Tobacco Documents Online

I I I I I I I I I I I I I 1 1 i CHAPTER 4 Biobehavioral Approaches to Smoking Control J. Allan Best, Patricia E. Wainwright, David E. Mills, and Susan A. Kirkland Smoking•related diseases ue such important causes of dissbility and paeaum nue death in developed countries that the control of cgarette smoldng could do cwre to impzove health and prolong life in these countties than any other singie action in the whole @eld of preventive ctedidne. (WHO Expert Com- auttee. 1979) Researchers involved in smoking control face the perpieadng question of why, given a reasonable understanding of the adverse effects of smok- ing, people still start and continue to smoke. Furthermore, although the majority of smokers express a desire to quit, those that attempt to do so u+e generally unsuccessful (Pechacek, 1979). Numerous programs have been developed that aim not only to aid smokers in their attempts to break the habit but also to prevent the onset of smoking in individuals who represent a population at risk. Unfortunately, these efforts-both smoking prevention and smoking cessation-have been notoriously in- effective (Bernstein & McAlister, 1976; Best & Bloch, 1979; Evans, Hen- derson, Hill, & Raines, 1979; Flay, d'Avernas, Best, Kersell, & Ryan,1983: Leventhal & Qeary, 1980; Lichtenstein & Danaher, 1976; Pechacek, 1979; Thompson, 1978). At best, recent interventions such as social-influences prevention programs and nicotine-replacement-therapy cessation strate- gies appear promising but have not been fully evaluated. I. AGa &rt. Peincir E. W4xwxi&, Lkvod E. Mills. arrd Suaae A. Kfskiend. • Departamnt of HuM stndip, Unive::ity of waterioo, waterloo. Ontuio. Canada N2L 3GL 63 ~ a ~ C~? O . ~

I I I I I I I I I I I I I I 1 64 J. ALLAN sEST n cl. Reviewers argue persuasively that our failure to develop effective intervention methods stems from inadequate analyses of the behavior (Best & Hakstian, 1978; Lichtenstein, 1982; Pechacek & Danaher, 1979; Pomerleau, 1980, 1981). Indeed, our understanding of smoking behavior lags far behind our knowledge of its physiological effects. Clearly, ciga- rette smoking involves more than a physiological addiction to the phar- macological components of tobacco. Motivational and situational factors have been implicated, but there is a paucity of empirical data addressing the interplay of these and other social, psychological, and pharmacologi- cal factors. There exist even fewer data characterizing the dynamic pro- cesses of behavior change involved in smoking prevention and cessation. Factors influencing cigarette smoking appear to vary as the individual moves from initiation to maintenance, to cessation, and resumption or relapse (Best & Hakstian, 1978; Best, Owen, & Trentadue, 1978; Le- venthal & Cleary, 1980; Ontario Council of Health, 1982; Pechacek & McAlister, 1979). Thus, if we are to develop effective interventions for smoking control we must address the following issues: 1. Smoking initiation: What factors influence experimentation and initiation of smoking, and how might the relative contributions of these factors change over the course of the initiation process? Simply, the question is, Why do individuals begin smoking? 2. Smoking maintenancr: Once a pattern of regular or chronic smoking has developed, what are the factors that maintain this behavior? That is, why do people continue to smoke, often even in the presence of compelling reasons for change? 3. Smoking reductions: Given a stabilized rate of smoking, what are the acute effects of smoking reduction? In essence, the question is, What is smoking withdrawal? 4. Maintenance of smoking reduction and/or recidivism: If significant smoking reduction is achieved, what are the factors that might precipitate relapse in the long term? In other words, why do people often fail in their efforts to maintain a reduced (or zero) rate of smoking? Note that we speak of smoking reduction to indude two perhaps quite different phenomena: cutting down on smoking and quitting en- tirely. Available research on smoking reduction processes refers almost exclusively to cessation, despite the fact that many smokers try to cut down rather than quit. Intervention programs can be designed to elimi- nate the behavior entirely or to change the behavior to reduce potential risk. Less hazardous smoking might be achieved through substance changes (switching to lower tar and nicotine cigarettes), through other C? ~ CD . ~ t7T I

I I I I I I I I I I I 1 I I ,~0~. APPROACHES TO SMOIGIVG CONTROL 65 ~,,,,,or changes (reducing the number of cigarettes smoked per day or ~u~g the yield of each cigarette by taking fewer puffs, inhaling less a ~ p~)tive feas b~tq and health c seq ences of thesevtwo~modes of ~ge is not currently available (Ontario Council of Health, 1982). Un- derstandably, then, health professionals might argue that cessation is a ~etter p~gram objective than reduced exposure. For the purposes of this ,,,c.ssion, however, we will use the general term smoking reduction to refer to either condition. •rne purpose of this chapter is to discuss the possible interplay be- ,h,,en behavioral and biological factors that influence smoking and ,he,eby contribute to our understanding of why interventions succeed or ~, First, a brief description of three relevant models of learning will be p~nted. Second, two central biobehavioral issues will be discussed: ~e pharmacological aspects of nicotine in smoking, and self-regulatory od rompensatory mechanisms controlling smoking behavior. Third, we ,,,ill propose preliminary models of the processes smokers encounter. Current behavioral approaches to smoking prevention and reduction will ye discussed in light of these accounts. Finally, implications for both ,,MTention and research will be outlined. Smoking Behavior and Learning Three distinct models of learning, well known to psychologists, have been related to smoking behavior. Social learning theory (Bandura, 1977) buiids on operant conditioning (Skinner, 1953) to describe social infiu- ences and reinforcement mechanisms. Associative learning theory (Hunt, Matarazzo, Weiss, & Gentry, 1979) attempts to explain the habit aspects of chronic smoking. Classical conditioning theory (Siegel, 1982; Stewart, deWit, & Fikelboom,1984) provides a way of integrating pharmacoiogical and physiological factors in a comprehensive learning-theory analysis of smoking behavior. Interested readers are referred to thoughtful discus- sions by others of the different ways in which diverse biobehavioral theories and data can be integrated around smoking (cf. Ashton & Step- cey,1982; Best & Hakstian, 1978; Glad, Tyre, & Adesso, 1976; Leventhal & Cleary, 1980; Mausner & Platt, 1971; Pechacek & Danaher, 1979; Pomerkau,1980, 1981; Raw, 1978; Russell, 1974). Different kinds of learning may occur and operate simultaneously. Indeed, the learning principles described above do not constitute pure examples of one form of learning, and the three theories outlined share some common features. For example, all are concerned with discrete ' ZZ Q W CU I

I I I I I I I I I I I I I I I 66 J. ALLANti BEST et at environmental situations and a scientific analysis of environment-behav- ior relationships. All focus attention on the antecedents and conse- quences of smoking. In addition, however, each theory has distinguishing features, and each has a relatively distinct focus. Operant and Social Learning Theories Behavior that produces positive results tends to be repeated. Thus, smoking is an operant to the extent that it acts to produce effects people desire-feeling accepted by peers, relief from boredom, relaxation-pos- itive reinforcers that contribute to the development and maintenance of smoking behavior. A primary reinforcer is one that is unlearned. For example, the act of smoking may produce biological effects that are natu- rally reinforcing to the organism (see the following section on the role of nicotine). Smoking also may be maintained by secondary reinforcers. For example, if an individual finds some social situations stressful and re- peatedly smokes to relieve that stress, the cigarette itself may come to serve as a secondary reinforcer with the ability to reduce stress in subse- quent social situations. Stimuli preparatory to the act of smoking (e.g., the sight of a cigarette) function as secondary reinforcers for behavior preceding them (e.g., picking up a pack of cigarettes) as well as discrimi- native stimuli that set the occasion for behavior to follow them (e.g., lighting a cigarette and inhaling). These chains of smoking beha:-ior de- velop such that smoking is controlled by both cues (events signaling potential reinforcement for smoking) and consequences (the process of reinforcement itself). From this perspective, an understanding of smok- ing requires both an analysis of the cues niggering smoking (e, g. , sight of other smoking, physiological reactions associated with stress) and the consequences of smoking (e.g., relief from boredom, feelings of sodal belonging) (Best & Hakstian, 1978). Note that a range of social, psycho- logical, and biological events can serve as both cues and consequences. Negative reinforcement refers to situations in which a behavior is strengthened, not because it produces positive consequences, but be- cause it prevents or relieves negative consequences. As smoking be- comes established, various states of discomfort become associated with nonsmoking, providing an escape-avoidance contingency for smoking behavior. Once smoking comes to relieve withdrawal, it may, through the process of stimulus generalization, come to relieve other disphoric states as well, for example, anger, tension, and boredom (Best & Haks- tian, 1978; Tomkins, 1966). Thus, smoking may serve as a generalized primary and secondary reinforcer providing both positive and negativE I

I I I I I I I I I I I I I I I a,ua1~HAyZpRAL APPROACHES TO SMOKNG CON'TROL 67 t,foKement over a remarkably wide range of life situations (Pomer- ~ 1980). ~eau'principles of operant learning suggest that the chains of behavior ,rvolved in smoking are strengthened in several ways. Because the act of -mo~g over the years is performed tens of thousands of times, in a .1de variety of situations, the conditioned links between smoking cues ,,d behavior will become very strong indeed. Repeated, consistent rein- rorcement will serve to strengthen links; the complexity of cue and conse- quence relationships will make the behavior highly resistant to change. Two more concepts are central to initiation and reduction processes. p.st, social learning theory emphasizes modeling as an important mech- a,,m (Bandura, 1977). Thus, children "learn" smoking behavior by watching parents and peers doing it, in addition to the learning that may occw through reinforcement as they begin to experiment with smoking (fyy et a1., 1983). Second, smoking reduction involves self-control. If smoking reliably produces immediate, powerful reinforcers, then not Smokina in a smoking situation requires an act of self-control-postpon- ;ng immediate satisfaction, or reinforcement, for long-term benefit. Be- havioral self-management theory and procedures can be brought to bear ;n developing "nonsmoking behaviors" -behavioral responses to smok- Ing situations that serve as operants and thus, as strengthened with time, rome to produce reinforcement without smoking (Best, 1980). Associative Learning Theory Hunt and his colleagues (Hunt et ai., 1979) discuss the concepts of associative learning and habit as they relate to health behavior. The theory begins by assuming operant learning principles and then extends these in several significant ways. A habit is defined as "a stable pattern of behavior marked by automaticity and unawareness and influenced by associative learning as well as reinforcement" (Hunt et ai.,197'9, p.112). A two-process theory of learning is thus proposed. Smoking behavior origi- nally develops through operant learning. However, as it becomes over- learned and habitual, associative learning gradually takes over. The behavior becomes increasingly automatic, occurring without much awareness, and is performed so quickly that there is little opportunity for motivationalarousal and reinforcement. Smoking becomes progressively indepaldent from reinforcement, to the point of being "functionally au- tonomous," characterized by continued maintenance of behavior pat- terns long after the apparent disappearance of original reinforcing I

I I I I I I I 1 I I I. I I I I I I I I 68 ALLAN BEST et al. consequences. As a result, habitual smoking displays a strong resistance to extinction. Classital Conditioning Theory In recent years, classical conditioning theory has enjoyed growing popularity as a model for understanding addiction and its development (Pomerleau, 1980, 1981; Siegel, 1982; Solomon, 1977; Solomon & Corbit, 1973, 1974; Stewart et al., 1984; Ternes, 1977). Again, the model takes nothing away from operant and associative learning accounts of smok- ing; rather, it extends the models to account for additional aspects of smoking behavior. Classical conditioning accounts of smoking are distinguished by a concern with stimulus-stimulus relationships, that is, an understanding of how antecedents or cues come to trigger a seemingly reflexive action. The theory has particular value in explaining drug tolerance and with- drawal phenomena. Solomon's (Solomon, 1977; Solomon & Corbit,19T3, 1974) opponent process theory of acquired motivation builds on classical conditioning to explain addictive phenomena. The model assumes cen- tral nervous system mechanisms which operate homeostatically to regu- late affect. A stimulus, such as nicotine, may elidt an unconditioned alpha-process. The alpha-process will in turn trigger an opposite beta- process to return the organism to homeostasis. The key postulate of opponent process theory is that the beta-process will become condi- tioned to stimulus aspects of the drug-taking (smoking) situation. Thus, cues associated with smoking (e.g., the sight of a cigarette, a stressful situation) come to trigger the opponent beta-process. Environmental events, particularly those that affect mood or signal performance de- mand, modulate the reinforcement value of nicotine self-administration. As a consequence, more of the drug, in this case nicotine, will be required to produce an alpha-process sufficiently strong to offset the beta-process and produce central nervous system (CNS) gratification. In other words, tolerance will develop. If an individual does not smoke in the presence of conditioned smoking eues, there will be no alpha-process to balance the conditioned beta-response, homeostasis will be disrupted, and aversive physiological reactions (withdrawal) will ensue. Smoking as an operant thus is positively reinforced by producing pleasure and negatively rein- forced by terminating withdrawal, setting up an addictive cycle whereby the individual smokes as much as to avoid withdrawal as to produce pleasure. 4

I I I I I I I I I I I I I I I I I I B10BEIjAVIORAL APPROACHES TO SMOKIIYG CONTROL 69 ;viost opponent process research has been done with addictive drugs other than nicotine, notably opiates, and primarily using animal models (gee Siegel. 1982, for a review). Though the data seem to fit classical conditioning principles and the opponent process theory rather well, it ma~ be the case that this is not the complete story. Recent data from ~al experiments indicate that in many cases, rather than opposing the effe~cts of the drug, many conditioned drug effects directly mimic the positive affective aspects of drug use (reviewed by Stewart, et a1., 1984). The,e authors propose that the self-administration of opiates and stimu- lots is maintained by the effect of conditioned stimuli in producing a state sirr~ to that produced by the drug and thereby "priming" the organism to respond to drug-related stimuli and to increase drug-seeking yehavior. This mechanism may be more useful in explaining the develop- ment of dependence in situations in which the pattern of use would not support the development of physiological withdrawal symptoms. An ;mportant aspect of the model of conditioned drug effects is that it sug- gests how environmental stimuli can precipitate aaving for a drug even aher an extended period of abstinence has ended physical dependence. Biobehavioral Processes in Smoking Many of the processes discussed above assume that there is an al- tered physiological state induced by smoking. In this section we will consider these pharmacological effects, followed with a discussion of how such effects might influence self-regulatory behavior. Pharnracology of Nicotine The products of combustion of tobacco include various "tars" and cubon monoxide in addition to the alkaloid nicotine. Although the other components do have physiological actions and contribute to the long- term health hazards associated with smoldng, the acute effects of inhal- ing cigarette smoke have been mainly attributed to nicotine. The free base, which is present under alkaline conditions, is lipid-soluble and therefore is readily absorbed through the ceil membranes of the lungs, sidn, bueral and nasal mucosa, gastrointestinal tract, bladder, and renal tubules. This has important clinical implications in that nicotine levels in the body can be altered by manipulating the pH at sites of absorption or aceedon. For example, nicotine chewing gum is buffered at an alkaline I

I I I I I I I I I I I I I I I ?p J. ALLAN BFST Ce atl pH to enhance absorption (Russell, Raw, & Jarvis, 1980) and the urine can be acidified to enhance excretion (Beckett, Rowland, & Triggs, 1965; Schachter, Kozlowski, & Silverstein, 1977). The liver microsomal enzyme system is the major site of metabolism, with 80-90% of the compound being modified prior to excretion by the kidneys. The major metabolites are cotinine and nicotine-N-oxide, which are pharmacologically inactive (Russell & Feyerabend,1978). A consider- able proportion of nicotine taken orally is metabolized during its first passage through the liver before ever becoming available for distribution by the systemic dreulation, accounting for the low activity of nicotine administered in the form of oral tablets. There is some evidence that chronic dgarette smoking may increase the activity of the hepatic meta- bolic system, leading to an increase in the deactivation of nicotine as well as of other drugs (Beckett & Triggs,1967). This will lead to an inaease in dosage requirements to compensate for an increase in the breakdown of the drug and constitutes the basis for the phenomenon known as meta- bolic tolerance (Goodman Gilman, Goodman & Gilman, 1980). The plasma half-life (t1a) of nicotine is dose to 2 hours (Benowitz, Jacob, Jones, & Rosenberg, 1982). Accordingly, observations have been made that blood nicotine concentrations accumulate for 4-6 hours with regular smoking, and elevated concentration persist in the blood over- night (Benowitz, Kuyt, & Jacob, 1981). The actual concentration attained will be a function of the dosage size and interval as well as the plasma half-life. It is appropriate to note here that there will be differences among individuals with respect to the rate of nicotine metabolism. These differences arise as a function of various factors such as genetic variation (including sex, age, and health status), as well as prior exposure to drugs or chemicals. There is some evidence that aging and declining metabolic rates may be reflected in declining rates of smoking (Garvey, Bosse, & Seltzer, 1974). If the motivation of smoking behavior is that of maintaining the plasma nicotine concentration within certain limits, this can be achieved most efficiently by a dosage regimen that provides an initial high "load- ing" dose, followed by smaller maintenance doses at appropriate inter- vals. It is possible that individual variability in metabolic processes as well as differences in tissue sensitivity with respect to the effects of nico- tine are factors contributing to the individual differences in smoking schedules. The preferred method of nicotine self-administration is through in- halation. Because of the large surface area of the Iungs accompanied by the slightly alkaline pH of the surface fluids, absorption of nicotine is I

I I I I I I I I I I I I I I I I I Bj08E?Lp,VIORAL APPROACHES TO SMOICNG CONTROL '1 yoth rapid and efficient. It has been shown that the inaease in plasma nicotine concentration as a result of repeated inhalations of cigarette smoke is similar to that following a "bolus" intravenous injection of a ;XWar dose (Russell & Feverabend, 1978). This means that with succes- ;lve inhalations the tissues will be intermittentlv exposed to peak concen- tratnons of nicotine that are much higher than those that would be at- tained were the same dose to be administered continuously. It has been shown that the response of cells to the same dose of nicotine differs as the mode of administration is varied (Armitage, Hall, & Sellers, 1969). Other methods of nicotine administration, such as nicotine chewing gum, are ,u,able to simulate this pattern of exposure, and this may explain why agarette smoking is more popular than chewing tobacco or taking snuff. The nicotine in the blood travels from the lungs to the heart and thence to the entire body. It is estimated that it takes 7-8 seconds for the nicotine from a single puff to reach the brain (cited in Russell & Feyera- bend, 1978). Nicotine is not distributed evenly throughout all the body tissues and is actively sequestered by the nervous system, leading to concentrations in the brain much greater than those in the blood. How- ever, because of its high lipid solubility, nicotine can diffuse freely out of cells; thus, it also leaves the brain quickly and so the maintenance of these high concentrations will depend on replenishment from the con- centrated "boli" obtained through repeated inhalations. Many of the physiological actions of nicotine can be attributed to its structural simiiarity to the neurotransmitter acetylcholine and its conse- quent ability to act on certain of the cholinergic receptors (reviewed in Goodman Gilman et al.. 1980). Acetylcholine is the neurotzansmitter re- leased by preganglionic fibres to all ganglia in the autonomic nervous system and the adrenal medulla. In addition, it is released by post- ganglionic parasympathetic fibres to effector organs as well as being the neurotransmitter of motor fibers to skeletal muscle. Receptors are regions on the presynaptic or postsynaptic cell membrane that recognize the structure of a drug in a specific way. An agonistic drug binds the receptor and activates the receptor complex which then mediates the biological effect. Conversely, an antagonist drug will block this effect. In some cases a drug will initially act as an agonist when it binds to the receptor, but if it is not readily displaced it will prevent the receptor from being further stimulated and thereby act as an antagonist. The choiinergic re- ceptors can be divided into two populations on the basis of their sensi- tivity to either nicotine or another naturally occurring alkaloid, muscarine. All the receptors responsive to acetylcholine in both the au- tonomic ganglia and in skeletal muscle are also responsive to nicotine I

I I J. ALLAN BEST et at. I I I I I I 1 I I I I I I 1 I and are termed nicotine receptors. Although both acetyicholine and nico- tine stimulate the receptors, nicotine, unlike acetyicholine, is not rapidly inactivated, so the combination with the receptor is of relatively longer duration. In fact, as the dosage inaeases, nicotine will eventually block the receptors so that at toxic doses no further transmission is possible. This accounts for the biphasic action of the drug wherein the initial effect is stimulatory but later on inhibition predominates. The peripheral effect of small doses of nicotine is that of activation of the sympathetic division of the autonomic nervous system. As the dose increases, stimulation of the sympathetic ganglia is followed by a similar effect on the parasympathetic ganglia, then later blockade of the para- sympathetic and then the sympathetic ganglia. Consequently, the physi- ological effects of nicotine are dose-dependent; however, those seen after cigarette smoking are generally those of sympathetic stimulation. Epi- nephrine is released from the adrenal medulla and norepinephrine from sympathetic nerve endings (Cryer, Haymond, Santiago, & Shah, 1976). This increase in catecholamine release has several biological effects, in- cluding vasoconstriction and an increase in heart rate, blood pressure, and blood sugar. As a result of the lowered peripheral blood flow, skin temperature decreases (Stephens, 1977). The influence of nicotine on the chemoreceptors in the carotid and aortic bodies leads to a modest in- aease in respiratory rate. It can also cause nausea and vomiting by stimu- lating the chemoreceptor trigger zone of the medulla oblongata and by activating the vagal reflexes involved in emesis. An interesting observa- tion is that, like acetylcholine, nicotine stimulates sensory receptors. This may contribute to the experimental finding that cigarette smoking is accompanied by a deaeased consumption of sweet-tasting, highly calo- ric foods (Grunberg, 1982). It is possible that this could explain, in part, the lower body weight commonly found in cigarette smokers and the fact that weight gain is often an unwanted consequence of smoking reduc- tion. Nicotine as an Addictiae Substance The effects of nicotine on the central nervous system are crucial to an understanding of smoldng behavior. Smoking has been descibed as an addictive or dependent behavior in that an individual will go to great lengths to continue smoking and will show difficulty in ceasing volun- tarily. The question is whether the pharmacological effects of nicotine act as reinforcers which serve to maintain the behavior. In this-context there are two types of reward to consider- either the positive reinforcing effect of nirntine through its effect on the central nervous system, or the nega- I