Tobacco Documents Online

I I I I I I I I I I I I I I An Analysis of the Addiction Liability of Nicotine Allan C. Collins, PhD SUMMARY. Even though the percent of adults who smoke hzs been reduced dramatically in the last 20 years, more than 50 million people continue to use tobacco on a daily basis. A majoriryef these people claim that they would like to stop smoking, but cannot. This review discusses the data suggesting that tobacco smokers are seek- ing nicotine. The behavioral effects elicited by nicotine and its rela• cive reinforcing properties, as well as thc development of tolerarce to nicotine and the role of nicotine in the tobacco withdrawal syn- drome are discussed. The data indicate that nicotine has addiction liability, but this liability is probably not equal to that of other agents such as cocaine and the opiates. rnrAOnvcrroN The health consequences of smoking have received a great deal of attention from the medical community ever since the first Sur- geon General's Report on Smoking and Health. The 20th Repon of the Surgeon General on the Health Consequences of Smoking (1988) is entitled "Nicotine Addiction."' The Report concludes that; Allan C. Collins is affiliated with the inuitute for Behavioral Geneiics and Department of Psyehototy, Campus Box 447, University of Colorado, Boulder, CO 80309. Dr. Collins is supported in part by a Research Scientist Development Award front the Notionat Institute on Drug Abuse (t)A-00I I6). C 1990 by The Haworsh Preu, Inc. All rights reserved.

i I I I I I I I I I I I I I 84 .4ddiction Porrnttal oJAbivtd Drugs and Drug Classei • Cigarettes and other fortns of tobacco ate addicting. • Nicotine is the drug in tobacco that causes addiction. • The pharmacological and behavioral processes that determine tobacco addiction are similar to those that determine addiction to drugs such as heroin and cocaine. The Surgeon General's Report is the most comprehensive review of the literature available today; the Report is over 500 pages long, and it reviews thousands of studies. Consequently, the present p2- per will not attempt to cover the entire literature. The reader is referred to the Surgeon General's Report for this comprehensive review. Rather, this paper will focus on the primary issues related to the potential role of nicotine in regulating tobacco use, and will attempt to discuss some of the areas of dispute generated by the Surgeon General's Report. EPIDEMIOLOGY OF TOBACCO USE The use of tobacco products, primarily cigarettes, is widespread in today's society. According to data in the 20th Surgeon General's Report, approximately 26.5% of all Americans 17 years and older were smokers in 1986. This represents 29,5% of all adult males and 23.8% of adult females. While the last 15-20 years have seen dra- matic reductions in the per cent of the population using tobacco (51.1% of adult males and 33.3% of females were smokers in 1965), these figures translate into over 50 million people who were smoken in the United States in 1986. Sex differences in smoking incidence have been reported in virtu• ally every analysis of tobacco use. However, the sex differential is narrowing. During the 1965-1985 time period, a 19.4% reduction was seen in the per cent of males 20 years of age and older who smoked whereas a modest 5.9% reduction was seen in women. In 1965, 51.3% of White males and 59.6% of Black males, 20 years of age and older, wcre smokers. By 1985, these values had been reduced to 31.8% and 40.6% for Whito and Black males, respec- tively. ln eontrast, 34.5el0' of White females and 32.7% of Black females were smokers in 1965 whereas Z8.3% and 31.6% of these same populations were smokers in 1985. This represents a modest

I I I I I I I I I I I I I I ,lllon C. Collins 83 6.21/'c decrease in the fraction of White women who smoke and an even more modest 1.1°,'c decrease in Black women. Apparently, smoking begins relativelv early in life. In 1965, 59.2% of all adult mates and 41.91"c of all adult females, age 20-24 were smokers. Consistent with the trends reported above, these val- ucs were reduced in 1985, with 31.0% of males and 32.1% of fe- males in this age group being listed as smokers. Only fragmentary data, using methods that varied from survey to survey, are available for younger age groups, but in the High School Class of 1986, 10,7170 of the males and 11.6''0 of the females reported using a half- pack or more on a daily basis. These trends clearly demonstrate that smoking develops in early adulthood and that the sex differential in tobacco use is disappearing or may even be reversiilg; ROLE OF NICO7TNE IN TOBACCO USE While the demographic data provide some indication of the inci- dence of smoking, these data provide little information about why people smoke or which of the many components of tobacco are being sought. Scientists know surprisingly little about why people smoke, but the evidence is very clear that nicotine is the primary agent that is sought when tobacco is used. Several lines of evidcnce indicate that nicotine is the most important psychoactive agent in tobacco. For example. cigarette smokers who use low-yield ciga- rettes smoke more cigarettes per day than do those who smoke high- yield cigarettes, and switching smokers from high•yield to low- yieid cigarettes results in an increase in the number of cigarettes smokcd.~ Alternatively, low-yield cigarette smokers increase the rate or depth of puffing, or duration of inhalation or they block filter vents so as to assure delivery of adequate amounts of nicotine." These observations have been offered as explanations for the fact that high-yield cigarette smokers maintain blood nicotine levels when the cigarette brand is changed to a low-yield brand. However, nicotine levels are not the only parameters that are reduced in low- yield cigarettes; reduction in tars and other components are also seen. This, and the fact that brand-switching studies have not al- ways resulted in changes in smoking behavior (see Gori and I

A6 .lddlrrin+r Pe(entiaf nJAbeued Drugs anE Drug Clcsses I I I I I I I I I I tynch') has resulted in some debate as to whether nicotine is the only agent in tobacco that is being sought by smokers. Perhaps the most convincing argument that nicotine is the sought-after agent in tobacco comes from the observation that pre- treatment with the nicotinic receptor antagonist, mecamylaminc, results in an increased consumption of usual brand cigarettes (see Pomerleau et al.,' 1987 and the studies cited therein). This in- creased consumption is accompanied by an increase in plasma nico- tine levels and probably represents an attempt by the smoker to overcome the functional blockade of brain nicotinic receptors. In contrast, both oral and intravenous nicotine administration have elicited decreased cigarette smoking in experimental settings where cigarettes were readily available.' These data clearly suggest that nicotine is an important factor influencing the amount of tobacco consumed on a daily basis. Psychoactive Effects of Nicotine Russell' has argued that smoking is a negatively reinforced be- havior; i.e., people continue to smoke to avoid withdrawal. This statement begs a very important issue: why do people smoke in the first place? Presumably, nicotine or tobacco is reinforcing, but it is not clear precisely how nicotine elicits its putative reinforcing ef- fects. Although the data suggest that smokers are seeking nicotine, it is not totally clear as to what behavioral action is being sought. At least theoretically, a psychoactive agent could be reinforcing either because it increases feelings of well being or because it decreases feelings of discomfort, pain or anxiery. Most smokers describe smoking as a pleasurable experience, but it is not totally clear that nicotine has euphoric effects in humans. This is the case because only a very limited number of direct investigations of the behavioral effects of "straight" nicotine have been reported. Johnston' in- jected nicotine intravenously into smokers and nonsmokers; smok- ers rated the response as pleasurable whereas nonsmokers rated the effects as unpleasant. More recently, Henningfield et at.' allowed human volunteers to lever press for the intravenous administration of nicotine. The subjects, most of whom were being treated for I

Attcn c. cottin.t s" I I I I I I I I I I I I I I I some form of substance abuse, rated the nicotine effects as eu- phoric, as defined by elevations in scores on the Morphine Benze- drine Group scale of the NIDA Addiction Research Center inven- tory. Those subjects with a drug abuse history also identified the nicotine injections as cocaine. Evidence for dysphoric actions of nicotine were also obtained in this study. These actions became more intense over the course of the experiment and served to limit the number of nicotine injections taken. Thus, it may be that nico- tine elicits pleasurable effects at lower doses and unpteasurable ef• fects at higher doses, but it should be noted that the experimental subjects continued to lever press for nicotine even after nausea de- veloped. Most of the subjects did not report a decreased desire to smoke while self administering nicotine, but in the three subjects whose smoking behavior was studied the number of cigarettes smoked and the number of puffs per cigarette were reduced during the self-administration test periods. The studies of Henningfield et al.• certainly suggest that intravenous nicotine infusion produces euphoria, but further experiments using additional doses and more subjects, particularly subjects who do not have a history of drug abuse, must be carried out before these findings can be accepted. Drugs, in general, can also be reinforcing because they decrease pain or anxiety. Nicotine's potential antittociceptive (pain•redue- ing) effects have been assessed in several experiments. For exam- ple, Fertig et at.' examined the effects of smoking on pain aware• ness (subject is aware of pain) and pain endurance (how long subject can endure the pain without responding) using habitual smokers who had been deprived of tobacco for an hour before test- ing. Pain was induced by placing the subject's forearm in cold (3') water, and latency between initiation of stimulation and pain per- ception and removal of the forearm from the water were recorded. Smoking increased the latency for both measures. This study also included an analysis of the effects of nicotine on pairi perception and response that used subjects who had been nonsmokers for a minimum of 1 year; nicotine was administered to these subjects in the form of snuff. Once again, latcncy to pain perception' was in- creased as was pain endurance. These studies clearly argue that, under some circumstances, nicotine may have antinociceptive ef- I

I I I I I I I I I I I I 88 Addirrion Porenuial oj.tbused Drugs ond Drug Classes fects in humans and are consistent with the observation that nicotire has antinnciceptive effects in animals.t0 Nicotine may also decrease anxiety, at least under some circum- stances. The use' of tobacco products clearly tnereases when srnok- ers are placed in stressful situations. Significant increases in smok- ing have been reported in response to a variety of laboratory stressors, including shock, public speaking, aversive white noise, and performing mental arithmetic with competitive pressure (see Pomerleau and Pomerleau" for a review). These observations have led to the suggestion that people smoke in an attempt to cope with stress. Although many smokers identify the production of a reiaxed state as a motive for smoking, early experiments on the relationship between smoking and anxiety were inconclusive. Problems in dem- onstrating antianxiety effects of nicotine in smokers may arise be- cause such studies usually use subjects who have been withdrawn from nicotine for several hours which may result in withdrawal- induced increases in anxiety. Pomerleau et al." assessed the effects of a nicotine-containing cigarette and a zero-nicotine cigarette en anxiety produced by presenting the subjects with unsolvable ana- grams. The subjects were minimally deprived smokers who re- frained from smoking for one hour before testing. The nicotine- containing cigarette elicited a measurable reduction in anxiety as measured by the Spielberger State Anxiety Inventory. Gilbett" re- viewed the literature in this area a decade ago and concluded that "too few studies have been reported and too many methodological problems are evident in the studies that are available to permit a definite statement about the role of nicotine in altering emotional behavior in humans." This situation has not changed remarkably since Gilbert's review was published. One of the major problems that hinders progress in this area is most researchers arc extremely reticent to give tobacco or nicotine to nonsmokers. Consequently, studies of the behavioral effects of nicotine in humans are frequently confounded by potential with- drawal symptoms. These problems are not encountered in labora- tory animal studies. Nicotine injection results in changes in condi- tioned suppression, emotional freezing and exploratory behaviors that may reflect anxiety-reducing actions in rats." Thus, it may be I

I I I I I I I .t11cn C. Collirv 89 that nicotine does, indeed, have antianxiety effects in humans, but much additional rescarch using appropriate controls such as non- smokers or ex-smokers may be required to assess fully the potential antianxicry effects of nicotine, In summary, even though most smokers claim to use tobacco because it gives them pleasure, the studies currently available do not provide unequivocal evidence for nicotine producing reward ei- ther via euphoric actions or through reduction of pain, anxiety or negative affect. Even though the study of Henningfield er al.' indi- cates that ex-cocaine users recognize nicotine as cocaine, it does not seem probable that a cocaine-like action is dominant in tobacco smokers, perhaps because smokers do not attain the necessary brain levels of nicotine to elicit cocaine-like actions. Similarly, questions regarding potential analgesic or antianxiety effects of nicotine per- sist, Consequently, the reasons why smokers find tobacco or nico- tine pleasurable remain unclear, FACTORS THAT INFLUENCE SMOKING I I I I I I I EaviroAmenta! Factors Tobacco products are readily available, yet not every person chooses to smoke or chew tobacco. This, of course, indicates that individuals differ either in the environmental or genetic factors that regulate smoking. As noted above, smokers generally increase ttxir use of tobacco products when placed in stressful environments. Similarly, it is very clear that an association exists between alco- holic beverage consumption and cigarette smoking. A greater frac- tion of the alcoholic population smoke than do nonalcoholics and cigarette smokers who have been diagnosed as alcoholics smoke more cigarettes per day than do nonalcoholic cigarette smokers (see Henningfield et at." and Mello et al.," for a review of this litera- ture). Laboratory studies utilizing alcoholic and nonalcoholic sub- I

I I I I I I I I I I I I I I 90 Addicrron Potentrol of Abuaed Drugs and Drug C1asSeJ jecis indicate that as alcohol consumption increases, so does to- bacco usc." " Genetic Factors 1n chapter 5 of the 20th Surgeon Gcneral's Report, considerable attention is given to the fact that individuals differ in smoking be- havior. This pattern is highly reminiscent of the pattern seen with alcohol. A host of studies have indicated that alcoholism is familial and adoption studies have demonstrated that the drinking behavior of the adoptee more closely resembles the drinking behavior of the biological parent. Although not nearly as well studied, several re- ports indicate that a genetic predisposition may contribute to to- bacco use. Fisher""" studied the smoking behavior of identical twins and noted that concordance for smoking behavior (whether both twins were smokers or nonsmokers) was greater in a population of monozygotic ( identical) twins than was the concordance in di7y- gotic twins. Fisher argued that the greater concordance for smoking behavior seen in the identical twin pairs indicates that tobacco use is influenced by genetic factors. Subsequent studies of smoking be- havior in monozygotic and dizygotic twins have supported Fisher's arguments (see Gurling et al.,'' for original contributions as well as a review of this literature). As is the case with alcohol use and abuse, the human genetic studies do not point towards a potential cause of smoking; the data indicate only that some genetic factor(s) either promote or inhibit tobacco use. ItEINFORCING PROPERT7ES OF NICO?7NE: FINDINGS FROM BASIC RESEARCH Primarily because of ethical considerations, very few investiga- tors 5ave attempted to assess the potential rewarding or reinforcing properties of nicotine in humans, Assessing the potentW reinforc- ing cffccts of drugs, in gcneral, has been dcvcloped to a high degree in laboratory animals. Three approaches have been used: the drug self-administration procedures which generally involve rewarding an animal for pressing the appropriate lever, conditioned place pref- ZND , C7 ~ CO ~ ~

I I I I I I I I I I I I I I I I Allon C, Collrxr 91 erence and drug discritnination. All of these strategies have been used to assess the potential reinforcing properties of nicotine, and equivocal results have been obtained. Self.Admr»istralion Studies Self-administration studies have classically been used to assess whether an agent has reinforcing properties. Most of these studies are done in animals, but occasionally humans are used. As men- tioned above, Henningfield et al.' have successfully demonstrated that human volunteers, most of whom were being treated for sub- stance abuse, will press a lever to attain an intravenous nicotine reward. These subjects described the effects elicited as pleasurable, and those with a substance abuse history likened the response to that elicited by cocaine, Studies of nicotine self-administration by animals show that nico- tine may act as a reinforcer, but its efficacy and the range of condi- tions under which it will be self-administcred are limited. Nicotine will be self-administered by rats, squirrel monkeys, rhesus mon- keys, baboons and dogs (see Henningfield' and Clarke'° for recent reviews). Nicotine self-administration is reduced or eliminated by treatment with high doses of the nicotinic antagonist meeamy- lamine.1' While these studies have all demonstrated that animals will self- administer nicotine, the evidence suggests that self-administration is not readily attained. In general, long periods of training are re- quired and response rates are quite low. In addition, factors other than drug or drug dose influence the response rate. For example, L~tng et al.u have shown that rats can be trained to self-administer nicotine only if they are maintained at a reduced body weight. Simi• lar}y, rats that are exposed to stress will self-inject nicotine more readily than will nonstresscd rats." Goldberg's=' studies of nicotine self-administration by squirrel monkeys are probably the most effective demonstrations that ani- mals will sclf-administer nicotine. These studies were published shortly after Griffiths et a!." had concluded that nicotine is a weak reinforcer when compared to other drugs of abuse. Goldberg and co-workers used second-order schedules of reinforcement (every I

I I I I I I I I I I I I I I I I 92 Addrerion Potenual of Abured Drugs osd Drug Clesses tenth response resulted in the presentation of a brief visual stimulus which was occasionally paired with an injection of 0.03 mgr3cg nic- otine). Most other investigators have used simple or continuous schedules of reinforcement. Using the second-order schedule pre- vented injections from occurring at interinjection intervals of 5 min or less. Too frequent injections, or too high a dose, results in nico- tine being aversive. Squirrel monkeys will also consume nicotine in high doses when it is given orally.ss However, this occurs only if noncontingent tail shocks are administered. In control sessions, subjects drink most of their fluid from water-only bottles. Risner and Goidberg-" were also successful in detecting signifi- cant self-administration of nicotine in beagle dogs using a proce- dure that involved a fixed ratio of 15 (15 lever presses for each drug injection) and a 4 minute time out period immediately after injec- tion. This study showed: peak rates of responding were about 0.3 responses/see (higher rates were maintained by cocaine); response rates increased with dose and then decreased at higher nicotine doses; and response rates for nicotine, but not cocaine, were re• duced by pretreatment with mecamylamine. All of these studies indicate that, under appropriate circum• stances, nicotine can serve as a reinforcer. However, the dose must be controlled and time between doses may be critical. In addition, the environment may also play a critical role in determining whether nicotine has reinforcing properties or not. The effects of environmental variables such as feeding status and stress on self= administration led Balfouc" to argue that nicotine does not act as a particularly effective positive reward in a "neutral" environment but can act as a positive reinforcer in situations in which it alleviates some unpleasant physiological response. Similarly, Henningfield, suggested "that nicotine serves as a reinforcer under a more limited range of conditions than do other reinforcers and that its strength is more related to environmental stimuli than is the case for other drugs of abuse." This reviewer is unaware of any published reports that have appeared recently that would alter the situation. Thus, nicotine can serve as a reinforcer in laboratory animals, as mea- sured by self-administration studies, but the range of conditions I