Is Behavioral Tolerance Learned?

MURIEL VOGEL-SPROTT, PH.D.

Both scientific and anecdotal evidence indicates that social drinkers can develop resistance (i.e., behavioral tolerance) to alcohol’s impairing effects over time. Although repeated exposure to alcohol is thought to explain tolerance development on a physiological level, the acquisition of behavioral tolerance appears to involve additional factors. In particular, learned associations between a drinker’s behavior following alcohol consumption and the subsequent consequences may play an important role. When favorable consequences result from displaying unimpaired (i.e., tolerant) behavior after drinking, a drinker learns to develop behavioral strategies to compensate for alcohol’s effects. In contrast, if a drinker does not receive a reward for unimpaired behavior—or finds that a more favorable outcome follows the display of intoxicated behavior—tolerance does not develop. Studies show that subjects also can develop behavioral tolerance to alcohol when they practice a task while impaired by factors other than alcohol or when they mentally rehearse task performance while under the influence of alcohol. KEY WORDS: AOD tolerance; reinforcement; AOD impairment; AODE (alcohol and other drug effects); behavior; AOD intoxication; expectancies; learning; context dynamics; literature review

he term “tolerance” refers to a conditions (i.e., homeostasis). A com- exposure to alcohol, giving their bod- reduction in the intensity of the pensatory response usually is assumed ies repeated opportunities to activate Teffect of alcohol (or other to become stronger each time a person increasingly strong compensatory drugs) over the course of repeated use. uses alcohol or other drugs and to responses. In addition, alcoholics may Thus, a person developing tolerance subside gradually during a period of display remarkable resistance (i.e., to alcohol must drink greater quanti- abstinence. tolerance) to the effects of drinking ties of alcoholic beverages to produce Research on the effects of alcohol alcoholic beverages in quantities that the same effect that had been previ- on cellular and neuronal functions has would greatly impair social drinkers. ously achieved at a lower consump- provided much information on where tion level. This phenomenon may and how alcohol acts in the brain to MURIEL VOGEL-SPROTT, PH.D., is a produce its physiological effects. In reflect the acquisition of tolerance to professor in the Department of alcohol’s physiological effects (e.g., contrast, no direct causal relationship Psychology at the University of reduced body temperature) as well as has been established between any Waterloo, Waterloo, Ontario, Canada. its behavioral effects (e.g., impaired specific biological change induced by motor coordination). Researchers alcohol exposure and a particular 1Throughout this article, the term “social generally believe that the physiologi- behavioral response (e.g., Hunt 1993). drinker” refers to a person who consumes cal actions of alcohol and other drugs Nonetheless, clinical observations alcohol but does not experience any alcohol- contribute to tolerance by triggering suggest that tolerance plays an impor- related problems and is not alcohol dependent. the body to produce opposite physio- tant role in alcohol abuse and addic- The terms “alcoholic” and “alcohol abuse” as used in this article encompass all levels of logical reactions in an effort to com- tion. For example, compared with problem alcohol use and do not refer to a pensate and restore stable internal social drinkers,1 alcoholics have more particular diagnostic system.

VOL. 21, NO. 2, 1997 161 These observations accord with the fully account for the development of MacNish’s notion that circum- notion that apparent sobriety after behavioral tolerance in social stances following drinking can affect drinking or an “ability to handle one’s drinkers. Thus, the question remains: behavioral tolerance has been liquor” (i.e., exhibit tolerance to alco- What additional factors in drinking advanced during the 20th century as hol’s behavioral effects) may be a situations affect the acquisition of well. Goldberg (1943) speculated that useful pathological diagnostic symp- tolerance to alcohol’s behavioral ef- “psychic compensation” contributed tom (American Psychiatric Assoc- fects among social drinkers? to the tolerance displayed by alco- iation 1994). This article focuses on an intrigu- holics. A few decades later, Dews The transition from social to abusive ing potential factor currently undergo- (1962) proposed that drinkers’ toler- drinking often occurs gradually over ing investigation in behavioral ance might result from new learned years of drinking sessions, however, neuroscience research—the idea that behavior that compensates for alco- and only in some drinkers. This finding environmental events associated with hol’s impairing effects. In addition, suggests that the acquisition of behav- a drinker’s behavior following alcohol MacAndrew and Edgerton (1969) ioral tolerance to alcohol in social consumption (e.g., positive or nega- noted that drinking orgies in aborigi- drinkers results from factors in addition tive consequences) play a role in de- nal cultures resulted in gross intoxica- to the extent of alcohol exposure. termining whether the drinker will tion in some societies but tolerance in One research-supported explana- develop behavioral tolerance (see box, others; they attributed this variance to tion involves environmental events p. 167). This research has been guided learned conformity to different cultur- that predict drug administration. by the theory that environmental ally specific standards of behavior Investigators have found that the events known to affect the learning of under alcohol. expectation of receiving a drug can new behavior likewise will influence Anecdotes about grossly inebriated affect tolerance. For example, studies behavioral tolerance to alcohol.2 This drinkers who become sober when they have shown that animals made toler- article first presents some historical believe it is important to do so also ant to a drug through repeated admin- background for theories on the devel- suggest that events after drinking istrations in a distinctive setting opment of behavioral tolerance, then alcohol can affect tolerance. This subsequently will display greater reviews contemporary findings and ability of drinkers to “sober up” ap- tolerance when the drug is adminis- their implications. (For a more de- parently provided the impetus for tered in this same setting as opposed tailed discussion of this work, see introducing the breathalyzer to mea- to a new one (see, for example, Siegel Vogel-Sprott 1992.) sure blood alcohol concentration 1989). Researchers have interpreted (BAC) levels for forensic purposes. these results in terms of associative Before the breathalyzer was available, HISTORICAL BACKGROUND learning: When distinctive events a medical examination of suspected reliably precede drug administration, Although the view that tolerance intoxicated drivers was required to they serve as a signal that provides a stems from alcohol-induced compen- support the charge. Goldberg and basis for expecting the drug. When satory reactions is prevalent in the Havard (1968) reported that these tolerance is established, this expecta- scientific literature, other possibilities clinical assessments were completely tion results in anticipatory compen- have received attention. The idea that unreliable. They noted that suspects satory reactions to reduce the drug’s environmental events following alco- faced with a doctor called in by the effect. Consequently, “tolerance is hol consumption contribute to toler- police often were capable of “pulling maximally displayed following ‘ex- ance development has roots in the themselves together” to pass all the pected’ drug administration but not 19th century, when opinion held that clinical tests. After the suspect satis- following ‘unexpected’ drug adminis- tolerance was largely under volitional fied the police physician and was free tration” (Siegel 1989, p. 116). control. For example, more than 100 to leave, however, signs of intoxica- People likely acquire tolerance to years ago, MacNish penned the fol- alcohol’s behavioral effects in drink- lowing observation: 2This article presents evidence that learning ing situations where reliable cues, plays a crucial role in the acquisition and The mind exercises a consider- display of behavioral tolerance to alcohol in such as liquor bottles, signal alcohol able effect upon drunkenness, and drinkers who have no alcohol-related prob- availability. Research indicates that may control it powerfully. When lems. For drinkers who have become physical- when alcohol is expected and in the company of a superior ly dependent on alcohol, however, the story received, however, a social drinker whom we respect, or of a woman may differ. Research cited in this article still leaves open the possibility that the behavioral may demonstrate behavioral tolerance in whose presence it would be tolerance of alcohol-dependent drinkers while performing one task but not indelicate to get intoxicated, a involves some alcohol-induced biological another (see, for example, Vogel- much greater portion of liquor changes. In other words, the processes initiat- Sprott 1979). Alcohol expectations may be withstood than in soci- ing behavioral tolerance to alcohol may be entirely different from those sustaining toler- arising from events that precede eties where no such restraints ance after physical dependence has been drinking therefore do not appear to operate (MacNish 1832, p. 45). established.

162 ALCOHOL HEALTH & RESEARCH WORLD Is Behavioral ToleranceRunning Learned? Heads

tion reasserted themselves. As a re- ment under the alcohol dose and as- Sprott 1981; Zack and Vogel-Sprott sult, the police frequently had to assist sess their development of tolerance by 1993). suspects from the station and escort noting any reduction in impairment as A drinker may readily discriminate them home. the alcohol dose is repeated over time. between the presence or absence of an Tolerance on the part of intoxicated One way to reward sober behavior immediate reward for sober perfor- suspects appears to depend on the is to pay drinkers whenever their mance, but this circumstance is unlikely expectation of a reward for sober be- performance while intoxicated match- to characterize social drinking situa- havior. The suspects exhibited sober es their level of achievement while tions. More commonly, behavior after behavior (i.e., tolerance) when they sober. In studies employing this re- drinking may result in a significant perceived a payoff for doing so but did ward treatment, researchers have consequence or outcome only when not act sober in the absence of a re- observed a progressive development such behavior deviates from a socially warding consequence for compensat- of tolerance with repeated drinking accepted standard (e.g., when behavior ing for alcohol’s effect. In this respect, sessions (e.g., Beirness and Vogel- is obnoxious). In these situations, the tolerance resembles a goal-directed Sprott 1984). In addition, they have consequence is likely to be negative. (i.e., instrumental) learning response found that no such increase in toler- Therefore, the absence of an unfavor- that becomes dominant when associat- ance occurs under control conditions able consequence serves as a reward for ed with a reward and extinguishes in which the reward is unrelated to displaying tolerance as sober behavior. when the reward is withdrawn. sober performance. Although social Withdrawal of the reward, however, drinking situations ordinarily do not means that nothing happens (i.e., no pay drinkers when they act sober, unfavorable consequence ensues), even EFFECTS OF REWARDING SOBER such behavior might result in verbal though the drinker meets the socially BEHAVIOR approval from others (e.g., the com- accepted standard of behavior. Using simple and complex psychomo- ment “good”). Some studies have Consequently, the drinker cannot dis- tor tasks as tests, researchers have compared the rewards of monetary tinguish the presence of a reward from studied the effect of rewarding sober payment versus more realistic verbal its absence, because both conditions are performance on the development of approval for sober performance in characterized by the lack of an aversive alcohol tolerance. Motor-skill tasks, terms of their tolerance-inducing consequence. such as tracking randomly presented effects (e.g., Sdao-Jarvie and Vogel- Learning studies indicate that re- visual targets, typically require eye- Sprott 1991). Interestingly, these warding a response by withholding an hand coordination of complex studies have found that both forms of aversive consequence results in the responses. Alcohol-induced impair- reward have similar efficacy. In gen- acquisition of a persistent response that ment of a psychomotor skill is charac- eral, a favorable outcome for sober is difficult to extinguish when the re- terized by reduced accuracy, slower behavior appears to enhance the de- ward is removed. Researchers have performance, or both. In studies using velopment of tolerance in social obtained similar results using this pro- motor-skill tasks, groups of social drinkers, and different types of favor- cedure to develop behavioral tolerance drinkers learn to perform a task, then able outcomes can accomplish the through repeated doses of alcohol: In a typically attend three or four weekly same end (see figure 1 for the results study by Zack and Vogel-Sprott (1995), drinking sessions in which they re- from one study). subjects performing a task received an ceive 0.62 gram of alcohol per kilo- In contrast, behavioral tolerance unfavorable verbal consequence (e.g., gram of body weight. (For a person disappears when the reward is with- the comment “bad”) when their perfor- weighing 154 pounds, this dose equals held (Mann and Vogel-Sprott 1981; mance deviated from a sober standard approximately three 12-ounce bottles Zack and Vogel-Sprott 1993). This of behavior (i.e., the subjects’ alcohol- 1 of 5-percent beer.) During each 2 ⁄2- finding is consistent with learning free level of proficiency on the task). hour session, the subjects perform a studies showing that subjects trained The absence of this consequence served task at regular intervals as their BAC to produce a particular response for an as a reward when they matched the rises to a peak (to approximately 0.08 immediate reward will extinguish this standard. Such training encouraged the percent, which is the BAC limit for response readily when the reward is subjects to act sober (i.e., it enhanced drivers in at least 14 States) and sub- withheld. After subjects had acquired their behavioral tolerance to repeated sequently declines. In these types of tolerance during drinking sessions in doses of alcohol). Furthermore, the studies, researchers determine alco- which they were immediately reward- subjects retained this tolerance well hol’s impairing effect by measuring ed for sober performance, for exam- during subsequent drinking sessions in how much the subjects’ performance ple, their tolerance vanished when the which all consequences for perfor- differs from their baseline proficiency reward was withheld on subsequent mance were withheld. on the task while sober. Each week of drinking sessions, even though they Taken together, evidence on the the study’s duration, the researchers continued to consume the same acquisition, extinction, and retention measure the subjects’ average impair- amount of alcohol (Mann and Vogel- of alcohol tolerance in social drinkers

VOL. 21, NO. 2, 1997 163 response to yield the same outcome each time. Applying this interpretation 4 to the evidence described thus far suggests that drinkers display greater tolerance in situations in which they yyy expect sober behavior to yield the most favorable outcome. 3 This conclusion also suggests that

yyy rewarding other types of behavior after drinking likewise would increase the occurrence of those behaviors. For yyy 2 example, a drinking situation in which flagrant intoxication is rewarded

yyy would be expected to lead to an inten- sification of behavioral impairment— a prediction that recently has been 1

yyy

Score Change (in seconds) confirmed (Zack and Vogel-Sprott in press). In this study, a group of sub- jects displayed intense impairment

yyy following repeated doses of alcohol 0 when impaired behavior was reward- ed, whereas another group displayed

yyy tolerance when the same reward was given for sober behavior. Loosely speaking, whether a drinker displays Favorable verbal feedback plus financial reward for sober performance tolerance or gross impairment as a Favorable verbal feedback for sober performance result of drinking alcohol apparently depends on which behavior he or she Imagined performing the task and receiving favorable verbal feedback and financial reward for sober performance expects will yield the more favorable outcome. This interpretation closely Received financial reward but did not know it was associated with sober parallels some learning theories that performance

y attribute the ability to adapt a goal- Received nothing (i.e., no outcome) for sober performance directed response to expectancies acquired from learned associations Figure 1 Alcohol-induced impairment of task performance. Completion times for between the response and its outcome subjects’ sober performance of an eye-hand coordination task were (e.g., Bolles 1979). obtained and used as baseline scores. Subjects then had a chance to overcome impairment (i.e., build tolerance) during three sessions of alcohol consumption followed by either performing or mentally rehearsing COMPENSATING FOR ALCOHOL the task. Next, all subjects performed the task after drinking. This graph EFFECTS shows the difference for each group between the average time needed to complete the final task and the average baseline score. Subjects in Substantial research indicates that a groups receiving actual or imagined rewards during previous sessions compensatory response that counter- displayed greater tolerance by achieving times closer to their sober acts the effects of a drug underlies performance (i.e., low score changes). Results indicate that actual or mental practice, when associated with a reward for unimpaired behavior, tolerance. Various studies (e.g., Siegel enhances behavioral tolerance. 1989) have shown that animals that have developed tolerance may display NOTE: Zero = Sober baseline score. a compensatory reaction when they SOURCE: Adapted from Sdao-Jarvie, K., and Vogel-Sprott, M. Learning alcohol tolerance by receive a placebo in the presence of mental or physical practice. Journal of Studies on Alcohol 53(6):533Ð540, 1992. p. 538. cues signaling drug administration (i.e., when they expect a drug but receive a placebo instead). In drinkers indicates that environmental outcomes figure 2). These results can be ex- who display alcohol-tolerant behavior, of alcohol-induced behavior are im- plained by associative learning: When some compensatory response also portant predictors of the degree of a reliable association exists between a may be operating to counteract alco- behavioral tolerance that social behavioral response and a particular hol’s disrupting behavioral effect. A drinkers will display (see flowchart in outcome, people learn to expect the compensatory response cannot be

164 ALCOHOL HEALTH & RESEARCH WORLD Is Behavioral ToleranceRunning Learned? Heads

Drinking

Which response leads to more favorable outcome?

Intoxicated Sober behavior behavior (tolerance)

No reward/ No reward Reward Reward no negative or reward is unrelated consequence

Learned Intoxicated tolerance Tolerance not developed behavior intensifies develops Subsequent drinking episodes

Presence/absence of

Subsequent drinking episodes Reward reward not Subsequent drinking episodes expected but not received distinguishable

Subsequent drinking episodes Tolerance developed and Tolerance sustained in subsequent extinguished drinking episodes

Figure 2 Current research suggests that drinkers will display intoxicated or sober (i.e., alcohol-tolerant) behavior following alcohol consumption, depending on which type of behavior they believe will lead to a more favorable outcome (i.e., a reward or positive consequence). The actual outcome will then influence the drinker’s behavior following subsequent drinking episodes. observed directly when alcohol effects indicated by faster, improved task will strengthen tolerance. (For a more are also present, but such a response performance. Because they all had technical learning-theory explanation theoretically should affect a drinker’s received their doses of alcohol and the of the contribution of these expectan- behavior by shifting it in a direction placebo in the same environment, the cies to tolerance development, see opposite to that attributed to the effect expectation of receiving alcohol could Bennett 1992.) of alcohol. For example, if a person have contributed to the compensatory performs a task more slowly after response the study subjects displayed. drinking alcohol, a compensatory Nevertheless, the groups that had LEARNING TO COMPENSATE FOR response should speed performance. received the reward and had shown the ALCOHOL EFFECTS Some researchers testing for a highest tolerance to alcohol displayed compensatory response have surrepti- a much greater compensatory Why should rewarding sober behavior tiously substituted a placebo for alco- improvement in performance com- strengthen a compensatory response? hol after a series of drinking sessions pared with the control groups (see Evidence exists that a reward provides in which subjects who were rewarded figure 3 for results from one study). drinkers with an opportunity to learn a for sober performance displayed toler- These findings are consistent with the new behavioral strategy to compensate ance and unrewarded control subjects theory that tolerance stems from a for the impairing effect of alcohol. did not (e.g., Sdao-Jarvie and Vogel- compensatory response. The findings Research on motor-skill learning Sprott 1991). After drinking a placebo, also indicate that both the expectation indicates that rewarding efficient per- all the subjects exhibited a compen- of receiving alcohol and the expecta- formance provides feedback that helps satory response to some degree, as tion of a reward for sober performance the learner identify the change in be-

VOL. 21, NO. 2, 1997 165 transferable to a similar task performed for the first time after alco- 0 hol consumption. Research confirms such skill transfer in drinking situa- tions that reward sober performance.

yyy When drinkers display tolerance to repeated doses of alcohol on one task, their tolerance transfers to a second -2 yyy task (Rawana and Vogel-Sprott 1985). Researchers also have directly tested the idea that new learned be- yyy havior contributes to alcohol-tolerant performance. Drug-free performance of a psychomotor task can be im- yyy paired by environmental factors, such -4

Score Change (in seconds) as lowered lighting or reduced visibil- ity of a moving target. Studies have shown that behavioral tolerance to alcohol increases when subjects receive prior training to overcome impairment induced by such environmental factors -6 (Zinatelli and Vogel-Sprott 1993; Easdon and Vogel-Sprott 1996). In Favorable verbal feedback plus financial reward for sober performance these studies, all subjects first prac- Favorable verbal feedback for sober performance ticed a task while drug-free, although some subjects did so under environ- Imagined performing the task and receiving favorable verbal feedback and financial reward for sober performance mentally induced impairment. The results showed that subjects receiving Received financial reward but did not know it was associated with sober practice in overcoming environmental performance y impairment later displayed greatly Received nothing (i.e., no outcome) for sober performance enhanced behavioral tolerance to alcohol. Apparently, practicing the task under environmentally induced Figure 3 Compensation for alcohol effects. In four sessions, subjects performed or impairment provided these subjects mentally rehearsed an eye-hand coordination task after consuming alcohol; their completion times (i.e., scores) were recorded. In a fifth with an opportunity to learn a com- session, the subjects performed the task but received a placebo instead pensatory behavioral strategy that of alcohol. The subjects’ compensation for the anticipated effects of they could then transfer to improve alcohol was measured by comparing their average speed at finishing the their resistance to alcohol-induced task with a score recorded earlier, when they were sober. All subjects impairment. These effects are analo- performed faster after drinking the placebo, but those who were least gous to cross-tolerance between two impaired after drinking (i.e., had become most tolerant) displayed the strongest compensation, indicated by the greatest decrease in their drugs; in this case, however, the scores. Results imply that tolerance to alcohol involves learning to cross-tolerance occurs between envi- compensate for its impairing effects. ronmental and alcohol impairment. Other research suggests that a NOTE: Zero = Sober baseline score. technique commonly referred to as SOURCE: Adapted from Sdao-Jarvie, K., and Vogel-Sprott, M. Learning alcohol tolerance by mental or physical practice. Journal of Studies on Alcohol 53(6):533Ð540, 1992. p. 538. “mental rehearsal” can build tolerance as an alternative to actually perform- ing a task after drinking alcohol havior required to perform more skill- behavioral strategy intended to com- (Annear and Vogel-Sprott 1985; fully (e.g., Schmidt 1988). The learner pensate for impaired behavior and Sdao-Jarvie and Vogel-Sprott 1986; maintains the behavioral changes that maximize reward. Vogel-Sprott et al. 1984; Zinatelli and yield a reward and discards those that If a person displaying tolerance Vogel-Sprott 1990). This technique, do not. In a similar way, a drinker’s (i.e., no impairment) in one often applied to improve motor skills progressive development of behavioral psychomotor task has learned a new in sports, involves imagining task tolerance as alcohol doses are repeated behavior to overcome alcohol’s ef- performance before putting it into may reflect the gradual acquisition of a fects, the learned behavior should be practice. To investigate the effect on

166 ALCOHOL HEALTH & RESEARCH WORLD Is Behavioral ToleranceRunning Learned? Heads

tolerance of mental rehearsal under Taken together, the findings indi- the influence of alcohol, subjects in cate that behavioral tolerance is CONTRIBUTORS TO the cited studies drank repeated doses learned. This conclusion has impor- TOLERANCE DEVELOPMENT of alcohol and either mentally tant implications in a variety of areas, rehearsed a task with an imaginary including the following: Physiological compensation. reward for sober performance or actu- Alcohol consumption induces ally practiced the task and received a ¥ Understanding behavioral tolerance. physiological reactions to com- reward for sober performance. After The development of behavioral pensate for its effects; these the treatments concluded and all tolerance to alcohol can be attribut- reactions strengthen with re- groups performed the task after drink- ed to normal learning processes. peated alcohol exposure and ing alcohol, both the mental-rehearsal Like any learned response, behav- contribute to tolerance. and the task-practice groups displayed ioral tolerance is displayed when it is reliably associated with a favor- complete tolerance (i.e., no impair- Learned expectation of receiving able consequence. Three exposures ment) under the influence of alcohol. alcohol. Reliable repetition of to alcohol under such conditions are Thus, behavioral tolerance to alcohol specific events preceding alcohol sufficient to develop behavioral apparently can be acquired either by consumption leads to the expec- tolerance in social drinkers (Sdao- mentally rehearsing or actually prac- tation of receiving alcohol; these Jarvie and Vogel-Sprott 1991; ticing sober performance after drink- expectations result in compen- Sdao-Jarvie and Vogel-Sprott ing alcohol. satory reactions that contribute 1992). Therefore, the environmental to tolerance. outcome expected by the drinker appears to be a far more critical CONCLUSIONS AND IMPLICATIONS Learned expectation of behavioral determinant of behavioral tolerance consequence. Favorable events Comments and observations spanning than a large number of alcohol reliably associated with sober more than a century have suggested exposures is. The process of learn- behavior lead to expectations of that drinkers retain some volitional ing about the relationship between a positive consequence for such control over the behavioral effects of behavior and outcomes in a given behavior; these expectations result situation continually occurs, and the alcohol. Contemporary research now in the development of behavioral associations that will be learned can indicates that learned associations strategies to overcome alcohol- be changed by altering the environ- between alcohol-induced behavior and induced impairment and contrib- mental consequence of a response. its environmental consequences may ute to tolerance. In other words, behavioral tolerance account for this volitional control. The can be controlled by controlling its important factor promoting behavioral environmental consequence. tolerance in a drinking situation ap- ¥ Attributing responsibility for be- pears to be the association between ¥ Using behavioral tolerance to as- havior after drinking. The finding the display of sober behavior and a sess alcohol abuse. Many people that social drinkers will display favorable environmental outcome. assume that drinking must cease either tolerance or significant im- Studies reviewed in this article show for tolerance to subside. Although pairment, depending on which they that drinkers acquire and display be- this requirement may apply to expect to yield a more favorable havioral tolerance in drinking situa- alcohol tolerance observed at cellu- consequence, suggests that these tions in which they expect sober lar and neuronal levels (i.e., physi- expectations also may influence performance to yield a positive out- ological tolerance), a period of other types of activities performed come or reward. Subsequently with- abstinence is not needed to reduce under the influence of alcohol, holding the reward so that the behavioral tolerance. Despite con- including antisocial activities. This drinkers’ expectancy is not confirmed tinued alcohol use, tolerant perfor- idea has implications particularly will extinguish their tolerance, even mance displayed by drinkers will for alcohol abusers who engage in though they continue to drink alcohol. extinguish when they do not expect harmful, violent, or obnoxious In contrast, drinkers retain tolerance or do not receive a favorable out- behavior after drinking. Like social well in drinking situations that pro- come. This apparent adaptive flexi- drinkers, most alcohol abusers are vide no clues to indicate that a re- bility of a drinker’s behavioral not physically dependent on alco- warding outcome is removed. Thus, response to alcohol implies that hol, and the line between the two the endurance of behavioral tolerance clinical assessments of behavioral groups can be difficult to draw. appears to depend greatly on main- tolerance are unlikely to be a reli- Alcohol abusers who engage in taining the expectation of a favorable able diagnostic symptom of a problem behavior, however, con- outcome for sober performance after drinker’s alcohol abuse or risk of tribute to a long-standing societal drinking alcohol. physical dependence. problem commonly attributed to

VOL. 21, NO. 2, 1997 167 alcohol’s intoxicating chemical REFERENCES reinforcement. Drug and Alcohol Dependence action. For example, in 1991 a man 16(1):75Ð83, 1985. American Psychiatric Association. Diagnostic SCHMIDT, R.A. Motor Control and Learning: A was acquitted on a sexual assault and Statistical Manual of Mental Disorders, charge by a Quebec court judge on Fourth Edition. Washington, DC: the Behavioral Emphasis. 2d ed. Champaign, IL: Human Kinetics, 1988. the grounds that the defendant Association, 1994. SDAO-JARVIE, K., AND VOGEL-SPROTT, M. probably was too intoxicated to ANNEAR, W.C., AND VOGEL-SPROTT, M. Mental Mental rehearsal of a task before or after know what he was doing when he rehearsal and classical conditioning contribute to ethanol tolerance in humans. Psychopharma- ethanol: Tolerance facilitating effects. Drug and broke into a locked home and at- cology 87(1):90Ð93, 1985. Alcohol Dependence 18(1):23Ð30, 1986. tacked a 65-year-old, wheelchair- SDAO-JARVIE, K., AND VOGEL-SPROTT, M. BEIRNESS, D., AND VOGEL-SPROTT, M. Alcohol bound woman. Pardoning tolerance in social drinkers: Operant and Response expectancies affect the acquisition antisocial activities that occur classical conditioning effects. Psychopharma- and display of behavioral tolerance. Alcohol: An cology 84(3):393Ð397, 1984. International Biomedical Journal under the influence of alcohol on 8(6):491Ð498, 1991. the grounds that alcohol is respon- BENNETT, R.H. The role of learning in alcohol SDAO-JARVIE, K., AND VOGEL-SPROTT, M. sible may be counterproductive, tolerance in humans. In: Watson, R.R., ed. Drug and Alcohol Abuse Reviews: Volume 3. Alcohol Learning alcohol tolerance by mental or however, because it fosters the Abuse Treatment. Totowa, NJ: Humana Press, physical practice. Journal of Studies on Alcohol expectation of minimal or nonexis- 1992. pp. 233Ð255. 53(6):533Ð540, 1992. tent adverse consequences for SIEGEL, S. Pharmacological conditioning and BOLLES, R.C. Learning Theory. 2d ed. New antisocial behavior. The removal of York: Holt, Rinehart and Winston, 1979. drug effects. In: Goudie, A.J., and Emmett- Oglesby, M.W., eds. Psychoactive Drugs: a penalizing consequence makes DEWS, P.B. Psychopharmacology. In: Bachrach, Tolerance and Sensitization. Clifton, NJ: the expected outcome of such be- A.J., ed. Experimental Foundations of Clinical Humana Press, 1989. pp. 115Ð180. havior more favorable than it Psychology. New York: Basic Books, 1962. pp. 423Ð441. VOGEL-SPROTT, M. Acute recovery and would be otherwise. As a result, tolerance to low doses of alcohol: Differences the undesirable behavior is poten- EASDON, C.M., AND VOGEL-SPROTT, M. Drug- in cognitive and motor skill performance. free behavioral history affects social drinkers’ Psychopharmacology 61: 287Ð291, 1979. tially more likely to occur. tolerance to a challenge dose of alcohol. Journal of Studies on Alcohol 57(6):591Ð597, 1996. VOGEL-SPROTT, M. Alcohol Tolerance and In summary, the research reviewed Social Drinking: Learning the Consequences. GOLDBERG, L. Quantitative studies of alcohol New York: Guilford Press, 1992. in this article suggests that the conse- tolerance in man. The influence of ethyl alcohol quences drinkers learn to expect will on sensory, motor, and psychological functions VOGEL-SPROTT, M.; RAWANA, E.; AND WEBSTER, R. Mental rehearsal of a task under ethanol influence their behavior. Theoretically, referred to blood alcohol in normal and habituated individuals. Acta Physiologica facilitates tolerance. Pharmacology Biochemistry then, alcohol-related antisocial behav- Scandinavica 5(Suppl. 16):1Ð128, 1943. and Behavior 21(3):329Ð331, 1984. ior might be reduced by cultivating ZACK, M., AND VOGEL-SPROTT, M. Response GOLDBERG, L., AND HAVARD, J.D.J. Research on the expectation that socially accept- the Effects of Alcohol and Drugs on Driver outcomes affect the retention of behavioral able drinking behavior yields a more Behavior and Their Importance as a Cause of tolerance to alcohol: Information and incentive. favorable outcome. This may be Road Accidents. Paris: Organization for Psychopharmacology 113(2):269Ð273, 1993. Economic Cooperation and Development, 1968. achieved by programs and policies ZACK, M., AND VOGEL-SPROTT, M. Behavioral tolerance and sensitization to alcohol in that advocate responsible drinking HUNT, W.A. Neuroscience research: How has it contributed to our understanding of alcohol humans: The contribution of learning. Experi- behavior and consistently penalize abuse and alcoholism? A review. Alcoholism: mental and Clinical Psychopharmacology antisocial behavior under the influ- Clinical and Experimental Research 3(4):396Ð401, 1995. ence of alcohol. ■ 17(5):1055Ð1065, 1993. ZACK, M., AND VOGEL-SPROTT, M. Drunk or sober? Learned conformity to behavioral MACANDREW, C., AND EDGERTON, R.B. Drunken Comportment: A Social Explanation. standards. Journal of Studies on Alcohol, in press. ACKNOWLEDGMENTS Chicago: Aldine Publishing, 1969. ZINATELLI, M., AND VOGEL-SPROTT, M. Learned MACNISH, R. The Anatomy of Drunkenness. 4th tolerance to alcohol: Mental rehearsal with Research support was provided by the ed. Glasgow: W.R. McPhun Press, 1832. p. 45. imagined consequences. Alcoholism: Clinical and Natural Sciences and Engineering Experimental Research 14(4):518Ð521, 1990. MANN, R.E., AND VOGEL-SPROTT, M. Control of Research Council of Canada. The alcohol tolerance by reinforcement in nonalcoholics. ZINATELLI, M., AND VOGEL-SPROTT, M. author thanks Dr. Mark Fillmore and Psychopharmacology 75:315Ð320, 1981. Behavioral tolerance to alcohol in humans is Professor David A. Sprott for providing enhanced by prior drug-free treatment. RAWANA, E., AND VOGEL-SPROTT, M. The Experimental and Clinical Psychopharma- suggestions on the manuscript. transfer of alcohol tolerance, and its relation to cology 1(1Ð4):194Ð199, 1993.

168 ALCOHOL HEALTH & RESEARCH WORLD RESEARCH UPDATE

THE NEUROBIOLOGY OF TOLERANCE, DEPENDENCE, AND REINFORCEMENT ALCOHOLISM IN GENETICALLY Historically, animal models of alcoholism have been used most extensively to study alcohol tolerance and physical SELECTED RAT MODELS dependence (see Kalant et al. 1971 for a seminal review in this field; see Kalant 1993 and Hoffman and Tabakoff Robert B. Stewart, Ph.D., and Ting-Kai Li, M.D. 1996 for recent reviews on the mechanisms of tolerance and dependence). Tolerance to alcohol occurs when, fol- lowing chronic consumption, higher doses of alcohol must Rats selectively bred for their tendency to drink be ingested to achieve a given effect. Consequently, re- large or small quantities of alcohol are a useful searchers believe that tolerance accounts for increases in model for investigators examining the possible the amount of alcohol consumed over time. Physical de- neurobiological processes underlying alcoholism. pendence is indicated by signs of withdrawal resulting Studies with the alcohol-preferring (P) and alcohol- from the absence of alcohol in the body when drinking is nonpreferring (NP) and the high-alcohol-drinking discontinued. Because alcohol withdrawal symptoms— (HAD) and low-alcohol-drinking (LAD) pairs of rat which range from anxiety, tremors, hypothermia, and sleep disturbances to hallucinations and seizures—are unpleasant lines developed at Indiana University have (i.e., aversive), researchers hypothesize that physically illustrated differences in several behavioral and dependent people drink to avoid or alleviate these symp- neurobiological characteristics associated with toms. In fact, the development of tolerance and physical alcohol consumption. Specifically, compared with dependence are considered hallmarks of alcoholism. alcohol-avoiding rats, rats with an affinity for These two processes, however, cannot account for the alcohol have a greater sensitivity to the stimulatory initiation of alcohol drinking or explain why relapse occurs effects of low to moderate doses and a reduced in abstinent alcoholics long after the signs of physical sensitivity to the negative effects of high doses. dependence have disappeared. Thus, researchers continue Rats that voluntarily drink large quantities of to investigate additional behavioral and neurobiological factors that may underlie alcohol use. Recent studies have alcohol also acquire tolerance to alcohol’s focused on a process called reinforcement. In behavioral aversive effects. In addition, these rats differ from psychology, reinforcement refers to the connection be- their alcohol-avoiding counterparts in the levels of tween a behavior and a stimulus whereby the chance of several chemical mediators (i.e., neurotrans- repeated behavior (e.g., alcohol-seeking) is enhanced if the mitters) found in the brain, including serotonin, behavior results in obtaining a reinforcing stimulus (e.g., dopamine, gamma-aminobutyric acid (GABA), and the desirable effects of drinking an alcoholic beverage). the endogenous opioids. KEY WORDS: animal strains; The biological basis of alcohol and other drug reinforce- selective breeding; AOD preference; amount of AOD ment appears to involve the interaction of these substances with specific systems in the brain that regulate “natural” use; AOD tolerance; neurotransmitters; reinforcement; reinforcing and motivated activities such as eating, drink- drug therapy; literature review ing, and sex (Wise 1980; Koob and Bloom 1988).

nimal models have been critical to many areas of SELECTIVE BREEDING PROGRAMS research, including the investigation of the behavioral Aand neurobiological processes that may underlie Animal studies of the relationship of reinforcement process- alcohol abuse and alcoholism. The use of animals, rather than es to alcoholism initially were hampered by the fact that humans, in research has two advantages: (1) animal models most laboratory animals, such as rats and mice, will not allow a high degree of experimental control not possible with voluntarily consume alcohol in quantities sufficient to pro- human subjects (i.e., scientists can focus solely on alcohol’s duce significant pharmacological effects (Cicero 1979). To effects without the interference of confounding factors that overcome this problem, researchers have found numerous may accompany alcoholism in humans, such as liver damage, environmental manipulations that increase the rates of alco- poor nutrition, or psychiatric disturbances) and (2) animal models permit the use of invasive procedures. This article ROBERT B. STEWART, PH.D., is an assistant scientist in the describes the findings of studies on rats that have been spe- Department of Psychology, Purdue School of Science, cially bred for their tendencies to drink either large or small Indiana University/Purdue University at Indianapolis, quantities of alcohol. In particular, the article focuses on Indiana. characteristics associated with high and low levels of alcohol TING-KAI LI, M.D., is a distinguished professor of medicine drinking that have been investigated in the specially bred and biochemistry at the Indiana University School of lines of rats developed at Indiana University. Medicine, Indiana University, Indianapolis, Indiana.

VOL. 21, NO. 2, 1997 169 RESEARCH UPDATE hol self-administration in laboratory animals (Meisch 1984; Much variability exists in the amount of alcohol consumed Samson et al. 1988). For example, the feeding-induced by individual rats, however. A small percentage of rats drinking procedure (Meisch 1976) involves feeding animals within a given population will drink relatively large such as rats or mice all or part of their daily ration of dry amounts of alcohol, and a small percentage will drink pelleted chow either during or immediately before daily relatively little. Selective breeding capitalizes on this varia- drinking sessions. This feeding results in thirst, and the tion in preference for alcohol over water and has resulted animals subsequently drink considerable amounts of the in the development of lines of rats that will consistently fluid made available to them. Initially, the fluid is only self-administer large or small quantities of alcohol when water, but alcohol solutions are then presented in gradually given continuous access to two bottles, one containing a increasing concentrations over several sessions. Finally, the 10-percent alcohol solution and the other containing water food is no longer presented during the drinking sessions, yet alone. Rats bred for their high affinity for alcohol typically the intake of the alcohol solution remains elevated. consume more than 5 grams of alcohol per kilogram (g/kg) A second environmental manipulation is called the of body weight per day, whereas rats bred for a low affinity sucrose-fading procedure (Samson 1986). In this proce- for alcohol typically ingest less than 1 g/kg per day. dure, animals are first trained to press a lever to access a Several pairs of rat lines have been produced through ge- sweet sucrose solution containing no alcohol. Over the netic selection for alcohol preference/aversion, including course of several daily drinking sessions, the sucrose con- the University of Chile UChA/UChB lines (Mardones and centration is gradually reduced while alcohol is added at Segovia-Riquelme 1983), the Finnish Alko alcohol-prefer- increasingly higher concentrations. Finally, the fluid con- ring and alcohol-avoiding (AA/ANA) lines (Eriksson sists of an alcohol solution with no sucrose, and high alco- 1968), the Sardinian sP/sNP lines (Fadda et al. 1989), and hol intake is maintained. the Indiana University alcohol-preferring and -nonprefer- In addition to such environmental manipulations, genet- ring (P/NP) lines and high- and low-alcohol-drinking ic manipulation also has been an effective approach to (HAD/LAD) lines (Lumeng et al. 1995). animal studies of alcohol reinforcement, particularly the An important criterion for the scientific usefulness of use of selective breeding programs.1 This approach springs these genetically selected rat lines is maintenance of the directly from the first experiments on rodent alcohol con- preference for (or aversion to) alcohol through successive sumption. The oldest and most straightforward method for generations. Figure 1 indicates that in the P/NP rat lines, measuring voluntary alcohol self-administration in rats is daily alcohol consumption has been relatively stable since to offer a continuous choice between an alcohol solution the eighth generation of selective breeding. The P rats, for and water (Richter and Campbell 1940). Although widely example, consistently drink more than 5 g/kg per day, result- used, this so-called two-bottle preference method has been ing in blood alcohol concentrations (BAC’s) of up to 0.2 severely criticized, because the average (i.e., mean) daily percent and the development of alcohol tolerance and physi- cal dependence following periods of chronic alcohol self- dose of alcohol consumed by groups of “normal,” or stock, administration (Li et al. 1988; Li and McBride 1995; laboratory rats is not high enough to produce significant Lumeng et al. 1995). Selectively bred rats, such as the P rats, levels of alcohol in the blood or brain. In other words, satisfy Cicero’s (1979) rigorous criteria for an animal model stock rats can metabolize alcohol (i.e., break it down and of alcoholism (see box, p. 173). In addition, rats selectively eliminate it from the body) faster than they consume it. If bred for alcohol preference add experimental evidence to the the rate of alcohol consumption does not exceed the rate of importance of genetic factors in determining the risk for alcohol elimination, then the amount of alcohol in the alcoholism in humans (Cloninger 1987) and provide an blood and brain can never achieve significant levels. Thus, opportunity to determine whether a genetic basis exists for it is not surprising that stock rats do not display tolerance, the association between high alcohol consumption and other physical dependence, or overt intoxication with such low behavioral and neurobiological characteristics. levels of alcohol intake. More important, researchers can- not determine whether the rats consume alcohol for its pharmacologic effects on the central nervous system CHARACTERISTICS ASSOCIATED WITH HIGH (CNS) or for other reasons, such as to alleviate hunger or AND LOW ALCOHOL DRINKING thirst or simply for its taste or smell. The low mean alcohol intake by stock rats reflects the The following survey of research findings emphasizes fact that most rats within a given population avoid alcohol. studies of the original pair of rat lines selectively bred at Indiana University, the P/NP rats, as well as a second pair 1In selective breeding programs, animals from a genetically heteroge- of rat lines, the HAD/LAD rat lines, which were developed neous population with a certain desired characteristic (e.g., a tendency to to replicate and confirm the research findings obtained drink alcohol) are mated and their offspring are screened for that charac- with the P/NP lines. Unless otherwise noted, the research teristic. Those offspring bearing the desired characteristic are then bred. If the characteristic is a heritable trait, the process of selection is repeated in described was carried out by Li and colleagues at Indiana successive generations to produce lines of animals that “breed true” for University. For more specific references, extensive reviews the characteristic. and bibliographies on these and other rat lines are provided

170 ALCOHOL HEALTH & RESEARCH WORLD RESEARCH UPDATE by Li and colleagues (1988, 1993), Li and McBride (1995), and Lumeng and colleagues (1995). Males Females 10 Alcohol-Related Traits 8 The amount of alcohol that an animal consumes is con- trolled in part by two competing factors: (1) the reinforcing 6 effects that encourage intake (e.g., euphoria or, conversely, 4 the alleviation of dysphoria or negative emotional states, such as anxiety) and (2) the aversive effects that limit in- 2 (g/kg body weight) take (e.g., unpleasant taste, motor-skills impairment, or Daily Alcohol Intake negative physical reactions, such as dizziness or vomiting). 0 Whether alcohol is reinforcing or aversive depends to some S-8 S-16 S-20 S-31 S-8 S-16 S-20 S-31 Generation of Selective Breeding extent on the amount of alcohol intake. Low doses of alco- hol generally are reinforcing, but high doses tend to be P rats NP rats aversive. Numerous studies have been conducted to char- Figure 1 Mean free-choice alcohol consumption (grams acterize the reinforcing and aversive effects of alcohol in of alcohol per kilogram [g/kg] of body weight the P/NP and HAD/LAD rat lines (see Lumeng et al. 1995 per day) by recent generations of selectively for a review). bred alcohol-preferring (P) and -nonpreferring Although differences in the rats’ alcohol intake in the (NP) rat lines. Males and females of both lines show similar rates of alcohol consumption. two-bottle preference test suggest that alcohol is more reinforcing for P and HAD than for NP and LAD rats, recent studies provide additional evidence for this conclu- sion. Rats can be trained to press a lever or perform some oral consumption as factors that may explain the line dif- ferences in the reinforcing effects of alcohol. other work to obtain alcohol (see figure 2), a paradigm 2 known as operant responding, and the alcohol “reward” Increases in spontaneous motor activity and other signs received as a result of the rats’ correct action can increase of behavioral arousal following drug administration are strongly associated with the reinforcing effects of many the frequency of the operant response (i.e., it is reinforc- drug classes, including stimulants, opiates, and sedative- ing). In both P and HAD rats, operant responding is main- hypnotics. Interestingly, low doses of alcohol also produce tained by the delivery of alcohol over a wide range of increases in spontaneous motor activity shortly after alco- experimental conditions. For example, the period of daily hol administration in P and HAD rats, but not in NP and alcohol availability may be limited to, say, 30 minutes LAD rats. In addition, studies using techniques such as (Schwarz-Stevens et al. 1991) or it may be continuous electroencephalography (EEG), a method of measuring (Files et al. 1993). Several methods of initiating the oper- brain electrical activity, also provide evidence that alcohol ant response for alcohol have been investigated, including produces more arousal in P than in NP rats. the feeding-induced drinking and sucrose-fading proce- Because alcohol usually is consumed orally, it is of dures (Schwarz-Stevens et al. 1991; Ritz et al. 1994). interest to determine whether P and HAD rats differ from Alcohol concentrations as low as 1 percent and as high as NP and LAD rats in their reactions to alcohol’s flavor (i.e., 40 percent maintain responding in P and HAD rats. In taste reactivity). A variance in avidity for alcohol may contrast, responding by NP and LAD rats under the same result from different preferences for the taste of alcohol conditions either is not maintained or is much lower than solutions, for example. To test the taste reactivity of the responding by their alcohol-preferring counterparts. selectively bred rats, Kiefer and coworkers (Bice and Alcohol also functions as a reinforcer by nonoral routes Kiefer 1990; Kiefer et al. 1995) placed drops of alcohol of administration for P, but not NP, rats. For example, P solution into the mouths of rats who had never previously rats will learn the correct operant response that results in consumed alcohol (i.e., alcohol-naive rats) and noted their alcohol administration through a tube directly into the facial responses, which were then quantified to measure stomach (i.e., intragastric administration). Conversely, how much the rats liked or disliked the flavor of the solu- stock rats will engage in intragastric alcohol self-adminis- tion. The investigators did not find any differences in taste tration only after developing physical dependence during a reactivity between the alcohol-preferring (P and HAD) and period of forced alcohol infusion. P rats, but not NP rats, -nonpreferring (NP and LAD) rat lines on initial exposure. also will learn to perform the correct response that admin- Next, the researchers gave the rats a two-bottle preference isters alcohol through a tube directly into the ventral test with alcohol solution and water for 3 weeks, and pre- tegmental area of the brain, a region implicated in rein- forcement by alcohol and other drugs. By bypassing the 2“Spontaneous motor activity” describes the measured amount of move- mouth, these methods of alcohol consumption eliminate all ment and exploration that takes place when rats are placed in an enclosure of the behaviors and stimuli (e.g., taste) associated with consisting of an open area or “field” surrounded by four walls.

VOL. 21, NO. 2, 1997 171 RESEARCH UPDATE

lowing chronic drinking. For example, in one study, P rats were given continuous access to an alcohol solution and water for 32 days. During this period, the rats increased their alcohol consumption by about 50 percent (see figure 3), indicating the development of tolerance. Following the period of oral alcohol self-administration, these rats, along with alcohol-naive P rats serving as control subjects, un- derwent conditioned taste aversion trials in which they drank a sweetened fluid they had never previously tasted, then immediately received an injection of alcohol. The injected doses were sufficiently high to produce condi- tioned aversive effects, as indicated by the rats’ avoidance of the sweetened fluid upon subsequent exposure. The alcohol-exposed P rats, however, exhibited an attenuated Figure 2 A female alcohol-preferring (P) rat presses a lever in an operant chamber. Each time the lever conditioned taste aversion relative to the alcohol-naive P is pressed, 0.1 milliliter of alcohol solution is control rats. Thus, P rats developed tolerance to alcohol’s delivered into a well to the right of the lever. P aversive CNS effects just as they had developed a toler- rats work to obtain alcohol solutions at concen- ance for alcohol’s flavor in the taste reactivity tests. This trations as high as 40 percent (the typical tolerance to aversive CNS effects could contribute, at least concentration of unmixed hard liquors, such as in part, to the rats’ high alcohol intake. straight whiskeys). In addition to the rats’ acquired reduction in sensitivity to the CNS effects of alcohol (termed “neuronal” or “func- Photograph by Maggie Johann Stewart tional” tolerance), prolonged periods of alcohol self-ad- ministration also increase the rate of alcohol metabolism in the liver (termed “metabolic” tolerance) in P rats. Rats of dictably, the P and HAD rats consistently drank more alcohol than did the NP and LAD rats. Following the two- the NP line do not self-administer sufficient quantities of bottle preference test, a second taste-reactivity test was alcohol to develop metabolic tolerance. A comparison of given, and results indicated that alcohol had become more alcohol-naive P and NP rats, however, found no differ- palatable to the P and HAD rats (although not to the NP ences in the rates at which alcohol is metabolized in the and LAD rats) during the period of oral alcohol consump- liver and eliminated from the body. Consequently, when tion. The alcohol-preferring rats maintained this increase in the same amount of alcohol is administered to P and NP alcohol palatability even after 1 month of alcohol absti- rats, both lines achieve the same BAC levels. Thus, the nence. Thus, preference for the taste of alcohol is not an divergent drinking levels and reactions to alcohol seen in P inherited characteristic in the alcohol-preferring rat lines; and NP rats apparently are not attributable to differences in instead, this taste preference is acquired through alcohol- alcohol metabolism or elimination, but to differences in drinking experience. neuronal sensitivity to alcohol. Some of the CNS effects of alcohol, especially at high A series of studies has shown that P rats can develop doses, are aversive or dysphoric. Rats selectively bred for acute tolerance to a single sedative-hypnotic dose of alco- high and low alcohol preference have been tested for their hol more quickly and/or to a greater extent than NP rats. sensitivity to these negative effects. Among the most use- That is, P rats recover more quickly than NP rats on a num- ful testing methods is the conditioned taste aversion proce- ber of tests measuring the depressant effects of alcohol, dure, wherein an animal receives a large, presumably including motor impairment, lowered body temperature, aversive, dose of alcohol by injection at approximately the and regain of righting reflex (i.e., sleep time). Using a same time that it receives a particular food or other taste behavioral measure of alcohol-induced motor impairment, stimulus. In normal rats, pairing a taste stimulus and an researchers observed that the tolerance P rats develop to a aversive dose of alcohol will cause the rat to avoid that single dose of alcohol can persist for as long as 10 days, taste in the future. Similarly, the conditioned place aver- whereas such tolerance in NP rats, which is weaker in the sion procedure creates aversion to a particular location by first place, dissipates within 3 days. Differences in initial placing the rat there as it experiences the aversive effects sensitivity and acute tolerance also have been found in of an alcohol injection. In both types of studies, P rats are other alcohol-preferring and -nonpreferring rodent lines less sensitive than NP rats to alcohol’s aversive effects. and strains, such as the AA/ANA rat lines (Le and Furthermore, P rats with histories of oral alcohol con- Kiianmaa 1988) and the C57BL/DBA mouse strains sumption, compared with alcohol-naive rats, experience (Tabakoff and Ritzmann 1979). This finding indicates a less motor-impairing and aversive effects from high doses strong association between tolerance and high voluntary of alcohol, suggesting that they developed tolerance fol- alcohol consumption.

172 ALCOHOL HEALTH & RESEARCH WORLD RESEARCH UPDATE

Traits Not Directly Related to Alcohol In addition to alcohol-related traits, researchers are interest- CRITERIA FOR AN ANIMAL ed in detecting other behavioral and biological differences MODEL OF HUMAN ALCOHOLISM between rats genetically selected for high or low alcohol preference, with the hope that such differences may provide The criteria for an animal model of human further clues to the genesis and perpetuation of alcoholism. alcoholism are as follows: Thus far, however, relatively few studies of this type have been conducted. ¥ Given a choice between an alcohol solution Among existing studies, results indicate that P rats ex- and another solution (such as water), the ani- hibit higher spontaneous motor activity than do NP rats mal must voluntarily consume alcohol in when placed in a new environment, but no difference ap- an amount sufficient to produce meaningful pears between the lines when the environment is no longer blood alcohol concentrations (BAC’s). The novel. These observations concur with the high novelty- animal should drink alcohol solely for its seeking personality characteristic that is noted in certain pharmacological effects, not for its caloric types of human alcoholics (Cloninger 1987). P rats also value or its taste or smell. seem to be more anxious than NP rats on a number of be- havioral tests of anxiety, which accords with the notion that ¥ Following a period of chronic alcohol con- alcohol may be self-administered to relieve tension and sumption, the animal must develop tolerance, anxiety. as demonstrated by a reduction in the effects In addition, P rats exhibit a higher preference than NP of the same dose of alcohol and the same BAC. rats for oral consumption of highly palatable, nondrug solu- tions, such as sucrose or saccharin, but intake of plain water ¥ Following a period of chronic alcohol con- and of sour and bitter-flavored solutions does not differ sumption, the animal must develop alcohol between P and NP rats. Because a preference for sweets dependence, as demonstrated by behavioral highly correlates with high alcohol intake in numerous and biological responses characteristic of rodent lines and strains, investigators have suggested that acute alcohol withdrawal and confirmation of the same, or overlapping, brain mechanisms may be in- alcohol’s ability to act as a reinforcer (i.e., its volved in the reinforcement mediated by some drugs (e.g., ability to increase the chance that alcohol- alcohol) and other palatable substances (e.g., chocolate). seeking behavior will occur).

SOURCE: Adapted from Cicero, T.J. A critique of animal analogues of alcoholism. NEUROBIOLOGICAL DIFFERENCES ASSOCIATED WITH In: Majchrowicz, E., and Noble, E.P., eds. Biochemistry and Pharmacology of HIGH AND LOW ALCOHOL DRINKING Ethanol. Vol. 2. New York: Plenum Press, 1979. pp. 534Ð535. Biological studies of the P/NP and HAD/LAD rat lines have focused on the study of chemical mediators known as neuro- neurotransmitter systems in alcohol reinforcement. Endo- transmitters3 (i.e., neurochemistry) and the identification and genous opioids also act as inhibitory neurotransmitters and mapping of groups of neurons that seem to have similar func- are released in response to stresses such as injury, childbirth, tions (i.e., neuroanatomy). Neurochemical studies implicate a and vigorous exercise. In addition, opioids play a role in subset of neurotransmitters with special roles in the control of eating and drinking behaviors and, like GABA, appear to alcohol-seeking behavior: 5-hydroxytryptamine (5-HT), also interact with dopamine and other neurotransmitter systems called serotonin; dopamine; gamma-aminobutyric acid involved in alcohol reinforcement. (GABA); and the body’s own opiatelike substances, the en- Several brain structures appear to participate in a postu- dogenous opioids (i.e., endorphins). lated brain “reward pathway,” including the ventral tegmen- The 5-HT system appears to be a key component in the tal area, raphe nuclei, lateral hypothalamus, olfactory regulation of food consumption and mood as well as the tubercle, nucleus accumbens, and medial prefrontal cortex development of alcohol tolerance. In addition, 5-HT modu- and other limbic areas (see glossary, pp. 177Ð179). The lates the release of dopamine, thereby directly affecting the function of this neural pathway system is to regulate behav- dopamine system. In turn, the dopamine system plays a iors motivated by “natural” rewards such as food, water, major role in motor activity, drug reinforcement, and the and sex. Scientists believe, however, that alcohol and other motivation to engage in several other behaviors that may be drugs of abuse (e.g., cocaine and morphine) function as considered reinforcing or rewarding, such as eating and sex. reinforcers by imitating, facilitating, or sometimes blocking GABA differs from 5-HT and dopamine in that it is not the various neurotransmitters involved in this system. Based confined to certain neurons and pathways forming a system. Rather, it is found throughout the brain, conveying inhibito- 3For definition of this and other technical terms used in this article, see ry signals and perhaps interacting with dopamine and other central glossary, pp. 177Ð179

VOL. 21, NO. 2, 1997 173 RESEARCH UPDATE

An abnormality in one of the major components of the brain reward circuitry—the ventral tegmental area-nucleus 35 accumbens dopamine system—also has been associated with high alcohol preference. Specifically, scientists have 30 observed low levels of dopamine and chemicals associated 25 with its breakdown in the nucleus accumbens and anterior striatum of P and HAD rats in the absence of alcohol. This 20 finding is of interest, because abused drugs from many pharmacological classes (including stimulants, opiates, and 15 sedatives), as well as low doses of alcohol, stimulate the Liquid Intake release of dopamine in the nucleus accumbens. The P rats 10

(milliliters per 24 hours) may be particularly sensitive to this alcohol-induced 5 dopamine release in the nucleus accumbens. That is, alco- hol consumption by P rats may be, in a sense, a regulatory 0 action aimed at increasing the rats’ abnormally low levels 4 8 12 16 20 24 28 32 of dopamine in the accumbens. Because the 5-HT system Days of Alcohol Access plays a role in regulating the dopamine system in the brain 10% alcohol solution Water reward pathway, the decreased 5-HT innervation noted in alcohol-preferring rat lines also may affect the function of Figure 3 Volume of water (open circles) and concurrently the dopamine system in these rats. available 10-percent alcohol solution (closed circles) consumed by 18 alcohol-preferring rats In addition to stimulating dopamine release in the nucle- during 32 days of chronic alcohol drinking. Data us accumbens, alcohol’s actions on neuronal activity also shown are averages for consecutive 2-day stimulate GABA receptors. Noting that anxiety-reducing ¨ periods. The increase in alcohol intake over drugs such as the benzodiazepines (e.g., Valium ) produce successive days in these rats is consistent with their effects by facilitating nerve signal transmission at the development of tolerance. synapses using the neurotransmitter GABA, researchers theorize that alcohol may produce its rewarding and anxi- ety-reducing effects via GABA neurons as well. Interest- on neuropharmacological studies and on studies in which ingly, studies in the P/NP and HAD/LAD rat lines have rats learn an operant response to electrically self-stimulate demonstrated a higher density of axon terminals containing their brain “reward” areas, the neurotransmitters 5-HT, GABA in the accumbens of the rats with high alcohol dopamine, GABA, and the endogenous opioids all have preference. This suggests a potential for increased GABA been implicated in the circuitries of the brain reward path- activity in P and HAD rats in an area of the brain involved way (Wise 1980; Koob and Bloom 1988). in alcohol reinforcement. One of the most consistent neurochemical and neuro- The endogenous opioid systems also are involved in the anatomical findings observed in P/NP and HAD/LAD rats regulation of alcohol drinking, as evidenced by the ability is a deficiency of 5-HT in the alcohol-preferring lines of opiate drugs to alter alcohol consumption. The endoge- (McBride et al. 1991; Li and McBride 1995). Compared nous opioid systems have been studied in P/NP rats with rats that drink little alcohol, the levels of 5-HT in rats (Froehlich and Li 1993) and in the Finnish AA/ANA lines that drink large amounts of alcohol are significantly re- (Nylander et al. 1994). The high- and low-alcohol-drinking duced in several brain regions, including the frontal cortex, rat lines differ in opioid activity in the absence of alcohol hippocampus, corpus striatum, thalamus, hypothalamus, as well as in alcohol-stimulated opioid activity in the nu- pons-medulla, and nucleus accumbens. These regions are cleus accumbens and pituitary gland (Froehlich and Li involved either in the brain reward pathway or in neural 1993; Nylander et al. 1994). However, the brain reward processes that are relevant to alcohol-seeking behavior pathway of the Finnish AA rats does not appear to have (e.g., learning, memory, and tolerance development pro- low levels of dopamine or 5-HT (Korpi et al. 1988) as is cesses). Closer examination of the neurons in some of these the case with the selectively bred P and HAD rats and brain regions (such as the frontal cortex, nucleus accum- other rodent strains that consume large amounts of alcohol bens, and hippocampus) suggests that these differences may (see Li and McBride 1995 for a review). be caused by a relative scarcity of 5-HTÐcontaining axons. The apparently discordant findings from the comparison Interestingly, research has shown that the decrease in 5- of P/NP and AA/ANA rat pairs may result from variances HTÐcontaining axons in P rats results in compensatory up- in the foundation stocks from which the two pairs of genet- regulation of 5-HT receptor activity (McBride et al. 1991; ically selected lines were derived. If so, selective-breeding Li and McBride 1995). That is, the number or sensitivity of experiments may achieve the same endpoints (e.g., high the target 5-HT receptors apparently increases to make up and low alcohol preference) by altering different brain for the reduction in the availability of 5-HT. mechanisms. Such findings may reflect the multiplicity of

174 ALCOHOL HEALTH & RESEARCH WORLD RESEARCH UPDATE mechanisms that may contribute to high alcohol intake. cal manipulations decrease alcohol’s reinforcing effects. The Indeed, the challenge to alcohol researchers and clinicians fact that drugs such as fluoxetine and naltrexone reduce is that alcoholism results from the interaction of many alcohol intake in both rodents and humans supports the biological factors (inherited and environmental) and is not predictive validity of the use of genetic animal models for a unitary phenomenon. evaluating therapies that can potentially reduce or prevent excessive alcohol consumption. ■

SIGNIFICANCE REFERENCES Studies with rats selectively bred for alcohol preference or nonpreference support several hypotheses on factors that ALTSHULER, H.L.; PHILLIPS, P.E.; AND FEINHANDLER, D.A. Alteration of may be associated with alcoholism in humans. For exam- ethanol self-administration by naltrexone. Life Sciences 26:679Ð688, 1980. ple, the demonstrated sensitivity of alcohol-preferring rats BICE, P.J., AND KIEFER, S.W. Taste reactivity in alcohol preferring and to the stimulatory effects of low to moderate alcohol doses nonpreferring rats. Alcoholism: Clinical and Experimental Research is in agreement with the contention that these effects are 14(5):721Ð727, 1990. important in the initiation and maintenance of alcohol CICERO, T.J. A critique of animal analogues of alcoholism. In: drinking. The alcohol-preferring rats also show an innate Majchrowicz, E., and Noble, E.P., eds. Biochemistry and Pharmacology insensitivity to the aversive effects of alcohol at high dos- of Ethanol. Vol. 2. New York: Plenum Press, 1979. pp. 533Ð560. es, which may tend to limit the amount of alcohol con- CLONINGER, C.R. Neurogenetic adaptive mechanisms in alcoholism. sumed by rats that are normal or genetically selected for Science 236(4800):410Ð416, 1987. low alcohol preference. Tolerance development to these ERIKSSON, K. Genetic selection for voluntary alcohol consumption in the aversive effects, which occurs to a greater extent in the albino rat. Science 159(3816):739Ð741, 1968. alcohol-preferring rats, also may encourage increased alco- FADDA, F.; MOSCA, E.; COLOMBO, G.; AND GESSA, G.L. Effect of hol intake. Furthermore, behavioral research with selec- spontaneous ingestion of ethanol on brain dopamine metabolism. Life tively bred rat lines indicates that individual differences in Sciences 44(4):281Ð287, 1989. responsiveness to alcohol can be heritable traits and that FILES, F.J.; ANDREWS, C.M.; LEWIS, R.S.; AND SAMSON, H.H. Effects of these animal models are valuable tools for investigating ethanol concentration and fixed-ratio requirement on ethanol self- neural mechanisms relevant to alcohol-seeking behavior. administration by P rats in a continuous access situation. Alcoholism: These animal models also have provided the underpin- Clinical and Experimental Research 17(1):61Ð68, 1993. ning for a new direction in the treatment of alcoholism and FROEHLICH, J.C., AND LI, T.K. Opioid peptides. In: Galanter, M., ed. alcohol abuse: testing pharmacotherapies related to the pos- Recent Developments in Alcoholism, Vol. 11: Ten Years of Progress. New tulated neural reward mechanisms. The 5-HT and opioid York: Plenum Press, 1993. pp. 187Ð205. systems have long been implicated as having a role in alco- HOFFMAN, P.L., AND TABAKOFF, B. Alcohol dependence: A commentary hol drinking (Myers and Melchior 1977; Altshuler et al. on mechanisms. Alcohol and Alcoholism 31(4):333Ð340, 1996.

1980), and basic research findings, such as the line differ- KALANT, H. Problems in the search for mechanisms of tolerance. In: ences in the brain reward systems of selectively bred rats, Taberner, P.V., and Badawy, A.A., eds. Advances in Biomedical Alcohol have provided a theoretical basis for clinical studies on the Research: Proceedings of the 6th ISBRA Congress. New York: Pergamon effects of drugs that may influence activity in these systems. Press, 1993. pp. 1Ð8. Various agents that alter 5-HT, dopamine, GABA, and opi- KALANT, H.; LEBLANC, A.E.; AND GIBBINS, R.J. Tolerance to, and oid functioning decrease alcohol consumption in animal dependence on, some non-opiate psychotropic drugs. Pharmacological models, including selectively bred alcohol-preferring rats Reviews 23(3):135Ð191, 1971. (see Lumeng et al. 1995 for a review). For example, fluoxe- KIEFER, S.W.; BADIA-ELDER, N.; AND BICE, P.J. Taste reactivity in high tine (Prozac¨) is an antidepressant drug that inhibits the alcohol drinking and low alcohol drinking rats. Alcoholism: Clinical and reuptake of 5-HT by the neurons that secrete it and thereby Experimental Research 19(2):279Ð284, 1995. facilitates 5-HT activity. Fluoxetine has been found to sig- KOOB, G.F., AND BLOOM, F.E. Cellular and molecular mechanisms of drug nificantly reduce alcohol intake in populations of heavy dependence. Science 242(4879):715Ð723, 1988. drinkers (Naranjo et al. 1986), but clinical trials to date have KORPI, E.R.; SINCLAIR, J.D.; KAHEINEN, P.; VIITAMAA, T.; HELLEVUO, K.; not shown fluoxetine to be effective in treating alcoholism AND KIIANMAA, K. Brain regional and adrenal monoamine concentrations (Litten et al. 1996). Nevertheless, some alcoholic subtypes, and behavioral responses to stress in alcohol-preferring AA and alcohol- such as those with comorbid depression, may respond favor- avoiding ANA rats. Alcohol 5(5):417Ð425, 1988. TM ably to fluoxetine. The opioid blocker naltrexone (ReVia LE, A.D., AND KIIANMAA, K. Characteristics of ethanol tolerance in or Trexan¨) also has been tested in clinical trials with alco- alcohol drinking (AA) and alcohol avoiding (ANA) rats. Psychophar- holics (O’Malley et al. 1992; Volpicelli et al. 1992). macology 94(4):479Ð483, 1988. Subjects receiving naltrexone showed decreases in the mean LI, T.K., AND MCBRIDE, W.J. Pharmacogenetic models of alcoholism. number of drinking days per week, frequency of relapse, Clinical Neuroscience 3(3):182Ð188, 1995. desire to drink (i.e., craving), and the alcohol-induced sub- LI, T.K.; LUMENG, L.; DOOLITTLE, D.P.; MCBRIDE, W.J.; MURPHY, J.M.; jective “high.” Such results suggest that these pharmacologi- FROEHLICH, J.C.; AND MORZORATI, S. Behavioral and neurochemical

VOL. 21, NO. 2, 1997 175 RESEARCH UPDATE associations of alcohol-seeking behavior. In: Kuriyama, K.; Takada, A.; intake by serotonin uptake inhibitors. Journal of Clinical Psychiatry 47(4, and Ishii, H., eds. Biomedical and Social Aspects of Alcohol and Suppl.):16Ð22, 1986. Alcoholism. Excerpta Medica International Series. Amsterdam: Elsevier Science Publishers B.V., 1988. pp. 435Ð438. NYLANDER, I.; HYYTIA, P.; FORSANDER, O.; AND TERENIUS, L. Differences between alcohol-preferring (AA) and alcohol-avoiding (ANA) rats in the LI, T.K.; LUMENG, L.; AND DOOLITTLE, D.P. Selective breeding for alcohol prodynorphin and proenkephalin systems. Alcoholism: Clinical and preference and associated responses. Behavior Genetics 23(2):163Ð170, 1993. Experimental Research 18(5):1272Ð1279, 1994.

LITTEN, R.Z.; ALLEN, J.; AND FERTIG, J. Pharmacotherapies for alcohol O’MALLEY, S.S.; JAFFE, A.J.; CHANG, G.; SCHOTTENFELD, R.S.; MEYER, problems: A review of research with focus on developments since 1991. R.E.; AND ROUNSAVILLE, B. Naltrexone and coping skills therapy for Alcoholism: Clinical and Experimental Research 20(5):859Ð876, 1996. alcohol dependence: A controlled study. Archives of General Psychiatry 49(11):881Ð887, 1992. LUMENG, L.; MURPHY, J.M.; MCBRIDE, W.J.; AND LI, T.K. Genetic influences on alcohol preference in animals. In: Begleiter, H., and Kissin, RICHTER, C.P., AND CAMPBELL, K.H. Alcohol taste thresholds and B., eds. The Genetics of Alcoholism. New York: Oxford University Press, concentrations of solution preferred by rats. Science 9:507Ð508, 1940. 1995. pp. 165Ð201. RITZ, M.C.; GARCIA, J.M.; PROTZ, D.; AND GEORGE, F.R. Operant ethanol- MARDONES, J., AND SEGOVIA-RIQUELME, N. Thirty-two years of selection reinforced behavior in P, NP, HAD, and LAD rats bred for high versus of rats by ethanol preference: UChA and UChB strains. Neurobehavioral low ethanol preference. Alcoholism: Clinical and Experimental Research Toxicology and Teratology 5(2):171Ð178, 1983. 18(6):1406Ð1415, 1994.

MCBRIDE, W.J.; MURPHY, J.M.; GATTO, G.J.; LEVY, A.D.; LUMENG, L.; AND SAMSON, H.H. Initiation of ethanol reinforcement using a sucrose- LI, T.-K. Serotonin and dopamine systems regulating alcohol intake. In: substitution procedure in food- and water-sated rats. Alcoholism: Clinical Kalant, H.; Khanna, J.M.; and Israel, Y., eds. Advances in Biomedical and Experimental Research 10(4):436Ð442, 1986. Alcohol Research: Proceedings of the Fifth ISBRA/RSA Congress. Toronto, SAMSON, H.H.; PFEFFER, A.O.; AND TOLLIVER, G.A. Oral ethanol self- Canada, 17Ð22 June 1990. Oxford: Pergamon Press, 1991. pp. 411Ð416. administration in rats: Models of alcohol-seeking behavior. Alcoholism: Clinical and Experimental Research 12(5):591Ð598, 1988. MEISCH, R.A. The function of schedule-induced polydipsia in establishing ethanol as a positive reinforcer. Pharmacological Reviews SCHWARZ-STEVENS, K.; SAMSON, H.H.; TOLLIVER, G.A.; LUMENG, L.; AND 27(4):465Ð473, 1976. LI, T.-K. The effects of ethanol initiation procedures on ethanol reinforced behavior in the alcohol-preferring rat. Alcoholism: Clinical and MEISCH, R.A. Alcohol self-administration by experimental animals. In: Experimental Research 15(2):277Ð285, 1991. Smart, R.G.; Cappell, H.D.; Glaser, F.B.; Israel, Y.; Kalant, H.; Popham, R.E.; Schmidt, W.; and Sellers, E.M., eds. Research Advances in Alcohol TABAKOFF, B., AND RITZMANN, R.F. Acute tolerance in inbred and selected and Drug Problems. Vol. 8. New York: Plenum Press, 1984. pp. 23Ð45. lines of mice. Drug and Alcohol Dependence 4:87Ð90, 1979.

MYERS, R.D., AND MELCHIOR, C.L. Alcohol and alcoholism: Role of VOLPICELLI, J.R.; ALTERMAN, A.I.; HAYASHIDA, M.; AND O’BRIEN, C.P. serotonin. In: Essman, W.B., ed. Serotonin in Health and Disease: Naltrexone in the treatment of alcohol dependence. Archives of General Volume 2. Physiological Regulation and Pharmacological Action. New Psychiatry 49(11):876Ð880, 1992. York: Spectrum Publications, 1977. pp. 373Ð430. WISE, R.A. Action of drugs of abuse on brain reward systems. NARANJO, C.A.; SELLERS, E.M.; AND LAWRIN, M.O. Modulation of ethanol Pharmacology, Biochemistry, and Behavior 13(Suppl. 1):213Ð223, 1980. Second Edition of the AOD Thesaurus Now Available

The Alcohol and The Alcohol and Other Drug Thesaurus: A Guide to Concepts and Termi- Other Drug nology in Substance Abuse and Addiction is an invaluable reference guide to Thesaurus the language and nomenclature of the AOD field. Designed for use by A Guide to Concepts and Terminology in Substance Abuse and Addiction researchers and practitioners, the second edition provides information in 4

Volume 2: Annotated Hierarchy volumes on more than 10,000 concepts and 2,300 definitions, as well as an Second Edition 1995 extensive network of cross-references.

developed jointly by A vital indexing tool for information specialists and database managers, the

National Institute on Alcohol Center for Substance Abuse and Alcoholism Abuse Prevention AOD Thesaurus also provides the necessary vocabulary for searching National Institutes Substance Abuse and Mental Health of Health Services Administration NIAAA’s Alcohol and Alcohol Problems Science Database (ETOH) and for Public Health Service

VICES SER ¥ U N SA A M free-text searching other databases. U H

D

N

A

H

T

L

A

E

H

F

O

T U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES

N

E

M

T

R A P D E

To order, contact: NIAAA Research Library, c/o CSR, Incorporated, Attn.: Publication Distribution Center, P.O. Box 10686, Rockville, MD 20849Ð0686. Telephone: (202) 842Ð7600; fax: (202) 842Ð0418; Web: http://www.niaaa.nih.gov/publications/thes.htm

176 ALCOHOL HEALTH & RESEARCH WORLD