Human Dependence on Tobacco and Opioids: Common Factors


Recent years have seen increasing acceptance of the notion that tobacco is an addictive or dependence-producing substance, particularly as it is used in cigarette smoking. This idea is supported by the observations that tobacco serves as a reinforcer (i.e., it maintains behavior leading to its use) and that most people who smoke cigarettes would like to quit but cannot, even in the face of well documented health risks and economic sacrifices (Surgeon General’s Report 1979). The term “drug dependence” suggests that (1) the drug serves as a reinforcer, (2) behavior occurs which is maintained by the opportunity to take the drug, and/or (3) other reinforcers are sacrificed as a consequence of taking the drug (). Many cigarette smokers in some degree satisfy these criteria for drug dependence ().

Since cigarette smoking has only recently been conceptualized as an instance of drug dependence, it should be useful to systematically compare cigarette smoking with another more thoroughly studied dependence process such as opioid dependence or narcotic addiction. At first blush, cigarette smoke and opioid drugs appear to produce vastly differing pharmacological and behavioral effects: large doses of opioids can produce a debilitating sedation that is not produced by heavy cigarette smoking. However, these differing direct drug effects may be only marginally relevant to the ongoing dependence processes per se, and certain functional similarities in the two forms of dependence suggest that opioid dependence may, in fact, provide a useful and valid conceptual model to which cigarette smoking may be compared.

The purpose of this chapter is to compare the functional similarities between tobacco and opioid dependence. Relevant experimental data, clinical observations, and epidemiological findings will be discussed under the organizational framework shown in list:

  • Patterns of Use
  • Personality Characteristics and Social Factors
  • Physiologic Dependence
  • Deprivation Effects
  • Tolerance
  • Dose Effects on Drug Intake
  • Reinforcing Efficacy and Dependence Liability
  • Response Requirement
  • Conditioning Factors
  • Antagonist Administration Effects
  • Preloading Effects
  • Relapse Patterns Following Abstinence Treatment
  • Feeding Behavior Effects

Patterns of UK

In both cigarette smoking and opioid dependence, use of the respective drug occurs on a regular daily basis and, given adequate supplies, self-imposed abstinence is infrequent. This overall pattern of use is distinct from that of many other drugs of abuse (e.g., the sporadic use of the hallucinogens or the use of psycho-motor stimulants in which periods of self-administration are broken by periods of abstinence). With bath tobacco and opioids, simple exposure to the drugs (“experimentation”) frequently leads to chronic use (). In fact, while exact figures vary, it has been estimated that 85% of adolescents who smoke one or more cigarettes become compulsive smokers (). Similarly, with opioids, it has been found that a high percentage of experimental users become dependent users, e.g., 97 percent in a study by Robins and Murphy (). From this perspective, both drugs have a high “dependence liability” or “addictive potential.” A difference in this regard is that most cigarette smokers are compulsive daily users (about 95%) whereas current data suggest that a substantial portion of the total population of opioid users are not compulsive daily users (that is, they are “chippers”). With both tobacco and opioids, certain routes of administration are preferred (smoking and injecting, respectively) but other routes or forms of the drug will be substituted if the preferred one is precluded. For example, most tobacco users are cigarette smokers (Surgeon General’s Report 1979) but some smokers will change to chewing tobacco or snuff if their occupation does not permit smoking (). With individual opioid users for whom the intravenous route is preferred, other routes and drug forms also will suffice (e.g., oral methadone and smoking of opioids).

Finally, when asked to give reasons for their smoking behavior, the answers obtained from most cigarette smokers may be categorized as follows: (1) smoking for the “pleasurable-relaxing” effects is the most common reason; (2) smoking for the stimulating effects is next most common; (3) smoking to “reduce negative feelings” or to “relieve anxiety” is the third most common (). This constellation is more similar to that reported by opioid users (cf. Dr. Charles Haertzen, personal communication) than it is, for instance, to that reported by amphetamine users, in which stimulation is the foremost reason for drug-taking behavior.

Personality characteristics and social factors

Social pressure from both peers and family members is critical in initiating and terminating the process of dependence to both tobacco and opioids. Specifically, there is a high probability that friends and family users will share the same pattern of drug use (). Additionally, a prime indicator of treatment success for both cigarette smoking and opioid dependence is the presence of friends and/or peers who have been successfully treated for their dependency (). That there are commonalities in the personalities of tobacco and opioid users is suggested by the fact that most opioid users (about 95 percent) are also cigarette smokers (). As groups, users of different drugs may be characterized by particular constellations of social and personality variables, and these constellations show greater overlap across certain drug classes than others. In this respect, psychological characteristics of opioid users () show considerable overlap with those of cigarette smokers (). Particular points of similarity include an increased prevalence of antisocial and psychopathic tendencies, rebelliousness, anxiety, repressed hostility, and extroversion. Additionally, in both cigarette smokers and opioid users, there is evidence that experimentally elicited aggressive responses are attenuated by use of cigarettes in cigarette smokers () and opioids in opioid users ().

Human Dependence on Tobacco and Opioids: Physiologic Dependence

Deprivation effects

Deprivation of opioids and tobacco increases the tendency of humans and animals to self-administer opioids and of humans to smoke cigarettes. While deprivation of opioids in an opioid user, and possibly deprivation of tobacco in a cigarette smoker usually results in the onset of a withdrawal syndrome, deprivation is, operationally, a temporal manipulation which may increase the reinforcing efficacy of a substance regardless of whether or not a withdrawal syndrome also happens to occur. In clinical studies, a sensitive measure of the deprivation effect is the probability that the drug will be self-administered. With cigarettes this effect was demonstrated in our laboratory when cigarette smokers were deprived 0, 1, or 3 hours and then given access to cigarettes (). Figure For each of three subjects the man number of seconds from the start of the session until the first puff occurred is shown as a functions of hours of deprivation of smoking shows that latency to the first puff following access to cigarettes was inversely related to the duration of the deprivation period. Curiously, a subsequent study showed that “anticipated deprivation” did not produce measurable changes in the smoking of a single cigarette when subjects were given a cigarette and were informed that after smoking that cigarette they would be required to abstain for 0, 1, or 3 hours (). One measure of deprivation is desire to smoke, and several cigarette smoking studies have shown (as noted in the Physiologic Dependence section) that strength of the desire to smoke is a direct function of the deprivation period (). With the opioid drugs it is well known clinically that the probability of self-administration is a direct function of the deprivation period, though this effect is usually considered to reflect the onset of physiologic withdrawal symptoms (). Similarly, several human studies on opioid withdrawal effects have shown that self-reported craving strength is a direct function of the deprivation period ().


Tolerance to toxic or aversive effects of both tobacco and opioids is thought to be important in the ontogeny of dependence. Tolerance may also be a determinant of levels of drug intake. When tolerance is suspected to have occurred at the cellular level it is of additional significance since it may be part of the phenomenon of physical dependence (), and hence share a role in the maintenance of the self-administration behavior (). Tolerance to the various effects of opioids has been extensively studied in both animals and humans (). Tolerance to the effects of nicotine, and to a lesser extent, cigarette smoke, have also been studied in both humans and animals (). The extent to which there are similarities and differences in the development of tolerance to tobacco as compared to the opioids must await further studies. However, it is possible that tolerance to certain effects of smoking may occur more rapidly than opioid tolerance. For example, it is known that tolerance to cardiovascular effects of nicotine can develop within a few hours when nicotine is injected intravenously every 20-30 minutes and that the development of this tolerance is more pronounced in smokers than in nonsmokers (). In our laboratory at the Addiction Research Center, preliminary data indicate that tolerance to certain effects of cigarette smoking, (e.g., attenuation of the patellar reflex and subjective responses) may be lost overnight and gained after a few hours of smoking, while tolerance to other effects may be more slowly acquired and more slowly lost. Interestingly, while tolerance to the initial nausea and dysphoria are thought to be important in the acquisition of smoking, even chronic cigarette smokers whom we have tested usually show these symptoms when they are given a high nicotine cigarette to smoke as their first cigarette of the day and only to a lesser extent when given an identical cigarette to smoke after several hours of normal smoking.

Dose effects on drug intake

Drug dose is an important pharmacologic variable that can determine rate of self-administration and quantity of drug obtained. If the rate of drug self-administration is an inverse function of the unit dose, and total drug intake remains constant across doses, then the organism is “regulating” its drug intake and “titration” or “compensation” is said to have occurred. In animal studies of both intravenous opioid self-administration () and intravenous nicotine self-administration (), drug intake is a direct function of drug dose. That is, except at high doses which have “rate-limiting” effects, drug intake regulation is poor at best. This relationship is distinct from that obtained in studies of intravenous psychomotor stimulant self-administration (e.g., amphetamine or cocaine) where dose regulation is more precise ().

In clinical studies on the effects of drug dose, findings with tobacco are mixed. When nicotine content of cigarettes is varied, findings are similar to those obtained in the animal studies described. That is, nicotine intake increases as a direct function of nicotine dose except at the highest doses, at which rate of self-administration declines sharply (). Dose compensation is much more striking when amount of cigarette smoke is manipulated as may be accomplished by varying cigarette size () or the concentration of the cigarette smoke is varied (). Figure 2 shows that when tobacco product concentration was decreased across sessions, from 100% (no. 0) to 10% (no. 4), number of puffs taken per 3-hour session doubled in 3 subjects tested. Expired air carbon monoxide levels confirmed that measured changes in puff parameters plus unmeasured but likely changes in inhalation parameters resulted in good tobacco smoke dose compensation by these subjects. Thus, while manipulations of cigarette dose may produce good titration, manipulations of nicotine content do not produce reliable changes in rate of self-administration and hence titration (). Comparable studies of the effects of dose manipulations on rate of opioid self-administration in humans have not been conducted. However, clinical studies in which humans are permitted to self-regulate their analgesic drug (opioid) intake indicate that humans are sensitive to drug dose manipulations and suggest that moderate intake compensation occurs ().

Reinforcing efficacy and dependence liability

Retrospective analyses suggest that cigarettes and opioids have high dependence liability; that is, a single exposure to either cigarettes or opioids is often followed by the development of a pattern of compulsive use (a notable exception being therapeutic administration of opioids in clinical settings). Furthermore, once compulsive use develops, users of both opioids and tobacco emit large amounts of work, spend considerable sums of money, and endure sacrifices in health and other areas to maintain their self-administration behaviors. A historical perspective illustrates a similarity with regard to the abuse potential of tobacco and opioids: Cocteau’s dictum regarding opium smoking, that “he who has smoked will smoke” is equally true with regard to tobacco ().

One approach to providing information about the relative dependence liability of drugs is to examine their efficacy in maintaining drug self-administration behavior in laboratory animals. Studies to date suggest that both opioids and nicotine (as well as cigarette smoke) do maintain self-administration in animals; however, opioids appear to be more efficacious reinforcers than nicotine or tobacco smoke. Specifically, while many studies have shown that opioids, delivered intravenously, intramuscularly, and orally, serve as effective reinforcers for animals (), studies involving intravenous self-administration in rats, monkeys, and baboons suggest nicotine to be an equivocal re-inforcer when compared to other drugs of abuse, including opioids (). However, it would be premature to pass final judgement on the results of the animal studies of nicotine and/or tobacco smoke self-administration, firstly, because the route of administration that is preferred by humans (inhalation) has not been extensively used with animals. Just as it required many years to develop a preparation in which orally delivered ethanol served as a potent reinforcer for animals (), it may take a long time to develop the appropriate procedures for studying tobacco use in animal preparations (for a promising development in tobacco smoke self-administration by monkeys). Secondly, the animal nicotine self-administration data are of questionable relevance to the reinforcing efficacy of cigarettes since it is clear that nicotine is not the sole determinant of smoking rates () and that noninhalation routes of nicotine administration are not equivalent to nicotine administration via cigarette smoking (). Finally, there have been no clear demonstrations that intravenously delivered nicotine is an effective reinforcer for humans ().

Another approach to providing information about the dependence liability of drugs is to conduct human studies and systematically evaluate self-reports of subjective “likingw or “drug satisfaction” (). The validity of this approach is suggested by the similarities in the human findings, animal self-administration findings, and in the epidemiological reports of drug abuse (). Intravenous injections of opioids in most addict subjects (), or of nicotine in cigarette smokers () are reported to be pleasurable. With both opioids and tobacco, studies have demonstrated an additional similar relationship: as dose of opioids, intravenous nicotine or cigarettes is increased, subjective reporting indicates that satisfaction also increases ().

Response requirement

Response requirement, also referred to as “response cost,” may be defined as the amount of behavior reauired to obtain a reinforcer. With regard to drug self-administration by humans, response cost can be defined as the amount of effort required to obtain the drug, or as the monetary value of the drug when monetary earning is proportional to work output. Economic theory uses the concept of “elasticity” to describe the extent to which consumption of a commodity varies with the price of that commodity. From a common perspective of drug addiction, opioids and cigarettes might be viewed as relatively inelastic commodities in dependent persons, i.e., that as price increases, consumption would remain relatively constant. In fact, however, both opioid demand and cigarette consumption have proved to be relatively elastic in that consumption decreases when price increases (). This is not to say that increasing the price or response repuirement for cigarettes and opioids does not result in an increase in net expenditure or response output. Response output does increase – it just does not keep pace with the requirements for maintaining a constant level of intake.

A clear experimental demonstration of the interactions between response requirement and intake of methadone or cigarettes was shown in preliminary studies by Bigelow and his co-workers. In these studies, response requirement was defined as the number of lever pulls per delivery of a methadone dose () or a cigarette (). As response requirement increased, for either cigarettes or methadone doses, response rate was an increasing or inverted U-shaped function, and the number of cigarettes or methadone doses obtained decreased. These findings are consistent with epidemiological findings which showed that for both opioids and cigarettes, increased prices result in increased spending but decreased intake (). Analogous results have been obtained in animal drug self-administration studies using opioids (), but these procedures have not been applied in animal studies of cigarette smoking or intravenously delivered nicotine.

Conditioning factors

Conditioning of both the operant type and the respondent (or Pavlovian) type is thought to occur as an integral part of the dependence process with both cigarette smoking and opioid dependence. Specifically, the development of conditioned stimuli, discriminative stimuli, and conditioned responses may contribute to maintenance of the pattern of compulsive use and facilitate relapse following a period of abstinence. For instance, clinical lore suggests that environmental stimuli previously associated with smoking are likely to evoke craving responses and increase the probability of smoking when these stimuli recur (). Therefore, in most smoking treatment programs it is recommended that the abstaining smoker try to avoid environmental stimuli which are highly associated with smoking, e.g., having visual access to cigarettes, social and drinking situations, etc (). Systematic studies are needed to determine if these environmental stimuli elicit withdrawal type responses in a manner similar to the elicited opioid withdrawal described below. It is known that desire to smoke cigarettes may persist for several years after smoking was terminated and that formerly high probability smoking situations are particularly effective at evoking the craving responses (). Finally, a preliminary study by Gritz () has shown that sioht and smell of tobacco smoke are important determinants of smoking rate, demonstrating that tobacco self-administration, like opioid self-administration, may be influenced by external stimulus factors.

With regard to opioid dependence, the evidence that conditioning factors -play a critical role has grown since the notion was first postulated by Wikler (). Recent studies by O’Brien and his colleagues have demonstrated. that opioid withdrawal can occur as a conditioned response to administration of placebo in patients who have previously received naloxone injections (). In another study from the same laboratory, it was demonstrated that subjective and physiologic responses which are normally elicited by opioid administration could also be elicited by presentation of heroin-related stimuli or by the self-administration of placebo in patients with histories of hydromorphone injections (). Analogous findings have been obtained in animal studies (). These studies are important in that they demonstrate that stimuli previously associated with drug administration or drug withdrawal may attain functional roles in the dependence process via conditioning (learning) mechanisms. While further experimental data are required for a more definitive conclusion, it is clear that conditioning factors may be important controlling variables which are common to both opioid and cigarette dependence.

Antagonist administration effects

One factor that distinguishes cigarette smoke from substances such as alcohol, barbiturates, and food is that the primary pharmacologically active constituent (nicotine) has a specific cellular site of action (viz., nicotinic receptors). It is well known that opioids are also receptor-specific. Self-administration of both opioids and cigarette smoke may be influenced by administration of pharmacologic antagonists. Clinical administration of opioid antagonists (e.g., naltrexone) to human opioid users decreases opioid self-administration). The limited available data regarding nicotine antagonist administration showed that mecamylamine (a centrally acting nicotinic blocker) administration to human cigarette smokers produced increases in smoking during weekly 2-hour sessions: it was not determined whether or not continuous antagonist administration ultimately would have reduced smoking rates (). Pentolinium (a peripherally acting nicotinic blocker) did not affect smoking rates. In a study of cigarette smoking by monkeys, mecamylamine (but not the peripherally acting hexamethonium) reduced overall levels of smoking over the course of several weeks (). A caveat with regard to the interpretation of results of antagonist administration in cigarette smoking studies is that, strictly speaking, there is not a tobacco antagonist; rather, there are nicotine antagonist drugs. Administration of nicotine antagonists () or of opioid antagonists () to animals which are intravenously self-administering nicotine or intravenous opioids, respectively, decreases the self-administration behaviors.

The effects of opioid antagonists in blocking or reversing the responses produced by opioids have been extensively studied and reviewed for both humans () and animals (). Preliminary studies of antagonism of the effects of nicotine in animals () and the effects of smoking in humans () indicate similar antagonist blockade and reversal of effects. A noteworthy difference is that opioid antagonists may precipitate withdrawal in opioid-dependent organisms, while no similar phenomenon has been demonstrated in organisms chronically exposed to tobacco smoke or nicotine.

 Preloading effects

In human research, acute preload administration of opioid drugs or tobacco products (e.g., nicotine or cigarette smoke) decreases subseauent administration of opioids or cigarettes, respectively. A good clinical example of this opioid preload effect is the use of methadone to treat illicit opioid dependence (). Jones and Prada () showed that methadone administration to patients who were given the opportunity to obtain intravenous hydromorphone (Oilaudid) produced decreases in self-administration of the opioid. Of six subjects tested, 3 completely stopped working for hydromorphone while the other 3 worked intermittently for hydromorphone. These findings are compatible with those obtained in studies of cigarette smoking in which preloading subjects with cigarette smoke produces a decrease in subseauent smoking (). Nicotine preloading given either orally () or intravenously () also may produce decreases in smoking, although these kinds of preloading manipulations produce weaker and less consistent decrements in smoking than when cigarette smoke preloading is done (). These results show that nicotine is not the sole determinant of cigarette smoking. Consistent with these experimental findings are the modest rates of therapeutic success of preload types of treatment for cigarette smoking (e.g., nicotine-containing chewing gum) which are similar to the modest rates of success of methadone programs for opioid dependence when methadone is dispensed to a heterogeneous population of opioid users.

These drug preloading effects have also been studied in animals where it has been demonstrated that opioid preloads usually () but not invariably (), reduce subsequent opioid self-administration by animals — these opioid findings are consistent with those obtained in human studies. Similarly, one study has demonstrated decreases in cigarette smoking rates in monkeys which occurred when nicotine was added in the monkeys’ drinking water ().

Relapse patterns following abstinence treatment

Hunt and his co-workers have shown that patterns of relapse to drug use following abstinence are similar for cigarette smoking, opioid dependence, and alcoholism (). During the first few months, roughly 70% of patients relapse. Subsequently, the rates of relapse approach asymptotically a level at which about 75% have relapsed, and the rest are still abstaining. These findings suggest an important commonality, but one whose mechanisms are not clear. Perhaps the protracted abstinence syndrome (Physiologic Dependence section), conditioned craving (Conditioning Factors section), or social and personality variables (Personality Characteristics and Social Factors section) are significant.

Feeding behavior effects

Roth opioids and tobacco can reduce feeding behavior and produce weight loss, and intake of both opioids and tobacco may be increased by food deprivation. The effects of food intake on opioid and nicotine self-administration have been experimently studied using animals. Meisch and his co-workers () showed that oral or intravenous etonitazine intake by rats was inversely related to body weight when body weight was manipulated by varying size of the daily food ration. A similar finding was obtained in rats which drank morphine solutions (). In a study of intravenous nicotine self-administration, nicotine was self-injected at significant levels when the animals were at 80 percent of their normal weights but not when the animals were at 100 percent body weight and allowed free access to food ().

In a clinical study of the effects of severe food deprivation, it was found that cigarette smokers smoked much more and that non-smokers learned to smoke. A similar finding was obtained with regard to coffee drinking (). Similarly, addicts sometimes report that when they are hungry (for food) they have a stronger craving for opioids and cigarettes and that they smoke more. Clinically, it has been observed that opioid dependence is frequently accompanied by nutritional deficiency, though it is not clear whether this effect is mediated pharmacologically or sociologically (). With regard to cigarette smoking, epidemiological data have shown that cessation of cigarette smoking is frequently accompanied by a gain (often excessive) in weight (). The possibility of a direct interaction between nicotine obtained by cigarette smoking and appetite has been experimentally demonstrated by Perlick () who showed that subjects who were given low- or no-nicotine cigarettes to smoke ate twice as many jelly beans as subjects who were given high nicotine cigarettes to smoke.

Discussion and conclusions

The behavior of cigarette smoking, as it occurs in many cigarette smokers, may be properly regarded as an instance of drug dependence or as an addiction. As a form of drug dependence, cigarette smoking bears striking similarities in its functional characteristics to the prototypic form of drug dependence – opioid dependence or narcotic addiction. The extent to which similar controlling variables pervade tobacco and opioid dependence may not be readily apparent when only the commonly described features of cigarette smoking and opioid dependence are considered. However, the idea that these two seemingly diverse kinds of drug dependence share some common features is not a new one: comnonalities between tobacco smoking, opioid use (e.g., opium smoking), and alcoholism have been noted for several hundred years (). The present paper has extended these observations by systematically comparing functional relationships found in cigarette smoking and opioid dependence which are empirically based on laboratory and clinical data. The research reviewed shows that many fundamental commonalities exist between tobacco and opioid dependence, adding further support to the notion that cigarette smoking is an instance of drug dependence.

Some of the common functional relationships reviewed in this paper are not uniquely shared by cigarette smoking and opioid dependence. For instance, most forms of drug and substance abuse can probably be reduced by increasing the response requirement necessary to maintain the dependence (). Other commonalities are less widely shared: the similarities in cigarette smoking and opioid dependence noted in the pattern of chronic daily use, receptor specificity, and the role of physical dependence do not appear to be shared with most other forms of drug or substance abuse.

In this paper we have explicitly avoided equating cigarette smoking with nicotine dependence or nicotine self-administration. This approach is consistent with a conservative evaluation of the available data which show that, while nicotine accounts for many of the effects produced by cigarette smoking, clinical and experimental manipulations of nicotine administration do not affect cigarette smoking to the degree that would be predicted if nicotine were the only factor controlling cigarette smoking.

The comparison of the functional similarities between tobacco and opioid dependence has been made possible largely by the application of the methodology of behavioral pharmacology to the analysis of drug dependence. Future basic science research will undoubtably point out further functional similarities and dissimilarities between cigarette smoking and other forms of drug and substance abuse. Such research will ultimately provide a thorough analysis of the dependence process, per se, and will have important implications for the treatment of drug dependence.


Selections from the book: “Behavioral Pharmacology of Human Drug Dependence”. Travis Thompson, Ph.D., and Chris E. Johanson, Ph.D., eds. Presents a growing body of data, systematically derived, on the behavioral mechanisms involved in use and abuse of drugs. National Institute on Drug Abuse Research Monograph 37, July 1981.