, the most noteworthy being piperidine (Von Euler 1944; Abood et al. 1961). Although piperidine possesses tranquilizing (Abood et al. 1961) and soporific actions (Dolezalova et al. 1973), it does not produce the prostration syndrome seen with intraventricularly administered nicotine. It does not appear from the present studies that the behavioral 147

effects of nicotine are associated with changes in the levels of brain met-enkephalin. Another finding that would tend to exclude the involvement of endogenous opioid peptides in the action of nicotine is that naloxone did not prevent or modify the behavioral effects of nicotine. Furthermore, morphine itself did not mimic or prevent the effects of nicotine. Finally, some comment should be made concerning the inability to demonstrate specific nicotine binding to neural tissue, whereas both specific and stereospecific binding is readily demonstrable to glass. One possibility is that the sites are either destroyed during disruption of the tissue or that released endogenous substances interfere with binding. Another is that the binding affinity is too low to detect with the 3H-nicotine used, a problem that might be rectified with material having a 100-fold greater specific activity. The fact that the response to nicotine is so fleeting (often lasting only 10-30 seconds) is su~gestive of a relatively low affinity. Since nicotine would appear to be highly potent when injected intraventricularly, the receptors may be close to the site of administration. Since many drugs bind stereospecifically to glass, it is difficult to attach much significance to the fact that nicotine also binds stereo- specifically. It is of interest, nevertheless, that the data with the glass filters correlate well with the in vivo effects of the various nicotine derivatives. Conceivably, the technique could be useful as an initial screening for nicotine agonists or antagonists. ACKNOWLEDGIIkVT This research was supported by a grant from the Council of Tobacco Research and USPH grant DA 00464. REFEREAICFS Abood, L.G.; Lowy, K.; Tometsko, A.; and Booth, H. Electro- physiological, behavioral and chemical evidence for a noncholin- ergic, stereospecific site for nicotine in rat brain. J_ Neuro- sci, 3:327-333, 1978. Abood, L.G.; Rinaldi, F.;and Eagleton, V. Distribution of piperidine in the brain and its possible significance in behavior. Nature, 201-202, 1961. Dolezalova, H.E.; Giacobini, E.; Seiler, N.;and Schneider, H.H. Determination of piperidine in snail brain. Brain Res, 55: 242- 244, 1973. 148 TIMN 0152513

Domino, E.F. Some comparative pharmacological actions of'(-)- nicotine, its optical isomer, and related compounds. Tobacco Alkaloids and Related Corrpounds, 1965. Domino, E.F.,and Yamamoto, K. Nicotine effects on the sleep cycle of cat. Science, 150: 637-638, 1965. Goodman, F.R.,,and Weiss, G.B. Alterations of 5-hydrox,ytryptamine -14C efflux by nicotine in rat brain area slices. Neuropharmacol 12: 955-965, 1973. lIa.ll, G.H.,and Turner, D.M. Effects of nicotine on the release of 3H-Noradrenaline from the hypothalamus. Brit J Pharmacol, 21: 1829, 1972. Hokfelt, T. et al. Aminergic and peptidergic pathways in the ner- vous system with special reference to the hypothalamus. In: Reichlin, S., Baldessarini, R.J., and Martin, J.B., eds. The Hypothalamus. New York: Raven Press, 1978. Larson, P.S.,and Silvette, H. Tobacco, Experimental and Clinical Studies. Supplement 3. Baltimore: Williams and Wilkins, 1975. Murphee, H.B.; Pfeiffer, C.C.;and Price, L.M. EEG changes in man following smoking. Ann N.Y. Acad Sci , 142: 245-260, 1967. Pfeiffer, C.C. Personal conmunication. Philips, C. The EEG changes associated with smoking. Psych _o- physiology, 8: 64-74, 1971. Von Euler, U.S. Piperidine as a normally present constituent of brain. Acta P siol Scand, 8:380-384, 1944. Wei, E.,and Loh, H. Physical dependence on opiate like peptides. Science, 193: 1262-1263, 1976. AUTHORS L.G. Abood, Ph.D.; K. Lowy, M.D.; and H. Booth Center for Brain Research University of Rochester Medical Center Rochester, New York 14642 149

Chapter 12 Tolerance to the Effects of Tobacco Murray E. Jarvik, M.D., Ph.D. The question of whether cigarette scpking or other forms of tobacco use can be considered an addiction comparable to addiction to her- oin or alcohol always sparks a lively debate. For one thing, there appear to be different styles of tobacco use. For another, there is some question about how often heroin or alcohol use results in addiction. And, finally, there is not even uniform agreement on the meaning of "addiction." The definition I prefer is that of Jaffe (1975): "Addiction...(is)...a behavioral pattern of compulsive drug use, characterized by overwhelming involvement with the use of a drug, the securing of its supply, and a high tendency to re- lapse after withdrawal." Tolerance and dependence are two phenomena which are closely re- lated to addiction to heroin and alcohol. In this paper I want to consider evidence concerning tolerance to tobacco and its components. Since prominent theories of physical dependence consider tolerance to be the basis of this state (Jaffe and Sharpless 1968; Martin 1968; Goldstein et al. 1974), this paper may be considered comple- mentary to that of Shiffman (this volume), who deals with the,de- pendence question. I can think of no example of a drug to which dependence occurs which does not also involve tolerance. On the other hand, tolerance may occur without dependence, for example, to phenothiazines, or to antihistami.nes. Tolerance is manifested by a decreasing response to repeated ad- ministration of the same dose of a drug, or by the requirement for increasing doses in order to elicit the same response. Tolerance to different effects of a given drug may vary widely. Of the many thousands of substances contained in tobacco and to- bacco smoke (Schmeltz and Hoffman 1976), three are of primary biological importance: "tar," carbon monoxide, and nicotine. There is evidence that tolerance can develop to the effects of each of these,and their interaction has scarcely been studied. Although there is evidence that tolerance may develop to other components such as acetone and phenol, it is unclear how much they contribute to the pharmacological actions of cigarettes. 150 TiMN 0152515, ,

Three kinds of tolerance are apt to occur with tobacco use as with other drug use: (1) drug dispositional or metabolic tolerance; (2) tissue or pharmacodynamic tolerance; and (3) behavioral tole- rance. The first refers to methods that the body uses to eliminate the drug or to deactivate it. For most chemicals derived from tobacco,the liver is the organ most heavily responsible for detox- ifying or transforming them into inactive and eliminable forms. The kidney is also important,especially for alkaloids whose water solubility varies with the pH of the solution. The second refers to changes in the ability of receptors to be activated by the drug at its final site of action. The third refers to the way in which the subject using the drug changes his behavior to adapt to the effects which the drug repeatedly produces. TAR "Tar" is defined as the total particulate matter (TPM) collected by a Cambridge filter after subtracting moisture and nicotine. The polycyclic aromatic hydrocarbons are generally blamed for a substantial portion of the carcinogenic activity of "tar." They are also powerful enzyme inducers and are undoubtedly responsible for much of the tolerance produced by smoking to themselves and a variety of other compounds. The tar content of cigarette smoke for all brands is determined yearly by the Federal Trade Commis- sion which publishes a listing, along with nicotine content. Tar and nicotine tend to covary and thus their effects may be con- founded. Obviously, tar is obtained in the smoke from pipes and cigars,but not from chewing tobacco and snuff. The latter do not deliver pyrolysis products such as carbon monoxide and may thus be somewhat safer than smoked.tobacco. It is known now that hepatic microsomal enzyme formation is in- duced by a number of carcinogens in the tar fraction of cigarette smoke,including benzpyrene (Oates et al•1975). This means that smokers are rendered tolerant to both the therapeutic and toxic effects of a wide variety of drugs (Swett 1974). Even the enzymes in platelets are activated (Horns et a1.1976). Paradoxically, just because of this type of enzyme induction, cigarette smokers are more resistant to the effects of carcinogens but not enough to prevent a wide variety of cancers. Some examples of the effects of induction of microsomal enzymes are cited by Hunter and Chasseaud (1976). Arylhydrocarbon hydroxy- lase is regularly induced by smoking. Benzpyrene hydroxylase and aminoazo dye N-methylase were higher in placentae of pregnant smoking women than of nonsmokers. Metabolism of benzodiazepines, of propoxyphene, pentazocine and phenacetin is increased in smokers. Xanthines such as theophylline are also metabolized more quickly in smokers (Powell et al. 1977). 151 TIMN 0152516

CARBON MONOXIDE While levels of carbon monoxide achieved in the human body follow- ing cigarette smoking do not appear to be higher enough to materi- ally affect psychological performance (Stewart 1975), the long term effects may be quite significant. There are some (McGill 1977) who contend that the atherogenic effects of cigarettes are largely due to carbon monoxide. Indeed,the chronically high levels of carboxyhemoglobin found in smokers can induce both a polycythemia and increase in hemoglobin levels. These compensatory changes en- able the smoker to cope with the oxygen deficit produced by cigar- ettes. Furthermore, since the latent oxygen carrying capacity of the blood is increased by smoking, sudden cessation should make a smoker more capable of functioning at high altitudes for a short while than a nonsmoker. NICOTINE For most drug dependencies it has been shown that pharmacodynamic tolerance is more important than drug dispositional tolerance. For example, animals and humans tolerant to alcohol and barbitu- rates show significantly less sedation and ataxia than do non- tolerant subjects at the same blood concentrations. Since the liver and the kidneys largely control the blood concentration, one would have to assume that tolerance must be occurring in the brain. Methods of determining blood levels of nicotine with gas chromatography or radioimmuloassay are too recently developed to have produced many studies in tolerance. In a recent study, Jones and colleagues (1978) showed that smokers could tolerate higher intravenous doses of nicotine than nonsmokers. Smokers reported generally pleasanter effects from 700 mcg. of nicotine intravenously than nonsmokers from 300 mcg. In addition to the chronic tolerance demonstrated by the smokers, both smokers and nonsmokers manifested acute tolerance. When the same dose of nicotine was injected at about hourly intervals,each succeeding injection elicited less of a response. This resembles tachyphylaxis,and it may represent occupation of the receptor by the drug or by some other substance such as a catecholamine re- leased by the nicotine, or it may imply depletion of catecholamines. The fate of 1 mg of nicotine base injected intravenously in humans (actually as nicotine hydrogen tartrate) was intensively investi- gated by Beckett (1971). They found that smokers excrete nicotine significantly faster than nonsmokers. None of the smokers reported any nausea from the nicotine injections,but this was reported in varying degrees by all nonsmokers. Haines et al.(1974) reported that the plasma concentrations of nicotine were actually higher in smokers than nonsmokers one minute after smoking,but these results were confounded by the fact that nonsmokers were instructed to smoke cigarettes. Obviously smokers were able to inhale more effectively than nonsmokers in part because they had acquired tolerance to the aversive effects of cigarette smoke on the respi- ratory passages. Indeed,some of the tolerance that smokers show to 152 I J 11TIMN 0152517

w cigarette smoke may be correlated with diminished function of the respiratory epithelium and possible depression of taste and smell (Kittel 1970). The phenomenon of tolerance to tobacco and particularly to nico- tine is reviewed by Larson and Silvette (1961, 1968, 1971, 1975) in their four volumes on tobacco. There is a fair sized litera- ture indicating that tolerance to nicotine can be developed in invertebrates, fish, frogs and birds. It has also been demon- strated in mice, rats, dogs and in one donkey forced to inhale smoke through a nasal catheter. In the latter study a donkey initially exposed to the smoke from 20 cigarettes daily showed marked impairment of bronchio-trachial transport of radioactive iron oxide particles for several months. However, by the end of eight months of exposure compensation had occurred and transport was quite normal. Westfall and Brase (1969, 1971) showed in rats that the tolerance to nicotine-induced elevations of urinary catecholamines is due to increased enzymatic metabolism of these amines. Both monamine oxidase and catechol-o-methyl transferase activity of the liver were increased but nicotine oxidases were unchanged. Schechter and Rosecrans (1972) demonstrated behavioral tolerance to the effects of nicotine in the rat. Nicotine (0.4 mg/kg) was used as the discriminative stimulus in a T-maze. After a crite- rion of eight out of ten correct responses was attained, the animals were given four daily injections of nicotine for five days. The authors found that tolerance occurred,because discrimination became much more difficult as time proceeded. Tolerance to the effects of cigarette smoke was noted in dogs given cigarette smoke via tracheostomy (Hamnond et al. 1970). In our laboratory Stolerman et al.(1973) showed the development of both acute and chronic tolerance in rats. Nicotine administered intraperitoneally to experimentally naive rats depressed activity in a Y-shaped runway in a dose-related manner. After a single intraperitoneal dose of nicotine, acute tolerance to the depres- sant action of a second dose developed with a definite time course, becoming maximal after two hours and wearing off after about eight hours. Repeated intraperitoneal doses of nicotine (3 times daily for 8 days) elicited chronic tolerance, which persisted for at least 90 days after the end of regular treatment with the drug. Tolerance was also produced when nicotine was administered in rats' drinking water and through reservoirs implanted subcutaneously. It appears9then,that tolerance to nicotine in rats can develop quickly, may be measured easily, and persists for prolonged periods after withdrawal. In these experiments rapid withdrawal of nico- tine did not produce the signs of illness which morphine withdrawal regularly produced. The existence of prolonged tolerance to nico- tine in rats suggested that the same phenomenon might exist in man just as it does for opioids. If prolonged tolerance occurred in exsmokers, then it might facilitate relapse, because such individ- uals would be more resistant to the negative effects of cigarettes. 153 i TIMN 0152518

Stolerman et al•(1974) examined the interaction between pairs of injections of nicotine, which varied both in dose and in interval. Two measures of spontaneous locomotor activity of rats in a T-maze were taken: rears and entries. After a single treatment with nicotine acute tolerance developed as indicated by a shift of the dose response curve. The dose of nicotine required to produce a given decrement in activity was multiplied by a factor of about 2.4 when a delay of two hours was taken between the two injections. When the initial dose was varied,it was found that there was an optimal level for producing tolerance. Higher doses were less effective. An explanation for the relative ineffectiveness of the higher doses in producing tolerance is not available. A general debilitating effect of pretreatment with large doses doesn't seem to explain it, since rats given a saline challenge exhibited normal motor activity. Perhaps the debilitating effects of a large pretreatment dose and a challenge somehow summate. QUESTIONS FOR FURTHER RESEARCH The phenomenon of tolerance to the effects of tobacco products has been clearly demonstrated both in humans and in animals. Naturally enough.,most of the emphasis has focussed upon nicotine, but carbon monoxide and tar components also play an important role. As with all other drugs,tolerance varies with subjects and functions. Certain invertebrate forms which feed on the tobacco plant have a high genetically determined tolerance. It is reason- able to assume that even in humans some of the variance in re- sponse to tobacco is innately determined and may account for some of the high concordance in smoking behavior seen in identical twins. Other forms of tolerance are clearly the result of experi- ence and develop after exposure to tobacco products. Much more research needs to be done to determine the degree of tolerance which develops in different physiological and psychological func- tions after tobacco use. For example, it is evident that even in heavy smokers of long duration the heart rate speeds up after each cigarette. On the other hand, at least nausea and vomiting diminish and disappear with continuing moderate use of cigarettes. It would be very informative indeedd to know what changes take place at the putative sites of action of nicotine with chronic use. Do nicotinic synapses at ganglia change in the same way as nicotinic synapses in the brain? Do carbon monoxide and tar constituents have any action on these components or on enzyme systems elsewhere in the body? Answers to these questions will enable us to understand better the physiological basis of the smoking habit. 154

REFERENCES Beckett, A. H.; Gorrod, J. W.;and Jenner, P. The effect of smoking on nicotine metabolism in vivo in man. J Pharm Pharmacol, 23:62S-67S, 1971. Goldstein, A.; Aronow, L.; and Kalman, S. M. Principles of _D_~ru Action: The Basis of ~Pharm_a~cology. Second Edion. New York: John Wiley and Sons, 1 7 Haines, C. F.; Mahajan, D. K.; Miljkovic, D.; Miljkovic, M.; Vesell, E. S. Radioimmunoassay of plasma nicotine in habituated and naive smokers. Clin Pharmacol Ther, 16(6):1083-1089, 1974. Hammond, E.'C.; Auerbach, 0.; Kirman, D.; and Garfinkel, L. Effects of cigarette smoking on dogs. Arch Environ Health, 21: 740-753, 1970. Horns, D. J.; Gerrard, J. M.; Rao, G. H. R.; Krivit, W.; and White, J. G. Smoking and platelet labile aggregation stimulating substance (LASS) synthesizing activity. Thrombosis Research, 9(6):661-668, 1976. Hunter, J., and Chasseaud, L. F. Clinical Aspects of microsomal enzyme induction. In: Bridges, J. W., Chasseaud, L. F., eds. Progress in Drug Metabolism. Vol. 1. New York: John Wiley and Sons, 1975. pp. 12 -. Jaffe. J. H. Drug addiction and drug abuse. In: Goodman, L. S., and Gilman, A. The Pharmacolo ical Basis of Therapeutics. Sth Ed. New York: Macmillan Pub~'ishing Co., Inc., 1975. p. 285. Jaffe, J. H., and Sharpless, S. I. Pharmacological denervation supersensitivity in the central nervous system: a theory of physical dependence. Proc Ass Res Nerv Ment Dis, 46:226-246, 1968. Jones, R.T.; Farrell, T.R. III; Herning, R.I. Tobacco smoking and nicotine tolerance. In: Krasnegor, N., ed. Self-Administra- tion of Abused Substances: Methods for Stud . Natio'-'-iialnst~ltute on Erug ATiuse 12eseac onograT7O-bub. No. (AIb4)78-727. Washington, D.C.: Superintendent of Doctmients, U.S. Government Printing Office, 1978. pp. 202-208. Kittel, G. Moeglichkeiten der Olfaktometrie. Ermuedungsmessungen bei Rauchern. Z Laryng Rhinol Oto1,49:376-386, 1970. Larson, P. S.; Haag, H. B.; Silvette, H. Tobacco, Experimental and Clinical Studies. Baltimore: Williams and Wilkens Co., Mi . Larson, P. S., and Silvette, H. Tobacco, Experimental and Clinical Studies. Supplement I. Baltimore: Williams and Wilkins Co., 1968. 155

Larson, P. S., and Silvette, H. Tobacco, Experimental and Clinical Studies. Supplement II. Baltimore: Williams and Wilkins Co., 1971. Larson, P. S., and Silvette, H. Tobacco, Experimental and Clinical Studies. Supplement III.Baltimore: Williams and Wilkins Co., 1975.. Martin, W. R. A homeostatic and redundancy theory of tolerance to and dependence on narcotic analgesics. Proc Ass Res Nerv Dis, 46:206-225, 1968. - ^ McGill, H. C., Jr. Atherosclerosis: Problems in pathogenesis. Atherosclerosis Review, 2:27-65, 1977. Oates, J. A.; Azarnoff, D. L.; Cohen, S. N.; Melmon, K. L. Medicinal misadventures. Emer en Medicine, 7(7):115-118, 121, 125, 129, 133-134, 136-137,1 Powell, J. R.; Thiercelin, J. F.; Vozeh, S.;Sansom, L.; and Riegelman, S. The influence of cigarette smoking and sex on the- ophylline disposition. Am Rev Respir Dis, 116(1):17-23, 1977. Schechter, M. D., and Rosecrans, J. A. Behavioral tolerance to an effect of nicotine in the rate. Arch Int Pharmacodyn Ther, 195:52-56, 1972. ~ Schmeitz, I., and Hoffmann, D. Chemical studies on-tobacco smoke. XXOCVIII. The physicochemical nature of cigarette smoke. In: Wynder, E.L.; Hoffmann, D.; and Gori, G.B., eds. Proceedin s, 3rd World Conference Smok' and Health. Vol. 1: i ing the isk-Tor eitf smoTcer. ETicati-'on-Vo. (NIH) 76-1221. Washing- ton, D.C.: U.S. Government Printing Office, 1976. pp. 13-34. Stewart, R. D. The effect of carbon monoxide on humans. Annu Rev Pharmacol, 15:409-423, 1975. Stolerman, I. P.; Bunker, P.; and Jarvik, M. E. Nicotine tole- rance in rats: Role of dose and dose interval. Psychopharma- col_~o,ia, 34:317-324, 1974. Stolerman, I. P.; Fink, R.; and Jarvik, M. E. Acute and chronic tolerance to nicotine measured by activity in rats. Psychophar- macologia, 30:329-342, 1973. Swett, C., Jr. Drowsiness due to chlorpromazine in relation to cigarette smoking. A report from the Boston Collaborative Drug Surveillance Program. Arch Gen Psychiat, 31:211-213, 1974. 156 TI1MN 0152521 ;

Westfall, T. C., and Brase, D. A. Adrenal stimulation and mono- amine oxidase (MAO) activity following daily nicotine treatment. Fed Proc, 28(2):287, 1969. Westfall, T. C., and Brase, D. A. Studies on the mechanism of tolerance to nicotine-induced elevations of urinary catechola- mines. Biochem Pharmacol, 20:1627-1635, 1971. AUTHOR Murray E. Jarvik, M. D. Ph. D. Professor of Psychiatry and Pharmacology Neuropsychiatric Institute University of California School of Medicine 760 Westwood Plaza Los Angeles, California 90024 Chief, Psychopharmacology Unit Veterans Administration Hospital Brentwood Wilshire and Sawtelle Boulevards Los Angeles, California 90073 157 rTIMN 0152522

Chapter 13 The Tobacco Withdrawal Syndrome Saul M. Shiffman, M.A., C.Phil. Characterizing tobacco use as a dependence process necessarily raises the issue of tobacco withdrawal, as the presence of an absti- nence syndrome is crucial to the definition of drug dependence. In- deed, some of the initial reluctance to label tobacco as a dependence- producing substance rested on doubts concerning the existence of a tobacco withdrawal syndrome. This was the position taken by the Sur- geon General in 1964, when first alerting the country to the dangers of tobacco. Since then, there has been an accumulation of studies which suggest that withdrawal from tobacco does produce a variety of signs and symptoms which can be characterized as a tobacco with- drawal syndrome. Although the syndrome is variable and is only roughly described and understood, its existence is no longer a matter of great controversy. The focus has shifted from the somewhat seman- tic question of categorizing tobacco use to exploration of the particulars of tobacco dependence and withdrawal. This paper reviews the findings on tobacco withdrawal syndrome, focusing initially on a description of the syndrome, and then on factors affecting its severity and course. Finally, noting methodological issues which need to be considered, goals and directions for further research are discussed. SYMPTCt~r.S OF TOBACCO WITfIDRAWAL A number of physiological changes have been observed on withdrawal from tobacco. Decreases in heart raxe and diastolic blood pressure are observed as early as six hours after withdrawal (Murphee and Schultz 1968). These changes persist for at least three days and perhaps for thirty (Weybrew and Stark 1967; Glauser et al. 1970). Decreased excretion of both adrenaline and norepinephrine and vari- ous metabolic changes have also been observed (Myrsten et al. 1977; Glauser et al. 1970). These indications of decreased arousal have been supported by EEG studies showing a decrease in the peak alpha frequency and increases in low frequency activity (Knott and Venables 1977; Ulett and Itil 1969). Increases in sleeplike acti- vity have been observed in the EEG records of deprived smokers (Itil et al. 1971). It has also been demonstrated that, in sleep, 158 TIMN 0152523

deprived smokers show decreases in RFM latency and increases in REM time (Kales et al. 1970). In contrast, however, Knott and Venables (1978) report that deprived smokers showed hypersensitivity to low intensity stimuli, as indicated by increased amplitude and decreased latency in the evoked response. Thus, while withdrawal from tobacco appears to produce physiological changes suggestive of decreased arousal, this is not a uniform or easily characterized pattern. Another documented physical change resulting from abstinence is weight gain. Wilhemsen (1968) observed a weight gain of 2.4 kg in two months of abstinence, while Glauser et al. (1970) documented an increase of 3.2 kg in one month. In Fletcher and Doll's (1969) retro- spective study of British physicians, the respondents reported gain- ing as much as 5.3 kg on the average. Although the weight gain itself is thus well documented, it is controversial whether it re- sults from increased appetite and consumption or from metabolic changes (Glauser et al. 1970). It has been demonstrated that with- drawal impairs the performance of smokers on psychomotor tasks re- quiring vigilance and/or tracking (Heimstra 1973). Not surprisingly, deprived smokers do poorly in a driving simulation task (Heimstra et al. 1967). The effects of tobacco deprivation on verbal learning are complex, with indications of both facilitation and impairment. Kleinman, Vaughn and Christ (1973) found that 24-hour deprivation facilitated learning of easy paired associates, but impaired perform- ance with hard pairs. They interpret this as an indication of hyper- arousal, while Andersson (1975) reports mixed facilitation and im- pairment and interprets this as indicating shifting arousal. Myrsten et al. (1977) found verbal learning essentially unaffected over five days of abstinence. Thus, while psychomotor perfoYmance is reliably Maired by abstinence, little can be said about its effects on bal learning. Accompanying these objective changes iinn physiology and performance are subjectively reported changes in physical symptoms, arousal, and mood. These have been reported in studies of smokers sampled while actually undergoing withdrawal, as well as in retrospective studies of exsmokers up to 14 years after cessation. Although the specific symptoms reported differ, as does the percentage of absti- nent smokers reporting each symptom, a consistent pattern of symptoms can still be discerned. Cormnon among the physical symptoms reported are nausea, headache, constipation, diarrhea, and increased appetite. Also reported are disturbances of arousal, including drowsiness and fatigue, as well as insomnia and other sleep disturbances (e.g., Guilford 1966). Inability to concentrate is a common complaint and is consistent with objective assessments of the concentration of smokers in abstinence (Perlick 1977; Heimstra 1973). Thus, the ob- jective changes reviewed above appear to be reflected in the subjec- tive experience and self-reports of deprived smokers. Chan.ges in mood or affective tone are prominent among the symptoms reported in withdrawal from tobacco. Increased irritability and hostility are common sequelae of abstinence from tobacco (Friedman 1972; Mausner 1970). Increases in aggressive behavior and its cor- relates have also been observed,in abstinence (Schechter and Rand 159 ; TIMN 0152524

1974; Hutchinson and Fmley 1973). Increased anxiety is another af- fective change which is both reported and observed (Friedman 1972; Nesbitt 1973). In addition to reports of these symptoms under neu- tral conditions, several studies suggest that deprived smokers are more likely to respond with anxiety and irritability when stressed (Myrsten et al. 1972; Frankenhaeuser et al. 1971). This may only be true compared to nondeprived smokers, however, with smokers showing decreased vulnerability to these dysphoric reactions to stress (Heimstra 1973). In surtanary, withdrawal from smoking shifts the smoker's affective tone toward dysphoria, and increases anxiety and irritability. By far the most common and clinically most important symptom appear- ing following withdrawal from tobacco is craving for tobacco. The best estimates indicate that 90 percent of all smokers in withdrawal will crave (Guilford 1966). Moreover, among smokers who have been abstinent for five to nine years, one out of five report that they continue to have at least occasional craving for tobacco (Fletcher and Doll 1969). The importance of craving lies not in its univer- sality or persistence, but in its relation to the clinical goal of modifying smoking behavior. Indeed, the importance of the tobacco withdrawal syndrome in its entirety is based on its provocative role in causing relapse among abstinent smokers. 3M0KING CESSATION, RELAPSE, AIVI) MAINTENANCE Since tobacco use--particu7.arly cigarette smoking--was identified as a major cause of "preventable" morbidity and mortality, concerted attempts have been made to apply behavioral technology to the prob- lem of smoking cessation, and, indeed, have become very successful in producing cessation. At least one method--rapid smoking--is claimed to produce cessation in more than 90 percent of clinic pa- tients (Lichtenstein and Danaher 1976). Unfortunately, very little headway has been made in producing long term maintenance of nonsmok- ing. It has been estimated that, on the average, only about 25 per- cent of those who successfully quit smoking will remain abstinent for more than six months (fAmt and Matarazzo 1973). The focus of clinical efforts in the treatment of smoking thus needs to shift from producing cessation to encouraging maintenance and preventing relapse. It is self-evident that craving for cigarettes is a major cause of relapse. The urge to smoke, when it becomes stronger than the ex- smoker's determination to quit, leads to relapse. In a retrospec- tive study of treated smokers 18 months after treatment, Peterson et al. (1968) found that over half of those who had relapsed cited craving specifically as the cause. Other withdrawal symptoms also play a role in relapse. In Peterson et al.'s study, 3S percent of the recidivists cited anxiety as the cause of relapse, while 12 percent cited weight gain. In my anal- ysis of questionnaire data collected by C. Hammen (personal comrnmi- cation 1973), I have found that 61 percent of smokers who were ap- plying for treatment cited either anxiety or craving as the cause 160 4TIMN 0152525

for relapse in previous attempts. This is in line with estimates by Guilford (1966) and Burns (1969). By providing immediate aver- sive consequences for smoking cessation, withdrawal symptoms thus make continued abstinence difficult, and relapse probable. FACTORS AFFECTING THE ABSTINENCE SYNDROME Thus far, the emphasis has been on the uniformity of reports concern- ing the tobacco abstinence syndrome. The impression might be drawn that every smoker, upon withdrawal from tobacco, becomes irritable and anxious and is unable to think, work, sleep, drive, or carry on normal social discourse for want of a cigarette. Fortunately, this is not the case. The tobacco withdrawal syndrome is apparently quite variable in character, severity,•and duration. An examination of va- rious studies reveals large differences in the frequency with which particular symptoms are reported. Where Trahair (1967), for example, reports that sleep disturbance was experienced by only ten percent of exsmokers, Guilford (1966) reports that 31 percent of her subjects suffered from insomnia. Similar examples of variability are com- monly reported in the literature on smoking withdrawal. About 23 percent of abstinent smokers report no symptoms at all (Myrsten et al. 1977; Pederson and Lefcoe 1976; Wynder, Kaufman and Lesser 1967). While this has led some to conclude that there is no absti- nence syndrome (e.g. USDHEW 1964), this conclusion is contradicted by a mass of data, some of which has been cited above. A simpler conclusion is that the abstinence syndrome is quite variable. This impels us to tease out the causes of this variability. Re- search on the factors which affect the tobacco withdrawal syn- drome is the major focus in the remainder of this paper. Recent work in our laboratory at UCLA has focused on factors which may affect the tobacco withdrawal syndrome. We have developed a 25 item questionnaire which measures the five symptom clusters which comprise the syndrome: physical symptoms, psychological symptoms, arousal, appetite, and craving for cigarettes. (Our appetite scale subsequently proved to be unreliable and therefore does not figure in our analysis.) Subscales assessing each of the symptom clusters were created using factor analytic methods. The questionnaire was then used to study sources of variance in the tobacco withdrawal synftome in 40 smokers who were attempting to quit smoking. These subjects c6mpleted the questionnaire at the beginning of cessation and then four times daily thereafter. (Most of the data reported here are from a study published in part as Shiffman and Jarvik 1976.) Baseline Cigarette Consumption It is characteristic of withdrawal syndromes that their severity is dose dependent (Jaffe 1971). Therefore, it is expected that heavy smokers would report more severe withdrawal symptoms than light smokers. A comparison of subjects smoking a pack or more (+x=26.6) per day with those smoking less than a pack (x=15.7) per day, how- ever, revealed no overall differences in the severity of their symp- toms.l This replicates our previous finding that light and heavy 161

smokers do not differ in craving when deprived for 48 hours (Gritz and Jarvik 1973). In work currently under way on the effects of brief deprivation (e.g., a two hour lecture), we have found some dose-related effects on craving (Gritz 1978). The inconsistency of this effect in our studies parallels the state of the literature. Studies by Myrsten et al•(1977) and by Mausner (1970) also report finding no differences between light and heavy smokers. In contrast, Burns (1969) reports that subjects who re- ported withdrawal symptoms had smoked an average of 6.9 cigarettes/ day more than asymptomatic subjects (p<.01). Since this is smaller than the mean difference between the'light and heavy smoker groups in our study (11.1), this discrepancy is puzzling. Wynder, Kaufman and Lesser (1967) report that the proportion of abstinent smokers reporting more than one withdrawal symptom increases with baseline consumption. A chi squared analysis performed on their data reveals that this association is highly significant (x2=15.81, df=2, p<.001). Perhaps these discrepancies can be explained by the fact that Burns and Wynder et al.studied the number of symptoms, while we studied the severity of each symptom separately. Another possiblee confound is that, because smokers can vary their smoking consumption in other ways -- depth of inhalation, number of puffs, etc. -- cigarette consumption may actually be a very poor measure of dose. Also, differences in nicotine metabolism intro- duce variability in dose even among those who consume similar amounts of nicotine. Thus, estimating a smoker's dose may require measuring serum levels of nicotine or its metabolites. In the one study which has approached this, Zeidenberg et al.(1977) found a high and signi- ficant correlation between serum cotinine levels before treatment and self-reported "degree of difficulty" in smoking cessation. This result held for men only, however, perhaps because of additional va- riability in the cotinine levels or the self-reports of the female subjects. There is some indication that the severity of the absti- nence syndrome is dose-dependent, but much ambiguity remains. Be- cause dose dependency is so characteristic of withdrawal syndromes from other substances, establishing this effect for tobacco would be an important step towards an understanding of tobacco dependency. Further research into the relationship should probably proceed along the lines of Zeidenberg et al., using serum cotinine levels rather than cigarette consumption as the independent variable. Dependent. measures should include more refined instrtmients than Zeidenberg et al:s estimates of "difficulty" and should explore both the number of withdrawal symptoms and their severity. Diurnal Variations in Symptoms Since our subjects were providing reports of symptoms at regular intervals throughout the day, we were able to study diurnal vari- ations in symptcros. One puzzling finding which emerged was that, while light and heavy smokers reported the same average level of arousal, they differed significantly in the diurnal pattern of arousal. 162 t 1TIMN 0152527

. FIGURE 1 6.0 55 50 MEAN RATING 5.5 40 3.5 3.0 AROUSAL o•HEAVY SMOKERS .7 8 9 10 1112 1 2 3 4 5 8 7 8 9 10 11 AM NOON PM APPROXIMATE TIME OF DAY • Mean arousaZ ratings of Zight and heavy smokers for four times of day. Ratings are based on the average of four questionnaire items reZating to arousaZ or atertness (e.g., "Do you feeZ unusuaZZy tired?"). Each data point represents an average of 12 days' ratings. The curves for the two groups differ significantZy in Zinear and quadratic trends. Figure 1 plots arousal.by time of day for the two groups, with data points at 7 AM, noon, 7 PM, and 11 PM, approximately. The figure shows that the two groups report the same level of arousal upon awakening. Whereas the heavy smokers then increase in arousal dur- ing the course of the day, the light smokers maintain this minimal level of arousal throughout the day. Both groups show a drop in arousal at bedtime, with the light smokers continuing to report lower arousal. This result is difficult to interpret, particu- larly because this scale does not have a clear evaluative dimen- sion. It is not clear, for example, whether the heavy smokers are hyperaroused during the day, or the light smokers underaroused. At bedtime, it appears that the heavy smokers may be suffering insomnia, since they are reporting nearly as much arousal as they do upon awak- ening in the morning. This cannot be determined definitively from the data, however, and this effect remains to be explored. 163 TIMN 0152528 1

The other symptom cluster which shows significant diurnal variation is craving. Craving is at its lowest point when the subject wakes up, gradually rising to a peak in the evening, then falling again at bedtime. FIGURE 2 43 42 MEAN RATING + OF CRAVING 4.1 o-----a 3.9 3.9 44 k60 F30 h40 101-30 h20 AM NOON PM A.PPROXIMATE TIME OF DAY CRAVING ..UNPLOTTEO DATA POINTS FROM MEADE & WALD (1977) t 0I 4 7 6 9 10 II 12 1 2 3 4 3 6 7 9 910 11 IO 7 6 3 NUMBER OF CIGARETTES 4 PER HOUR 3 2 I O ,9moking, craving, and reZapse as a function of time of day. Rat- ings are based on the average of seven questionnaire items reZating to craving for cigarettes (e.g., "Do you have an urge to smoke a cigarette right now?"). Note simiZartities among these funetions, aZZ of which peak in the evening. Data on reZapse from retrospective interviews (Lichtenstein et aZ. 1977). Data on smoking from obser- vation of cZericaZ personneZ,(Meade and WaZd 1977). Data on craving from our questionnaire study of smokers in ruithdravaZ. Figure 2 shows this diurnal variation in craving. Plotted on the same axis is data collected by Meade and Wald (1977) on the ciga- rette consumption of British office workers at similar times of day. The figure shows that craving in abstinent smokers and ad lib smoking have the same diurnal function. A third function plotted on the same axis draws on data reported by Lichtenstein, Antonuccio, and Rainwater (1977), who surveyed a group of smokers who relapsed following a period of abstinence. The curve shows the percentages of recidivists who said they had first relapsed in the morning, afternoon, and evening. This function parallels the diurnal functions of ad lib smoking and craving in withdrawal. Thus, there is a consistent function which describes three differ- ent stages of the habit and its control (unrestricted smoking, ab- 164 SMOKING, CRAVING, AND RELAPSE TLMN 0152529 ~

_1 stinence, and relapse). The meaning of the underlying function has not been determined. Two different types of explanation are plausible. One focuses on diurnal variation in the internal en- vironment of the smoker, suggesting the influence of some metabolic factor with diurnal variation2. The other explanation focuses on the diurnal variation in the social environment: e.g., the timing of work, meals, social contact, recreation, and so on, which affects craving for tobacco. Researchh which accurately measures craving and relates it to environmental stimulus events and circadian variations in the internal environment could decide between these explanations. A more comprehensive understanding of how craving varies with stimu- lus events and with time of day might prove helpful in designing in- terventions which help prepare the smoker to cope with his/her craving. Time Course and Duration While the time course of the abstinence syndrome following abrupt withdrawal from other dependence-producing substances has been sys- tematically studied (Jaffe 1971), assessment of the course of the tobacco withdrawal syndrome is made difficult by the subtlety and variability of the symptoms (USDHEW, 1964). The onset of the syndrome appears to be rapid, with changes in mood (Schechter and Rand 1974) and performance (Myrsten et a1.1972) evi- dent as early as two hours after withdrawal. However, these early effects are not easily distinguishable from the absence of nicotine effects or the effects of simple frustration. Our research focused on the course of withdrawal symptoms over the first two weeks of ab- stinence. Figures 3 - 6 show the progression of the four symptom categories -- craving, arousal, physical symptoms, and psychological symptoms -- over a twelve-day period. (These curves are for all subjects and therefore obscure some im- portant group differences which are discussed belbw.) Note that these symptoms all decrease at first and then either level off or rise again in the second week of abstinence. (The quadratic curvature of each curve is statistically significant.) This initial pattern of symptoms is consistent with Herold's (1967) findings concerning the percentage of deprived smokers reporting craving, irritation, tiredness, and headache in the first five days following abrupt withdrawal. Each symptom shows a sharp decline during this period, and all show signs of leveling off by day ten. Other studies also report data suggesting a similar decrease in symptoms over time. (Lichtenstein, Antonuccio and Rainwater, personal commmica- tion 1978; Guilford 1966). 165 TIMN 0152530

FIGURES 3 and. 4 2sr PHYSICAL SYMPTOMS x 2.8 \ MEAN RATING 27 x (ALL SUBJECTS) 2.6 2.5 4.8 4.6 MEAN , RATING 4.4 (ALL SUBJECTS) 42 z 2-3 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE PSYCHOLOGICAL SYMPTOMS 4.0 2-3 4-5 5 6-7- , 8-9 10-11 12-13 DAYS IN ABSTINENCE Mean reported severity of physicaZ'and psychoZogicaZ symptoms over 12 days of rvithdramaZ. After an initiaZ decrease, these symptoms ZeveZ off. Ratings are based on the average of three items reZating to physicaZ symptoms and four items reZating to psychoZogicaZ dis- comfort, respectiveZy (e.g., "Is your heart beating faster than usuaZ?"; "Are you more nervous than usuaZ?"). Days have been paired on ordinate in order to smooth the curves. Note smaZZ magnitude of changes in ordinate scaZes. 166

FIGURES 5 and 6 DROWSINESS ("AROUSAL" INVERTED) 4.3r 4.4 MEAN 4.5 RATING (ALL SUBJECTS) 4.6 4.7 4.8 t ~--f 2-3 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE 4.6 4.4 MEAN 42 RATING (ALL SUBJECTS)4A 3.8 x \x CRAVING " 2-3 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE Mean drowsiness and craving ratings after 12 days of rvithdrarvaZ. Note initiaZ drop and subsequent leveling off (Fig. 6) or rise (Fig. 5). Drowsiness scale is simply an inversion of arousal ratings. Days have been paired in order to smooth the curves. Note small magnitude of changes in ordinate scales. 167

Thus, after a marked decline in the first week, the tobacco withdrawal syndrome becomes increasingly less yielding. In- deed, figures 3-6 suggest that an aggravation of symptoms oc- curs after the second week of abstinence, and leaves open the question of the syndrome's subsequent course. Estimates of the tobacco withdrawal syndrome's duration have been made in retrospective studies which ask exsmokers to recall how long their discomfort or "difficulty" lasted. However, these stu- dies produce contradictory findings. Burns (1969) reports a range from one to twelve weeks, and Wynder, Kaufman, and Lesser (1967) report that most s toms were gone after four weeks. In contrast, Mausner (1970 reports that, of the exsmokers who ventured an estimate, fully two-thirds stated that their difficulty had lasted between one month and five years! In another retrospective study, 21 percent of the sample of exsmokers reported at least intermittent craving for ciga- rettes five to nine years after cessation (Fletcher and Doll 1969). Thus, the duration of the tobacco withdrawal syndrome appears to be extremely variable, and no definitive estimate is yet available. Degree of Deprivation Even with continued use, reduction in the dose of a dependence- producing substance typically results in the emergence of a withdrawal syndrome (Jaffe 1971). It has been shown that smokers changed to low-nicotine cigarettes often report the gamut of withdrawal symptoms described above (Schachter 1977; Finnegan, Larson,'and Haag 1945). In order to study the with- drawal syndrome in partially deprived smokers, we divided our sample into two groups: one was composed of totally abstinent smokers who quit "cold turkey" and remained abstinent for the entire period of the study; the second group consisted of partially abstinent smokers who reduced their cigarette con- sumption by an average of 60 percent. With the exception of the craving scores--which were lower in the cold turkey sub- jects--we found no overall differences in the severity of the reported symptoms. We found that the two groups did differ in the course of the withdrawal syndrome. Figures 7-T-hs ow the progression of physical symptoms, craving, and psycholo- gical symptoms as abstinence proceeded. In each case, the totally abstinent subjects show a greater reduction of symptoms early in the abstinence period. 168 TAN 0152533

FIGURE 7 C RAV ING 4,8 44 x x_",. . 4.0 x. PARTIALLY ABSTINENT Ss MEAN RATING 3.6 o= TOTALLY ABSTINENT Ss 3.2 2.8 ~--ffJ- 2-3 i 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE Mean craving scores for partiaZZy and totaZZy abstinent smokers over 12 days of abstinence. Note the sharp decZine of craving in subjects who quit"coZd turkey" and remained abstinent. Days have been paired in order to smooth the curve. 169 ~TIMN 0152534

FIGURES 8 and. 9 3.0 2.9 2.8 MEAN RATING 2.7 2.6 2.5 PHYSICAL SYMPTOMS x= PARTIALLY ABSTINENT Ss o=TOTALLY ABSTINENT Ss . ~/'- --- - I 2-3 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE 4.8 4.6 MEAN 4.4 RATING 4.2 4.0 PSYCHOLOGICAL SYMPTOMS ~ 2-3 4-5 6-7 8-9 10-11 12-13 DAYS IN ABSTINENCE Mean reported severity of physicaZ and psychoZogicaZ symptoms for partiaZZy and totaZZy abstinent smokers over 12 days of abstinence. Note the sharp decZine in symptom severity reported by the totaZZy abstinent sub,jects, who initiaZZy report more severe symptoms. Note smaZZ magnitude of changes in ordinate scate. 170 TIMN 0152535

Notice that early in abstinence the two groups are reporting simi- lar levels of discomfort, but that four days later the ccaRpletely abstinent group is reporting less craving and fewer physical and psychological symptoms. Thus, abrupt and total withdrawal from tobacco is associated with a withdrawal syndrome that is no worse than that seen in partial abstinence, and which subsides more quickly. One might speculate that continued smoking results in a prolonga- tion of the abstinence syndrome, while complete abstinence re- sults in a relatively rapid decrease in symptoms. Such causal interpretations of our data must be made with great caution, of course, because cur groups of partially and totally abstinent subjects were formed post hoc. Therefore, it may be that the course of the withdrawal s~i7irane is determined by some other factor, and the degree of abstinence is in turn determined by sane combination of these factors. Nevertheless, there are findings in our study and others which at least suggest that the degree of abstinence does indeed control the course of the withdrawal syndrome. As figures 7-9 show, the two groups typi- cally begin abstinence with very similar levels of discomfort, and only diverge after a few days of abstinence. Moreover, in- formal analysis of our data suggests that craving rises when a subject has been abstinent and then begins to smoke again.• More systematic observation of this phenomenon was made by Lichten- stein, Antonuccio, and Rainwater (personal communication 1978) in a retrospective study of relapsed smokers. They found that 53 percent of their subjects reported that their craving de- creased with time, as long as they were abstinent. In contrast, 42 percent reported that their craving was increased the day after a relapse episode (only 18 percent reported decreased craving). Similar phenanena are well known in other dependency disorders; the alcoholic's proverbial "one drink" is an example. Thus, craving leads to smoking and smoking leads to craving in a cycle of dependence. Gradual Reduction and Chronic Withdrawal This has important implica.ticns for methods of smoking cessation. Despite the usefulness of gradual withdrawal in other dependency disorders, and despite the congruence of this method with sound behavioral principles, there is considerable evidence suggesting that gradual withdrawal from tobacco is associated with treatment failure (Guilford 1966; USDHEW 1975; Mausner 1970). Our findings.suggest an explanation for this discrepancy. Par- tial abstinence from smoking leads to more, rather than less, discomfort in withdrawal, since the witTidraw`al syndrome experi- enced by those who "cut down" on smoking (by 60 percent) is as severe as that in an abrupt cessation group and lasts longer. The result is that a partially abstinent smocer is in a chronic state of withdrawal. That such a condition is possible is con- firmed by Perlick (1977) who observed "restrained smokers" who were keeping their cigarette consumption chronically at about 171 _ 0152536

50 percent of their normal intake. These smokers were compared to unrestrained smokers while smoking high and low nicotine cigarettes and no cigarettes. Perlick found that, even when allowed to smoke high nicotine cigarettes, the restrained smokers showed as much irritability as normal smokers deprived of cigarettes. Moreover, they also showed impairments of con- centration that were even greater than those observed in heavy smokers. In other words, they showed signs of a severe absti- nence syndrome even while smoking "ad lib." Thus, the ciga- rettes indulged in by smokers who attempt to "cut down" may serve only to prolong their withdrawal by intermittently re- inforcing their symptoms and smoking behavior. Typically, this chronic state of withdrawal leads to relapse and return to baseline rates of smoking. Although this explanation is plausible and fits the data avail-able, it must be treated with caution pending further research, Since all of the research relies on smokers who have chosen whether to quit cold turkey or by gradual reduction, there is still the possibility that smokers in some way predisposed to experience a protracted withdrawal syndram disproportionately choose the gradual reduction method. What is needed is experi- mental research in which smokers are randcan].y assigned to.cold , turkey or gradual reduction groups, and the effects on the course of the abstinence syndrome evaluated. Another direction for new research might be to determine the threshold for the onset of the abstinence syndrome in gradual reduction. Perhaps there is some rate or degree of reduction which would not precipitate withdrawal, so that a smoker could be weaned from tobacco. In addition to aPrate of reductio' n'~ parameter, the onset of severe withdrawal may also be controlled by the absolute dose, as well. It has been reported that in gradual reduction programs, smokers tend to meet plateaus or "stuck points " at certain levels of consumption, even when the rate of reduction is sl av and controlled by the smoker. Levinson et a1.(1971) report that the most difficult "stuck point" was at 12 cigarettes.per day, and that most of the premature terminations of treatment occurred at this level, An intriguing possibility is that this constitutes scme minimum dose below which withdrawal is precipitated. In any case, the relationship between degree of tobacco deprivation and the emer- gence of withdrawal symptoms deserves further study. Other Factors Which May Affect the Abstinence Syndrome In addition to the factors already cited, the tobacco with- drawal syndrome may be affected by a number of other variables whose influence remains to be determined. One could speculate, for example, on differences between types of smokers in the se- verity, pattern and/or course of abstinence. A study by Ikard and Tomkins (1973) suggests rather tautologically that "addictive smokers" experience more severe craving. The smokers in this study were only deprived of tobacco,for three hours, however, 172 TIM~ 0152537

so that the effects of this typology on the clinical abstinence syndrome are still essentially unknown and deserving of study. Other individual difference variables also deserve study. A smoker's smoking history, for example, may affect the withdrawal syndrome, especially considering such variables as previous at- tempts to quit and the reason for failure. Since the symptoms of withdrawal are relatively ill-defined, the smoker's expecta- tions and set are probably related to his/her experience of ab- stinence, as is his/her motivation to quit (cf. Barefoot and Girodo 1972). There is another major factor whose relationship is potentially imporeant but unexpected: Fragmentary evidence suggests that the abstinence syndrome is more severe in women than in men. Unfortunately, relevant data are too seldom analyzed for this sex difference. For example, Guilford (1966) reports data separately by sex, but does not.submit it to statistical analysis of the sex difference. Yet, of 18 major symptoms reported by her subjects in the first four days of abstinence, 15 show some sex difference. Among these 15, 13 are more frequently reported by women. This difference is statistically significant (sign test, N=15, r-`-2, p<.005). Moreover, a chi square analysis of retrospective data collected by Hammen (1973 personal communication) shows a trend (p<.10) for women to report more frequently than men that their past relapses were due to anxiety or craving. Data reported in a number of other studies line up in the same direction, though the effect fails to reach significance in the individual studies (Trahair 1967; Peterson et al. 1968; Wynder, Lesser,. and Kaufman 1967). It seems likely, then, that women suffer from a more severe abstinence syndrome than men. The importance of this finding lies in its possible relation'to another sex difference in smoking cessation: it is well established that women are more likely to fail in smoking cessation efforts (Gritz 1978). Guilford (1966) has also presented data suggesting that the rela- tionship between withdrawal symptoms and failure in smoking ces- sation is stronger for women than for men. Thus, women experi- ence more discomfort in withdrawal and are more affected by it in their attempts to quit smoking. It seems likely that this is at least partly responsible for their reduced rates of successful cessation. Nor are organismic variables the only variables relevant here. The method used to achieve cessation may well have an effect on the subsequent withdrawal syndrome. Environmental factors, such as the smoker's social environment, are potentially powerful de- terminants of the smoker's experience of withdrawal. Many smo- kers cite a stressful event or situation, rather than an internal feeling, as the precip'tant of relapse Hammen,personal communi- cation 1973). These and other events,such as social drinking, may produce conditioned craving and are to be considered high risk situations for relapse (Lichtenstein, Antonuccio, and Rainwater 1977). Thus, in addition to the few factors whose in- fluence on the tobacco withdrawal syndrome is known, there are many other potentially important variables whose effects remain to be determined. 173 TIMN 0152538

METHODOLOGICAL ISSUES The influence of the factors discussed above on the tobacco with- drawal syndrome also touches on methodological issues. Since these variables represent significant sources of variance in the with- drawal syndrome, they must be taken into account when withdrawal effects are assessed. Research in this area must distinguish between partial and.total abstinence, for example, if the results are to be consistent. This distinction is blurred in many studies, mak- ing the results difficult to interpret or compare. Similarly, data should be reported separately for men and women and for light and heavy smokers. Since some symptoms vary with time of day and probably with stimulus events, they should, where pos- sible, be measured repeatedly and at fixed times during the day. Other methodological issues in research on the tobacco withdrawal syndrome involve the measuring process or operation itself. Be- cause there is no accepted standard, no two studies in the litera- ture use the same assessment device, making the results impossible to compare across studies. Some studies use ratings of symptom severity, where otheis simply use a present/absent symptom check- list method and still others an open-ended query. Consequently, results may be reported as mean ratings or as percentage of sub- jects reporting a given symptom. Often, no distinction at all is made among various symptoms,and subjects are simply asked whether they experienced an symptoms, or, even more globally, whether they experienced " ifficulty" in giving up tobacco. This is a major factor contributing to the variability of estimates of the prevalence and duration of the withdrawal syndrome. Much more clarity would result if the assessment instruments were made homogeneous. A rating scale which taps several symptom clusters, such as the one we have used in our studies, would probably meet the requirements of most investigators. Some measurement difficulties which arise in the study of with- drawal symptoms are kin to the difficulties inherent in measure- ment and operationalization of psychological constructs. Take, for example, the mood changes that are reported in withdrawal. Colloquial terms such as "nervous" and "irritable" are most na- tural and are therefore useful in obtaining self-reports. Un- fortunately , they are also quite ambiguous. It is unclear, for example, whether they refer to tonic, baseline states or to predispositions to overreact to certain stimuli. Does a smoker in withdrawal show higher overall levels of anxiety or a prone- ness to react with greater anxiety when stressed? These two interpretations imply different measurement operations, and measurements of both varieties appear in the tobacco withdrawal literature as measurements of "anxiety" or "irritability," with different findings resulting. In order to avoid the embiguity of colloquial language, one can turn away from self-report methods, which require its use, to observational studies which rely on more defined measurement operations. Behavioral correlates of anxiety can be observed, 174 JIMN 0152539

for example. As one moves further away from colloquial self-de- scription, however, the interpretation of the data becomes in- creasingly problematic. Physiological measures such as galvanic skin response are far removed from the affective state, "anxiety;' and furthermore do not correlate well with other measures of "anxiety." A tradeoff is inevitable. The more operational the measurement device, the less related it seems to the underlying psychological construct, especially when the latter is a subjec- tive state. Differences in approach to this dilemma have resulted in different conclusions about the tobacco withdrawal syndrome. In a number of studies reported by Heimstra (1973), deprived smokers performed presumably stressful tasks and rated their mood before and after the task. Results showed that deprived smokers responded much the way nonsmokers did, but that undeprived smokers showed sig- nificantly less response to the stressor. This suggests that smoking protects against stress-induced mood changes, but that withdrawal has no effect on mood. In contrast are a series of studies in the Schachter laboratory (Nesbitt 1973; Silverstein 1976; Schachter 1978) which use the number of shocks that a sub- ject will tolerate as a measure of his/her response to stress. In these studies, undeprived smokers behave like nonsmokers and deprived smokers show increased vulnerability to stress. This suggests that smoking per se has no effect on reactions to stress, but that withdrawal results in overresponsiveness to stress. Thus, self-report and indirect objective measurement of with- drawal symptoms sometimes produce inconsistent and even directly contradictory results.3 This reflects a general problem in psy- chological assessment which will not be easily solved. However, investigators studying the withdrawal syndrome should be aware of the dilemma, and perhaps try to obtain measurements at both ends of the spectrum wherever possible. The cited studies by the Heimstra and Schachter groups also touch on another issue of methodology: control groups. Data from de- prived smokers are most interpretable when they can be compared to data from smokers in an undeprived state and from nonsmokers. Ideally, this allows one to tease out the effects of smoking e~r se, the effects of abstinence, and the effects of smokerJ nonsmoker differences. An interesting and novel idea would be to introduce a deprived nonsmoker group which is frustrated in some way. This would control for the effects of simple frustration, which is conceptually distinct from a withdrawal syndrome, but has swne similar effects on mood. Quite often, the comparison between nondeprived and deprived smokers is made within subjects. This requires that baseline data be collected while the subject smokes normally. This is difficult to achieve in clinical settings, where smokers often begin to modify their smoking behavior before formally beginning treatment. In these cases, baseline measurements should begin as early as possible while instructing the subject to maintain 175 ~TIMN 0152540

usual smoking behavior. Where several staggered changes in smoking behavior take place -- as in gradual reduction programs -- multiple baselines should be considered. In a truly ideal world, one would also compare a smoker's responses in withdrawal to the presmoking baseline. As this is impossible, one cannot currently distinguish between the effects of simple return to baseline upon discontinuation of nicotine and true withdrawal effects, which involve an overshoot of predrug baseline. If one assumes that smokers and nonsmokers were similar prior to tobacco use (a tenuous assumption), then the nonsmoker group provides a between group comparison with predrug baseline. Most of the research on withdrawal symptoms in long term absti- nence from smoking consists of retrospective questionnaire studies in which successful exsmokers are queried about their experiences. Almost all of the studies which have a more con- trolled methodology are confined to short periods of deprivation in the laboratory, and therefore have limited generalizability. The retrospective studies, while more relevant to the experience of a real life user giving up tobacco, suffer from the failings of retrospective and post hoc correlational designs. For example, while such studies show that successful abstainers report fewer withdrawal symptoms than do their relapsed colleagues, this may be an artifact of the retrpspective viewpoint. Exsmokers may simply "recall" finding withdrawal easier, whereas recidivists justify their relapse with recollections of serious discomfort. The risk of such bias, while present, is not nearly as great when self-reports are collected during the actual period when the subject is undergoing withdrawal. More such prospective, real time studies of long term abstinence in clinical settings are needed. Such data could readily be collected as part of any clinical or experimental smoking cessation study. Investi- gators in the field of smoking cessation should document the effects that withdrawal from tobacco has on their subjects. In this manner, a great deal could be learned about the tobacco withdrawal syndrome at very little cost. DIRECTIONS FOR RESEARCH ON THE TOBACCO WITHDRAWAL SYNDROME The relationship between withdrawal symptoms and smoking cessa- tion efforts is what lends importance to research on the tobacco withdrawal syndrome. To the extent that avoidance of withdrawal symptoms motivates continued smoking, the problem of smoking control becanes one of understandirig and controlling the with- drawal syndrome. Especially as our clinical goals shift away from smoking cessation er se to the prevention of relapse,- the withdrawal syndrome will have to become a major focus of our efforts. And it is this clinical public health goal which should, in turn, determine the direction of research on the to- bacco withdrawal syndrome. The major foci of such research should be (1) to clarify the"relation between the withdrawal syn- drome and the outcome of smoking cessation efforts; (2) identify risk factors relevant to the withdrawal syndrome; and'(3) formu- 176 T'jMN 0152541

late interventions which affect the withdrawal syndrome, with a view toward encouraging maintenance of abstinence. Although the association between withdrawal symptoms and failure in abstinence from smoking is well established in retrospective studies, this 'relationship deserves further exploration. Data on withdrawal symptoms should be collected during the period of initial withdrawal and then analyzed retrospectively for differ- ences between eventual successes and failures. Parameters which modify the association between symptoms and outcome should also be identified. A symptom might have differing predictive power among different subpopulations, for example. The Guilford (1966) study is a model for this type of investigation, but more of the same type are needed in order to understand how the tobacco withdrawal syndrome affects smoking cessation. Research should continue to study those factors which have al- ready been tentatively identified as determinants of the severity of the tobacco withdrawal syndrome. Besides leading to a better understanding of the process underlying dependence on tobacco, this would identify populations at risk for severe withdrawal symptoms. The,in£luence of factors such as sex, for example, needs to be better documented and explored. Since the smoking behavior of men and women differs in a number of important re- spects--e.g., dose, age at initiation, etc. (USDHEW 1975)-- studies are needed to analyze whether the sex difference is reducible to these other variables. More generally, the search for other factors affecting the with- drawal syndrome needs to be expanded. Cne strategy for such a preliminary search would'be to analyze data retrospectively, observing the severity of the syndrome, and then attempting to account for it by multiple regression on other variables. Among the other variables whose influence warrants examination is the method of smoking withdrawal. The differences between"cold turkey"and gradual withdrawal should be studied in'randomiz.ed experiments. The effects of different antismoking therapies should also be explored. This could be easily accomplished if investigators and clinicians in this field agreed to measure and report withdrawal symptoms on a standard instrument. A quasi-experimental comparison among methods would then grow naturally out of the accumulated literature. The influence of environmental factors on withdrawal and relapse also warrants investigation. One area of investigation that has already been suggested involves explanations of diurnal varia- tions in craving in terms of stimulus events occurring at dif- ferent times of day. More broadly, the influence of environ- mental stimuli in producing conditioned withdrawal symptoms-- e.g., Wikler (1965)--should be explored with a view toward identifying high risk situations. This has particular rele- vance to the clinical problem of preventing relapse. If high risk relapse situations are identified, they would provide 177

natural targets for interventions, Some high risk situations might be entirely avoidable, and methods might be devised for reducing the risk of relapse in others. Thus, a better understanding of the tobacco withdrawal syndrome could be applied to increasing the probability of success in smoking cessation efforts. Knowledge about individual differ- ences in risk for severe withdrawal symptoms, for example, might be used to screen patients for treatment, with high risk patients receiving additional attention. If more is learned about the in- fluence of cessation methods as well, perhaps matching of treat- ment methods and patient populations could be used to increase treatment success or cost efficiency. The treatment of choice might depend on the patient's withdrawal syndrome risk factors. Interventions which directly attack withdrawal symptoms need to be developed and evaluated. Manipulation of patients' expecta- tions or attributions of withdrawal symptoms might be one way in which to reduce their severity and affect relapse (cf. Barefoot and Girodo 1972). If withdrawal symptoms are indeed aggravated by conditioned responses to particular stimuli, it might be possible to decondition or countercondition these responses as part of the therapy. Patients might also be pro- vided with coping skills to deal with unavoidable withdrawal symptans. Smokers who suffer from inability to concentrate might be taught to enhance their concentration, for example. Also, progressive relaxation training or other anxiety and af- fect management techniques might prove useful to those who suffer from anxiety or irritability in withdrawal. For those smokers for whom weight gain is a major concern, weight control measures might be incorporated into a smoking control program, with the aim of improving maintenance of nonsmoking. Controlling the withdrawal syndrome may be the key to controlling relapse. SUtiMARY AND CONCLUSION There is an identifiable syndrome which occurs on withdrawal drawal from tobacco. The tobacco withdrawal syndrome is charac- terized objectively by changes in the EEG and cardiovascular function, by decrements in psychomotor performance, and by weight gain. Subjective symptoms of irritability, anxiety, in- ability to concentrate, and disturbances of arousal are charac- teristic of tobacco users in withdrawal, and intense craving for tobacco is almost universally reported. Although reports vary regarding the prevalence, severity, and course of these symptoms, there is nevertheless sufficient consensus to justify the conclu- sion that a withdrawal syndrome occurs in habitual tobacco users. Among the factors which affect the severity of the tobacco with- drawal syndrome are the smoker's sex and habitual dose. Other factors, such as motivation for smoking and motivation for 178 ~TIMN 0152543

quitting, remain to be investigated. At least one symptom -- craving for tobacco -- shows diurnal variation. The early course of the withdrawal syndrome depends on the smoker's de- gree of deprivation. Totally abstinent smokers show a marked decrease in all symptoms in the first week of abstinence, with subsequent leveling off or increase in symptoms. Smoking among partially abstinent smokers appears to prevent this early drop and to maintain symptoms at their high initial levels, prolong- ing the withdrawal symptoms. This may explain the ineffective- ness of gradual withdrawal as a method of smoking cessation. More research on the tobacco withdrawal syndrome is warranted because of the relation between withdrawal symptoms and relapse in smoking cessation. Research in this area must consider a number of methodological issues, including awareness of sources of variability in symptoms. Standardized assessment of symptoms would be a boon to research on the abstinence syndrome, making results more comparable across studies. Much could be gained from simply assessing and reporting withdrawal symptoms in smoking cessation studies. Most studies in the literature are based on retrospective self-reports of exsmokers, with all the limitations attending to such data. More prospective, randomized experiments using control groups and multiple measures are needed. Ultimately, the goal of research on the tobacco withdrawal syn- drome should be the development of applications in clinical tech- nology for smoking control. Currently, three out of four smokers who succeed in stopping smoking will relapse, most of them be- cause of withdrawal symptoms. A better understanding of tobacco withdrawal phenomena may help us to identify high risk populations of smokers and match them to maximally efficient treatments. A major focus for research should be on developing effective treat- ments which specifically attack tobacco withdrawal symptoms and thereby discourage relapse. The health of millicns of people depends on it. FOOTNOTES 1We did find a rather puzzling difference between these two groups in their diurnal pattern of arousal, with some sugges- tion that heavy smokers may be more prone to insomnia. See the section on diurnal variations. 2Schachter, Silverstein and Perlick (1977) present such an expla•• nation of smoking rates based on urinary pH, which affects the excretion of nicotine. Their explanation, however, leads them to predict that smoking rates will be ~hi ~h~er in the morning, and they present data suggesting this is t~ie case. Meade and Wald's (1977) data are probably more reliable in estimating ad lib smoking rates, as they are based on a larger sample and more systematically study the effects of time of day. 3Although they do illustrate the point here, these two sets of studies are not entirely contradictory. In the Heimstra (1973) 179 544 1IMN 0152

studies, the smokers perform like the nonsmokers on psycho- motor tasks, with the deprived smokers showing impairment, suggesting a withdrawal effect but no effect of smoking per se. Also, the study by Perlick (1977) in the Schachter la- oratory, shows a similar pattern of results with self-report ratings of response to a stressor. REFERENCES Andersson, K. Effects of cigarette smoking on learning and retention. Psychopharmacologia, 41:1-5, 1975. Barefoot, J. C.,and Girodo, M. The misattribution of smoking cessation symptoms. Canad J Behav Sci, 4(4):358-363, 1972. Burns, B.H. Chronic chest disease, personality, and success in stopping cigarette smoking. Brit J Prev Soc Med, 23:23-37, 1969. ~ Finnegan, J. K.; Larson, P. S.; and Haag, H. B. The role of nicotine in the cigarette habit. Science, 102:94-96, 1945. Fletcher, C.,and Doll, R. A survey of doctors' attitudes to smoking. Brit J Prev Soc Med, 23:145-153, 1969. Frankenhaeuser, M.; Myrsten, A-L; Post, B.; and Johansson, G. Behavioural and physiological effects of cigarette smoking in a monotonous situation. Psychopharmacologia, 22:1-7, 1971. Friedman, J. Psychopharmacol'ogical aspects of cigarette smoking.ihpublished thesis, University of Melbourne,Australia, 1972. Frith, C. D. Smoking behavior and its relation to-the smoker's immediate experience. Br J Soc Clin Psychol, 10:73-78, 1971. Glauser, S.C.; Glauser, E. M.; Reidenberg, M. M.; Rusy, B. F.; and Tallarida, R. J. Metabolic changes associated with the cessation of cigarette smoking. Arch Environ Health, 20: 377-381, 1970. Gritz, E. R. Women and smoking:A realistic appraisal. Paper delivered at the International Conference on Smoking Cessation, New York, 1978. Gritz, E. R. The "intermission effect": Short-term deprivation from smoking. Paper delivered at the 86th Annual Conference of the American Psychological Association, Toronto, 1978. Gritz, E. R.,and Jarvik, M. E. Preliminary study: Forty-eight hours of abstinence from smoking. Proceedings, 81st Annual Convention, American Psychological Association, 1973, pp.1037-1040. 180 jTIMN 0152545 ~

Guilford, J. S. Factors Related to Successful Abstinence From Smoking. Pittsburgh, Pa.: American Institutes for Research, 1966. Hall, G. H.yand Morrison, C. F. New evidence for a relationship between tobacco smoking, nicotine dependence and stress. Nature, 243(5404):199-201, 1973. Hammen, C. Personal communication, 1973. Heimstra, N. W. The effects of smoking on mood change. In: Dunn, W. L.,ed. Smoking Behavior:h4otives and Incentives. Washington, D. C. : Winston an Sons, pp. 197-207. Heimstra, N. W., Bancroft, N. R. and Dekock, A. R.. Effects of smoking upon sustained performance in simulated driving task. Ann N Y Acad Sci, 142:295-307, 1967. Herold, R. Report regarding our effort to carry out the five-day plan for nicotine withdrawal in the transformer factory "TCarl Liebknecht" at Berlin-Oberschone-Weide. Unpublished mimeo, 1967. Hunt, W. A.,and Matarazzo, J. D. Three years later: Recent developments in the experimental modification of smoking behavior, J'Abnorm Psychol, 81(2): 107-114, 1973. Hutchinson, R. R.,and Emley, G. S. Effects of nicotine on avoidance, conditioned suppression, and aggression response measures in animals and man. In: Dunn, W. L., ed. Smoking Behavior: Motives and Incentives. Washington, D. C. : Winston an Sons, ons, 1973, pp. T71-196. Ikard, F. F.,and Tomkins, S. The experience of affect as a determinant of smoking behavior: A series of validity studies. J Abnorm Psychol, 81(2):172-181, 1973. Itil, T. M.; Ulett, G. A.; Hsu, W.; Klingenberg, H.;and Ulett, J. A. The effects of smoking withdrawal on quantatatively analyzed EEG. Clin Electroenceph, 2(l):44-51, 1971. Jaffe, J. H. Drug addiction and drug abuse. In: Goodman, L. S. and Gilman, A., eds. The Pharmacolo ical B-asis of Therapeutics. Fourth Edition. New York: Macmillan, 1971. pp. 276-313. Kales, J. D.;Allen, C.;Preston, T. A;; Tan, T-L. Changes in REM sleep and dreaming with cigarette smoking and following withdrawal. Psychcph siology, 7:347-348, 1970. Kleinman, K. M.; Vaughn, R. L.; and Christ, T. S. Effects of cigarette smoking and smoking deprivation on paired-associate learning of high and low meaningful nonsense syllables. Psychol Rep, 32:963-966, 1973. Knapp, P.H.; Bliss, C. M.; and Wells, H. Addictive aspects in heavy cigarette smoking. Am J Psychiatry, 119:966-972, 1963. 181 { TIMN 0152546

Knott, V.J., and Venables, P.H. Stimulus intensity control and the cortical evoked response in smokers and nonsmokers. Psycho- physiology, 15(3):186-192, 1978. Knott, V. J., and Venables, P. H. EEG alpha correlates of non- smokers, smokers, smoking, and smoking deprivation. Psycho- physiology, 14(2):150-156, 1977. Levinson, B. L.;Shapiro, D.;Schwartz, G. E.;and Tursky B. Smoking elimination by gradual reduction. Behav Ther, 2:477-487, 1971. Lichtenstein, E. ; Antonuccio, D. 0. ; and Rainwater, E. Personal communication, 1978. Lichtenstein, E.; Antonuccio, D. 0.; and Rainwater, E. Unkicking the habit: The resumption of cigarette smoking. Paper presented at the annual conference of the Western Psychological Association, 1977. Lichtenstein, E.jand Danaher, B. G. Modification of smoking behavior:A critical analysis of theory, research, and practice. In: Hersin, M.; Eisler, R. M.; and Miller, P. M., eds. Progress in Behavior Modification. Vol III. New-York: Academic Press, 1976. pp. 70-1 Mausner, J. S. Cigarette smoking among patients with respiratory disease. Am Rev Resp Dis, 102:704-713, 1970. Meade, T. W.jand Wald, N. J. Cigarette smoking patterns during the working day. Brit J Prev Soc Med, 31:25-29, 1977, Murphee, H. B.eand Schultz, R. E. Abstinence effects in smokers. Fed Proc, 27:220, 1968. Myrsten, A-L; Elgerot, A.; and Edgren, B. Effects of abstinence from tobacco smoking on physiological and psychological arousal levels in habitual smokers. Psychosom Med, 39 (1):25-38, 1977. Myrsten, A-L; Post, B.; Frankenhaeuser, M.; and Johansson, G. Changes in behavioral and physiological activation induced by cigarette smoking in habitual smokers. Psychopharmacologia, 27: 305-312, 1972. Nesbitt, P. D. Smoking, physiological arousal, and emotional response. J Pers Soc Psychol, 25:137-145, 1973. Pederson, L. L.,and Lefcoe, N. M. A psychological and behavioral comparison of ex-smokers and smokers. J_ Chron Dis, 29:431-434, 1976 . 182 !TIMN 0152547

, Perlick, D. The withdrawal syndrome: Nicotine addittion and the effects of stopping smoking in heavy and light smokers. Unpublished disseration. New York: Columbia University, 1977. Peterson, D. J.;, Lonergan, L. H.;Hardinge, M. G.;and Teel, C. W. Results of a stop-smoking program. Arch Environ Health, 16:211- 214, 1968. Phillips, C. The EEG changes associated with smoking, Psychophysiology, 8(1):64-74, 1971. Ryan, F. J. Cold turkey in Greenfield, Iowa: A follow-up study. In: Dunn, W. L. , ed. Smoking Behavior: Motives and Incentives. New York: Wiley, 1973. pp. 231-242. - Schachter, S. Pharmacological and psychological determinants of smoking. Ann Intern Med, 88:104-114, 1978. Schachter, S. Nicotine regulation in heavy and light smokers. J F.x~ Psychol (Gen), 106(1):5-12, 1977. Schachter, S., Silverstein, B., and Perlick, D. Psychological and pharmacological explanations of smoking under stress. J Exp Psychol (Gen), 106(l):31-40, 1977. - Schechter, M. D., and Rand, M. J. Effect of acute deprivation of smoking on aggression and hostility. Psychopharmacologia, 35: 19-28, 1974. Shiffman, S. M.,and Jarvik, M. E. Smoking withdrawal symptoms in two weeks of abstinence. Psychopharmacology, 50:35-39, 1976. Silverstein, B. An addiction explanation of nicot~.n ' e-induced relaxation. Unpublished dissertation. New York: Columbia University, 1976. Trahair, R.C.S. Giving'up cigarettes: 222 case studies. Med J Aust, 1:929-932, 1967. - Ulett, J. A. , and Itil, T. M. Quantitative electroencephalogram in smoking and smoking deprivation. Science, 164:969-970, 1969. U.S. Department of Health, Education, and Welfare, PHS, Center for Disease Control, 1976. Adult Use of Tobacco - 1975. U. S. Department of Health Education and Welfare, Smoking and Health. Report of the Advisory Committee to the Surgeon General of-t-Fe Public Health Service, 1964. Weybrew, B. B., and Stark, J. D. Ps cholo ical and Ph siolo ical Changes Associated with Deprivation rom S_moc_i.ngU. . ava Submarine and Medical Center Report No. 490., 1967. 183 jTIMN 0152548

1Vikler, A. Conditioning factors in opiate addiction and relapse. In: Wilner, D. M. , and Kassebaum, G. G. , eds. Narcotics. New York: McGraw-Hill, 1965. pp. 85-100. Wilhemsen, L. One year's experience in an anti-smoking clinic. Scand J Res Dis, 49:251-259, 1968. Wynder, E. L.; Kaufman, P. L.; and Lesser, R. L. A short-term follow-up study on ex-cigarette smokers, with special emphasis on persistent cough and weight gain. Am Rev Resp Dis, 96:645-655, 1967. ~ Zeidenberg, P.; Jaffe, J. H.; Kanzler, M.; Levitt, M. D.; Langone, J. J.; and Van Vunakis, H. Nicotine: Cotinine levels in blood during cessation of smoking. Comp Psychiatry, 18:93-101, 1977. AUTHOR Saul M. Shiffman, M.A., C. Phil. Department of Psychiatry School of Medicine Neuropsychiatric Institute and Department of Psychology University of California Los Angeles, California 90024 184

Part IV Implications and Directions for FUture Research TIMN 0152550

Chapter 14 Implications and Directions for Future Research Norman A. Krasnegor, Ph.D. A new Surgeon General's Report on Smoking and Health, which was is- sued in January 1979, carefully documents tTie scientific data on the health risks associated with cigarette smoking. A major addi- tion to that document is a series of five chapters on the behavioral aspects of smoking. While much new information on this latter topic is presented, one is struck by the imbalance between what is known about the consequences of smoking and the behavior itself. More specifically, it is clear that research on factors involved in the establishment, maintenance, and cessation of cigarette smoking is only in its formative stages. Due to the implications of cigarette smoking behavior for the pub- lic health and the view that smoking is the prototypical dependence process, the National Institute on Drug Abuse has established a high priority initiative designed to acquire new lmowledge on the behavioral, biological, and psychosocial aspects of this type of substance abuse. The Institute's programmatic goal is focused upon the systematic compilation of basic and applied scientific data on variables which are causally related to the acquisition of the behavior and the observed chronic usage patterns. Such information can serve as a basis for designing feasible and effective treatment strategies and enhance our understanding of dependence associated with substance use. A RFSEARCH AGENDA on the basis of the papers presented in this monograph and my in- volvement in the preparation of NIDA's contribution to the Surgeon General's Report, I should like to provide a selected research agen- da composed of topics which the National Institute on Drug Abuse views as programmatic priorities. One caveat should be noted, how- ever. The list developed below is meant to be a guide for research- ers. It is not intended to be all-inclusive or exhaustive. The order of presentation does not imply a rank order among priorities. In general, the NIDA research program is focused upon cigarette smok- ing behavior and the dependence process associated with it. We 1 186

have interest in each stage of the behavior's natural history (ex- perimentation, establishment, maintenance, cessation, and relapse). We view the problem area broadly and are desirous of supporting a multidisciplinary approach to undertaking investigations in this research domain. Accordingly, we would encourage investigators from the biological, behavioral, and social sciences to apply their methodologies to the field of smoking behavior research in order that a comprehensive and balanced understanding can be achieved. RESEARCH PRIORITIES 1. Peer Pressure one of the most often cited reasons that people, particularly ad.o- lescents, begin smoking is that friends and acquaintances influence those not smoking to take up the behavior and conform to group norms. Peer pressure has been identified by Kandel as a psychoso- cial variable involved in experimentation with and use of marihuana. Other researchers have emphasized peer pressure as a causative fac- tor in other health-risk behaviors (such as reckless driving, ex- cessive drinking, etc.). Studies which undertake prospective investigations of peer pressure as this construct relates to cigarette smoking should be initiated. Both laboratory and field experiments should be carried out to de- termine the contribution of peer pressure to the initiation and maintenance of smoking behavior. 2. The Role of Nicotine One of the central questions in the field of smoking behavior is the role played by nicotine. Do people smoke cigarettes to maintain a certain level of the drug? What we know at present is that nicotine can be discriminated by animals, it is intravenously self-administered by rats, it has a central nervous system effect, and has been demonstrated to be ti- trated by smokers. In addition, the work of Abood and Lowy suggests the existence 'of a specific nicotine receptor. The methodologies of Hanson et al. and Abood and Lowy offer exciting possibilities for studying the reinforcing effects of nicotine with- in the context of both an experimental analysis of behavior and neuropsychopharmacology. We would encourage more research along similar lines. In the area of self-administration, we have particular interest in developing animal models which employ the inhalation route of ad- ministration. This is the case because nicotine passes most rapid- ly into the brain via the lungs, and it may be that the reinforcing efficacy of nicotine is enhanced when administered via this route. Studies which explore the central site of action of nicotine and drugs which block its effect are of high programmatic interest to 187 Tj1VN 0152552 ' ~~.~

T the Institute. Such research could help to elucidate the neuro- psychological and biochemical bases'for the reinforcing effects of the drug. 3. Withdrawal Another important area of research is the characterization of symp- toms associated with cessation of smoking. Dr. Shiffman has pro- vided an excellent overview of the literature and a surtgnary of his own research. In general, our interest in withdrawal stems from the anecdotal observation that smokers who relapse often claim that they take up smoking after cessation because they cannot cope with the withdrawal. Many questions need to be addressed. For exam- ple, is there a characteristic withdrawal pattern associated with cessation? How does this vary with number of years that one has smoked prior to cessation? How does withdrawal vary with the strength of the cigarettes that were smoked? Is there an acute withdrawal pattern? 4. Behavioral Pharmacology of Smoking Our interest in this topic stems from the view that collecting descriptive and quantitative information on the smoking act itself can be useful in designing effective treatment. While there are some data on topography of smoking (Dr. Lee Frederiksen), relatively few experiments have been conducted to determine the rate of puf- fing, volume of puffing, interpuff interval, etc. We would encour- age studies on the relationship of such parameters to smoking his- tory, nicotine content of cigarettes, stimulus control, etc. 5. Prolonging Abstinence At present, there are a variety of methods (e.g., rapid smoking, warm air, etc.) which can be employed to produce cessation. In general, those smokers who finish such treatment do stop. The prob- lem is that the relapse rate is extremely high. Between 70-80 per- cent of those who stop are likely to take up smoking again within a year. Thus, we would encourage research designed to discover procedures which will lead to a lengthening of abstinence. 6. Objective Methods for Validating Self-Reports There are many reports in the literature on incidence and prevalence of cigarette use and on the evaluation of treatment efficacy. Un- fortunately, the analysis and conclusions are often based on self- reports only. While some studies do use significant others to cor- roborate self-reports, few have employed biological assays to vali- date such subjective data. Work on biological assays such as analy- sis of breath for CO content and blood for thiocyanate levels is just getting under way. We would encourage more research on these two biological assays and development of others to help supplement self-reports with objective measures. 188 ;TIMN 0152553

7. Longitudinal Studies of Smokers Many of the research findings which appear in the literature are based upon short term studies of smoking. While there are some exceptions, the general picture suggests that followup in treatment studies is conducted for up to one year. Yet more recent experi- ments show that a minimum of two years of followup is necessary to evaluate treatment efficacy for smoking cessation programs. We would encourage researchers to employ long term followup designs when evaluating efficacy of modalities for cessation. In addition, longitudinal studies should be designed to investigate the natural history of spontaneous quitters. Thus, there are reports that 4 to 5 million smokers each year stop using cigarettes on their own. We know virtually nothing about such people or their success at achieving and maintaining abstinence. 8. Treatment Research While there have been many different techniques to help people stop smoking, the literature generally shows that the best that can be achieved is a 40-50 percent abstinence rate at the end of two years. This result represents a 10-20 percent improvement over the general finding of a 30 percent success rate. New and innovative techniques, particularly in the context of well-designed multi-modal treatment approaches, should be carried out. Such re- search should include within the design appropriate control groups, random assignment, objective measures of cigarette use (CO, thio- cyanate, etc.), and longitudinal followup. 9. Drug/Cigarette Smoking Interactions Anecdotal observations suggest that alcoholics and heroin addicts smoke at a high rate, particularly during periods of abstinence. fixperiments designed to study the interaction of cigarette smoking with alcohol drinking reveal that the rate of consumption of both substances increases when both are used together, compared to the situation when each is used independently. Whether this is an in- teraction based on pharmacological or behavioral mechanisms is not currently known. However, we would encourage investigators to study the interaction of cigarette smoking with the self-adminis- tration of a variety of drugs. Such work would help to elucidate the role that cigarette smoking may play in the maintenance of other drug use. We hope that this brief listing of selected research topics will aid scientists who are interested in the field to focus their in- vestigations. Further information pertaining to the programs on cigarette smoking may be obtained by contacting the National Insti- tute on Drug Abuse, Division of Research, 5600 Fishers Lane, Rockvi7le, Maryland 20857. 189 14 ~TIMN 0152554

parlicipanla in Sy=osium on Cigarette Smoking as a Dependence Process June 19, 1978 Leo G. Abood, Ph.D. John A. Rosecrans, Ph.D. Center for Brain Research University of Rochester Medical Center Department of Pharmacology hiCV Station, Box 726 Virginia Commonwealth University Rochester, New York 14642 Dorothy E. Green, Ph.D. Richmond, Virginia 23298 Michael A. H. Russell, M. B. ,MIl2C ; Consulting Research Psycholo- gist :42CPsych Addiction Research Unit 3509 N. Dickerson Street Arlington, Virginia 22207 Harley M. Hanson, Ph.D. Institute of Psychiatry The Maudsley Hospital 101 Denmark Hill London, S.E.S, England Merck Institute for Therapeu- tic Research 26-208 West Point, Pennsylvania 19486 Daniel $orn, Ph.D. R.D. 1 Box 182 Prenchtown, New Jersey 08825 Stanley Schachter, Ph.D. Department of Psychology 1400 Schernerhorn Columbia University New York, New York 10027 Saul M. Shiffman, M.A., C.Phil. Jerome H. Jaffe, M.D. College of Physicians and Department of Psychiatry School of Medicine Surgeons of Columbia University Neuropsychiatric Institute and Department of Psychology 722 West 168th Street New York, New York 10032 Murray E. Jarvik, M.D., Ph.D. Department of Psychiatry and Pharmacology School of Medicine Neuropsychiatric Institute University of California - L.A. 760 Westwood Plaza Los Angeles, California 90024 John A. O'Donnell, Ph.D. Department of Sociology University of Kentucky Lexington, Kentucky 40506 University of California - L.A. 760 Westwood Plaza Los Angeles, California 90024 Charles R. Schuster, Ph.D. Departments of Psychiatry and .Pharmacological and Physiologi- cal Sciences University of Chicago Pritzker School of Medicine 950 East 59th Street Chicago, Illinois 60637 190 ;TIMN 0152555

monograph series While limited supplies last, single copies of the monographs may be obtained free of charge from the National Qearinghouse for Drug Abuse Information (NCDAI). Please contact NCDAI also for informa- tion about availability of coming issues and other publications of the National Institute on Drug Abuse relevant to drug abuse research. Additional copies may be purchased from the U.S. Government Printing Office (GPO) and/or the National Technical Information Service (NTIS) as indicated. NTIS prices are for paper copy. Microfiche copies, at $3, are also available from NTIS. Prices from either source are subject to change. Addresses are: NCDAI National Clearinghouse for Drug Abuse Information Room 10-A-56 5600 Fishers lane Rockville, blaryland 20857 GPO NTIS Superintendent of Ibcuments National Technical Information U.S. Government Printing Office Service Washington, D.C. 20402 U.S. Department of Commerce Springfield, Virginia 22161 1 FINDINGS OF DRUG ABUSE RESEARCH. An annotated bibZiography of NIMH- and NIDA-supported extranruraZ grant research, 1964-74. VoZume Vol.l: 1, 384 pp., VoZzone 2, 377 pp. GPO out of stock NTIS PB #272 867/AS $14 Vol.2: GPO Stock #017-024-0466-9 $5.05 NTIS PB #272 868/AS $13 191 TIMN 0152556

2 OPERATIONAL DEFINITIONS IN SOCIO-BEHAVIORAL DRUG USE RESEARCH 1975. Jack Elinson, Ph.D., and David Nurco, Ph.D., editors. Task Force articles proposing consensual definitions of concepts and terms used in psychosociaZ research to achieve operational comparability. 167 pp. GPO out of stock NTIS PB #246 338/AS $8 3 AMINERGIC HYPOTHESES OF BEHAVIOR: REALITY OR CLICHE? Bruce J. Bernard, Ph.D., leditor. Articles exarnining the relation of the brain monoamines to a range of animal and human behaviors. 149 pp. GPO Stock #017-024-00486-3 $2.25 NTIS PB #246 687/AS $8 4 NARCOTIC ANTAGONISTS: THE SEARCH FOR LONG-ACTING PREPARATIONS. Robert Willette, Ph.D., editor. ArticZes reporting current alternative inserted sustained-reZease or Zong-acting drug devices. 45 pp. GPO Stock #017-024-00488-0 $1.10 NTIS PB #247 096/AS $4.50 5 YOUNG MEN AND DRUGS: A NATIONWIDE SURVEY. John A. O'Donnell, Ph.T1., et al. Report of a nationaZ survey of drug use by men 20-30 years old in 1974-5. 144 pp. GPO Stock #017-024-00511-8 $2.25 NTIS PB #247 446/AS $8 6 EFFECTS OF LABELING THE "DRUG ABUSER": AN INQUIRY. Jay R. Williams, Ph.D. Analysis and review of the literature examining effects of drug use apprehension or arrest on the adolescent. 39 pp. GPO Stock #017-024-00512-6 $1.05 NTIS PB #249 092/AS $4.50 7 CANNABINOID ASSAYS IN HUMANS. Robert Willette, Ph.D., editor. Articles describing current deveZopmen'ts in methods for measuring cannabinoid levels in the human body by immunoassay, liquid and duaZ column chromatography and mass spectroscopy techniques. 120 pp. GPO Stock #017-024-00510-0 $1.95 NTIS PB #251 905/AS $7.25 8 RX 3x/WEEK LAAM - ALTERNATIVE TO METHADONE. Jack Blaine M.D., and Pierre Renault, M.D., editors. Comprehensive summary o~ development of LAAM (Levo-aZpha-acetyZ methadoZ), a new drug for treatment of narcotic addiction. 127 pp. Not available from GPO . NTIS PB #253 763/AS $7.25 Pierre Renault M.D, editors. Pro ress re ort o deveZo ent, ~d reb~malZ andA ebNni aZ stuZd'1e of naZtrexoneo a ~ew~drug~or eatment of narcotic addiction. 182 pp. GPO Stock #017-024-00521-5 $2.55 NTIS PB #255 833/AS $9 10 EPIDEMIOLOGY OF DRUG ABUSE: CURRENT ISSUES. Louise G. Richards, Ph.D., and Louise B. Blevens, editors. Conference Proceedings. Excmtin- ation of methodological problems in surveys and data collection. 259 pp. ,GPO Stock #017-024-00571-1 $2.60 NTIS PB #266 691/AS $10.75 11 DRUGS AND DRIVING. Robert Willette, Ph.D., editor. State-of-the- art review of current research on the effects of different drugs on performance impairment, particularly on driving. 137 pp. GPO Stock #017-024-00576-2 $1.70 NTIS PB #269 602/AS $8 192 TIMN 0152557

12 PSYCHODYNkMICS CF DRUG DEPENDENCE. Jack D. Blaine, M.D., and Demetrios A. Julius, M.D., editors. A pioneering coZlection of papers to discover the part played by individual psychodynamics in drug de- pendence. 187 pp. GPO Stock #017-024-00642-4 $2.75 NTIS PB #276 084/AS $9 13 COCAINE: 1977. Robert C. Petersen' Ph.D., and Richard C. Stillman, M.D., editors. A series of reports developing a picture of the extent and limits of current knowledge of the drug, its use and misuse. 223 pp. GPO Stock #017-024-00592-4 $3 NTIS PB #269 175/AS $9.25 14 MARIHUANA RESEARCH FINDINGS: 1976. Robert C. Petersen, Ph.D., editor. Technical papers on epidemiology, chemistry and metaboZism, toxicological and pharmacologicaZ effects, learned and unlearned be- havior, genetic and immune system effects, and therapeutic aspects of marihuana use, 251 pp. GPO Stock #017-024-00622-0 $3 NTIS PB #271 279/AS $10.75 15 REVIEW OF INHALANTS: EUPHORIA TO DYSFUNCTION. Charles Wm. Sharp, Ph.D., and Mary Lee Brehm, Ph.D., editors. A broad review of inhalant abuse, including sociocultural, behavioral, clinical, pharma- cologicaZ, and toxicological aspects. Extensive bibliography. 347 pp. GPO Stock #017-024-00650-5 $4.25 NTIS PB #275 798/AS $12.50 16 THE EPIDIIKIOLOGY OF HEROIN AND OTHER NARCOTICS. Joan Dunne Ritten- house, Ph.D., editor. Task Force report on measurement of heroin- narcotie use, gaps in knowledge and how to address them, improved re- search technologies, and research implications. 249 pp. GPO Stock #017-024-00690-4 $3.50 NTIS PB #276 357/AS $9.50 17 RESEARCH ON SMOKING BEHAVIOR. Murray E. Jarvik, M. D. , Ph. D. , et al., editors. State-of-the-art of research on smoking behavior, including epidemiology, etiology, socioeconomic and physical conae- quences of use, and approaches to behavioral change. From a NIDA- supported UCLA conference. 383 pp. GPO Stock #017-024-00694-7 $4.50 NTIS PB #276 353/AS $13 18 BEHAVIORAI, TOLERANCE: RESEARCH AND TREATMENf IMPLICATIONS. Norman A. Krasnegor, Ph.D., editor. Conference papers discuss theoretical and empirical studies of nonpharmacoZogie factors in development of tolerance to a variety of drugs in animal and hurnan subjects. 151 pp. - GPO Stock #017-024-00699-8 $2.75 NTIS PB #276 337/AS $8 19 THE INTERNATIONAL CHALLENGE OF DRUG ABUSE. Robert C. Petersen, Ph.D., editor,. A monograph based on papers presented at the World Psychiatric Association 1977 meeting in Honolulu. Emphasis is on emerging patterns of drug use, international aspects of research, and therapeutic issues of particular interest raorlciwide. In Press 193 1 ' TIMN 0152558

20 SELF-ADMINISTRATICN OF ABUSED STiRSTANCES: METHODS FOR STUDY. Norman A. Krasnegor Ph.D., editor. Papers from a technicaZ review on methods used to stud'y seZf-administration of abused substances. Dis- cussions include overview, methodological analysis, and future planning of research on a variety of substances: drugs, ethanol, food, and tobacco. 246 pp. Not available from GPO NTIS PB #288 471/AS $10.75 21 PHENCYCLIDINE (PCP) ABUSE: AN APPRAISAL. Robert C. Petersen, Ph.D., and Richard C. Stillman, M.D., editors. Monograph derived from a technical review to assess the present state of knowledge about phen- cycZidine and to focus on additional areas of research. Papers are aimed at a professional and scientific readership concerned about how to cope with the problem of PCP abuse. 313 pp. GPO Stock #017-024-00785-4 $4.25 NTIS PB #288 472/AS $11.75 22 QUASAR: OUANTITATIVE STRUCTiJRE ACTIVITY RELATIONSHIPS OF ANAL- GESICS, NARCOTIC ANTAGONISTS, AND HALLUCINOGENS. Gene Barnett, Ph.D. Milan Trisc, Ph.D., and Robert E. Willette Ph.D., editors. Reporls an interdiscipZinarg conference on the molecular nature of drug-receptor interactions. A broad range of quantitative tech- niques were appZied to questions of molecular structure, correZa- tion of molecular properties with biological activity, and moZe- euZar interactions with the,receptor(s). 487 pp. GPO Stock #017-024-00786-2 $5.25 NTIS PB #292/265/AS $15.00 Coming Issues: 24 SYN'IHETIC ESTIMATES FOR SMALL AREAS: WORKSHOP PAPERS AND DISCUS- SION. Joseph Steinberg, editor. Papers from a workshop consponsored by NIDA and the National Center for Health Statistics on a class of statistical approaches that yield needed estimates of data for States and local areas. Methodology and applications, strengths and weak- nesses are discussed. 25 BEHAVIORAL ANALYSIS AND TREATMEMf OF SUBSTANCE ABUSE. Norman A. Krasnegor, Ph.D., editor. TeehnicaZ review papers present corronon- aZities•in the behavioral analysis and treatment of substance abuse. Specific behavior therapeutic approaches are presented for the treatment of drug addiction, tobacco smoking, drinking, and obe- sity. An agenda for future research needs is included. 194 TIMN 0152559

DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE PUBLIC HEALTH SERVICE ALCOHOL, DRUG ABUSE, AND MENTAL HEALTH ADMINISTRATION 5600 FISHERS LANE POSTAGE AND FEES PAID ROCKVILLE, MARYLAND 20857 U.S. DEPARTMENT OF H.E.W, HEW 396 OFFICIAL BUSINESS THIRD CLASS 300 i t Penalty for pr va e use, $ NOTICE OF MAILING CHANGE BULK RATE • Check here if you wish to discontinue receiving this type of publication. O Check here if your address has changed and you wish to continue receiving this type of publication. (Be sure to furnish your complete address including zip code.) Tear off cover with address label still affixed and send to: Alcohol, Drug Abuse, and Mental Health Administration Printing and Publications Management Branch 5600 Fishers Lane (Rm. 6C-02) Rockville, Maryland 20857 DHEW Publication No. (ADM) 79-800 Printed 1979 i TIMN 0152560