Brnirl R~xwrdr Rt~~~icrc~.s, 18 ( 1993) 247-29 10 1W3 Elsc~ier Science Publishers B.V. All rights resewed 0165-O 173/93/$M.()()

247

Full-length reviews

The neural basis of drug craving:an incentive-sensitization theory of addiction

Terry E. Robinson and Kent C. Berridge

Department (If P.++~holog~ and Netrroscitwce Program, The Unil *ersity of Michigan, Neuroscience Idmratory Building , Ann Arbor ,MI 46104-1687 (USA) .

(Accepted 20 April 1993)

Key words: Drug addiction; Brain; Dopamine; Incentive motivation; Sensitization; Neuroadaptation; Nucleus accumbens; Striatum

This paper presents a biopsychological theory of drug addiction, the ‘Incentive-Sensitization Theory’. The theory addresses three fundamentalquestions. The first is: why do addicts crave drugs? That is, what is the psychological and neurobiological basis of drug craving? The second is:why does drug craving persist even after long periods of abstinence? The third is whether ‘wanting’ drugs (drug craving) is attributable to ‘liking’drugs (to the subjective pleasurable effects of drugs)? The theory posits the following. (1) Addictive drugs share the ability to enhancemes~~tetencephalic dopamine neurotransmission. (2) One psychological function of this neural system is to attribute ‘incentive salience’ to theperception and mental representation of events associated with activation of the system. Incentive salience is a psychological process thattransforms the perception of stimuli, imbuing them with salience, making them attractive, ‘wanted’, incentive stimuli. (3) In some individuals therepeated use of addictive drugs produces incremental neuroadaptations in this neural system, rendering it increasingly and perhaps permanently,hypersensitive (‘sensitized’) to drugs and drug-associated stimuli. The sensitization of dopamine systems is gated by associative learning, whichcauses excessive incentive salience to be attributed to the act of drug taking and to stimuli associated with drug taking. It is specifically thesensitization of incentive salience, therefore, that transforms ordinary ‘wanting’ into excessive drug craving. (4) It is further proposed thatsensitization of the neural systems responsible for incentive salience (for ‘wanting’) can occur independently of changes in neural systems thatmediate the subjective pleasurable effects of drugs (drug ‘liking’) and of neural systems that mediate withdrawal. Thus, sensitization of incentivesalience can produce addictive behavior (compulsive drug seeking and drug taking) even if the expectation of drug pleasure or the aversiveproperties of withdrawal are diminished and even in the face of strong disincentives, including the loss of reputation, job, home and family. Wereview evidence for this view of addiction and discuss its implications for understanding the psychology and neurobiology of addiction.

CONTENTS

Summary.............................................................................................

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248

247

2. The Incentive-Sensitization Theory of Addiction: an overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249

%Theoriesofaddiction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2503.1. Negative reinforcement views of addiction (escape from distress) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250

3.1.1. Problems with negative reinforcement views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 5 13.2. A positive reinforcement view of addiction (pleasure-seeking) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . 252

3.2.1. Problems with a positive reinforcement/euphoria view of addiction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2523.3. Requirements of an Incentive-Sensitization Theory of Addiction . . . . . . . . . . . . . . . . . . . . . . . . . . _ . . . . . . . . . . . . . . . . . . . . . 2 5 5

4. Evidence for the Incentive-Sensitization Theory of Addiction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564.1. Criterion 1: there should be a common neural system affected by many different addictive drugs . . . . . . . . . . . . . . . . . . . . . . . . 2564.2. Criterion 2: the repeated administration of different addictive drugs should render a common neural system hypersensitive in a

gradual and incremental fashion, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564.2.1. Sensitization to the psychomotor-activating effects of addictive drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564.2.2. Sensitization to the incentive motivational effects of drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2574.2.3. The neural basis of behavioral sensitization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

Correspondence: T.E. Robinson, Neuroscience Building, The University of Michigan, 1103 East Huron Street, Ann Arbor, MI 48104-1687, USA.Fax: (1) (3 13) 936-2690; e-mail: GB92@um.cc.umich.edu.

248

4.3. Criterion 3: sensitization-related neuroadaptations should be very long-lasting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2584.4. Criterion 4: the expression of sensitization-related neuroadaptations should be amenable to conditioned stimulus control. . . . . . . 2594.5. Criterion 5: the role of mesotelencephalic dopamine systems in incentive motivation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2604.6. Criterion 6: the effects of dopamine are on incentive salience, not pleasure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

4.6.1. Incentive salience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2614.6.2. Introspection into ‘wanting’ and ‘liking’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2624.6.3. Evidence that mesotelencephalic dopamine mediates incentive salience, not pleasure. . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

5. Elaboration of the Incentive-Sensitization Theory and its implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2655.1. The independence of drug craving from drug pleasure and withdrawal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2665.2. The development of addictive behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2675.3. Relapse: drug-induced drug craving. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2675.4. Relapse: interactions between different drugs and the effects of drug-related stimuli. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2~5.5. Relapse: the role of stress. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2695.6. Individual differences in the propensity to addiction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2695.7. Implications of the Incentive-Sensitization Theory for therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

5.7.1. Extinction training. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2705.7.2. Pharmacotherapeutic approaches. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

5.8. Relationship between incentive sensitization and other views of addiction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

6. Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2726.1. Role of control and intermittency of drug administration in sensitization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272

6.1.1.Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2726.1.2. Intermittency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

6.2. Specific motivational effects of dopamine blockade. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2736.3. The neural substrate of drug ‘liking’ (pleasure). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2746.4. Dopamine, sensitization and incentive salience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2746.5. Tolerance to drug pleasure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2756.6. The compulsive nature of addictive behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2766.7. Benzodiazepines and sedative-hypnotics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2766.8. The role of dopamine in mediating the effects of conditioned incentives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2766.9. Stress, aversive stimuli and stimulant-induced psychoses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2776.10. Dopamine antagonists and therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

7. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

A firm comiction of the material reality of Hell nel’er prellented mediesal Christians from doing what their ambition, lust or co! c’tousnesssuggested. Lung cancer, traffic accidents and the millions of miserable and misery-creating alcoholics are facts er*en more certain tlrrrll was, inDante’s day, the fact of the Inferno. But alI such facts are remote and unsubstantial when compared with the near , felt fact of a crar-ing. here andnow, for release or sedation, for a drink or a smoke.

(Aldous Huxley, The Doors of Perception, 1951)

1. INTRODUCTION

There are three major features of addictive behaviorthat need to be explained by any adequate theory ofdrug addiction *. The first is drug craving?, by whichwe simply mean intensely ‘wanting’? drugs’84. Al-though drug addiction? is defined as a pattern of2‘compulsive drug-taking behavior’, drug taking doesnot in itself constitute addictive behavior. Only whenthe repeated self-administration of drugs leads to apattern of compulsive drug-seeking and drug-takingbehavior, which occurs at the expense of most otheractivities, is a person said to be addicted7”*‘48. Tounderstand addiction, therefore, we need to under-

stand the process by which drug-taking behavior evolvesinto compulsive drug-taking behavior. Presum:My thistransformation in behavior occurs because adhcts de-velop an obsessive craving for drugs, a craving that is SO

irresistible that it almost inevitably leads to drug seek-ing and drug taking. It is difficult, of course, to providean adequate definition of subjective terms, such as‘wanting’ and craving184*“6’, but clinical experience sug-gests that drug craving is fundamental to addiction; itcannot be ignored. Any satisfactory account oC addic- ’tion must explain: why do addicts want or cr;i\‘c’ drugsso much?

Drug addiction is also “a chronic relapsing horder”(ref. 148, p. 522). The second major fe;tture Tut’ addic-

* Many of the terms in this iirticle are used in different ways by different authors ;tnd they art’ not always clearly dcfincd. To avoid mbiguifywe halve provided a glossary with definitions of n~;rny of the most problematic terms. Thus, a refcrc‘ncc’ to the glossary, which will bc indicatedby the symbol, t, refers the rcadcr to the definition of ;I term. The reader may not always agree with a given definition, but we hope at leastthis makes it clear what WC mean.

tion that must be explained, therefore, is: why drugcr;lving often persists or can be rcinstatcd, long afterthe discontinuation of drug use. An understanding ofthe propensity to relapse will be critical not only forunderstanding the process of addiction, but in develop-ing effective therapies.

A third feature of drug addiction that requires ex-planation is that, as drugs come to be ‘wanted’ more-and-more, they often come to be ‘liked’ less-and-less.That is, as craving for drugs increases the pleasurederived from drugs often decreases. Why is this? Whatis the relationship between ‘wanting’ drugs and ‘liking’drugs and does this relationship change during addic-tion?

The purpose of this article is to present a biopsycho-logical theory of addiction, an Incentive-SensitizationTheory, that addresses these issues *. The paper isorganized into four parts. In Part I the. theory issummarized to give a brief overview of its major fea-tures. In Part II the theory is put into a broadercontext by critically discussing other theories of addic-tion. specifically negative reinforcement (e.g., with-drawal avoidance) and positive reinforcement (e.g.,pleasure-seeking) theories. In doing so it is argued thattheories based on the concepts of negative or positivereinforcement? do not adequately explain the key fea-tures of addiction discussed above. In Part III researchfindings that support the concept of an Incentive-Sensitization Theory are reviewed. Finally, in Part IVthe theory is elaborated and its implications discussedin greater detail.

2. THE INCENTIVE-SENSITIZATION THEORY OFADDICTION: AN OVERVIEW

The Incentive-Sensitization Theory of Addictionposits that addictive behavior is due largely to progres-sive and persistent neuroadaptations caused by re-peated drug use. It is, if you will, a ‘neuroadaptationistmodel’. It is proposed that these drug-induced changesin the nervous system are manifest both neurochemi-tally and behaviorally by the phenomenon of ‘sensitiza-

. tion’, which refers to a progressive increase in a drugeffect with repeated treatment 272v291. These sensitiza-tion-related neuroadaptations have not been consid-ered in previous theories of addiction. In fact, untilrecently, the phenomenon of sensitization usually wasnot mentioned in books and articles on addiction and if

249

sensitization was mentioned, it was referred to only inpassing, as part of a more extensive discussion oftolerance. Nevertheless, it is proposed here that thedefining characteristics of addiction (craving and re-lapse) are due directly to drug-induced changes in thosefunctions normally subserved by a neural system thatundergoes sensitization-related neuroadapta tions.

The neural system that is rendered hypersensitive(‘sensitized’) to activating stimuli is hypothesized tomediate a specific psychological function involved inthe process of incentive motivation: namely the %ttribu-tion of incentive saliencef to the perception and mentalrepresentation of stimuli and actions. This makes stim-uli and their representations highly salient, attractiveand ‘wanted’. It is the activation of this neural systemthat results in the experience of ‘wanting’, and trans-forms ordinary stimuli into incentive stimuli.

Sensitization of this neural system by drugs resultsin a pathological enhancement in the incentive saliencethat the nervous system attributes to the act of drugtaking. The co-activation of associative learning directs.the focus of this neurobehavioral system to specifictargets that are associated with drugs and leads to anincreasing pathological focus of incentive salience ondrug-related stimuli. Thus, with repeated drug use theact of drug taking and drug-associated stimuli, gradu-ally become more and more attractive. Drug-associatedstimuli become more and more able to control behav-ior, because the neural system that mediates ‘wanting’becomes progressively sensitized. ‘Wanting’ evolves intoobsessive craving and this is manifest behaviorally ascompulsive drug seeking and drug taking. Therefore,by this view, drug craving and addictive behavior aredue specifically to sensitization of incentive salience.

But ‘wanting’ is not ‘liking’. The neural systemresponsible for ‘wanting’ incentives is proposed to beseparable from those responsible for ‘liking’ incentives(i.e., for mediating pleasure) and repeated drug useonly sensitizes the neural system responsible for ‘want-ing’. Because of this, addictive behavior is fundamen-tally a problem of sensitization-induced excessive‘wanting’ alone. This is in contrast to ‘pleasure-seek-ing’ theories of addiction, which explicitly assume thatthe incentive motivational properties of drugs are duedirectly to their subjective pleasurable effects; i.e., theirability to produce positive affective states. In colloquiallanguage, it is usually assumed that addicts ‘want’drugs because they ‘like’ drugs and the more they ‘like’

’ This paper is not a comprehensive review of the primary research literature on addiction and addictive drugs. We cite review articles tosupport specific points in many instances. Readers should consult these review articles for more extensive lists of citations to the primaryliterature.

250

them fhc more they should ‘want’ them. In this tradi-tional view ‘wanting‘ and ‘liking’ drugs are necessarilyconncctcd. The Incentive-Sensitization Theory isunique, however, bec~~sc we propose the progressiveincrease in drug ‘wanting’ that characterizes addictionis not accompanied by an increase in the pleasurederived from drugs. Repeated drug use does not sensi-tize neural systems responsible for the subjective plea-surable-l- effects of drugs, only those responsible forincentive salience - transforming ‘wanting’ into crav-ing.

In addition, the neuroadaptations underlying behav-ioral sensitization are long-lasting and in some casesthey may be permanent. It is hypothesized that it is thepersistence of sensitization-related neuroadaptationsthat renders addicts hypersensitive to drugs and todrug-related stimuli, even after years of abstinence. Itis the permanence of sensitization that is thought torender drug-related stimuli so effective in precipitatingrelapse, even in detoxified, ‘recovered’ addicts.

FinalIy, it is hypothesized that the neural substratefor incentive-sensitization (that is the neural system(s)that normally attributes salience to incentive stimuliand becomes sensitized by addictive drugs) is themesotelencephalic dopamine system. Sensitization re-sults in an increase in the responsiveness of thedopamine system to activating stimuli, such that acti-vating stimuli produce a greater increase in dopamineneurotransmission in sensitized than in non-sensitizedindividuals. The relationship between changes indopamine neurotransmission, the subjective pleasur-able effects of drugs and incentive salience, whichoccurs during addiction according to the Incentive-Sensitization Theory, are illustrated schematically inFig. 1.

The Incentive-Sensitization Theory of Addiction willbe discussed in much greater detail later. Before that,however, we need to address the features of addictivebehavior that are not adequately explained by othertheories of addiction and thus require explanation by anew theory. The most widely accepted theories ofaddiction presently fall into two classes: negative rein-forcement models (e.g., drugs are taken to avoid thesymptoms of withdrawal) and positive reinforcementmodels (i.e., drugs reinforce self-administration behav-ior by producing pleasure). There have been manypapers describing the strengths and short-comings ofnegative and positive reinforcement views of addic-tionX6~326*363*365 and it is not necessary to review thisentire literature in great detail here. Instead, negativeand positive reinforcement models will be only brieflysummarized and their difficulty in explaining severalkey features of addiction (e.g., craving and relapse) will

8. Incentive SalienceA. Affective Experience m Drug-Related Stimuli0 Pleasurable (Euphoria) m Other Stimuli

Initial Drug Addict DrugAddict Rest meto Drug-Re -Ited

Before Drug Response Response Stimuli

be emphasized, primarily to provide a comparison M iththe Incentive-Sensitization Theory. We want to stress,however, that none of these views are mutually exciu-sive. Pleasure-seeking, escape from distress and inctmtive-sensitization probably each play some role indrug-taking behavior.

3. THEORIES OF ADDICTION

3.1. Negative reinforcement views of addiction (escapefrom distress)

Historically, the aversive consequences of discontin-uing drug use (the withdrawal syndrome) have been a

251

dependence 335y372, For example, Ternes et al.335 foundthat in cynomolgus monkeys the opioid, hydromor-phone, maintained self-administration at doses thatproduced neither tolerance nor physical dependence,the latter indicated by the absence of any effect of analoxone challenge. Similarly, Lamb et a1.192 recentlyreported that former heroin addicts, who showed nowithdrawal symptoms upon a naloxone challenge, nev-

central focus of research on addiction, in part becausemany early studies were on opiates, which produceclear tolerance and physical withdrawal symptoms. Thisresearch emphasized the action of drugs as negativereinforcers t64,148,180,227,312,326,337,349,365 . To paraphraseWise and Bozarth365, negative reinforcers sustain be-havior (drug seeking and drug taking in this case) notbecause of the state they produce, but because of thestate they alleviate. According to this negative rein-forcement view of addiction drug use is maintainedbecause the aversive symptoms associated with with-drawal are alleviated by the drug. Addictive drugs thatdo not result in overt physical withdrawal symptoms,such as cocaine and the amphetamines, are thought toact as negative reinforcers by alleviating a‘ psychologi-cal distress syndrome’ produced by the discontinuationof drug use (ref. 105 for example). In addition, previ-ously neutral environmental stimuli associated withwithdrawal can themselves come to elicit withdrawal-like symptoms, by secondary conditioning305v306*357.

ertheless worked at highceive a low-dose injection

rates (lever pressed) to re-of morphine.

- Maximal periods of drug self-administration often donot coincide in time with periods of maximal with-drawal distress3? This lack of correlation betweenwithdrawal distress and drug-seeking behavior is alsoevident in comparisons made across drug classes. Jaffe(ref. 149, p. 9) notes that, although the severity of thewithdrawal syndrome associated with different drugsvaries dramatically, ranging from very subtle physiolog-ical signs to life-threatening consequences, “there islittle correlation between the visibility or physiologicalseriousness of withdrawal signs and their motivationalforce” in maintaining addictive behavior.- There are many drugs used medically that producewithdrawal syndromes but “are not typically self-ad-ministered for non-medical purposes”, including “cer-tain tricyclic antidepressants (imipramine, amitripty-line), anticholinergics and K-opioid agonists” (ref. 149,p. 9).- There are numerous reports that the “relief ofwithdrawal is minimally effective in treating addiction”(ref. 365, p. 470 for references).- There is a high tendency to relapse even after anextended period of abstinence from drugs, long afterovert withdrawal symptoms have subsided. This is usu-

Thus, drugs may not only alleviate ‘primary’ with-drawal symptoms, but also the conditioned withdrawalsymptoms induced by exposure to drug-related stimuli.A second negativesometimes used to

reinforceme nt view is that drugs are‘self-medicate’, relieving preexistent

symptoms such as pain, anxiety or depression thatoccur in life independent of drug use (ref. 170 forexample).

The traditional focus on withdrawal and tolerancewas driven by the assumption that these processes arecritical for the development and maintenance of addic-tive behavior. It is now clear, however, that the avoid-ante of withdrawal is not the most important factor inthe development or maintenance of addictive behavior,although certainly the avoidance of withdrawal maymotivate drug-seeking and drug-taking behavior inqome instances. A number of leading figures in drugaddiction research have noted, for example, that*‘physical dependence is neither a necessary nor suffi-cient condition for addiction” (ref. 365, p. 470), that“for rats and monkeys physical dependence is neither a-a

ally explained in the con text of conditioned withdrawaleffects, whereby environ mental stimuli associated withwithdrawal come to elicit withdrawal-like symp-toms 306357 . There are, however, a number of problemswith this explanation. (1) At least a third of opiateaddicts deny that they experience conditioned with-drawal symptoms when they are exposurcd td drug-re-lated stimuli .5o (2) Although many opiate addicts expe-rience conditioned withdrawal symptoms very few citethis as the reason for resuming drug use2*‘. There is, infact, a poor correlation between craving and with-drawal signs”‘. Even withdrawal-like physiologicalsymptoms induced by drug-associated cues (e.g., tem-perature, skin resistance, heart rate) are not highlycorrclatcd with reports of subjcctivc statcS”. (3) Stewartet al. (ref. 326, p. 258) have argued th;it “attempts todemonstrate such conditioned withdrawal symptomsincrease the probability of drug taking and relapse inanimals have been unsuccessful” (ref. 329 as well).

rlecessary nor a sufficient condition for opiates to actifs reinforcers” (ref. 290, p. 186) and that “physicaltjependence is currently viewed not so much as a directcause of drug dependence but as one of several factorsthat contribute to its development” (ref. 148, p. 527). AIlumber of critiques of negative reinforcement theories( If addiction have been published”6~149~32”.“h”*365 and themajor shortcomings of negative reinforcement theoriesin explaining addiction are brie fly summarized below.X1.1. Probkms with nega tiue reinforcement views-- Both people and animals will self-administer opioidsin the absence of withdrawal symptoms or physical

252

- A n umber of researchersporte d craving for some drugs

!I ave noted that) such

self-re-is often

highest irnmcdiately after drug ;Idrninistr~~tion, whenthe drug is producing subjective pleasure (a ‘high’) andwithdrawal symptoms are eliminated or are at theirweakest 50,s0,96,9& 15(),2 1s. If drug craving were due to adesire to relieve withdrawal symptoms (ref. 64 forexample), it would be expected that craving woulddramatically decrease when the drug alleviates with-drawal and is producing pleasure, not be sustained oreven increase.- Finally, animals will avidly self-administer a varietyof drugs directly into brain regions that do not producewithdrawal symptoms3? Furthermore, the infusion ofdrugs into these same brain regions can ‘prime’ orreinstate responding in animals in which drug respond-ing has been extinguished3*? These studies have estab-!ished that the incentive motivational effects of cen-trally applied drugs can be dissociated from their nega-tive reinforcing or withdrawal-related effects.

For the reasons summarized above it is now gener-allv accepted that the negative reinforcing effects ofMdrugs are not necessary for the development and main-tenance of addictive behavior. Escape from distresscannot explain the defining characteristics of addiction,craving and relapse. We do not wish to imply, however,that withdrawal plays no role at a!! in the maintenanceof addictive behavior. The alleviation of withdrawaldistress may indeed sometimesand drug-taking behavior. In

motivate drug-seeaddition, the positive

.king

reinforcing effects of drugs may be enhanced whendrugs are given during periods of withdrawaldistress86*290. However, relief from withdrawal symp-toms cannot be the sole cause or even the primarycause of drug craving and compulsive drug-taking be-havior.

3 2. . A positive(pleasure-seeking)

reinforcement

In part because ofreinforcement theories

view

the shortcomings of negativeof addiction more recent for-

of addiction

mulations have focussed on the role of drugs as posi-tive reinforcerst148*326*349*363J65. Most drugs that areself-administered by people also act as positive rein-forcers for animals. Thus, a positive reinforcementview of addiction posits that drug self-administration ismaintained because of the state drugs induce, notbecause they alleviate an unpleasant state326*363V365.

But to state that addictive drugs are positive rein-forcers does not explain addiction. As pointed out byWise and Bozarth (ref. 365, p. 472): “To assert that alladdictive drugs are reinforcers is to do little more thanredefine the phenomenon of addiction.“... “To identify

a drug as reinforcing goes no further than to identifythe drug as addicting, because it is the common obst;r-vation of habitual self-administration that serves as ti:ebasis for most definitions of both drug reinforcemt .Itand drug addiction. A theory of addiction based on tileconcept of reinforcement would have to identify X-tions of drugs that arc operationally independent Ifself-administration habits in order to offer insight as owhy drugs are addictive.” That is, positive reinforc :-ment is merely a description of a behavioral effect, n)tan explanation of the effec?“‘. The critical questio ISare, why are some drugs positively reinforcing (i. k.,what specific actions of drugs are positively reinfo: Z-ing) and why do drugs become more effective rel i-forcers as addiction develops? It is usually assum dthat drugs act as positive reinforcers because th ‘y

produce pleasure. Thus, Wise and Bozarth state “ ’“the only existing positive reinforcement view of add :-tion that might qualify as an explanatory theory ider! i-fies positive reinforcement with drug euphoria. In t! isview drugs are addicting (establish compulsive habi 5)because they produce euphoria or positive affect” (rc.f.29 and ref. 213, p. 474). Similarly, Stewart et a!.’ 26argued that compulsive drug use is maintained Ijyappetitive motivational states generated by the abili ..yof drugs to produce positive affective statesf.

There are, however, a number of problems with t! ehypothesis that the subjective pleasurable (hedoni .:)effects of drugs are either necessary or sufficient omotivate compulsive drug-seeking and drug-taking b :-havior. As put by Dews’* over 15 years ago, “it w issupposed that the prediction of addiction liability w tsessentially equivalent to prediction of euphoriger~ icpower. As with most self-evident ideas, the mere rn; t-ter of there being essentially no evidence in favor of itand much against it, had little effect on its acceptance”(p* 75).3.2.1. Problems with a positive reinforcement /euphoraview of addiction- If the positive reinforcing effects of drugs are primar-ily due to their ability to produce pleasurable affect?estates (euphoria?) and if this is sufficient to produceaddictive behavior, the subjective pleasurable effects ()fdrugs must be enormous. Indeed, the subjective p!e+surable effects of drugs would have to be so potent th;itjust the memory of drug experiences would be sufbcient to evoke compulsive drug-seeking and drug-tai;-ing behavior. Although addictive drugs can indeedproduce extremely pleasant affective states (anon++mous, personal communication) it is difficult to belielethat this property of drugs alone is sufficient to accourltfor addiction. For one, there is no clear relationshipbetween the ability of individual drugs to produ(:e

253

euphoria and their addictive potential. For example,nicotine is considered highly addictive, but nicotinedoes not produce marked euphoria or other stronghedonic states (ref. 266 for example). Also, many ad-dictive drugs can actually produce strong dysphoricstates, especially with initial use. Second, it could beargued that in addicts the magnitude of the negativeconsequences of continued drug use often far outweighthe magnitude of drug pleasure or the memory of drugpleasure. In fact, to most people (including many ad-dicts), the negative consequences of continued druguse, including loss of health, friends, family, home andlob, seem enormous relative to the pleasure derivedFrom drugs. Falk et al.@ have pointed out (p. 58): “theapparent irrationality of these activities [drug use]. Theactivities seem to produce more harm than benefit forthe individual. How could creatures have evolved suchpowerful, wasteful and even self-destructive propensi-ties? Not only are the activities apparently irrational,but also an apparent disparity exists between the im-mediate consequences of the behavior and its strength.The rush of an intravenous injection is transient andwith street-quality heroin, rarely dramatic. Yet, thedrug somehow can support day-long hustling and de-termine a whole subculture”.

Indeed, addicts will sometimes report that they aremiserable, that their life is in ruins and that the drug isnot even that great anymore - but they still want it!Addicts themselves often are bewildered by the inten-sity and irrationality of their own ‘wanting’. It is diffi-cult to explain this situation by just evoking short-last-ing drug pleasure. If the incentive motivational effectsof drugs are due only to their ability to produce tran-sient pleasure, but the aversive consequences of contin-ued drug use eventually come to far outweigh thepleasurable effects - and, if drug taking behavior ismaintained by simple contingencies of reinforcement,self-administration behavior should extinguish. But itusually does not (although see Falk et al? for adiscussion of the unique effects of intermittent sched-ules of reinforcement)..- A positive reinforcement/euphoria view of addictiondoes not adequately explain drug craving or relapseelicited by environmental stimuli associated with drugtaking. Both Stewart et a1.326 and Wise and Bozarth”“’have argued convincingly that drug-related stimuli canevoke ‘drug-like’ effects that serve to motivate furtherdrug-seeking and drug-taking behavior. They recentlytermed this view a ‘proponent-process theory’“2Y, incontrast with the ‘opponent-process’ view associatedwith negative reinforcement models1”2~“12. Stewart andWise32” argue “it is drug-like processes rather thandrug-opposite processes that whet the appetite and

stimulate renewed responding. In this view it is the‘taste’ of the drug or the experience of stimuli that -through Pavlovian conditioning - cause drug-like cen-tral effects that motivate drug intake in experiencedsubjects” (p. 80).

The question remains, however, what exactly is this‘drug-like process’ ? One possibility is that it resemblesthe positive affective state, the pleasurable state,evoked by the drug itself. That it is equivalent to whathas been called a conditioned ‘high’50. For example,Stewart et al.326 say: “Conditioned drug effects thatmimic the unconditioned drug effects, as are condi-tioned positive affective states, are elicited by the envi-ronment where these drugs are experienced.” (p. 264,our italics). In this view drug-associated stimuli mayevoke ‘conditioned pleasure’, which reminds the addictof the even greater pleasure of the drug itself, thusmotivating the individual to once again obtain thedrug362.

Addicts do report conditioned ‘highs’, as in theexample of the oft cited ‘needle freak’. In laboratorystudies, however, self-reports of conditioned ‘highs’occur much less frequently than self-reports of eitherconditioned craving or conditioned withdrawal-likesigns50922g. Self reports of conditioned craving are espe-cially frequent. This suggests that conditioned cravingis dissociable from conditioned ‘highs’ and therefore,in many instances, drug craving is not caused by aconditioned ‘high’. How then is craving explained inthe context of a positive reinforcement/euphoria viewof addiction?

A second possibility is that relapse in a ‘recovered’addict is triggered by cues that evoke an explicit mem-ory or representation of past drug experiences. Unlikea ‘conditioned high’, which is accompanied by an affec-tive experience similar to that produced by the drugitself, an explicit memory need not be pleasant in itself.It recalls past pleasure in a cognitive form, as a seman-tic proposition or as a conscious image of the act ofdrug taking and spurs the addict to attempt to regainthe remembered experience of pleasure once again.

An ‘explicit memory of past pleasure’ interpretationof relapse is not implausible and may describe someinstances of relapse. But an interpretation that positsexplicit memories of taking drugs to be a sufficientcause for relapse finds it difficult to explain why re-lapse occurs only when it does. No addict who relapsesafter months or years of abstinence could possibly nothave remembered drug cxpcriences many times before.During the process of withdrawal every addict mustoften recall and dwell upon memories of the drugexperience. Once wit hdr;\w;tl is successfully enduredthe circumstances of d;lily living would cause orlc to

253

sometimes remember earlier times, when life was dif-ferent and drugs wet-c the focus. Even without anyparticular cue or reminder, the mere process of freeassociation would often call to mind scenes from ear-lier life, including upon occasion, drug-related cxpcri-ences. Why should the explicit memory of a drugexperience suddenly be sufficient to trigger relapse,when a person has had innumerable previous memo-ries of drug experiences without relapse?

An ‘explicit memory of past pleasure’ interpretationmight reply that, because relapse is often triggered byparticular situations that have been paired with druguse in the past, these situations evoke a memory that ismore vivid than all the memories that have comebefore. Whether this is true or not is an empiricalquestion. Certainly, vivid memories can be triggered byassociative cues and there is ample evidence that asso-ciative context is a powerful modulator of conditionedbehavior; although whether context modulates explicitmemories in this way is less clear (ref. 259 for example).However, images or other forms of conscious remem-bering that occur during withdrawal, daily life or freeassociation might also be expected to be fairly vivid atleast some of the time. It is not intuitively obvious thatthese memories should necessarily differ in vividness orin any other subjective quality that might explain whysome conscious memories provoke relapse when othersdo not. In other words, an ‘explicit memory’ hypothesisplaces an extraordinary explanatory burden on thecrucial assumption that relapse-provoking memoriesare qualitatively different from the myriad other mem-ories of drugs that do not provoke relapse. As far as weknow, there is no evidence to support this assumption.- The most compelling evidence against the idea thatdrug taking is necessarily motivated by the subjectivepleasurable effects of drugs comes from studies show-ing that drug self-administration can be maintained inthe absence of subjectiLve pleasure; that is, subjectivepleasure is not necessary to maintain drug-seeking anddrug-taking behavior. A striking example of a dissocia-tion between the incentive motivational effects of mor-phine and the subjective pleasurable effects of mor-phine was reported recently by Lamb et al.192. Theseresearchers found that opiate ‘postaddicts’ would work(press a lever) to get an injection of a low dose ofmorphine, despite the fact that four of five peoplecould not distinguish the subjectice effects of the mor-phine from the placebo - but the placebo did notreinforce lever pressing (ref. 151 as well). In otherwords, people ‘self-administered’ a low dose of mor-phine and not the placebo, but reported that neitherthe drug nor the placebo produced pleasure; there wasno subjective difference between them. Similar effects

have been reported by Fischman and Foltin941”s inlaboratory studies of cocaine self-administration betI;+ior in humans. Cocaine users reliably choose a 10~dose of cocaine (4 mg) over placebo, although this dt seproduces no self-reported subjective effects or card ‘O-vascular effects. In addition, Fischman and Foltii?report that within-session tolerance to many of thecardiovascular and subjective (euphoric) effects p: O-duced by higher doses of cocaine is not accompanldby changes in drug-taking behavior; that is, withir: aself-administration session a dissociation develops :-:z-.-tween the subjective ’ effects of cocaine and coca%self-administration behavior.

On the basis of their study Lamb et a1.*92 conclut’ .*d“that the reinforcing effects of morphine can occur in Qeabsence of self-reported subjective effects and thus, lonot appear to be causually related to drug-liking weuphoria” (p. 1172, our italics). Similarly, when ask. :dto speculate what maintains the self-administration ;3fcocaine in the absence of subjective pleasure Fisi I?-man”’ replied:

“I think cocaine is maintaining their behavior! Ifyou want to know what the subjects say about th .irself-administration of these low doses, they tell 1 lethat they were not choosing cocaine over place! !.I.They often insist that they were sampling equ; lyfrom each of the two choice options and both wt Feplacebo. On the other hand, if you look at the d: tafrom that session, you see that they were choosr lgthe low dose (as low as 4 mg) or the dose with IO

measurable effect.“... “I do not believe that meas r-ing subjective effects provides us with the inforr J-tion about ‘what’ is maintaining their cocaine-t4 k-ing. The best we can say at this point is that it is r lecocaine that is maintaining cocaine-taking” (p. 173Of course, these studies also suggest that a memc,.y

or representation of the subjective pleasurable effec isof morphine or cocaine is not required to sustil Indrug-taking behavior either, because if that were tritethen there should be some subjective difference ‘nmemories during low-dose morphine or cocaine cot I-pared to the placebo and there was none. Clearly the; tis a difference between low-dose morphine or cocair, t:and the placebo, but the data suggest it is not asubjective difference; it is not explicit and does mthave access to conscious awareness.

Although the effects described by Lamb et al.192 ar:dFischman and Foltin95 are particularly striking, dissoc*i-ations between the subjective pleasurable effects ofdrugs and drug-taking behavior have been noted prei.*i-ously. For example, Falk et al? review studies showi!lgthat “the subjective effects produced by a drug do notnecessarily predict whether the drug actually will be

255

self-administered” (p. 58). In addition, Katz and Gold-berg’65 describe experiments that suggest “the rein-forcing effects and the subjective reports by humanv$unteers are not functionally equivalent entities” (p.24)

Similarly, studies in rats on the affective vs. reinforc-ing properties of opiates suggest these actions may bemediated by separable neural systems205*355. This wassuggested, for example, by White et a1.355 on the basisaf studies on the effects of morphine in a runway task,i.n which food was available in the goal box. They found-hat morphine acted as a strong positive reinforcer,:eading to faster and faster running speeds. But at the-ame time animals learned to avoid morphine-paired*hod. Thus, “the aversive effects of morphine wereCtccompanied by positive reinforcement, a paradox that7s difficult to understand” (ref. 355, p. 66). White et;tl.355 suggested an explanation for this paradox may be“that the reinforcing effects of morphine” do “notdepend upon the affective properties of the drug, butthat the drug directly activate(s) a neural mechanism ofI einforcement, which facilitates learning independantlyof the animal’s affective state”.

The studies cited above are very important becausetheir findings directly contradict the central premise of:: positive reinforcement/euphoria view of addiction.In our terms, they establish that the incentive motiva-Conal effects of drugs are not directly attributable tot!leir subjective pleasurable effects: that is, drug ‘want-i lg’ is not equivalent to drug ‘liking’.

In summary, both negative reinforcement (e.g., with-drawal avoidance) and positive reinforcement/euphoria (pleasure-seeking) views of addiction haveCfficulty explaining a number of important features ofaddictive behavior. Any plausible new theory of addic-tion needs to address the same issues and better ex-rlain them. Specifically, an adequate theory of addic-ton must explain:(i) What accounts for drug craving elicited by drug-as-sjciated stimuli, if craving is not causally related tocb)nditioned withdrawal signs, conditioned ‘highs’ ortile explicit memory of past pleasure?(:1) Why is craving sometimes highest immediately afterdrug. administration, when subjective pleasurable ef-ftbcts are still predominant?(3) Why does obsessive craving for drugs persist in thefitce of enormous negative consequences associatedwith continued drug use, and relatively modest subjec-tive pleasurable effects?(4) How can low doses of drugs, which do not producediscernible subjective pleasure or physical dependence,maintain drug-seeking and drug-taking behavior?(5) Why is relapse such a prevalent and persistent

feature of addiction, even in ‘recovered’ addicts? .(6) Why can relapse be precipitated by so many differ-ent stimuli (drugs, environmental stimuli associatedwith drugs, mood changes)?

3.3. Requirements of an Incentive-Sensitization Theory ofAddiction

Neither traditional positive reinforcement nor nega-tive reinforcement theories of addiction provide com-pelling answers to the questions listed above. We willargue below that the Incentive-Sensitization Theory ofAddiction does. The Incentive-Sensitization Theory wasintroduced as a ‘neuroadaptationist’ model. It positsthat repeated intermittent drug use results in incre-mental and persistent changes in a neural system thatmediates craving for drugs; to be more precise, in aneural system responsible for the attribution of incen-tive salience (not pleasure) to stimuli. We first need toask, therefore, whether there is any experimental evi-dence that repeated exposure to addictive drugs canproduce neuroadaptations that meet the relevant crite-ria. The criteria required of such neuroadaptations inorder for the theory to be true include the following:(1) To the extent that the excessive incentive salienceelicited by drugs is mediated by a common substrate,there should be a common neural system affected bymany different addictive drugs.(2) To explain the progressive development of addictivebehavior repeated drug administration should renderthis neural system hypersensitive in a gradual andincremental fashion.(3) To explain the persistence of relapse these drug-in-duced neuroadaptations should persist for very longperiods of time (if not permanently) following thediscontinuation of drug use.(4) To explain the role of drug-associated stimuli inrelapse the expression of these neuroadaptations shouldbe susceptible to conditioned stimulus or environmen-tal control.(5) To explain drug rraving the activation of this neuralsystem should mediate the incentive motivational ef-fects of drugs and drug-related stimuli and the neu-roadaptations produced by drugs should potentiatethese motivational effects.(6) To explain the dissociation between drug ‘wanting’and drug ‘liking’ this neural system should not directlymediate the subjective pleasurable effects of drugs orthe subjective plcasure associated with other stimuli.

Evidence that the rcpcatcd use of a number ofdifferent addict ivc drugs does indeed produce neu-roadaptations that meet each of thcsc criteria is pre-sented next.

256

auxoesthesia IhO, 162,20 1,257,272,29 1,297,354. We will use the

term ‘sensitization’ here.Psychomotor stimulant drugs, such as amphctaCnecocaine, have been used in the majority of stdies

4. EVIDENCE FOR THE INCENTIVE-SENSITIZATIONTHEORY OF ADDICTION

oron drug-induced sensitization, and the effects of tftesed rugs have been well characterized 188 7250 s *257,269,272 _920 i*- 354 .For example, the acute administration of a IOV. -to-

4.1. Criteriorr 1: there sCtorrM he n common rwrrral systemaffected by many difft>ren t acidictic v drugs

Addictive drugs represent a diverse group of com-pounds that markedly differ in their behavioral andneurochemical actions. Nevertheless, there is increas-ing evidence that many addictive drugs share the abilityto enhance mesotelencephalic dopamine neurotrans-mission73*365. This evidence has been reviewed recently;!nd need not be reiterated here. Suffice it to say thatthe following drugs have been reported to increasedopamine neurotransmission in the nucleus accumbensand dorsal striatum: amphetamine (‘speed’), cathinone(‘Khat’), cocaine (‘coke’; ‘crack’), methamphetamine(‘ice’; ‘crystal meth’), methylenedioxyamphetamine(MDA; ‘the love drug’); methylenedioxymetham-phetamine (MDMA; ‘ecstasy’), methylphenidate,ethanol, fentanyl (‘China white’), methadone, mor-phine, nicotine and phencyclidine (PCP; ‘angel dust’).This common action of diverse drugs is consistent withthe hypothesis that mesotelencephalic dopamine sys-tems mediate, at least in part, the incentive motiva-tional properties of many different drugs of abuse362v363.Although it cannot be said that there is a single neuralsystem that is affected by all addictive drugs, dopaminesystems and their associated structures are affected bymost 179.363.365

.

moderate dose of amphetamine or cocaine pro&‘psychomotor activation’, characterized by an incrt

ulation), rearHigher doses

f iiseing behi& 1 iorin locomotor act

I beand rotationaivity (ambhavior 267. result in the

emergence of focussed stereotyped behaviors, suc:~ asrepetitive head movements and sniffing and a resui’antdecrease in locomotion and rearing254y284. The re-peated intermittent administration of a constant, i -Ia-tively low dose, produces a progressivedrug-induced locomotor stimulation with

inincreasi.each SW : es-

sive administration. Repeated administration of a I-’ od-erate dose will come to elicit the stereotyped beh;* -.iortypical of a higher dose, even though it produced : dy

was given.in behavior

Furtherrr ’can come

jre,m-

der strong conditioned stimulusture of sensitization is discussed

controlin moreincrementalactivation’

ambulation the first time itsensitization-related changes

and this .<a-detail be’.jw.

t he gradual and‘psychomotor

inch .iseand the

gly stereotyped behavior,‘behavioral sensitization

only comparable stimulant effect that has been ch:! -ac-

In summary, it isin drug-induced

th. t is9. i’he

emergence of increasinusually referred to as

terized in humans is sensitization to the psychotog niteffects of amphetamine and cocaine250~272~280~297.

Behavioral sensitization is produced by the repe tedadministration of many different psychomotor sti-mulants, including the amphetamines8~188~272~292, co-caine160*250, methylphenidate40~177~301, fencamfam .ne3and the endogenous trace amine, phenylethylamil 235.The phenomenon is not limited, however, to classicalpsychomotor stimulants. Other drugs, not traditioridllyconsidered psychomotor stimulants, also produce :~y-chomotor activation, enhance dopamine neurotr .ns-mission and produce behavioral sensitization7s’.36saThese include: opioids’1*153*302, nicotine25*52~17**‘86,phencyclidinell6*117,147,225, ethanol5*,61,114,20*,2O9 ;i.nd

MDMA316 (cf. Note 7 in Ch. 6).

4.2. Criterion 2: the repeated administration of differentaddictive drugs should render a common neural systemhypersensitive in a gradual and incremental fashion

Drug effects are known to change when drugs aregiven repeatedly, and some of these changes are knowninvolve adaptations in neural elements mediating spe-cific drug effects“? Much of the emphasis in the pasthas been on homeostatic neuroadaptations thought tounderly the development of tolerance and to con-tribute to withdrawal symptoms. But as discussed above,tolerance and withdrawal do not account for the defin-ing characteristics of addiction and therefore, toleranceand withdrawal-associated neuroadaptations do notmeet Criterion 2 under discussion here.4.2.1. Sensitization to the psychomotor-activating effectsof addictive drugs. However, some effects of drugs arenot decreased, but are actually increased by repeatedintermittent drug administration. Indeed, for a givendrug, some effects may decrease (show tolerance) whilesimultaneously other effects increase. This latter phe-nomenon has been referred to as behavioral sensitiza-tion, behavioral facilitation, reverse tolerance and

Repeated intermittent treatment with an addic -ivedrug not only produces sensitization to that drug, but

produce cross-sensitization to other drilgs.Althoughcross-sensi

thetizat

literature is not entirely consistcion has bee,n reported between drug)

nt,in

may also

the same class (e.g., amphetamine and cocaine) .lndbetween drugs in different classes (e.g., stimulants Jndopioids)l59*16(W*. Furthermore, cross-sensitization isalso found between drugs and stress, which led to the

257

suggestion that drugs may induce sensitization by theiraction as stressors8*9*249. Evidence for cross-sensi tiza-tion between drugs and stress comes from studiesshowing that animals pretreated with drugs like am-phetamine, cocaine or morphine are later hyperrespon-sive to stress and vice versa, animals sensitized to stressare hyperresponsive to a drug challenge**9~1~*1~*196*268.Ben the repeated administration of exogenous corti-iosterone is reported to increase the locomotor re-:ponse to a later challenge with amphetamine’* (seer~lso Note 9 in Ch. 6).42.2. Sensitization to the incentive motivational effects(If drugs. Of more direct relevance to the process of;Iddiction are recent studies on sensitization to theincentive motivational properties of addictive drugs.‘I’here has been relatively little research on this topic,but there have been a number of recent experiments,ilsing either self-administration procedures or the con-(Iitioned place preference paradigm, which suggest thatl)rior exposure to amphetamine, cocaine or morphinejjroduces sensitization to the incentive motivationalc:ffects of these drugs.

In one of the first reports of this kind Woolverton et;il 373 found that rhesus monkeys would self-administer;t ;ow dose of methamphetamine only after they hadreceived a regimen of non-contingent injections ofrnethamphetamine. That is, the threshold dose neces-:ary to sustain self-administration was lowered byrnethamphetamine pretreatment, suggesting “an in-creased sensitivity to the reinforcing properties of theclrug” (p. 740). There are now a number of similarI eports in ratsz41. For example, Piazza and his co-\vorkers2429243 have reported that d-amphetamine pre-treatment, which induces behavioral sensitization, facil-itates the later acquisition of an amphetamine self-ad-lninistration habit, especially in animals not predis-posed to self-administer amphetamine. Similarly, pre-treatment with cocaine facilitates the later acquisitionof a cocaine self-administration habit 13’.

Sensitization has also been observed with the condi-tioned place preference paradigm, in which the placewhere a drug is experienced becomes preferred by an:!nimal in subsequent choice tests. For example, Lett195t.xamincd the influence of amphetamine, cocaine ormorphine pretreatment on the later acquisition of aconditioned place preference produced by the sameclrug. For all three drugs, pretreated (sensitized) ani-mals showed a significantly enhanced conditioned placeJjreference, relative to control animals. A similar effecthas been rcportcd following pretreatment withmorphine102 or ethanol’14.

Cross-sensitization also occurs in these situations.Animals pretreated with amphetamine show ah cn-

hancedtreated

place preference for morphine,with morphine show an enhanced

animals pre-place prefer-

ence for cocaine and animals pretreated with morphineshow an enhanced place preference for amphet-amine195. These latter findings suggest that the en-hancement of the conditioned place preference is notdue to tolerance to the drug’s aversive properties,because cross-tolerance does not occur between thestimulants and morphinelg5. Similarly, cross-sensitiza-tion has been found using self-administration proce-dures. Noncontingent pretreatment with amphetamine,caffeine or nicotine facilitates the later acquisition ofcocaine self-administration13”-138*282. An especially in-triguing example of cross-sensitization between opiatesand amphetamine was reported recently by Cunning-ham and Kelley .62 These researchers found that therepeated intra-accumbens application of a mu receptoragonist (morphine or DAMGO) for 4 days potentiated(sensitized) the ability of systemic amphetamine tolater enhance responding for a conditioned reinforcer;i.e., a light/tone previously

It was mentioned abovethe psychomotor-activating effects of drugs cannot only between drugs, but between drugs and

paired with food.that cross-sensitization to

occurstress.

Therefore, the effects of prior stress on the incentivemotivational effects of drugs are also of interest. In-deed, prior stress (tail pinch) facilitates the acquisitionof an amphetamine self-administration habit ‘? In fact,a number of potentially stressful environmental manip-ulations, such as social isolation or prenatal stress, arereported to increase sensitivity to amphetamine andfacilitate amphetamine or cocaine self-administrationbehavior6~v~52Ble

In conclusion, these studies establish that not onlyare the ‘psychomotor-activating’ properties of addictivedrugs sensitized by repeated drug administration, buttheir incentive motivational properties are sensitized aswell. Animals sensitized to amphetamine. cocaine ormorphine later show an enhanced preference for anenvironment associated with drug administration andanimals sensitized to amphctaminc or cocaine showenhanced vulnerability to acquire a drug self-adminis-tration habit.motivationalstimuli) could have a profound influence on the devel-opment of addictive behavior. With more-and-moredrug experience the incentive value of the act of drug-taking and of drug-related stimuli would be progres-sively enhanced, which would increase the prob;lbilityof repeating drug-seeking and drug-takiilg behavior inthe future (although, see Note 1 in Ch. 6 for a discus-sion of the role of response contingency in drug scnsiti-

Obviously,properties

sensitization toof drugs (and

the incentivedrug-related

zat ion).

a dopamine receptor antagonist, which prevents ‘Ihe.induction of behavioral sensitization to meth;:m-phctaminc, also attenuates the dopaminergic rcsporjseto a methamphetamine challenge assessed with mh-o-dialysis”‘. Although there is no convincing evidcilcefor sensitization-related changes in dopamine recei*tor

behGorn1 studies summarized above strongly suggestthat the repcatcd administration of many differentaddict iw drugs produces gradual and incremental neu-roadaptat ions that render animals hypersensitive tothese agents. The behavioral studies also provide primafacie evidence that sensitization-related neuroadapta-t ions involve a hypersensitivity in nlesotelcnccpl~alic

binding2” , there is elet t rophysiological evidence foincreased sensitivity of nucleus accumbens neuron

doprlmine systems. A change in dopami ne neurotr;tns-mission is implicated for a number of i-e asons. First,

iontophoretically applied dopamine in cocaine sc?lsi-tized rats’24*““4. It is possible, therefore, that sensi+:La-tion produced by coca& and amphetamine is :+oaccompanied by changes in the transduction ofdopamine receptor-mediated events’“~’ 12~124~140~26s~27~~.

The biophysical basis of sensitization-rela:r:dchanges in dopamine neurotransmission is not kno if,.There have been a number of hypotheses propo:..>d,including changes in autoreceptor sensitivity, than ,:es

the behaviors thatknown to involve

are sensitized by addictive drugs arcactivation of mesotclencephalican

dopaboth

mine systems. ThereItor-act

IS considerable evidenceivating effects and the in

that.cen-the psychomo

tive motivational effects of many of these drugs re-quires the integrity of mesotelencephalic dopaminesystems, especially dopamine projections to the ventralstriatum? Second, the activation of dopamine sys-tems appears to be necessary to induce sensitization.The sensitization produced by amphetamine, cocaineor morphine is prevented by co-treatment withdopamine antagonists12’*187*344*345*352 (see ref. 160 for areview) and amphetamine sensitization is prevented bya 6-OHDA lesion293. Third, the application of am-phetamine or morphine directly into the ventraltegmental area, where dopamine cell bodies are lo-cated, induces sensitization’? Fourth, a local ‘chal-lenge’ injection of amphetamine- into the lateral

in theing to

intraneuronal distribution of dopamine le *i -dse, changes in the dopamine ‘P-enhanced relea

take carrier and changes in transduction mecba-nismsh, 1609 188,226,269,354 . It is not known, however, wheti>trthe behavioral sensitization produced by diffeinvolves the same or different cellular and

rent dr:. gsmolecij 3r

changes. Even the processes involved in the induct! Jnof sensitization differ from those involved in the X-pression of sensitization 160~233~269. Furthermore, sen5 ti-zation-related changes in

phasized here becausedopa.mine systems have bcfn

dopa mine systems ha veonlye mventricle 2’s

behavor nucleus accumbens 178,233 evokes a sensi-

treated previouslybeen studied in any detail. Butother neurotransmitter systems

neuroadaptationsthat interact M

in;thtized ioral response in animals

with systemic amphetamine. dopamine systems must also be considered. For ex;i:n-ple, glutamate systems have been implicated in recntstudies showing that glutamate antagonists, I.kedopamine antagonists, prevent the induction of sen) ‘ti-zationl6lql63yl64,369

.

Perhaps even more importantly, behavioral sensiti-zation is accompanied by changes in mesotelencephalic

’ dopamine activity 1~*257~269*272~354. For example, am-phetamine sensitization is accompanied by an increasein amphetamine-stimulated dopamine release fromstriatal tissue in vitro48*176*178*271*358+375. More recently,

In summary, there is considedence in support of Criterion 2.

rable experimental t vi-The repeated adm.ir is-

in vivo microdi&lysis studies have shown that althoughamphetamine sensitization is accompanied bY

tration of many differehavioral sensitization

nt addictive drugs produces 1-jnsitizationand behavioral se

e-is

changes in the basal extracellular concentration ofdopamine, it is associated with an enhanced dopamineresponse to a drug challenge144~166*231,269~274~370 (cf. ref.295). Even the local application of amphetamine intothe ventral tegmental area sensitizes the dopaminerelease produced by a subsequent systemic challengewith amphetamine347. A similar enhancement indopamine response has also been reported in associa-tion with sensitization to cocaine5*157*167*236,238 (cf. ref.296), ethano124, nicotine25*10’V123 (cf. ref. 65), mor-phine 154q160 (cf. ref. 155), phenylethylamine’8’ andmethylphenidate”’ and following cross-sensitizationbetween different drugs4*159*166 and between drugs andstress 156,159,160,3 15,320,358

. Furthermore, co-treatment with

associated with hypersensitive mesotelencephalic dol:a-mine systems.

4.3. Criterion 3: sensitization-related neuroadapta tiersshould be very long-lasting

One of the most striking characteristics ofsensitization is its persistence. A single injt>c-tion of amphetamine, cocaine or morphine indu-ces behavioral sensitization lasting for weeks tomonths40, 197,236,237,267,302 and animals sensitized with L+calating doses of amphetamine remain behaviorallyhypersensitive to an amphetamine challenge for atleast 1 year232. In fact, Paulson et al.232 found that ratswere just as sensitized a year following the discontinua-

259

tion of amphet!Grd of their

tamine pretrelife-span, as

atment, which is over onethey were at 2-4 weeks.

7’hese findings suggest that after at least some pre-treatment regimens the neuroadaptations responsiblefor behavioral sensitization to amphetamine may beessentially permanent. Similarly, behavioral sensitiza-tion in rats is reported to persist for over 3 monthsfollowing pretreatment with cocaine3W, for over 50days following pretreatment with methylphenidate3”’and for over 8 months following pretreatment withnlorphine’2*‘7*302.

It is not known if sensitization to the incentiveniotivational properties of drugs persists for as long ass :nsitization to their psychomotor-activating effects,but this is obviously an important issue for the Incen-tlve-Sensitization Theory. To the extent that sensitiza-tion to the psychomotor-activating effects and the in-ccntive motivational effects of drugs have a commonnatural basis we would expect both effects to showc Jmparable persistence. Neither have neurochemicals udies been conducted as long as a year following drugpretreatment. But sensitization-related changes indlpamine systems have been reported to persist forn:onths after withdrawa126”9272. The available evidences lggests, therefore, that the neuroadaptations underly-il g behavioral sensitization meet the criterion of per-s; jtence.

4 4. Criterion 4: the expression of sensitization-relatedvuroadapta tions should be amenable to conditionedsI imu Ius control

The Incentive-Sensitization Theory of Addictionp jsits that drugs sensitize a neural system that medi-a:es ‘wanting’. It is also hypothesized that associativepr-occsses focus exaggerated ‘wanting’ (craving) specifi-c tlly onto drug-related stimuli. This implies that theb .:havioral expression of sensitization-related neuroad-aj)tations should be strongly influenced by associativefil ctors.

Indeed, the environmental context in which drugsa:-e experienced can have profound effects on thed~:velopment and expression of sensitization. Stewart 323recently reviewed the literature on the conditionedstimulus control of sensitization and the reader is re-fcrrcd to this paper for a more detailed and excellentd%cussion of such issues. In brief, Stewart32-7 pointso:rt that the conditioned stimulus control of scnsitiza-tL)n can take one of two basic forms, depending tosome extent on experimental design. In one design thedrug (the UCS), which produces a pharmacologicalelt‘ect (the UCR), is given only in association withu:?ique environmental cues (CS). After rcpeatcd pair-ing of the UCS and CS, the CS alone can acquire the

ability to elicit drug-like responses (CR). For example,after repeated administration of a dose of am-phetamine that produces locomotor hyperactivity, justplacing an animal in the environment in which it previ-ously received amphetamine is sufficient to produceconditioned locomotor hyperactivity, in the absence ofany drug. A number of researchers have reporteddrug-environment conditioning of this type and havesuggested these conditioned effects contribute to thedevelopment of sensitization’2’*2H”*“2” (see ref. 338 for areview).

A second type of conditioned stimulus control ofsensitization is the situation where, after explicit pair-ing of a drug and specific test environment, animals areadministered a ‘challenge’ injection of the drug ineither the drug-paired environment or in a new envi-ronment. In this case one observes the effect of the CS(the environment) on the response to the UCS (thedrug). In some experiments of this type the expressionof sensitization has come under complete conditionedstimulus control. For example, Post et al? reportedthat rats given ten daily injections of 10 mg/kg ofcocaine in a test environment showed a progressivelygreater behavioral response (locomotor activation) tothe drug, but animals given daily injections of cocainein their home cage did not show evidence of behavioralsensitization when subsequently challenged with co-caine in the test environment. Similarly, Vezina andStewart jd4 reported that repeated injections of mor-phine into the ventral tegmental area produced evi-dence of sensitization on a subsequent challenge testonly when rats were tested in the environment wherethey received morphine. In addition, the conditionedstimulus control of sensitization may not only enhancethe expression of sensitization in drug-paired environ-ments, but may also inhibit the expression of scnsitiza-tion in environments that predict the absence of thedrug. For example, Vezina and Stewart3” found th;itthe locomotor response to an intra-ventral tcgmentalarea challenge injection of morphine in explicitly un-paired animals was significantly dcprcsscd relative tosaline-pretreated controls (also see ref. 328). Ncverthe-less, scnsitizat ion is not only a conditioned response,even t bough the expression of scnsit ization can comeunder strong conditioned stimulus control, a point thathas been made by a number of rtutl~c)rs’“.““~17~.312~31h.

Our discussion of the condit ioncd stimulus controlof scnsit ization has focus4 thus far on the effects ofdrug-associated st irnuli OII t hc subscqucnt rcsponsc todrugs. It is important to know, howcvcr, whether cnvi-ronmental stimuli associated with drugs can also influ-ence the response to other inccntivc stimuli; evennon-drug-related stimuli. This issue has reccivcd ;iI-

most no cspcrimcntal attention, despite its theoretical signals predicting the availability of food, water of- aimportance. Ncvcrt hcIcss, it was addressed in one study sexual partner activate brain dopamine systems, in-by Mitchell and Stewart ‘lx. In this study male rats weregiven ntorphinc either in a test arena or their homecage, every other d;iy for four injections. AU rats werethen placed in the test arena in the presence of areceptive female. Rats pretreated with morphine in thetest arena showed ntore frequent female-directed bc-haviors than either rats pretreated with morphine intheir honte cage or saline-pretreated controls. Theenhancement of sexual behavior produced by thedrug-associated environment did not invo.lve changes incopulatory behavior per se, but only in the appetitivebehaviors leading to copulation, including more fre-quent “pursuit of the female, anogenital explorationand partial mounts and... shorter tatencies to initiatecopulation” (p. 367). That is, the female appeared tobe a more salient incentive stimulus in male rats testedin the presence of morphine-associated cues. Theseresults suggest that sensitization to drugs may changeneural systems that not only modulate the incentiveproperties of drug-associated stimuli, but the incentive

creasing dopamine ncurotransmission in the vent ratstri;atunlS 1,66,126,24h*37he Addictive drugs also incre;+edopaminc neurotransmission’“. Indeed, in the case ofdrugs, a direct action on dopamine systems alone issufficient to motivate behavior. Animals will work %rmicroinjcctions of drugs directly into appropriate p jr-tions of the mesotelencephalic dopamine system ‘?Furthermore, microinjections of amphetamine ordopaminc directly into the nucleus accumbens fat ii-tates responding for conditioned incentive stiml ;ii(conditioned reinforcers) - stimuli that have acqui :dincentive properties by association with a natural .n-centivet~~92h2,263 . In short, a common neural currei .cyfor many incentives appears to be activation of mes(“e-lencephalic dopamine systems (see Note 4 in Ch. 6“

Consistent with this idea are studies showing t’latthe motivational properties of natural incentives ;andaddictive drugs are attenuated by decreasing dopam neactivity ‘ho. Antagonist drugs that prevent the activat _>nof dopamine receptors or the complete destructiorl ofmesotelencephalic dopamine projections by net: so-chemically selective toxins, impair the instrumer.tafperformance of animals for food, for drugs and :‘orelectrical brain stimulation. A great deal of effort (as

been directed towards ascertaining whether the s .p-pression of dopamine neurotransmission prods *eschanges in behavior because of effects on the con rolof movement or because of effects on incentive mot ca-tion and a variety of experimental paradigms have bl. endeveloped to distinguish between motor and mot a-

properties of ‘natural incentives? This is importantbecause it implies that drug-associated stimuli maypotentially influence a wide range of motivated behav-iors.

To summarize thus far, we have addressed the firstfour criteria, and reviewed evidence to establish that:(1) Many different addictive drugs activate a commonneural system, the mesotelencephalic dopamine sys-tern; (2) Repeated administration of many addictivedrugs causes dopamine systems to become hypersensi-tive and this is accompanied by a grad ual and incre-mental increase (sensitization) in the psychomotor-

tional effects (see Note 2 in Ch. 6). It is now genel Jlyaccepted that dopamine antagonists can have effi ctson behavior that are truly motivational. This is no: tosay they may not also have effects that are ‘motor’, Jutin many cases the effect on behavior is precisely M hatone would expect if dopamine antagonism reduced :hemotivational properties of incentives82*360~36’.

In summary, the large literature on the role of

activating anddrugs; (3) The

incentive motivational properties ofneuroadaptations underlying sensitiza-

tion are extremely persistent; and (4) The expression ofsensitization is subject to conditioned stimulus control.

We next need to address Criteria 5 and 6. Criterion5 requires that the neural system sensitized by re-peated treatment with addictive drugs normally medi-ate the incentive motivational effects of drugs anddrug-related stimuli. Criterion 6 requires that this neu-ral system not mediate the subjective pleasurable ef-fects of drugs or the pleasure associated with otherstimuli.

dopamine in mediating the incentive motivationaI ef-fects of drugs and other stimuli satisfies Criteriort 5.What of Criterion 6, which requires that dopamine ,lotmediate the subjective pleasurable effects of drugs orthe pleasure associated with natural incentives.

4.6. Criterion 6: the effects of dopamine are on incentive4.5. Criterion 5: the role of mesotelencephalicsystems in incen tirle motiua tion

dopamine salience, not pleasureThe evidence that brain dopamine systems mc di-

ate the incentive motivational effects of natural incisn-tives and addictive drugs provided the basis forWise*s359,360,361 provocative anhedonia hypothesis: thehypothesis that mesotelencephalic dopamine systems

There is a wealth of evidence implicating dopaminesystems in the incentive motivationalt effects of drugs,as well as of food, sex and other natural incen-tives 19,MvW, I3 1,262,263,356,365,368 (cf. ref. 181). For example,

2 6 1

mediate the subjective pleasure produced by food,drugs, electrical brain stimulation, etc. and thatdopamine antagonists suppress the pleasure producedby these agents. The anhedonia hypothesis provided aparsimonious explanation of dopamine’s motivationalctffects by equating it with the psychological process ofsubjective pleasure *.

If the motivational properties of natural incentive:.timuli or drugs depended only on their ability toproduce subjective pleasure, then the effects ofdopamine manipulations on motivated behavior wouldbe compelling evidence for the anhedonia hypothesis.1Ve suggest, however, that incentive motivation de-pends on a number of additional psychological pro-c’esses that interact with pleasure: including associativelearning and the attribution of incentive salience tolbxternal events and their representations2’y2! It is thetbntire complex of pleasure, learning and incentive:!aliencet together that comprise the process of incen-i ive motivationt.4 6.2. Incentive salience. Of particular importance to;he Incentive-Sensitization Theory of Addiction are theaelative roles of ‘wanting’ and ‘liking’ in incentive

hnotivation. The idea that there may be a psychological‘lrocess (and neural substrate) responsible for ‘want-ng’ incentives that is dissociable from the psychologi-

l:al process (and neural substrate) responsible for ‘lik-ng’ incentives has not been considered explicitly in

Grevious theories of incentive motivation30q339v* *. BasedI>n a series of studies on the relationship between tastenleasure and appetite, such a dissociation was recentlyproposed 27s28. Berridge and his colleagues hypothe-sized that a psychological process specifically involving:-he attribution of salience to incentive stimuli (incen-Live salience) results in the experience of ‘wanting’.This view of incentive motivation posits that salience.jttribution is a specific psychological process that is;rctivated normally in conjunction with pleasure (‘lik-ing’) and associative learning in the creation of newincentives. As new incentives are acquired particularstimuli that allow an individual to recognize an incen-tive (e.g., the sight of food; in nature, a flowering plantthat signals the availability of food; or, in the labora-tory, a tone that predicts food), become associated withthe pleasure food produces by the process of classical

conditioning . The stimul us fea.tures of the incentivepredict the pleasure that will follow and may elicitconditioned pleasure. But conditioned stimulus fea-tures also become themselves the target of a separateand powerful motivational process - salience attribu-tion. Salience attribution transforms the sensory fea-tures of t .he inceperce PC which

ntive st‘grabs

:imulus into an especially salientattention’, be comes att ractive

and ‘wanted’ and thus guides behavior to the incentive.That is, new incentives become attractive in their ownright as conditioned incentives? (also called condi-tioned or secondary reinforcers).

Thus, the role of salience attribution in incentivemotivation is proposed to occur as the third stage of athree-stage process (see Fig. 2). First, pleasure is nor-mally activated by an encounter with a natural incen-tive, such as when an hungry animal eats food. In thenormal course of events, pleasure is a necessary step inthe establishment of a new conditioned incentive.However, pleasure (‘liking’) is not by itself sufficient tomotivate behavior 28. Pleasure by itself has no object oraction. Assignment of pleasure to something requiresassociative learning, which is the second stage in theformation of incentives.

If pleasure is assignedassociative lea rning, then

to an actionthe action or

come to predict pleasure or elicit pleNo doubt this often happens. But we

or stimulus bystimulus should

asure, on its own.would argue that

neither the experience of pleasure nor the expectationof impending pleasure by themselves constitute ‘want-ing‘27y28. ‘Wanting’ requires an additional process: theattribution of incentive salience to stimuli or actions.Stimuli that are attributed with incentive salience be-come attractive and demand attention. Like the sightof food to a starving person, they cannot be ignored.This does not necessarily make them ‘liked’; the sightof food may be irresistibly attractive to the starvingperson, but if out of reach it may torment rather thanplease. But the food is still much ‘wanted‘. In sum-mary, incentive motivation is proposed to involve threedistinct psychological proccsscs acting together; plea-sure, associative learning and illccrltir*e smlicrzce anddifferent neural systems are thought to be responsiblefor each27*28* * * *.

Changes in any one of these three processes that

* Note, however, that in discussing his anhcdonia hypothesis of dopamine blockade Wise”’ stated: “The anhcdonia hypothesis is mostvulnerable in its assumption that positive hedonic states such as pleasure or euphoria are attenuated by ncurolcptics. This is, for the mostpart, speculation” (p. 184).Although neither Bindra nor Toates dist inguishcd ‘wanting’ from ‘liking’, Toates’“’ did distinguish the associative of l wit nt ing/lik-ing’ from mere recall of past ‘wanting/liking’. For example, as in the Krieckhaus effect, a stinmlus that is ordinm-ily not ‘wntcd‘ or ‘liked‘may suddenly bccomc ‘wanted/liked’ on the basis of its prior associations when physiological state i s changcd*Ss.

*** For a discussion of evidence that most motivated bch avior is primarily controlled by incentive processes see Toatcs”?

const itutc inccnt ive mot iv3 tion would produce effectson instrurwnt;~l or ;lppctitivc behavior indistinguish-able, bv most mCasurcs, from changes in another.Mesotelencephnlic dopnminc does not seem to be criti-cal specifically for associative learning; i.e., for formingstim~il\is-stinl~llus associ;ltir~ns’“~~~“. For example,dopamine antagonists do not disrupt the learning ofassociations between a stimulus and clcctrical shock, asmeasured by a conditioned emotional response or bythe defensive burying paradigm”+23*141. It is more likely,therefore, that the effects of dopaminc on motivatedbehavior are due to effects on subjective pleasure orincentive salience. If mesotelencephalic dopamine sys-tems mediate the subjective pleasurable effects ofdrugs. then the addictive property of drugs such ascocaine, heroin, amphetamine, etc., which sensitizedopamine systems, might be best understood in termsof the pleasure these drugs induce. But if a sensitizeddopamine system enhanced the incentive salience ofthe act of taking drugs, rather than the subjectivepleasurable effects of drugs, thenunderlying sensitization would

the neuroadaptationscause drugs to be

‘wanted’ or craved, independently of the pleasure theyproduce.

Most animal studiesantagonists on motivated

on the effects of dopaminebehavior (summarized in Note

2 in Ch. 6) are equally consistent with the hypothesisthat dopamine mediates pleasure (the anhedonia hy-pothesis of Wise) and the hypothesis that dopaminemediates incentive salience*‘**“. Likewise, studies onthe effects of manipulating dopamine systems on theincentive properties of conditioned reinforcers are con-sistent with both interpretations of dopaminefunct ionHs*2”2. In these latter experiments animals mustlearn a new instrumental response (bar press) that isreinforced only by a stimulus (light/tone) that waspreviously associafood, water or adopamine neurotransmission

the incentive prope

ted with a primary incentive,mate. Manipulations that

potentiatein the ventral striatum

rties of conditioned rein-

isuch asncrease

forcers and manipulations that decrease dopamineneurotransmission in the ventral striatum block thesepotentiating effects44,127,168,169,264,333,334 Barry Everitthas suggested these studies support g view that, insome way “ventral striatal dopamine makes the worldbrighter” (Catecholamine Symposium, Amsterdam,1992). This view is compatible with the attribution ofincentive salience hypothesis proposed here. Neverthe-less, these studies could also be interpreted in supportof the ‘pleasure’ hypothesis and therefore, they do notresolve the issue.4.6.2. Introspection into ‘wanting’ and ‘liking’. At firstglance it seems it should be relatively easy to resolve

the issue by asking people to report what efkctdopaminc antagonist drugs have on the subjective cf-fects of addictive drugs. Indeed, Wise”” challen,:cdresearchers to test this hypothesis, arguing that ‘-“itwould bc procedurally simple for workers using r~u-roleptics with human subjects to determine the eff;:ctsof these drugs on the subjective effect of rewart’ingstimuli”. Despite his challenge, the relationship be- 1tween dopamine blockade and anhedonia rem-insequivocal and the data are inconsistent9X~298.

In light of the incentive salience hypothesis, 1:. iw-ever, assessment of the effects of dopamine block. :leon subjective pleasure in humans may not be as sin pieas Wise thought. In fact, it may be an especially dffi-cult task2*‘. Humans may not, under normal co1Ji-tions, be able to subjectively tell the difference 3e-tween the two psychological processes of ‘want ng’versus ‘liking’. S,tudies of introspection and self-reporthave established that humans often (a) report ~SVC!IO-mlogical ‘events’ that can be shown to have not 1: ip-pened; (b) strongly deny the existence of psycholog talevents that can be shown to have influenced tl.eirbehavior; and (c) confuse events that are not c:jn-netted**‘. Under many circumstances humans actt; lilyhave very little direct access into the nature of tl:sirown psychological processes. Rather, introspection tp-pears to interpret underlying processes in ways :iatseem plausible to the person2*“. Introspection does notreveal psychological processes directly. This imr liesthat a person might mistake a change in inter ivesalience for a change in pleasure (‘If I don’t war : it,then I must not like it’). It may be possible to di [in-guish between changes in ‘wanting’ and ‘liking’ byasking appropriate questions that tap into diffe .entaspects of a person’s reaction to an incentive, but thiswill have to be approached with great sophistic;! ionand caution.4.6.3. Erlidence that mesotelencephalic dopamine n: ..di-ates incentive salience, not pleasure. Although the md-ies cited above are equivocal, there are at least !ourlines of evidence which lead us to suggest that seCti-zation of dopamine neurotransmission produces cn-hanced incentive salience rather than enhanced $ea-sure. The evidence is based both on reports by hurTIanaddicts of their own experience and on studie:\ inanimals on the role of dopamine in sensory plea\ureand motivated behavior.- The first line of evidence comes from animal studiesthat have explicitly examined the role of dopaminc inmediating natural reactions to the sensory qualities oftastes. These studies have used the ‘taste reactivityparadigm’, which is based upon natural hedonic :Indaversive reactions (tongue protrusions, gapes, forelimb

263

B .

Pleasurable3lncent i ve -Stimulus

Subjective

fWanting

\Pl&surc

j AffectiveAct ions

SubjectiveCraving

Attraction

El ic i tConsumPtion

ActionsT3g. 2. A schematic illustration of a major existing model of the psychological processes that constitute incentive motivation (A, top) and our* hernative model, which proposes a separate process of incentive salience and accounts for the consequences of drug-induced sensitization (B,

; ottom). Panel A: the ‘Toates/Bindra model’ of incentive motivation, on which our model is based (adapted from Toates33’). By this model thesensory stimuli (CS and UCS) of incentive objects are both pleasant and attractive. Their ability to produce an incentive motivatibnai state (Ki) isi)artly dependent on memories of previous favorable experiences with them. Physiological states (such as hunger, thirst or withdrawal) selectively;jotentiate the ability of particular stimuli (related to food, water or dru gs) to evoke incentive processes: to become ‘wanted’ and ‘liked’ ( s e e‘roates33Y for a more complete description of this model). Panel B: i n our modified model of incentive motivation the psychological process (and,ieurai substrate) for pleasure (‘liking’) is separate from the psychological process (and neural substrate) responsible for incentive salience

( ‘wanting’). We further propose that the activation of mesoteiencephaiic dopamine systems plays a direcr role only in the process of ‘wanting’, via:he attribution of incen tive salience to the perception and representation of conditioned st imuii (as described by Berridge and Vaiensteir?‘; also>ee Note 4 in Ch. 6). In Panel B the portion of the model (i.e., the psychological process) that is sensitized by repeated drug administration ishighlighted’ within the shaded box. It is the hyperactivation of this specific psychological process (incentive salience), due to sensitization of itsleurai substrate by drugs, that results in the excessive attribution of incentive salience to drug-related stimuli. Whereas normal levels of incentive,aiience attribution results in normal ‘wanting’, we propose hyperactivation of this system results in excessive incentive salience attribution, whichs experienced as craving. Craving is pathologically intense ‘wanting’. The major difference between our model of incentive motivation and the

traditional model is that in ours the psychological processes and neural substrates responsible for pleasure (‘liking’) are separate from those forIncentive salience (‘wanting’). Thus, in our model ‘natural incentives’ (UCS stimuli) produce pleasure directly, but produce incentive salience and:iicit goal-directed approach behavior only indirectly (as indicated by the dashed arrow from ‘pleasure integrator’ to the ‘incentive salience.ittributor’). The direction of incentive salience attribution to stimuli that preceded or accompanied incentive salience activation is determined byassociative learning. Thus, activation of the incentive salience attributor by a UCS results in incentive salience being assigned to the perception of~:onditioned stimuli that were originally neutral (such as the sight of a syringe) and to their mental representations. This is what makesconditioned stimuli attractive and ‘wanted’ and able to elicit approach. Conditioned stimuli (and UCS’s) are always compared against pastassociative memories. Without the direction provided by associative learning, incentive salience could not be focussed upon any single target.Mthough diffuse attribution of incentive salience would be both psychologically and behaviorally activating, without associative direction it wouldroof be sufficient to guide behavior towards a specific goal. Familiar conditioned stimuli that have been paired with incentive salience attributionin the past are the target of incentive salience when encountered again, especially when an animal is in particular physiological states (indicatedby the arrow from ‘physiological drive cues’). Incentive salience assigned to conditioned stimuli must be further ‘reboosted’ each time they arepaired again with salience activation (indicated by the &shed arrow from the incentive salience attributer to associative learning). Disruption ofthis reboosting, by neuroieptics for example, can produce ‘extinction mimicry’ or decay of incentive value. Ordinarily, incentive salience isassigned only to stimuli that have been paired with pleasure. But brain manipulations (st~ch as drugs or electrical brain stimulation) maycircumvent pleasure, by activating the neural substrate of incentive salience directly. This will result in fhe attribution of incentive salience toassociated stimuli and actions and result in their becoming ‘wanted’, even in the absence of pleasure. This can be considered a kind of ‘shamreward’ (see glossary entry for reward). Sensitization of the neural substrate for incentive saiicncc will icad to pathological ‘wanting’ (craving) forstimuli :tssociated with its excessive activation (e.g., those involved in drug taking), even if this produces little or no pleasure. As mentionedabove, the direction of incentive salience by associative learning is the primary determinant of exactly which stimuli become craved. Thus, in theaddict, drug-p;iircd stimuli, which have been experienced repeatedly in association with the excessive stimulation of dopamine systems, becomethe nearly exclusive t;lrgcts for the attribution of irlcentive salience. Other contributions of associative icrlrning are also possible in this model.For example, the ple:tsure elicited by a UCS can change with repeated cxpcricncc, as when one dcvciops an apprcciativr paiitte for Sccltchwhiskey (this is indicated in the model by the arrow from learning to the ‘pleasure integraMr’). Also, a CS that hirs been rcpt’ittediy paired withpleasure can come to itself elicit subjective pleasure, as in the example of a conditioned ‘high’ rcportcd by ‘needle frciiks’ Gtrrows from the CS tothe ‘pieasurc integrator via associative learning). But we suggest thcsc effects are scpariite from the iittribution of incentive saiicnce and thatthey have only a relatively weak influence 011 nMivated behavior, ccmpmcd to the craving pmduced by the ;ittributic>n of excessive inccnlivesalience. i?naliy, we sugg:cst that none of the psychological proccsscs descrihcd in this motlci. cscept for subjective ‘wanting’ (craving) andsubjective plcasure, are apparent to conscious awareness. The interaction among incentive siilit‘nce, picasure i11111 associative learning is notavailable to introspection. 0niy the final products of the interaction arc intcrprcted by cognitive mechanisms (not shown in the figure. see Nisbcttand Wiison’Lx) as subjective ‘wanting’ and ‘liking’. For an addict , whc)sc IICJJlsilI SlJhStriJtt?S of inccntivc siilicnce have been scnsitizcd, the

subjecGve product is dominated by the intcnsc cxpericncc of drug crilving.

264

flails, etc.) that rats emit to tastes119. Much like thefacial expressions that human infants display to sweetor bitter tastes 319, these hedonic and aversive reactionsreflect the perceived pleasure or displeasure of a tastesensation. The reactions of rats to taste are altered bymany of the same things that control human percep-tions of taste pleasure. The sensory pleasure of sweet-ness to humans, for example, is enhanced by hungerand suppressed by caloric satiety41v4*. Hedonic reac-tions of ratshunger and

to sweet tastes are similarly enhanced bYsuppressed by satiety26v43. The taste pl ea-

sure of a palatable food for humans can be abolishedand replaced with subjective aversion by associativepairing of that food with gastrointestinal illness*‘*.Similarly, hedonic reactions of rats to sweetness areabolished and replaced by aversive behavioral reactionsafter associative pairing of that taste with illness1’8.Finally, drugs that affect opioid or GABA neurotrans-mitters can enhance or suppress the hedonic reactionsof rats to tastes in ways that should be expected basedon current theories of the role of these neurotransmit-ter systems in taste pleasure75~230~340.

Application of the ‘taste reactivity paradigm’ to therole of mesotelencephalic dopamine systems in tastepleasure and motivated behavior leads to the conclu-sion that dopamine systems do not mediate taste plea-sure, although they do mediate the incentive motiva-tional properties of foods. There are three lines ofevidence leading to this conclusion. (1) Dopamine an-tagonists do not decrease the sensory pleasure of tastes,measured by hedonic reactions, although they can de-crease their incentive value. Conversely, dopamine ag-onists do not increase the sensory pleasure of tastes,although they can increase their incentive value340. (2)A bilateral neurotoxic lesion (6-OHDA), which de-pletedate

S dopamine from the nucle us accumbens and cau-nucleus, does not diminish the hedonic evaluation

of tastes, even though it completely abolishes the moti-vation to eat and renders natural incentives ineffective -(ref. 28 and unpublished data). (3) Activation of themotivation to eat by electrical stimulation of the lateralhypothalamus, which is mediated in part by dopaminesystems, does not potentiate the hedonic evaluation oftastes*‘.

These experiments suggest, therefore, that the roleof dopamine systems in behavior motivated by food isnot to enhance the sensory pleasure of tastes. Or, putanothersystems

way, these experiments suggest that neuralmediating ‘wanting’ food can be dissociated

from neuraldopamine ac

systemstivates ’

mediating ‘liking’ food and thatwanting’. ‘This is what would be

expected if dopamine mediates the salience of incen-tive stimuli, rather than the sensory pleasure evoked by

incentive stimuli (Fig. 2). These findings are preciselywhat would be expected on the basis of our hypothesisthat dopamine systems mediate the incentive motiva-tional effects of drugs and are dissociable from otherneural systems that mediate the subjective ‘pleasurableeffects of drugs and other stimuli. This kind of disswi-ation would explain the findings reported by Fischmanand Foltin95 and by Lamb et aLJ9* (see above). YOUwill recall that in these studies a low dose injection ofcocaine or morphine motivated drug-taking behavior in‘postaddicts’, but did not produce self-reported subjec-tive pleasure.- The second line of evidence that dopamine mediatesincentive salience rather than sensory pleasure comesfrom a series of electrophysiological experiments onthe conditions under which dopamine neurons dis-charge in behaving animals288. When monkeys are firstexposed to a novel situation dopamine neurons dis-charge to new, unexpected stimuli that produce orient-ing behavior. But these neuronal and behwioral re-sponses soon habituate199. Dopamine neurons also re-spond when an animal encounters a natural incentive,such as when it touches food located out-of-sight or, ina learning task, when liquid in delivered to themouth199*287. However, when a neutral stimulus (e.g.,light) is paired with the availability of a natur:ll incen-tive, dopamine neurons soon stop responding to thenatural incentive and start to discharge most viiorouslyin response to the newly established conditioned incen-the stimulusf 199 (also see ref. 217). Dopamine neuronsdo not discharge when the animal actually tsrlts thefood, which they should if dopamine mediated thesensory pleasure associated with incentives (ref. 330 aswell). Similarly, Kosobud et al.“” reported in a recentposter that in rats trained to bar press for sucrose,VTA unit activity increased prior to the presentationof sucrose. Dopamine neurons did not increase activityafter sucrose presentation, when presumably the ani-mal would experience sensory pleasure. Finillly, theactivity of dopamine neurons can be dissociat4 fromnon-incentive aspects of a situation and from the de-tails of motor behavior, because their discharge> is notcoupled to “mnemonic or preparatory reprcsen t ;I tionaltask components” (ref. 198, p. 337), to the execution ofreaching movements to obtain and retrieve food or to alight unassociated with food1YY*‘“7*‘“9. Studies in catsalso suggest that VTA dopaminc neurons do not dis-charge in relation to most phasic r1iovt’inents2”“.-““.

In summary, dopamine neurons discharge cinderconditions consistent with an attribution of inc+c’ntivesalience hypothesis of dopaminc function”? Theychange their rate of discharge to a stimulus ~1s theincentive value of the stimulus cl~anges; as the stirnub

265

1mx)111cs more or less salient. In&cd, Schuk? con-cludcd that their c1cctrc,pl~ysiological cxpcrimcnts “arcconsistent with the conclusion that dopamine neuronsrespond specifically to snkr~ s~+lrtli that have alert-ing, arousing and attention-grabbing proper-tics” (p.1.34, our italics) #.- A third line of evidence that dopamine mediatesinccntivc salicncc rather than sensory pleasure comesfrom recent studies using high speed chronamperome-try to measure nucleus accumbcns dopamine activityduring i.v. self-administration of heroin or cocaine”s*‘72.In trained animals the first few drug injections greatlyelcvatcd a dopaminc-related electrochemical signal.However, detailed analysis of the time course ofchanges in the dopamine-related electrochemical sig-nJ, relative to subsequent injections, revealed that thedopamine-related signal increased prior to drug self-administration, peaked at the time a response wasinititated and then significantly decreased immediatelyfollowing drug infusion. This is consistent with the viewthat dopamine mediates the incentive salience at-tributed to a drug-associated stimulus (presumably thelever in this case), because as extracellular dopamineincreased drug ‘wanting’ would increase, to the pointth;it an animal would initiate another drug infusion.The results are not consistent, however, with the viewthiit dopamine mediates subjective pleasure (in whichcase dopamine should rise after drug administration).N&her are they consistent with the view that drugresponding is initiated by withdrawal symptoms associ-ated with dopamine depletion, as suggested, for exam-ple, by Dackis and Gold? A dopamine depletionhypothesis would predict that animals should initiate adrug infusion when the dopamine signal is at its lowest,not highest. Of course, a dopamine depletion/withdrawal hypothesis, such as that proposed by Dackisand Goldh4, also suffers from the shortcomings associ-ated with all negative reinforcement hypotheses ofaddiction, which were discussed above.- The fourth line of evidence that dopamine is morelikely to produce enhanced incentive salience thanenhanced drug pleasure comes from a consideration ofthe relative patterns of change in drug ‘wanting’ vs.drug ‘liking’ reported by human addicts during thegradual development of addictive behavior (Fig. 3).The pleasure induced by drugs does not increase if thedose of a drug is held constant over the course ofrepeated administrations. In fact, if pleasure changes

Y ---

incent ive Va lue

Subject ive PleasureI. r ’ I. I., -

Time -wFig. 3. A schematic illustration of the hypothetical relationshipbetween changes in the incentive value of drugs and drug-relatedstimuli (drug ‘wanting’) vs. the subjective pleasurable effects of drugs(drug ‘liking’) during the development of an addiction. The develop-ment of an addiction is characterized by an increasing dissociationbetween the incentive properties of drugs, which gradually increaseand the subjective pleasurable effects of drugs, which are shown hereto slightly decrease (develop tolerance; but also see text and Note 5

in Ch. 6).

at all it decreases with repeated administrations (al-though, see Note 5 in Ch. 6). If increased synapticactivity in sensitized dopamine systems were the neuralsubstrate of pleasure, a given dose should producemore and more pleasure with repeated experience,rather than less and less (or no change).

Although the pleasure produced by a drug does notincrease for human addicts, the craving for the samedrug does increase with repeated experience (Fig. 3).An increase in ‘wanting’ drugs, as evidenced by self-re-port and by progressively more compulsive drug-seek-ing and drug-taking behavior is, of course, the definingcharacteristic of drug addiction. An increase in drugcraving without an increase in drug pleasure cannot byexplained on the assumption that sensitization en-hances a dopaminergic mechanism mediating the sub-jective pleasure of drugs (see Note 3 in Ch. 6). But thisincreasing dissociation between drug ‘wanting’ and drug‘liking’ is precisely what would be expected if enhancedactivity at dopamine synapses were the neural sub-strate responsible for incentive salience (see Note 4 inCh. 6).

5. ELABORATION OF THE INCENTIVE-SENSITIZA-TION THEORY AND ITS IMPLICATIONS

In this section we will elaborate more precisely howthe major features of addiction are explained by theIncentive-Sensitization Theory of Addiction and dis-cuss implications of the theory. To summarize the

* SchultzZnn tentatively suggested that the common psychological process underlying the discharge of dopamine neurons may be “motivationalarousal”, but also noted this was not completely satisfactory because “dopamine neurons respond every few seconds to the same stimuli overseveral tens of trials -without habituation*~*275*2”7~289, whereas arousal should be a longer lasting state not repeatedly induced at such shortintervals” (p. 135). We suggest that the conditions under which dopamine neurons discharge are better described by the hypothesis thatdopamine neurons attribute incentive salience*‘?

266

central elements of the theory, we posit the following.(1) Many addictive drugs have in common the ability toenhance mesotelencephalic dopamine neurotransmis-sion and to engage related structures (but see Note 4 inCh. 6). (2) A psychological function of this neuralsystem is to attribute incentive salience to the percep-tion and mental representation of stimuli and actions,especially those that have been associated with activa-tion of the system; to cause them to become highlysalient, attractive and ‘wanted’. (3) Repeated and in-termittent administration of addictive drugs leads toincremental neuroadaptations in this neural system,which render it increasingly and perhaps permanentlyhyperresponsive (sensitized). (4) Associative control ofthis sensitized neural system causes tremendously en-hanced incentive salience to be attributed to the act ofdrug taking and to stimuli associated with drug taking(i.e., to the acts and stimuli most closely associatedwith hyperactivation of dopamine systems); and thus,in the addict, drugs come to be pathologically ‘wanted’

(craved). (5) Sensitization of the neural system respon-sible for incentive salience can motivate addictive be-havior (compulsive drug seeking and drug taking) inde-pendent of other motivating factors, such as the expec-tation of drug pleasure or the aversive properties ofwithdrawal. The associative targeting of sensitized in-centive salience to drug-related stimuli results in thepersistence of addictive behavior even in the face ofmany disincentives: for example, the loss of reputation,job, home and family (see Note 6 in Ch. 6).

We have argued that addictive behavior is motivatedby the excessive ‘wanting’ of drugs (drug craving).Therefore, the first major issue we need to address is:why do addicts crave drugs independently of drugpleasure and withdrawal? What is the neuropsychologi-cal process that results in obsessive craving for drugs,leading to compulsive drug-seeking and drug-takingbehavior, even when the drug may produce little plea-sure? The second fundamental question we need toaddress concerns why drug craving persists for so longafter the discontinuation of drug use and after thecessation of withdrawal symptoms; i.e., the nature ofrelapse. After this we will discuss the implications ofIncentive-Sensitization for individual differences in thepropensity to addiction, for therapy and the relation-ship between Incentive-Sensitization and other viewsof addiction.

5.1. The independence of dr-rrg craGg from drug plea-SW-C md withdrawal

The Incentive-Sensitization Theory of Addictionprovides an unique ncuropsychological cxpkation fordrug craving. Drug craving is the subjective experience

that accompanies the attribution of excessive levels ofincentive salience to drug-related stimuli (or their men-tal representations), due to sensitization of dopaminesystems (see Note 4 in Ch. 6). Thus, drug craving isconsidered by this hypothesis to be a psychologicalprocess that is distinct from conditioned withdrawalsigns and from either drug-induced pleasure or a con-ditioned ‘high’. Other views of addiction often considercraving to be identical with or the direct result of,either conditioned withdrawal or a conditioned ‘high’.For example, Childress et al.” state: “We have oftenused the terms ‘conditioned withdrawal’ and ‘con&tioned craving’ almost interchangeably, with the as-sumption that craving might be a form of mild with-drawal” (p. 38). But Childress et al?’ go on to say:“Our patients did not, however, always subscribe tothis position; reports of craving usually showed lowcorrelations with reports of withdrawal. To paraphraseone indignant user, ‘No, doe, craving is when you wantit - want it so bad you can almost taste it... but youain’t sick... sick is, well, sick’ “. To this addict with-drawal sickness clearly is separable from drugcraving .*I* Neither is craving equivalent to a condi-tioned ‘high’, because reports of conditioned ‘highs’are uncommon and are thus dissociable from the morefrequent occurrence of both conditioned withdrawalsigns and conditioned craving, as discussed earlier50*229.

The Incentive-Sensitization view of addiction is inagreement with the indignant user cited above; cravingis pathological ‘wanting’. It is not due to sickness. It isdistinct from both the unpleasant symptoms of with-drawal and from drug pleasure. This view is supportedby studies directly relating self-reported cr:iving in-duced by exposure to drug-associated stimuli to ratingsof withdrawal-like symptoms, drug-like effects and‘outcome expectancies’8”T253. Although exposure todrug-related stimuli produced a significant increase inself-reported craving, as well as drug-opposite anddrug-like effects “in a simple additive model the com-bined effects of positive outcome expectancies, cue-specific dysphoria and cue-specific drug-positive reac-tions were able to predict 28% of the variance incue-specific craving.“... “A much larger proportion ofthe variance in craving remains unexplained bv theseMfactors” (ref. 253, p. 1142-l 143). We suggest the rea-son there is only a weak relationship between thesevariables and drug crnving is that t/ley no IUV causecralkg. Craving is due to excessive activity in 3 sepa-rate and scnsit izcd rtcur-onaI system that rncdi;l tes theattribution of snlicncc to incentives. This is a nwona~systont that normally mediates the ‘wanting’ of thingsin the environrncnt. Athough this ncuronal systemusually functions in concert with neuron al systc ms that

267

medlink

iatethese

‘liking’) , in the add ict the normalsystems is disrupted and pat hologi-

cal levels of ‘wanting’ become dissociated from ‘liking’.We think this dissociation accounts for the unusualpsychological profile of an addict: intcnsc drug cravingseparated from the normal pleasures and punishmentsof life *.

In this light the irrationality of addictive behavior,which is discussed so eloquently by Falk et al.‘“, startsto make some ‘sense’. The irrationality of the behavioris due to an increasing dissociation between the inccn-tive properties of drugs (incentive salience) and theirsubjective pleasurable effects. Because the process ofsalience attribution can be activated, independent ofsubjective pleasure, incentive salience can be strongeven if pleasure is weak or absent. This is one reasonwhy there is not always a strong correlation betweenthe incentive motivational properties of drugs and theirhedonic properties. It is also why people will self-ad-minister low doses of drugs that do not produce subjec-t ive pleasure 95*192 Furthermore, the attribution of in-

centive salience is not a conscious process and theintrospective experience of ‘wanting’ qr craving is onlya person’s interpretation of the outcome of that pro-cess. Much of the time the attribution of incentivesalience may be more implicit than explicit***. Regard-less, the add.ict can be only subj ectively awa.re of theoutcome of excessive incen.tive salience at tribution,craving. The addict may have little insight into thereason for the craving and indeed, may himself bebewildered by its intensity. At a conscious level addictsmay recount all of the negative consequences of con-tinued drug use, deplore their situation, even commentthat the drug does not continue to give great pleasure- and not understand why their craving persists.

5.2. The der vlolmmt of addictil v bcha LiarThe Incentive-Sensitization Theory can explain why

the development of an addiction is typically a gradual,progressive process. The attribution of a high level ofincentive salicncc to drug-associated stimuli and thepleasurable effects of drugs, increase the probabilitydrug-related stimuli will attract attention and that drugswill be sought out in the future. If drug use continues,dopamine systems become progressively more sensi-tized. With each repetition greater and greater incen-tive salience is attributed to drug-related stimuli andthe associative pairing of drug-related stimuli with theintense activation of dopamine systems produced bydrugs leads to an increasing focus of salience attribu-tion upon just these stimuli. Thus, ‘wanting’ is gradu-ally transformed into craving, drugs become craved tothe relative exclusion of all else, and drug-associatedstimuli elicit this craving independent of any pleasurethey produce. In short, the developing addict comes to‘want’ drugs more and more because drug-relatedstimuli become imbued with greater and greater incen-tive salience, even though at the same time drugs maybe ‘liked’ less and less.

5.3. Relapse: drug-induced drug crakzgDrug craving sometimes remains high or is even

increased immediately after drug administration, whenthe drug is producing subjective pleasure, as has beenreported for alcohol, cocaine, heroin and hydromor-phone 80,96,150,2 15,325 This is the proverbial drink thatwhets the appetite and leads to relapse. Why shouldthis be? As pointed out earlier, this is not consistentwith a negative reinforcement view of craving329, be-cause the drug should eliminate withdrawal symp-toms **. Neither is it consistent with a pleasure-seek-

* Other brain manipulations also are known to ‘fracture’ behavioral or psychological subsystems whose operations are so intertwined that one’ssubjective experience is of only one process. An example is the phenomenon of ‘blindsight’350, which refers to the ability of people renderedblind by an occipital cortex lesion to accurately localize visual stimuli presented in their blind visual field, despite having no consciousawareness of perceiving any stimulus. People have no subjective experience of separate psychological processes (and neural systems)underlying the identification vs. the localization of visual stimuli. But an occipital lesion ‘fractures’ these processes, revealing two distinctpsychological processes where there appeared to be only one. Another example is the dissociation of declarative (explicit) and procedural(implicit) memory systems seen following damage to the medial temporal lobes 318 We are not subjectively aware that distinct neural systems.are involved in learning to solve puzzles (for example, the Tower of Hanoi puzzle) and learning facts (for example, learning a list of words).But following a bilateral medial temporal lobe lesion the ability to learn and remember facts is lost, whereas the ability to learn andremember puzzles is left intact; although the latter occurs in the absence of conscious awareness. We think that the dissociation between‘wanting’ and ‘liking’ seen in addicts represents a ‘fracture’ of psychological processes akin to ‘blindsight’ and the implicit/explicit memorydistinction. That is, repeated drug use changes the brain, as a lesion changes the brain, revealing two distinct psychological processes wherethere subjectively appeared to be only one. Furthermore, like the localization of visual stimuli and procedural learning, the attribution ofincentive salience is ‘implicit’; it often may occur in the absence of conscious awareness.

* * Except according to Solomon’s opponent-process theory 3L2, which is the only negative reinforcement theory that can successfully explainrelapse induced by re-exposure to the drug. According to Solomon’s theory, re-exposure to the drug elicits a moderate a-process or drug-likeeffect, which in turn triggers the still strong b-process or drug-opponent effect. The problems faced by the opponent-process theory, however,are: (1) it relies entirely upon withdrawal symptoms to motivate addictive behavior and is thus liable to the general criticisms that wedescribed for negative reinforcement theories of addiction; (2) it posits the growth of an opponent-like process during addiction, for whichthere is no direct evidence other than the phenomena of tolerance and withdrawal themselves (and which is contradicted by evidencediscussed above that the incentive properties of drugs show sensitization rather than tolerance); and (3) it posits the opponent-process to beelicited only and always by the a-process, whereas evidence exists that withdrawal and drug pleasure have separate, independent neuralsubstratesM5.

2 6 8

ing view, because a dose sufficient to produce subjec-tive pleasure should satisfy or at least reduce thecraving; not exacerbate it. Drug craving at the time ofdrug taking is consistent, however,Sensitization view. According to I

with an Incentive-ncentive-Sensitiza-

tion craving is the subjective experience associated withince ntive sa.lience attribution. Because many addictivedrugs increase dopaminecentive salience, one woulto produce drug ‘wanting’.

Indeed, there is considerable evidence that

activity, which produces in-d expect drug administration

re-ex-posure to drugs can reinstate compulsive drug-seeking

.and drug-taking behavior83v325*326. As pointed out byStewart et al.326: “ The idea that ingestion of a formerlyabused drug induces a strong motivational state orcraving for the drug and that it retains the ability toreinstate this craving over an indefinite period of absti-nence from the drug is not new. One of the basictenets of Alcoholics Anonymous (anonymous, 1939) isthat people who have at one time shown uncontrolleddrinking and physical dependence are permanently un-able to drink moderately; one drink is said to elicit anurge to have another” (p. 257). This phenomenon isusually explained by ‘priming’ and Stewart et a1.326have argued that ‘priming’ reinstates drug use because“the presence of the drug in the body (not its absence)activates appetitive motivational mechanisms that areinvolved in the reinitiation of drug seeking behavior”(p. 253). An involvement of dopamine in priming issuggested by reports that the infusion of morphinedirectly into the ventral tegmental area is sufficient toprime responding for i.v. heroin or cocaine, intra-accu-mbens amphetamine can prime responding for i.v.heroin325q326, and haloperidol prevents priming foramphetamine83. It is also relevant to note that cravingin long-term abstinent cocaine abusers (humans) hasbeen associated with elevated plasma and CSF levels ofthe dopamine metabolite, HVA173*206.

Stewart’s view is entirely consistent with the Incen-tive-sensitization Theory of Addiction presented here.We would add only two additional points. First, anIncentive-Sensitization view of addiction identifies the“appetitive motivational mechanism” mentioned byStewart et al.“*” specifically as incentive salience andnot, for example, drug pleasure. Second, we hypothe-size that the ability of drugs to produce incentivesalience is progrcssivcly increased (sensitized) by re-peated exposure to drugs because drugs sensitizemcsotelcnccphalic dopamine systems. Thus, in highlysensitized individuals, such as addicts, relapse is therule rather than the exception, especially after a prim-ing ‘ taste’, because this acts on a hypcrscnsitivc neuralsystem that mediates incentive salience - &citing

pathologically strong ‘wanting’ (craving) and thus re-lapse.

5.4. Relapse: interactions between different drugs and theeffects of drug-related stimuli

Not only can the preferred drug of abuse reinstateaddictive behavior, but often other drugs can as well;and addicts usually use more than one drug (which issignificant in itself). In animals too, priming can Occuracross drug classes. For example, i.v. amphetamine &bromocriptine can prime the self-administration ofheroin367 and i.v. morphine can prime responding inanimals trained to self-administer cocaine326. Further-more, dopamine systems have been implicated in prim-ing between drug classes. An intra-accumbens injectionof amphetamine, which selectively activates themesolimbic dopamine system, can prime respondingfor i.v. heroin327 and intra-ventral tegmental area mor-phine can prime responding for i.v. cocaine?

According to the Incentive-Sensitization view of ad-diction, drugs can prime responding for each otherbecause the same dopamine systems are activated byeach and dopamine mediates the incentive salienceattributed to many different drugs. Therefore, ifdopamine systems become sensitized by past drug useit would be expected that a second, novel drug wouldbe able to prime responding and precipitate relapse, aslong as the second drug also activates hypersensitivedopamine systems. This idea was proposed previouslyby Stewart and Vezina327, who argued that the abilityof opiates and stimulants to prime responding for oneanother “may be related to the ability of ophtes andstimulant drugs to cause sensitization” (p. 2$7), withindopamine systems. In support of this view, cross-sensi-tization has been reported to the psychomotc>r stimu-lant effects and to the incentive motivational effects ofa number of drugs, as well as in the ability ot‘ drugs toelevate dopamine neurotransmission (see hove forrefcrcnces).

Not only may re-exposure to drugs thems&es preycipitate relapse, but environmental stimuli iissociatedwith drugs are known to induce craving and precipitaterelapse in humanssO~**’ and prime drug resporlding inani*nals32s7326. This evidence has been re\ticwed byStewart et al.326 and need not he reiterated here. It isconsistent, however, with an Incentive-Sewitizationview of addiction. In this view scnsit ization of ;i neuralsubstrate responsible for inccntivo salience becomesexpressed as addictive behavior Iqcly by crhmci~gthe incentive properties of stinluIi associated withdrugs. Heightened inccntivc salience is fowssed onthcsc stimuli and their mental representations by asso-ciative learning processes and they become the elicitors

2 6 9

and objects of craving. Even in the absence of the drugand long after withdrawal signs have faded, drug-rc-lated stimuli remain potent conditioned incentives ableto elicit the attribution of incentive salience. Indeed, inthe contest of an Incentive-Sensitization view of addic-tion this is why environmental stimuli associated withdrugs are extremely effective in precipitating relapse in;lddicts (howcvcr, see Note 8).

This view is an alternative to both the conditionedl high’ interpretation and the ‘explicit memory intcr-prctation of relapse discussed above in the section onpositive reinforcement/euphoria theories. We suggestthe effects of drug-related stimuli on relapse are indc-pendent of both drug cue elicited feelings of pleasure(a conditioned ‘high’) and of the explicit memories ofdrug taking such cues might elicit. You will recall weargued that an ‘explicit memory’ hypothesis of relapseplaces an extraordinary explanatory burden on theassumption that relapse-provoking memories are quali-tatively different from the myriad other memories ofdrugs that do not provoke relapse. We suggest, to thecontrary, that conscious remembering in response tocues may be essentially similar to explicit memoriesthat have gone before, neither more vivid nor qualita-tively different. The difference in the processes trig-gered by an effective drug-paired context, which resultsin relapse when earlier memories did not, may be inassociative incentive systems that are not explicitlya\*ailable to consciousness.

Incentive salience is such an associatively triggeredprocess. It occurs in the absence of awareness and itsoperation requires no qualitative difference in explic-itly conscious memory in order to provoke relapse.Unlike explicit memories, the attribution of incentivesalience is an implicit process. It is governed by thelaws of associative learning and is influenced by factorsthat control other forms of implicit learning. Chiefamong these controlling influences is the gating role ofassociative context 259. Context refers to the entire con-figuration of situational stimuli in which the CS hasbeen learned. Associative context can modulate theeffectiveness of any CS. Why then should relapse occurat a particular moment, rather than during earliermemories or earlier encounters with drug-paired stim-uli? Presumably because of variations in the complete-ness of the’associative context. The greater the extentto which contextual factors, such as mood, environmentand other situational variables, mimic the context ofprevious drug taking, the more likely relapse will occur.

5.5. Relapse: the role of stressRelapse to compulsive drug use is not always precip-

itated by re-exposure to a drug or even by specific

environmental stimuli associated with drugs, but some-times by ill-defined environmental circumstances; in-cluding mood changes evoked by stress4’3**? A tradi-tional view of why stress may lead to relapse is that itprompts ‘escape’ from an unpleasant situation via drugtaking. An alternative possibility is that sensitization ofincentive salience could play a role in stress-inducedrelapse because addictive drugs and stress both acti-vate dopamine systems and both sensitize dopaminesystems 8,9,160,268 . As discussed above, animals previouslyexposed to drugs such as amphetamine, cocaine ormorphine are later hyperresponsive to stress and ani-mals exposed to repeated intermittent stress are laterhyperresponsive to the psychomotor stimulant and in-centive motivational properties of drugs243. Accordingto an Incentive-Sensitization view stress may inducecraving and relapse because, by activating dopaminesystems, stress would magnify the incentive salienceattributed to environmental stimuli. Environmentalstimuli that were especially potent as incentives, suchas drug-associated stimuli for addicts, would be thefocus of enhanced salience due to their associativehistory. Drug-associated events would become espe-cially craved again as a consequence of stress (see Note9 in Ch. 6).

It is interesting to speculate that the converse se-quence of events could also occur. That is, prior expo-sure to repeated intermittent stress may predisposesusceptible individuals to drug addiction by sensitizingthose neural systems that mediate the incentive moti-vational effects of drugs69*‘59v268. In such individuals theincentive motivational effects of an initial drug experi-ence may be significantly enhanced because of drugaction on a previously sensitized neural substrate. Thiswould increase the probability that these individualswould show subsequent drug-seeking and drug-takingbehavior. Indeed, experimental evidence for such aphenomenon has been reported by Piazza et al.243, whofound that past experience with stress (repeated tailpinch) facilitated the subsequent acquisition of am-phetamine self-administration behavior in rats.

5.6. Individual differences in the propensity to addictionThe last feature of addiction we will discuss con-

cerns the fact that the majority of people in thiscountry at some point experiment with drugs, but mostdo not become addicts. For example, over 55% of18-34 yr olds have at one time sampled illicit drugs(e.g., marijuana, inhalents, cocaine, heroin or hallu-cinogens; NIDA National Household Survey on DrugAbuse, 1991). Why do the vast majority of these peoplenot develop an addiction? Why are some individualsmore susceptible to addiction than others? Social fac-

270

tors are important, of course, but even persons fromvery similar backgrounds differ greatly in their ten-dency to develop addictive behavior. If, as proposedhere, drug-induced neuroadaptations underlying sensi-tization play a central role in the development ofaddiction: (1) there should be large individual differ-ences in the susceptibility to sensitization and (2) indi-vidual differences in the susceptibility to sensitizationshould be related to the propensity to addiction.

There are indeed enormous individual differences inthe susceptibility to sensitization, a point that has beenemphasized by a number of researchers190*194v268v294.Some of this individual variation is due to geneticvariation, because among animals there are markedstrain differences in the susceptibility to sensitization.Strain differences in both rats and mice have beenreported in the sensitization produced by repeatedtreatment with amphetamine110*194~2”8*300, cocaine304,ethano1~,‘42,‘40 and morphine302. There are alsomarked strain differences in mesotelencephalic dopa-minerelati

systems14ng strain

,91,145

differbut we know of no studies directly

ences in the susceptibility to sensi-tization to strain differences in mesotelencephalicdopamine systems. Nevertheless, it is important thatinitial studies with recombinent-inbred lines of micesuggest that the genetic determinants of acute respon-

aresiveness to drugsfor susceptibility to

dissociable fromsensitization304.

responsiblethoseMany behavioral

genetic studies on drug responsiveness have focussedon variation in the acut e response to drugs, not suscep-tibility to sensitization. But the Incentive-SensitizationTheory suggests that the susceptibility to sensitizationmay be most relevant for the development of addictivebehavior and therefore, information on genetic factorsleading to high susceptibility to sensitization may be ofparticular importance in understanding the genetics ofaddiction.

A number of other factors have been reported toinfluence individual differences in the susceptibility tosensitization including, agelf)o*l75v303, sex45*46,107,250,267,273

andWhe

hemispheri.ther the in

c differences in dopamine systems’?fluence of th ese variables on the sus-

ceptibility to sensitization is causally related to thepropensity to self-administer drugs or to related varia-tion in dopamine systems is not yet known, althoughcorrelative relations have been reported47~10H*‘~*294.

There arc, howcvcr, a number of intcrcsting studieson behavioral traits that do predict both the suscepti-bility to sensitization and the propensity to self-admin-ister amphctamine24’. For example, responsivity tonovelty is reported to predict susceptibility to sensitiza-tionJ33-135,242. Similarly, animals that eat and drink inresponse to electrical stimulation of the lateral hy-

pothalamus (not all do) show an enhanced susceptibil-ity to amphetamine sensitization220*221. Most impor-tantly, these same traits are correlated with a propen-sity to acquire amphetamine self-administration24’ andwith differences in dopamine dynamics in the nucleusaccumbens369’32. Individual differences in reactivity tonovelty,phetamithe responsiveness of the hypothalamo-pituitary-adre-nal (HPA) axis. An imals that show a high response tonovelty also show a prolonged elevation in plasmacorticosterone in this situation, relative to low re-sponders190*241. That the HPA axis may play a role indrug sensitization is suggested by experiments showingthat activatio n of the HPA axis is necessary to inducesens itization to amphetamine or stress54y55*69, perhapsby the action of corticosterone on glucocorticoid recep-tors26’ (cf. ref. 57) and by a report that repeatedexposure to exogenous corticosterone sensitizes ani-mals to a subsequent amphetamine challenge244.

in amphetamine sensitization, and in am-.ne self-administrat.ion may involve var iation in

5.7. Implications of the ken tice-Sensitiza tion Theoryfor Therapy

There is considerable interest in developing effec-tive therapies for the treatment of drug addiction, butthis has proven to be a very difficult problem. TheIncentive-Sensitization view of addiction may providesome insight as to why effective therapies have beenelusive, and potentially, may point the way to thedevelopment of more effective approaches.5.7.1. Extinction training. A recent trend in the psy-chotherapeutic treatment of addiction is based on therecognition that drug-conditioned stimuli are very po-tent in eliciting craving and precipitating relapse229.There have been attempts, therefore, to ‘extinguish’conditioned responses to such stimuli. Indeed, the re-peated presentation of drug-related stimuli, in a labo-ratory setting, results in a progressive declirlc in drugcraving elicited by drug-related stimulisO~“*“. It is inter-esting, however, that some of the autonomic responsesto such stimuli are more resistant to extinction than thesubjective effects, and non-specific changes in moodstate (especially anger) can rapidly reinstate condi-tioned stimulus-induced drug craving49.

In the context of the Incentive-Sensitization Theorythis might occur because the ncuroadaptations underlying sensitization persist, despite extinction of the con-ditioned stimulus control of sensitization. That is, theability of conditioned stimuli to control the cspressionof sensitization may be thought of as tc:u-ning-relatedncuroadaptations 1;lyered ‘on-top’ of the ncuroadapta-tions responsible for sensitization, but which do notdirectly alter or reverse the neuronal changes responsi-

2 7 1

blc for sensitization. Also, contextual factors that con-trol the associative attribution of incentive salicnccmay not transfer bctwccn the clinic and the street.Thus, extinct ion training may extinguish responses tospecific stimuli under specific circumstances, but othernon-target stimuli can still access sensitized neuralsystems mediating incentive salience, as can environ-mental stress. On the positive side, the fact that theexpression of sensitization can be brought under strongconditioned stimulus contro13*” suggests it should bepossible to develop learning strategies to control theoutput of sensitized neural systems. However, the per-sistence of the neuroadaptations underlying sensitiza-tion and their resistance to extinction”28, suggests thatcoping with addiction may be a very long, ongoingprocess. Of course, this has been recognized for manyyears by organizations like Alcoholics Anonymous. Forexample, Anonymous states: “We know that while thealcoholic keeps away from drink, as he may do formonths or years, he reacts much like other men. Weare equally positive that once he takes any alcoholwhatever into his system, something happens, both inbodily and mental sense, which makes it virtually im-possible for him to stop” (ref. 7, p. 22).5.7.2. Pharmacotherapeutic approaches. The Incentive-Sensitization Theory of Addiction also has implicationsfor the development of effective pharmacotherapies.The two major pharmacotherapeutic approaches atpresent either target the treatment of withdrawalsymptoms or involve drug substitution therapy (e.g.,methadone maintenance). The Incentive-SensitizationTheory predicts neither of these approaches will bevery successful in eliminating addictive behavior, be-cause neither target the fundamental neuroadaptationsunderlying sensitization. Of course, many years of ex-perience with opiate addicts have already shown thatthe alleviation of withdrawal is not an effective long-term solution for addiction363 and the drugs used insubstitution ‘therapies’ usually are addictive them-seIves363. Furthermore, some drugs used to treat with-drawal also may induce sensitization89~202~203~277.

An Incentive-Sensitization view of addiction sug-gests that to really ‘cure’ addiction agents need to bedeveloped that directly target and reverse the neuroad-aptations underlying sensitization. There are present-ly a number of agents known to prevent the develop-ment of sensitization, including dopamine antagonists(see above for references) and glutamate antago-nists 161,164,369 (cf. ref. 342). Unfortunately, these com-pounds do not reverse the neuroadaptations underlyingsensitization, but only prevent its development if theyare given every time the addictive drug is given. This isnot a practical approach for the treatment of addiction

(see Note 1 in Ch. 6). The Incentive-Sensitization The-ory of Addiction predicts that an especially effectivepharmacotherapcutic agent would reverse sensitiza-tion-related neuroadaptations. However, to our knowl-edge, no one has identified such a compound. Ofcourse, any rational drug design program will requirethat we know a lot more about the nature of sensitiza-tion-rclatcd ncuroadaptations than we know at pre-sent .

5.8. Relationship between incen tice sensitization and otheruiews of addiction

In closing, we? want to emphasize that theIncentive-Sensitization Theory of Addiction does notexclude other factors that contribute to drug-takingbehavior. For example, the Incentive-SensitizationTheory does not address a number of features of druguse, including why people experiment with drugs in thefirst place (experimental drug use), casual (not addic-tive) patterns of drug use or why people often usedrugs that do not usually lead to compulsive patternsof use (e.g., LSD; see also Note 7 in Ch. 6). TheIncentive-Sensitization Theory of Addiction does notpreclude a role for pleasure-seeking and withdrawalavoidance in drug-taking behavior. These differentviews of addiction are not mutually exclusive, unlessthey are taken as the sole explanations for addiction.There can be no doubt that addiction results from verycomplex interactions amongst social, cultural, eco-nomic, psychological and biological variables. Thesecomplex interactions determine whether experimenta-tion with drugs first occurs, whether further drug use issustained and whether drug use leads to addiction.Exactly which factors motivate behavior will vary overtime and across different drugs. The Incentive-Sensiti-zation view does not exclude the possibility, for exam-ple, that an explicit memory of drug pleasure or aconditioned ‘high’ could contribute to a desire to re-peat the drug experience in some situations, This maybe especially true early in the, development of anaddiction363. Early in the development of an addiction,before marked sensitization has occurred, the memoryof the subjective pleasurable effects of drugs could be amajor factor motivating drug-taking behavior. For ex-ample, it has been suggested that the initial subjectiveeffects of drugs can predict later drug habits’*‘. Alsodrug-taking behavior is influenced initially to a greatextent by social factors, such as peer pressure.

Neither does the Incentive-Sensitization view of ad-diction deny that the unpleasant symptoms of with-drawal could motivate drug taking in some individuals,under some circumstances, in order to relieve symp-toms. The role of withdrawal avoidance may vary greatly

272

from drug-to-drug. In some respects drug taking moti-vated by pleasure-seeking or withdrawal avoidanceseems ‘normal’; that is, the reasons for drug taking areunderstandable to a non-addict. For example, recentsuggestions that drug craving is caused by short-termdopamine depletion after drug use are based largely onthe assumption that craving is merely a rational re-sponse to withdrawal symptoms63*2041351. But by ourview this is not craving; nor is it the fundamentalproblem in addiction. Craving is obsessive, irrational,pathologically intense drug ‘wanting’ for no obviousreason, which leads to compulsive drug-seeking anddrug-taking behavior. Craving is difficult for both theaddict and the non-addict to understand and this iswhat we propose is primarily due to sensitization ofincentive salience.

The Incentive-Sensitization Theory of Addiction isalso compatible with the ‘Psychomotor Stimulant The-ory’ of Wise and Bozarth3? Wise and Bozarth365 ar-gued that addictive drugs have in common the abilityto induce ‘psychomotor activation’, which was pro-posed to be due to activation of a common biologicalmechanism associated with approach behavior and me-diated by dopamine. This biological mechanism isthought to be fundamental in producing rein-forcement 111,343 The fact that the psychomotor-activat-ing effects of addictive drugs are sensitized by repeateddrug administration is certainly consistent with the‘sensi tization’ componen t of the IncentiveSen sitiza.tionTheory. But we further specify here that the psycholog-ical process responsible for ‘reward-related’ psychomo-tor activation is the attribution of incentive salience.Although incentive salience may lead to locomotionand approach, because this psychological process makesstimuli in the environment more salient, attractive and‘wanted’, these functions may be separable260. Forexample, a brain manipulation could induce locomo-tion, perhaps by activating brainstem locomotor pat-tern generators, without producing incentive salience;and incentive salience could be attributed in the ab-.sence of locomotion (for example, in d rat renderedcataleptic by morphine, but who still acquires a condi-tioned place preference). Thus, in our view it is specifi-cally the sensitization of incentive salience that makesdrugs and drug-associated stimuli increasingly attrac-tivejust

and ‘wanted’.a corrclatc of

Increasedscnsi t izcd

psychomotor activation isincentive salience.

In closing, the ability of the Incentive-SensitizationTheory of Addiction to capture the ‘essence’ of addic-tive behavior (compulsive drug seeking and drug tak-ins> can be illustrated by a ‘thought experiment’. Forthe sake of argumenregarding th e criteria

t, imagine that our assumptionsfor an adequate theory of addic-

tion are correct and that the neural system involved inassigning incentive salience to drugs and drug-asswi-ated stimuli is indeed dissociable from those mediatingthe subjective pleasurable effects of drugs. Now hag-ine that repeated intermittent drug use causes gradualand incremental changes in the neural system responsi-ble for incentive salience, such that this neural systembecomes very hypersensitive (sensitized). Further imag-ine that the expression of this sensitized system isfocussed expressly on stimuli that have become associ-ated with its excessive activation, so drugs and drug-as-sociated stimuli become irresistibly attractive (‘wanted’)and thus able to control behavior. But the neuralsystem(s) responsible for the subjective pleasurable ef-fects of drugs either does not change or else becomeshyposensitive (tolerant). Finally, imagine that incentivesalience is attributed in the absence of consciousawareness. Now consider what a creature with thisbrain would be like. An addict, we think.

6. NOTES

6.1. Note 1. Role of control and intermitteucy of drugadministration in sensitization6.1.1. Control. Although there have been many reportsof sensitization to addictive drugs, it is important toacknowledge that in nearly all of these studies sensiti-zation was induced by non-contingent drug treatment.That is, the animal’s behavior had no influence onwhether it received a drug or not. It is known. however,that drugs can produce different effects depending onwhether they are given in a response-contingent orresponse-non-contingent manner3@‘*310. Also, the be-havioral sensitization to amphetamine or cocaine in-duced by footshock stress is influenced by whether ananimal has control in the situation. For example,MacLennan and Maier2’H.’ reported that lxhavioralsensitization did not occur in rats who could controlthe duration of footshock, but did occur in r;its receiv-ing an identical amount of shock, but who had nocontrol. It will be critical to determine, Wrefore,whether the response-contingent administr;ltion ofdrugs induces sensitization.

There have bocn very few studies directly addressingthis issue. We are aware of one report of senGtizationfollowing experience with cocaine self-adminiyt ration.Falk et al.‘” tcstcd for behavioral sensitization by chal-lenging animals with cocaine 7-10 days after the dis-cant inuat ion of an oral cocaine self-administration reg-imen (involving a schedule-induced polydipsi:) para-digm). Cocaine-experienced animals showed a markedshift to the loft ill the dose-response curve for cocaine-induced locomotor ;ict ivity. These dat;l show t tu t scnsi-

273

tizat ion to cocaine can occur following responsc-contingent drug adnlinistration, as well as foLIowingnon-contingent drug administration. It may also berelevant that in self-administration studies animals arcoften given non-contingent drug injections prior totraining or during shaping, because this facilitates theacquisition of a self-administration habit. It is possiblethe development of a drug self-administration habit isfacilitated by these procedures because they producethe kinds of sensitization-related neuroadaptations un-der discussion here (although one also has to considerthe possibility that prior non-contingent administrationproduces tolerance to the aversive properties of drugs).61.2. Intermitt4tzcy. On the other hand, sensitizationwas not found in a microdialysis experiment involvingcocaine self-administration’““. These researchers foundthat the ability of self-administered cocaine to elevateextracellular dopamine was actually decreased in drugexperienced rats. However, in this experiment the dial-ysis test was given 24 h after the last self-administra-tion session. Sensitization-related changes in dopamineneurotransmission often are not evident after suchshort periods of withdrawal, even following non-contin-gent drug administration1sX~176~“95~296~370. This is proba-bly because intermittency is a critical variable both ininducing sensitization and in its later expres-sion 232,248,268,272 If injections are given too close to-gether in time tolerance, rather than sensitization, usu-ally occurs. The development of sensitization is maxi-mized by spacing injections far apart in time (2-3 daysto a week). Similarly, if a challenge injection is givenwithin the first few days after the discontinuation ofescalating dose amphetamine treatment behavioralsensitization is not evident. But if a challenge injectionis given after a longer period of withdrawal, from 1week to 1 year, animals are markedly sensitized2”*.Thus, sensitization may not be apparent after self-ad-ministration regimens that allow animals to maintainelevated brain levels of a drug for prolonged periods oftime, especially if animals are tested soon after the endof a bout of self-administration.

such intermittent schedules of drug administration maygreatly e hancc the rein forcing propcrtics of drugs andcatalyze ‘drug overindu lgcnce in humans”. They s tate“when life’s crucial commodities are in short supplyand available only on intermittent, marginal schedules...drugs can become all-powerful in reinforcing efficacy”(ref. 87, p. 1506). We would only add that the neuralbasis of this effect, in the context of Incentive-Scnsiti-zation, may involve a facilitation of sensitization-re-lated neuroadap tat ions. W e w o u l dmittent pattern of d rug se1 f-admi

expect tha t an intcr-nistration, such as a

cycle of ‘runs’ and ‘crashes’, would produce sensitiza-tion of dopamine neurotransmission and incentive sali-ence 270.

4.2. Note 2. Specific motioational effects of doparnineblockade

The literature on the effects of dopamine antago-nists on motivated behavior has been reviewed exten-sively and the reader is referred to papers cited in thetext for comprehensive lists of citations. In brief, someof the most clearly motivational effects of dopamineantagonists include the following. (a) Mimicry of ex-tinction by dopamine antagonists, in which instrumen-tal responses for a food, drug or electrical brain stimu-lation decline only gradually after neuroleptic adminis-tration as though the reinforcer were no longer effica-cious (the decline in response is not right away anddoes not occur unless the animal is allowed to performthe task)“. (b) The associative reinstatement of instru-mental performance after a dopamine antagonist hassuppressed instrumental responding, by transferring theanimal to a separate task in which it has previouslybeen reinforced, but in which it has not yet experi-enced the drug, and in which it responds again at ahigh level even though the drug is still in effect (i.e., itmust learn to ‘extinguish’ again in the second task)‘O”.oC Transfer between real extinction and extinction

It is intriguing that intermittency is not only criticalfor inducing sensitization, but is thought to play animportant role in the development of many persistentand habitual behaviors, including addictive behavior?Of course, drug-taking behavior in human addicts isoften C :haracte rized by inte rmittency. Drugs are fre wquentlY taken in ‘r uns’ of self-administratio In, in.ter-

mimicry in ‘resistance to extinction’ paradigms, asthough a dopamine antagonist drug were perceived bythe animal as being similar to ‘no reinforcer’. (d)Reward+-specific ‘curve shift’ reductions in psy-chophysical paradigms that can distinguish betweenreductions in instrumental performance due to motorimpairment versus reductions due to decreased rein-forcement (refs. 78, 104 for example). (e) ‘Reducedpalatability’ patterns of decrease in food consumptionthat mimic those produced by manipulations of the

sper sed with ‘crashes’ lastingalso be imposed because

a few days. Intermittencyconsiderable amount of

sensory pleasure of the food: for exampledopamine antagonists mimic the effects

7 l o w doses ofa upon sucrose

time is required to obtain the money necessary to buydrugs. Faik and his colleagues86*87 have argued that

drinking that are produced by dilution of the sucrosesolution 13,53,67,106,308,374 .

274

6.3. Note 3. The neural substrate of drug ‘liking’ (plea-sure)

We have argued that mesotelencephalic dopamineprojections provide the neural substrate for drug‘wanting’ (via attribution of incentive salience), but notfor drug ‘liking’ (for the subjective pleasurable effectsof drugs). This naturally raises the question: what is theneural substrate for drug ‘liking’?; for the pleasurableaffective states produced by addictive drugs? We can-not provide a definitive answer to this question, exceptto rule out dopamine, but we can point to othercandidate neural systems. Chief among these are en-dogenous opioid neurotransmitter systems. Opioid ago-nists that increase motivated behavior towards food,such as morphine, do enhance the sensory pleasure offood as measured by the taste reactivity paradigm”.Similarly, activation of benzodiazepine-GABA systemswithin the brainstem enhances sensory pleasure by thetaste reactivity measure, although it is not yet clearwhether this depends on an interaction with brainopioid systems (ref. 340 for example). In summary, wedo not know the neural substrate of drug pleasure, butthese are a couple of candidate systems and futureresearch may reveal others.

projections to the ventral striatum (nucleus accumbens),Therefore, to the extent that the sensitization of inan-tive salience is mediated by a specific dopamine systemit is probably the dopamine projection system to theventral striatum. But we do not discount t h c possibilitythat other dopamine systems may play a role.

The second issue is whether only dopamine systemsare involved in the sensitization of incentive salience.We have focused on dopamine systems as the site ofsensitization-related neuroadaptations for the reasonsdescribed in the text. We acknowledge, however, thatthe neurobiology of drug craving involves much morethan just a simple sensitization-related increase indopamine neurotransm ission and th e neu ral substrate

much more complicatedof incentive salience is surelythan this.

For example, Wise and RompresM have cautionedthat “while the evidence is strong that dopamine playssome fundamental and special role in theeffects of brain stimula tion, psychomotor

rewardingstimulants,

opiates and food, the exact nature of that role is notclear. One thing is clear: dopamine is not the onlyreward transmitter and dopaminergic neurons are notthe final common path for all rewards” (p. 220). Weagree with Wise’s caution. To influence motivated be-havior via the attribution of incentive salience,dopamine systems must interact with many other neu-ral systems, especially those involved in hedonics andassociative learning. It is also possible that some incen-

6.4. Note 4. Dopamine, sensitization and incen tiuesalience

There are two issues regarding the neural system(s)responsible for sensitization of incentive salience thatrequire further discussion. The first concerns elabora-tion of exactly which of the many different mesotelen-cephalic dopamine projection systems is likely to medi-ate incentive salience. We are well aware that there aremultiple anatomically and functionally distinct mesote-lcncephalic dopamine projection systems. The extentto which the effects of manipulations of dopaminesystems on incentive motivation are due to an actionon any single one of these dopamine systems is notalways clear. The contribution, for example, ofdopaminc projections to the frontal cortex, septum,caudate, accumbens (core vs. shell), olfactory tubercleor amygdala and the extent to which there may beinteractions between these systems in the assignmentof incentive salience, is, for the most part, unknown.We hesitate, therefore, to prematurely assign all re-sponsibility for incentive salience or the sensitization ofincentive salience, to a specific dopamine projectionsystem. This is why we often use the broader term,mcsotclcnccph;llic dopamine systems. Ncvcrthcless, werecognize that most of the evidence linking mcsotelcn-cephalic dopamine systems to inccntivc motivation pri-marily implicates the so-called mcsolimbic dopamine

tives may not directly activatea1118’ and it is conceivable that

dopamine s\ stems at.some behavior may be

controlled by non-incentive based processes. such asSkinnerian S-R reinforcement, Hullian drive-rciSuctionor goal-directed computational procedures (e.g., Test-Operate-Test-Exit procedures*‘“) (see Toates’-‘9 for adiscussion of how to distinguish some of these alterna-t ives).

The phenomenon of behavioral sensitization alsosurely involves more complex ncuroadaptations thanjust an enhancement in dopamine release, although todate most studieso n dopamine. The

of sensi t iza .tion have focusscd onlycvid cnce that tlw induc-re is already

tiondrug

of sensi tizatha

tion involve s a diffcrc‘ntof scn

cellulx site ofsit ization’60*269,action n the expression

that cross-sensitization does not occur between all ad-dict ivc drugs1h0~269, that an action of drugs on neuralsystems other than dopamine is required to iJldUce

sensitizat ions4~161~1M a n d that t h e r e a r c scnsitiza-tion-rclatcd changes in other neurotransmitt~r sys-tems 122,235,3 13.3 14 . Similarly, Stewart “’ has argwd thatthe conditioned stimulus control of xnsitizatiw mayoccur at diffc ren t syn;\pt ic sit cs,specific actions of differcnt drugs.

275

that sensitization involves nctlrondaptations at multiplesites and in multiple nc’uI-otl-;lnsmittcr systcnts.

Nevertheless, in our prcscnt state of ignorance it is areasonable working hypothesis that the adaptations indopaminc systems described ttcrc, involving an en-hancenttnt in mcsotelencephalic dopantine neurotrans-mission, fornt a critical link in a chain of events leadingto drug craving in addicts. We readily acknowledge,however, that sonte addictive drugs could produce suchincentive-sensitization effects by an action on other, asyet unidentified, neural systents, including, for cxant-ple, output pathways front the ventral striatum to theventral pallidurn’““. On the other hand, it is also possi-ble that the dopantine-independent rewarding effectsof some drug treatntent regimens’“’ involve the activa-tion of this same dopantine-incentive salience circuitry,but at a later stage, ‘downstream’ from dopamine neu-rons. Regardless, we want to emphasize that the Incen-tive-Sensitization Theory of Addiction does not requirethat the sole or et*en primary site of drug-induced neu-roadapta tions responsible for crakg specifically be ondopamine neurons. If it is not, then our assignment ofsensitization of incentive salience to dopamine wouldbe incorrect. Nevertheless, the concept that drug crav-ing develops because of sensitization of incentivesalience could still be fundamentally correct, but itwould be mediated by another, as yet unidentifiedneural substrate.

6.5. Note 5. Tolerance to drug pleasureThe magnitude of the decrease in the subjective

pleasurable effects of drugs is illustrated as being rela-tively small in Fig. 3 because, although the develop-ment of tolerance to the euphoric effects of drugs iswidely accepted in the clinical literature, there is actu-ally very little objective evidence for this. The evidencethat is usually cited is that addicts tend to graduallyescalate their dosage with repeated drug use. “One ofthe most insidious aspects of drug abuse is the seem-ingly inexorable tendency for addicts to increase theirdrug consumption over time” (ref. 86, p. 81). But Falket al? point out there is very little evidence linkingescalation in dose with tolerance to the subjectivepleasurable effects of drugs. They state: “It is com-monly presumed that... as tolerance develops, moredrug must be ingested to satiate the addict’s need forthe drug. It is fascinating that there is little experimen-tal data relevant to this assumption and that whichdoes exist does not support” (p. 81). Indeed, there area number of reports that addicts continue to experi-ence euphoria even after years of drug use193~*11*213.Some studies have suggested there even may be anincrease in the pleasurable effects of morphine in

long-term addicts, because naive subjects usually ratethe effects of morphine as unpleasant, whereas experi-enced users (‘postaddicts’) overwhelmingly rate mor-phine effects as pleasant? This apparent increase inthe subjective effects of morphine is probably not due,however,tolerance

to sensitization to its euphoric effects, but toto its aversive effects; because ‘postaddicts’

also report a lower incidence of nausea and vomitingthan naive subjects.

Why then do addicts typically escalate their dose? Apossible alternative explanation to tolerance of eupho-ria is that ad diets increase dose to achieve the’intense (and more desirable) subje ctive effects

m0repro-

duced by larger doses. They are able to do this onlybecause tolerance develops to the aversive ‘side-ef-fects’ of drugs. That is, addicts increase their dosebecause they can, without the dire negative effectsexperienced by naive users. Doses that might be un-pleasant or even life-threatening in inexperienced users,are ‘tolerated’ by experienced users, because of toler-ance to many of the drug’s negat ive effectseffects on the autonomic nervous system.

‘9 including

On the other hand, short-term tolerance to theeuphoric effects of drugs may play a role in the escala-tion of dose seen when drugs are administered in a‘run’, as is often the case with amphetamine or co-caine. If a large supply of amphetamine or cocaine isreadily available addicts often readminister the drug assoon as the effects of the previous dose beginsdissipate. However, if successive administrations agiven too close together in time the positive effects

toreof

the drug may be ‘masked’ or suppressed by a transientdepression of brain ‘reward’ systems’05~174~204. Aspointed out by Stewart 324: “Tolerance to the rewardingeffects of opiates has been found in expe riments inwhich animals were exposed to the drugs continuouslyprior to place preference training299” (also see ref. 81)and in humans there is a “rapid within-session devel-opment of tolerance to the subjective mood effects ofcocaine, but this dissipated completely within 24 h9”’(also see ref. 95). Therefore, it may be that dose isescalated within a run to overcome this apparentshort-term tolerance to the pleasurable effects of thedrug. But this may not be relevant to the escalation ofdose seen over the long-term, that is, between runs.

In summary, although we indicate in Fig. 3 thatsome tolerance develops to the subjective pleasurableeffects of drugs we are aware that this is a complicatedandtive

largelypleasu

unresolved issue. But whether the subjec-rable effects show some tolerance or no

change with repeated drug administration, the devel-opment of an addiction is still characterized by anincreasing dissociation between ‘wanting’ drugs and

276

‘liking’ drugs. As Fig. 3 illustrates, ‘wanting’ drugs,produced by the attribution of incentive salience to theact of drug taking and to drug-related stimuli and theirmental representations, increases dramatically duringaddiction; ‘wanting’ evolves into craving. At the sametime ‘liking’ drugs does not increase.

6.6. Note 6. The compulsive nature of addictive behaviorIt is interesting to speculate that addictive behavior

may be so compulsive in nature because the neuroad-aptations underlying drug addiction are in some wayrelated to other obsessive-compulsive disorders. It iswell known that hyperactivity in dopamine systemsresults in behaviors with a highly stereotyped (compul-sive?) structure. For example, relatively low doses ofamphetamine or cocaine elicit high levels of locomotoractivity and the pattern of locomotion is abnormallystereotyped224~28s. At higher doses locomotor activity isdiminished as animals engage in highly stereotyped,narrowly focussed, repetitive behaviors254. Similarpatterns of stereotyped behavior (‘punding’) havebeen described extensively in human amphetamineusers2549286 A number of compulsive behavioral disor-ders have been linked to dysfunction in the striatum,including obsessive-compulsive disorder itself, Tour-ette’s syndrome, tic disorders’*~2s5~25” and Huntington’sDisease (ref. 59 and N. Wexler, personal communica-tion). Thus, there are a number of different disordersin which dysfunction in the striatum has been associ-ated with compulsive, repetitive thoughts (obsessions)and actions (compulsions). It is interesting to specu-late, therefore, that some of the neural changes under-lying drug addiction may be, in some respects, similarto those responsible for other obsessive-compulsivedisorders.

6.7. Note 7. Benrodiazepines and sedative-hypnoticsThere are drugs of abuse that do not seem to fit this

profile, namely the bcnzodiazepines (BZ) and barbitu-rates. These compounds have biphasic, effect.3 on be-

. havior, producing mild psychomotor activation at lowdoses and a marked depression of motor activity athigher doses365, but they may not increase dopaminencurotransmission. Although the psychomotor-activat-ing effects of a low dose of diazepam (0.25 mg/kg) hasbeen reported to require the activation of dopaminesystems31 l, microdialysis studies have found that di-azepam and midazolam decrease extracellulardopamine in the nucleus accun~bcns92~14’ (although inthese latter studies rclativcly high doses were used,> 0.5 mg/kg). It may be that the Incentive-Sensitiza-tion Theory does not account for why these particulardrugs are used recreationally. It should be noted, how-

ever, the BZ’s are not very addictive and in normalexperimental subjects they have “little or virtually noreinforcing effects” (ref. 56, p. 142). They do not pro-duce the compulsive pattern of drug-seeking anddrug-taking behavior characteristic of amphetamine,cocaine or the opiates3’! Whether the addictive poten-tial of alcohol can be accounted for by Incentive-Sensi-tization remains to be seen. As cited in the text, thereare reports that alcohol: (1) produces psychomotoractivation, especially in alcohol-preferring strains; (2)increases extracellular dopamine; and (3) producessensitization. But there have been very few studies,they are not all consistent and therefore, more work isneeded to resolve the issue.

6.8. Note 8. The role of dopamine in mediating theeffects of conditioned incentives

The view of how conditioned incentive stimuli evokerelapse proposed here is similar to that of Stewart etal 326, except we hypothesize drug-associated stimuli.evoke craving by activation of a sensitized neural sys-tem that specifically mediates incentive salience. LikeStewart et a1.326 we suggest that this neural systeminvolves mesotelencephalic dopamine projections to theventral striatum and that conditioned incentive stimuliact much like a ‘priming’ dose of a drug itself; produc-ing a small increase in dopaminergic activity. The dif-ference is that we specify the consequence of thisenhanced dopamine activity to be conditioned incen-tive salience, not necessarily a conditioned affectivestate, a ‘high’. We do not deny, however, that condi-tioned pleasure can be elicited seperately, presumablyvia associative activation of separate neural systems.

We need to acknowledge, however, that there hasbeen relatively little research on the role of dopaminein mediating the effects of conditioned incentive stim-uli, especially conditioned incentive stimuli establishedthrough their association with drugs. There is consider-able evidence that activation of the ventral stria-tal dopaminc system enhances responding for condi-tioned incentive stimuli established by pairing a neutralstimulus with a natural incentive, like food orwater44~127,1b~,Ih'),2~~L.323,334 (for reviews see refs. 85, 263).There is Aso a general consensus that dolxlminesystems are critical in the process by which sti-muli acquire conditioned incentive propertiesthrough their ;Issoci;ttion with natual inccntivt’s ord rugs2,20,21,32.7$,77,1 12,12’~,21~,222,344,3S2 (cf. refs. 207,260,317). It has been suggcstcd, however, that once ac-quired, conditioned incentive stimuli may activate be-havior indcpcndcntly of dopamine. This conclusion is

based on rcyorts that dop;uninc rcccptor blockade withpimozidc (or in one espcrimcnt, Moperidol) does not

277

prevent the conditioned psychomotor activation evokedby a conditioned stimulus associated with food’3Y, ord rugs20.2 1,352. although the absolute amount of psy-chomotor activation produced by a food-associated CSis reduced ‘-? ‘This idea is controversial, however, be-cause others have reported that dopamine antagonismdoes attenuate the expression of conditioned ‘prepara-tory behaviors’ (including locomotor activity) producedby a conditioned stimulus signaling food”‘*“**‘” (also seeref. 353), as well as conditioned responses establishedby drugs 76.77,1133130,247 Also, under some conditionsdopamine antagonists may induce gradual extinction-like effects on conditioned responding360*3”4. Such ef-fects may be interpreted by the hypothesis that lowdoses of dopamine antagonists impair the ‘reboosting’of incentive salience to established conditioned stimuli,which could occur each time an incentive stimulus isencountered. This ‘reboosting’ may be essential for themaintainence of a conditioned response*‘.

Direct measures of dopamine neurotransmission also. support the hypothesis that this neural system mediates

the incentive effects of conditioned stimuli3’. Condi-tioned stimuli predictive of food have been reportedto increase the discharge rate of dopamine neu-rons 183,199,287,289 9 to increase dopamine metabolism in

with psychomotor stimulants or opiates are reported toenhance dopamine metabolism’91*234*283, to elevate adopamine-related electrochemical signal associatedwith cocaine’ I5 and in the presence of the uncondi-tioned stimulus, to elevate dopamine in dialysate*‘*.

6.9. Note 9. Stress, aversiveduced psychoses l

stimuli and stimulant-in-

Not only do pleasant natural incentives, such asfood, water and access to a mate activate mesotelen-cephalic dopamine systems51*66~126~214*246~376, but so dosOme presumably unpleasant aversive events, includingclassical stactivating dopam

ressors . St.ine

ressorsprojec

are ptions

cortex and to the shellIntense aversive stimulatal dopamine system’

of the nucleus accumbens1*71*336.i may also activate the nigrostri-

ously neutral stimuli pa i

articulto the

arly effectivem e dial fron

in.tal

Conditioned stressors (Previ-red with an aversive event) can

activate mesotelencephalic dopamine systems as well,increasing dopamine metabolism in the frontal cortex’*”and the concentration of dopamine in nucleus accum-hens dialysate .377 What does this mean for the hypoth-

esis that dopamine mediates incentive salience andthat dopaminergic activation makes stimuli salient, at-tractive and ‘wanted’?

There are at least two alternative explanations thatcan reconcile stress-induced activation of dopamineand the incentive-sensitization hypothesis and untilfurther data are available the incentive salience hy-pothesis does not commit to either one. First, it ispossible that mesotelencephalic dopamine systems me-diate the salience of stimuli that signal unpleasantconsequences as well as those that signal pleasantones. The rustling noise that signals an approachingtiger should grab the attention no less than the sight ofdelectable food. The salience of both tiger and foodmay be mediated by dopamine systems, whereas thevalence of that salience (attractive incentive vs. fright-ening warning) may be determined by the coactivationof other neural systems. A second possibility is thatmoderate levels of dopamine activation, such as thatproduced by natural incentives and stressors (see abovefor references), always makes stimuli attractivelysalient, whereas even higher levels of dopamine activa-tion makes stimuli frightening.

Is it nonsensical to say that stress can make stimulimore attractive? Not at all. Stressors are known topotentiate behavior that is ordinarily incentive-based.For example, stress-induced feeding is a phenomenonthat has been well documented in both animals andhumans1”v223. Stress may cause mesotelencephalicdopamine systems to magnify the incentive salienceattributed to known incentives such as familiar foods,thus leading to increased eating. Furthermore, stres-sors themselves may sometimes fascinate and elicitapproach, rather than drive an individual away. Forexample, in a laboratory model of predator mobbingrats will repeatedly approach an electrified object thathas in the past given them a shock and will attempt tobury the offending object by pushing sand, etc. uponit245. Also, under particular conditions animals willwork (lever press) to deliver electric shocks to them-selves, shocks that are known to be otherwise aversive(ref. 86 for review). Indeed, rats will even bar press toself-administer corticosterone i.v., in doses that pro-duce a plasma concentration comparable to that seenduring mild stress and this is associated with an in-crease in nucleus accumbens dopamine neurotransmis-sion244. Furthermore, animals that are more prone toacquire a drug self-administration habit are more sen-sitive to the reinforcing effects of corticosterone than‘low risk’ animals. Corticosterone also produces a largerincrease in dopamine neurotransmission in ‘high risk’animals than in ‘low risk’ animals244. These examplessuggest that stress-induced dopamine activation may

278

indeed activate incentive processes - quite separatelyfrom their activation of pain, aversion or discomfort.Which of these alternatives best describes the role ofdopaminergic salience attribution during stress and inresponse to aversive stimuli will require further re-search to resolve.

In the examples above we suggest that the moderatelevel of dopamine activation produced by natural in-centives and even stressors, may increase incentivesalience. Likewise, we posit that a moderate level ofdopamine activation produced by addictive drugs en-hances incentive salience and the higher levels ofdopamine activation produced by. increasing doses ofaddictive drugs may progressively increase incentivesalience. The ability of addictive drugs to elevatedopamine neurotransmission beyond that which nor-mally occurs may be the feature of drugs that makethem such potent incentives. However, this may be trueonly up to a point. Exceedingly high levels of dopamineactivation may sometimes result in markedly aversiveexperiences. Although the attribution of incentivesalience can make stimuli in the environment ‘brighter’and more attractive, beyond a certain point the worldmay become too ‘bright’: stimuli may become confus-ing, distracting and potentially frightening. For exam-ple, the hallucinations and terror of amphetamine psy-chosis may reflect the excessive and indiscriminateattribution of salience to all stimuli in general, by awildly hyperfunctioning dopamine system. The sensiti-zation of dopamine neurotransmission may explain whythe propensity to amphetamine or cocaine psychosisusually develops in a progressive, sensitization-likefashion and why the susceptibility to stimulant-induced(or stress-induced) psychosis persists for years after thediscontinuation of drug use279*280.

The symptoms of stimulant-induced psychosis arevery similar to those seen in paranoid schizophrenia292and the suggestion that they are due to wildly excessiveincentive salience is consistent with some current hy-potheses regarding the nature of schizophrenia. It hasbeen suggested”“‘, for example, that, “mcsolimbicdopamine activation may regulate the extent to whichparticular types of environmental cues elicit or shapeappetitive behavior”... and a loss of this ‘gating’ func-tion by the overactivation of dopamine systems “mayresult in cognitive ‘flooding’, information overload andcognitive fragmentation in clinical states putatively as-sociated with dopamine overactivity [such as schizo-phrenia]37*32’~332” (p. 419).

6.10. Note 10. Dopanzine alztagonists and therapyAt first sight, it might appear that an implication of

the Incentive-Sensitization Theory is that an effective

treatment for excessive craving would be to blockdopamine receptors with a postsynaptic -antagonist suchas pimozide or haloperidol. However, it is nat at allclear that such a treatment would be effective. In fact,there are several reasons for doubting the usefulness ofdopamine antagonists as a treatment for addiction.

Even though the Incentive-Sensitization Theory pro-poses that excessive craving for drugs results directlyfrom sensitization of dopamine neurotransmission,dopamine antagonists may not be as useful in reducingthe expression of pre-established incentives (attribu-tion of incentive salience that is directed by existingassociations) as they are in blocking the acquisition of

new incentives (attribution of incentive salience to pre-viously neutral stimuli). An addict is already too latefor a treatment that blocks the acquisition of incentivesensitization to work. Only a treatment that blockedthe expression of sensitized incentive salience would behelpful.

A largedopamine

body of evidence from animal studies o fantagonist effects on incentive motivat ion

indicates that while the acquisition of, for example, aconditioned preference for an environment paired withdrug administration is nearly always blocked by adopamine antagonist, the behavioral expression of apreference that was previously conditioned is onlysometimes suppressed by the same drug (see the afore-mentioned Note 8). There are a number of possibleexplanations for the equivocal effects of dopamineantagonists on the expression of pre-existing incentives.First, it might be that dopamine antagonists actually doreduce the incentive salience of drug-paired condi-tioned stimuli, but that they also reduce the incentivesalience of all other stimuli. This would produce anabsolute reduction of all incentive motivation but wouldleave the relative incentive value of stimuli unchanged(ref. 139 for example). Although an addict might cravedrugs less after taking a dopamine antagonist, drugswould still be wanted more than anything else and drugseeking would still dominate behavior. Second, it mightbe that the neural processes that mediate the expres-sion of sensitized incentive salience truly are moreresistan t to the effects of dopamineare the n c ural processes which medi

antate

agonists thanthe establish-

merit of sensitization. If so, only very highncuroleptics, which might seriously disrupt

dosesmany

ofas-

pects of normal behavior, would be sufficient to sup-press drug craving. We do not know why this should betrue, but there arc a nunnber of possible mechanismsthat could explain it. For csanlplc, once cstablishcdscnsitiz;ltiont ions that

might bcextend to

associated with neuro;ldapta-systems ‘downstream from

dopaminc neurons thcmsclvcs. These neural systcrns

279

may continue to respond cscessivcly cvcn if dopaminencurotr~~r7sr1lissiC)I1 is reduced. On the other hwd, theacquisition of sensitization may dcpcnd more spccifi-tally upon dopamine activation alone.

Another possibility is that the acquisition vs. exprcs-sion of sensitization are mediated by differentdopaminc subsystc=ms that have, for example, differentdopaminc receptor subtypes, etc. Hiroi and White’“”noted that when a dopamine receptor antagonist failedto selectively decrease the expression of conditionedresponding (see the aforementioned Note 8) “ncuro-leptics with a higher affinity for D2 than Dl receptors”,were used and further noted that “a Dl antagonist,SCH23390 is equally effective in blocking acquisitionand expression of an amphetamine conditioned placepreference, whereas much higher doses of D2 antago-nists are required to block expression than acquisition.Thus, some workers may have failed to observe block-ing of the expression of learned behaviors because theyused an inappropriate dose range of D2 antagonists”(pp. 40/U).

Finally, even if neuroleptic drugs were effective atsuppressing drug craving at moderate doses, their use-fulness could be compromised by the possibility thataddicts would refuse to take them. Anecdotal evidenceabounds to suggest that neuroleptics are unpopulardrugs among patients who take them. Aside from theirpotential motor effects, it is not surprising that thisshould be so. By the incentive salience hypothesis, adrug that directly suppressed the attribution of incen-tive salience would make the world ‘less bright’. Eventhough the neural substrates of pleasure would not besuppressed, such a drug could produce ‘sham anhedo-nia’ - that is, the conscious inference by the addict thatpleasure was reduced via cognitive interpretation (“Idon’t want anything very much, therefore I must notlike anything”) - just as direct activation of incentivesalience should produce ‘sham reward’27.

7. GLOSSARY

Addiction. There has been considerable debate regard-ing the appropriate definition of drug addiction. Wewill use the term here in the sense proposed by aWorld Health Organization Expert Committee in1981 70,148 Drug addiction is defined as “a syndrome inwhich th, use of a drug is given a much higher prioritythan other behaviors that once had higher value.“...“In its extreme form [addiction] is associated withcompulsive drug-using behavior and it exhibits thecharacteristics of a chronic relapsing disorder” (ref.148, p. 522). The phrase ‘addictive behavior’ is used to

refer collectively to obsessive drug craving and to com-pulsive drug-seeking and drug-taking behavior.Addictir Te beiuwior. (see Addict ion)Appetitir le motil Mon. (see I ncen t ive motivation)Aversion. The subjective experience of a sensation asactively unpleasant or the underlying evaluative pro-cesses and neural mechanisms that directly producethis subjective experience. The opposite of pleasure oreuphoria (see Pleasure). Aversion results from an ac-tive evaluation of a sensation carried out by brainsystems. In the context of addiction, aversion can besynonymous with the symptoms associated with drugVTwithdrawal, inclu&g physical distress and dysphoria.Aversion can also refer to direct subjective effectsproduced by a drug that can be discriminated frompleasure by the user.Conditioned incentive stimuli. (see Incentives)CrarGzg and ‘Wanting’. These are used in accordancewith their usual English meaning, which for ‘wanting’refers to the subjective experience of needing or desir-ing something (“to feel a need or desire for”; RandomHouse Dictionary of the English Language, 2nd edn.,1987). We further propose, however, that this experi-ence is produced by the psychological process ofsalience attribution (see incentive salience), that is, theattribution of incentive salience to an external event orits mental representation. The process of incentivesalience attribution is pre-concious and only the resultof this psychological process is accessible to conscious-ness. When this occurs it is interpreted228 as a subjec-tive feeling of ‘wanting’. For our purposes craving and‘wanting’ differ only in magnitude: craving equals in-tense ‘wanting’lB4. In the addict, craving is the experi-ence associated with excessive incentive salience, whichresults from drug-induced sensitization of the neuralsystems that attribute salience to incentives.Dependence, drug. (see Addiction)Euphoria. (see Pleasure)Hedonics. (see Pleasure)Incentive motivation (al). A psychological theory (alsosee Fig. 2) of how goal-direction is controlled by thestimulus properties of the target30*339. Incentive moti-vation is one of a number of potential psychologicalmechanisms for controlling the direction of motivatedbehavior. Drive reduction and opponent processes areexamples of other potential mechanisms that mightcontrol behavior independently of incentive processes(see Toates339 for discussion and evidence).

Incentive motivation appears to be the chief mecha-nism that controls behavior directed towards naturalincentives such as food, water and a potential mate andtowards more artificial incentives, such as self-adminis-tered drugs and reinforcing electrical brain stimula-

ti*n30*339 In the context of addiction, incentive motiva-tion and’appetitive motivation are synonymous: that is,appetitive motivation works primarily through incen-tive processes339. Incentive motivation directed towardsparticular stimuli results from the outcome of a three-stage process. First, the neural substrates for pleasureare activated by the consequences of a particular act orevent. Second, pleasure is associated with the object,act, event or place in which pleasure occurs by theprocesses of classical associative learning. Third,salience is attributed to subsequent perceptions andmental representations of the associated object, act,event or place, by a separate neural system from thoseresponsible for the first two processes. This third pro-cess of salience attribution (incentive salience) is pro-posed here to involve dopamine. The attribution ofincentive salience causes the associated situation tobecome attractive and ‘wanted’ and it is this psycholog-ical process that produces the direct manifestation ofincentive motivation: goal-directed seeking and instru-mental behavior.Incentive salience. Refers to the attractiveness of exter-nal stimuli, events, places and their mental representa-tions; their ability to capture attention (also see Fig. 2).The term, incentive salience, applies always to theperception of external events and to internal represen-tations of those events. Incentive salience must beactively generated (attributed) by the brain and as-signed to particular perceptions and representations,based on their association with past activation ofmesotelencephalic dopamine systems. For any givenstimulus incentive salience will vary at different timesdepending upon changes in learned associations re-garding the stimulus, the internal state of the perceiverand most specifically, the degree of activation of thedopamine systems that mediate incentive salience. In-centive salience is one of a number of psychologicalmechanisms that can produce direct behavior (othermechanisms include drive reduction and goal-directedcomputational algorithms that do not depend on mod-ulated perception of the goal; see Toates339). Incentivesalience constit’i.tes one component of the complexprocess of incentive motivation. Although the assign-ment of incentive salience to an event normally istriggered by a pleasurable experience, manipulations ofdopamine systems can disconnect incentive saliencefrom pleasure and alter incentive salience indepen-dently. It is hypothesized that the attribution of incen-tive salience to an event or representation is an uncon-scious process; only the product of this process, theperception of the object as ‘wanted’, is interpreted andconsciously experienced.

Incentive stimuli (incentives). Stimuli that have heenattributed with incentive salience. The perception andmental representation of these stimuli are transformedas a consequence and as incentive stimuli they becomesalient, attractive, ‘wanted’, and approached. The termsnatural incentives and artificial incentives are also used,The 3-stage process described for incentive motivationevolved to enable animals and humans to recognizeand respond to ‘natural’ incentives. By natural incen-tives we mean stimuli such as food, water, social andsexual partners, thermal and tactile sensations, whichhave been endowed by evolution with the capacity toelicit pleasure and incentive salience under particularconditions (e.g., under certain hormonal conditions).Most natural incentives exert their effects via sensoryreceptors. ‘Artificial’ incentives, on the other hand,such as addictive drugs or electrical brain stimulation,bypass sensory receptors and activate the componentprocesses of incentive motivation more directly. Mostincentive stimuli are conditioned incentive stimuli:stimuli that have become incentives as a consequenceof associative learning during the three-stage processdescribed for incentive motivation. Conditioned incen-tive stimuli not only include the arbitrary lights andsounds used in laboratory experiments (e.g.. an audi-tory tone that signals food delivery), but also the stimu-lus configurations that must be learned through experi-ence that allow natural incentives to be recognized(e.g., the sight of a delectable food; the sound of aloved one’s voice). Conditioned incentive stimuli areoften referred to as conditioned rewards (see below),secondary rewards or secondary reinforcers (see below).‘Liking’. See Pleasure.Negative reinforcement. See Reinforcement).Pleasure and ‘liking’. These are used in accordancewith their usual English meaning, which refers to thesubjective experience of a sensation as pleasurable orhedonic and the underlying evaluative and neural pro-cesses that directly produce this subjective experience.The opposite of aversion (see above). Pleasure is USU-

ally the first stage of the larger process of incentivemotivation (together with incentive salience and asso-ciative learning) and serves as the normal trigger thatactivates components of associative learning antI ken-tive salience. By itself, however, pleasure is not quiva-lent to either reward or ‘wanting’: it is merely a subjec-tive cxpcricncc or feeling. The evaluation of the sensa-tion that produces pleasure is pre-conscious; only theproduct, the subjective pleasure, is experienced. In thepresent paper and in the context of addiction, the termpleasure is used synonymously with the terms cupho-ria, hcdonia or positive affective state.

Positiz ‘4 czfii’ctii *c stcz tc. (see Pleasure).Fositiw I’c’i~~~i)l-~~(‘1)zC’Ilf. (see Reinforccmcnt).R~~i~zf~)n~~~~rrr3lz t. A purely behavioral and descriptiveterm for the relationship between the occurrence of astimulus and changes in the subsequent probability ofthe behavior that preceded it. Reinforcement denotesa change in the probability of a behavior (increased ordecreased) that is contingent on presentation of stim-uli. Reinforcement does not offer either a psychologi-cal explanation or a physiological explanation of whythe probability of a behavior is changed; it merelynotes the existence of the change. Reinforcement canbe positive or negative. Positive reinforcement refers toincreases in the probability of emission of a behaviorproduced by subsequent presentation of a stimulus (thepositive reinforcer). In the context of addiction theterm positive reinforcement is sometimes used both inits proper descriptive sense and sometimes in the theo-retical or explanatory sense of reward (see below),where the pleasure produced by a drug is implicitlyassumed as a psychological explanation for the changein behavior. In order to avoid ambiguity in the presentpaper we use positive reinforcement only in its properdescriptive sense. Negative reinforcement refers to in-creases in the probability of emission of a behaviorproduced by subsequent omission or termination of astimulus. For example, in the context of addiction,drugs may act as negative reinforcers by relieving thedistress of drug withdrawal.&ward. The word reward is used in the literature inmany different ways and for the most part we avoid theterm. For us, the process of reward is essentially equiv-alent to the process of incentive motivation; that is,reward refers to the process of creating incentives (or,as a noun, a stimulus that triggers this process). Re-wards (or incentives) cause future behavior to bechanged in a goal-directed fashion so as to obtain againthe situation or stimulus that triggered the process.This process normally requires three separate stages.The first stage is the activation of pleasure by theconsequences of a particular act or event. In the sec-ond stage pleasure is associated with a mental repre-sentation of the object, act, event or place in whichpleasure occurred, by the process of classical (associa-tive) conditioning. The third stage involves the attribu-tion of incentive salience to subsequent perceptionsand representations of the associated object, act, eventor place, which causes them to become ‘wanted’. Stim-uli that signal the availability of the incentive becomeattractive. Acts that Ied to the situation in the past arelikely to be repeated. New acts, which the animal orperson can predict (cognitively) will lead to the incen-

281

tive in the future, are likely to be produced. If thethree stages of normal reward (pleasure, associativelearning, incentive salience) are separated, the processremains incomplete. Some separations have beenachieved by brain manipulations; others are usefulsimply as illustrative ‘thought experiments’. If the firststage of pleasure is activated without associative learn-ing or salience attribution, then it is merely an isolatedhedonic experience that remains unconnected to otherevents in the world or to subsequent behavior. If thefirst two stages occur alone so that pleasure is activatedi;! conjunction with associative learning only, then asso-ciative conditioning of pleasure will occur to the associ-ated events, but they will not be attributed with incen-tive salience. The events will become ‘liked’, but theywill not be ‘wanted’ (this separation may possibly beachieved by destruction of brain dopamine systems28).Conversely, if the third stage, incentive salience, isactivated alone, then ‘wanting’ arises in isolation fromother processes. This may be achieved, for example, bya stimulating electrode that directly activates braindopamine systems27. If salience attribution is magnifiedabnormally in conjunction with associative learning,but pleasure is not, for example, by the process ofdrug-induced sensitization discussed here, then the at-tribution of incentive salience becomes intensified andfocused narrowly on the stimuli and acts associatedwith drug administration and they become pathologi-cally ‘wanted’ (craved). This may be considered a typeof ‘sham reward’, which shapes instrumental behaviorand creates craving, but is dissociated from pleasure.Viewed from the outside, the behavior produced bysham reward and natural reward is identical on allmeasures of ‘wanting’ or instrumental performance.Only behavioral measures that are specifically sensitiveto pleasure - rather than wanting - will identify shamreward (without pleasure) as distinct from natural re-ward (triggered by pleasure).Salience. A salient stimulus is a stimulus that has been(or is being) attributed with incentive salience (seeabove). Salience refers to the feature(s) of the perceptor representation of a stimulus that makes it highlynoticeable and difficult to ignore. A salient stimulus isnot merely more obvious, but it also becomes ‘wanted’and attractive because, according to our usage, it isincentive salience that is attributed to the percept ormental representation (but see Note 9 in Ch. 6 onaversive salience).Salience attribution. (see Incentive salience; Incentivemotivation).Secondary reinforcers. (see Incentive stimuli)‘Wanting: (see Craving/wanting’).

282 .

Acknowledgements. Preparation of this manuscript and research bythe authors was supported by Grants from the National Institute onDrug Abuse (04294) to T.E.R. and from the National Institutes ofHealth to K.C.B. (NS23959). An earlier version of this theory wasfirst presented at a UCLA/NIDA conference on The BiologicalBasis of Substance Abuse, Santa Monica, CA, January Y-12, 1991.We are grateful to Aldo Badiani, Jill Becker, Craig Berridge, BarryEveritt, Tim Schallert, Jane Stewart, Frederick Toates, Keith Tru-jillo, Elliot Valenstein, Derek van der Kooy, Ian Whishaw and RoyWise for their helpful and thought-provoking comments on an earlierdraft of this paper.

REFERENCES

1 Abercrombie, E.D., Keefe, K.A., Di Frischia, D.S. and Zig-mond, M.J., Differential effect of stress on in vivo dopaminerelease in striatum, nucleus accumbens and medial frontal cor-tex, J. Neurochem., 52 (1989) 1655-1658.

2 Acquas, E., Carboni, E., Leone, P. and Di Chiara, G., SCH23390 blocks drug-conditioned place-preference and place-aver-sion: anhedonia (lack of reward) or apathy (lack of motivation)after dopamine-receptor blockade?, Psychopharmacology, 99(1989) 151-155.

3 Aizenstein, M.L., Segal, D.S. and Kuczenski, R., Repeatedamphetamine and fencamfamine: sensitization and reciprocalcross-sensitization, Neuropsychopharmacology, 3 (1990) 283-290.

4 Akimoto, K., Hamamura, T., Kazahaya, Y., Akiyama, K. andOtsuki, S., Enhanced extracellular dopamine level may be thefundamental neuropharmacological basis of cross-behavioralsehsitization between methamphetamine and cocaine - an invivo ‘dialysis study in freely moving rats, Brain Rex, 507 (1990)344-346.

5 Akimoto, K., Hamamura, T. and Otsuki, S., Subchronic cocainetreatment enhances cocaine-induced dopamine efflux, studiedby in vivo intracerebral dialysis, Brain Res., 490 (1989) 339-344.

6 Akiyama, K., Hamamura, T., Ujike, H., Kanzaki, A. and Otsuki,S., Methamphetamine psychosis as a model of relapse ofschizophrenia - a behavioral and biochemical study in theanimal model. In T. Nakazawa (Ed.), Taniguchi Symposia onBrain Sciences, Vol. 14, Biological Basis of Schizophrenia, JapanScientific Societies Press, Tokyo, 1991, pp. 169- 184.

7 Anonymous, Alcoholics Anonymous, Alcoholics AnonymousWorld Services, New York, 1976, 3rd edn.

8 Antelman, S.M. and Chiodo, L.A., Amphetamine as a stressor.In I. Crccse (Ed.), Stimulants: Neurochemical, Behalqioral andClinical Perspectives, Raven Press, New York, 1983, pp. 269-299.

9 AnteIman, S.M., Eichler, A.J., Black, C.A. and Kocan, D.,Interchangeability of stress and amphetamine in sensitization,Science, 207 (1980) 329-331.

10 Antelman, S.M. and Szechtman, H., Tail pinch induces eating insated rats which appears to depend on nigrostriatal dopamine,Science, 189 ( 1975) 73 1-733.

11 Babbini, M. and Davis, W.M., Time-dose relationships for Ioco-motor activity effects of morphine after acute or repeated treat-ment, Br. J. Pharmacol., 46 (1972) 213-2?4.

12 Babbini, M., Gaiardi, M. and Bartolet-ii, M., Persistence ofchronic morphine effects upon activity in rats 8 months afterceasing the treatment, Nprrr~,ph(Irrnac~lo~, 14 (1975) 61 l-61 14.

13 Bailey, C.S., Ilsiao, S. and King, J.E., Hedonic reactivity tosucrose in rats: modification by pimozide, Physiol. Behat!., 38(1986) 447-452.

14 Baker, H., Joh, T.H. and Reis, D.J., Genetic control of numberof midbrain dopaminergic neurons in inbred strains of mice:relationship to size and neuronal density of the striatum, Proc.Natl. Acad. Sri. USA, 77 ( 1980) 4369-4373.

15 Barnett, J.V., Scgal, D.S. and Kuczenski, R., Repeated am-phetamine pretreatment alters the responsiveness of striataldopamine-stimulated adenylatcr cyclase to amphetamine-induceddescnsi t iza t ion, J. Plrarmacd. E-up. Ther. , 242 ( 1987) 40-47.

16 Barr, G.A., Sharpless, N.S., Cooper, S., Schiff, S.R., Parcdes, W.and Bririgcr, W.l-I., Classical conditioning, decay and extinction

of cocaine-induced hyperactivity and stereotypy, Life sci, 33(1983) 1341-1351.

17 Bartoletti, M., Gaiardi, M., Gubellini, G., Bacchi, A. and Bab-bini, M., Long-term sensitization to the excitatory effects ofmorphine. A motility study in post-dependent rats, Neurophr.macology, 22 (1983) 1193-1196.

18 Baxter, L.R.J., Schwartz, J.M., Bergman, K.S., Szuba, M.p.,Guze, B.H., Mazziotta, J.C., Alazraki, A., Selin, C.E., Femg,H.K., Munford, P., et al., Caudate glucose metabolic rate changeswith both drug and behavior therapy for obsessive-compulsivedisorder, Arch. Gen. Psychiatry, 49 (1992) 681-689.

19 Beninger, R.J., The roIe of dopamine in locomotor activity andlearning, Brain Res. Rel?. , 287 (1983) 173- 196.

20 Beninger, R.J. and Hahn, B.L., Pimozide blocks establishmentbut not expression of amphetamine-produced environment-specific conditioning, Science, 220 (1983) 1304- 1306.

21 Beninger, R.J. and Herz, R.S., Pimozide blocks establishmentbut not expression of cocaine-produced environment-specificconditioning, Life Sci., 38 (1986) 1425-1431.

22 Beninger, R.J., MacLennan, A.J. and Pine], J.P., The use ofconditioned defensive burying to test the the effects of pjmozjdeon associative learning, Pharmacol. B&hem. Behw., 12 (198())445-448.

23 Beninger, R.J., Mason, S.T., Phillips, A.G. and Fibiger, H.C.,The use of conditioned suppression to evaluate the nature ofneuroleptic-induced avoidance deficits, J. Pharmad. fip. mr.,213 (1980) 623-627.

24 Benjamin, D., Grant, E.R., Goldstein, K.R. and Pohorecky,L.A., Sensitization to the dopamine release-enhancing effects ofethanol demonstrated in male long-evans rats, SOC. Neurosci.Abstr., 18 (1992) 1431.

25 Benwell, M.E.M. and Balfour, D.J.K., The effects of acute andrepeated nicotine treatment on nucleus accumbcns dopamineand locomotor activity, Br. J. Pharmacol., 105 ( 1992) 849-856.

26 Berridge, K.C., Modulation of taste affect by hunger, caloricsatiety and sensory-specific satiety in the rat, AppcW 16 (1991)103-120.

27 Berridge, K.C. and Valenstein, E.S., What psychological processmediates feeding evoked by electrical stimulation ot‘ the lateralhypothalamus ?, Beharq. Neurosci., 105 (1991) 3-14.

28 Berridge, K.C., Venier, I.L. and Robinson, T.E., T:i\ttl reactivityanalysis of 6-hydroxydopamine-induced aphagia: implications forarousal and anhedonia hypotheses of dopamine function, Be-harv. Neurosci., 103 (1989) 36-45.

29 Bijerot, W., Addiction to pleasure: a biological and social-psy-chological theory of addiction. In D.J. Lettieri, M. Sayers andH.W. Pearson (Eds.), Theories on Drug Abuse: Seltwd Contem-

porary Perspectirqes, National Institute on Drug Ahusc. Rockville,1980, pp. 246-255.

30 Bindra, D., How adaptive behavior is produced: a iwceptual-motivation alternative to response reinforcement, 1h hw. BrainSci., 1 (1978) 41-91.

31 Blackburn, J.R., Pfaus, J.G. and Phillips, A.G., Dopwjine func-tions in appctitive and defensive behaviours, Prog. .t’wobioL,39 (1992) 247-279.

32 Blackburn, J.R., Phillips, A.G. and Fibiger, H.C., Dop:t;nine andpreparatory behavior: I. Effects of pimozide, Bt*hw. .vc’urosci.,101 (1987) 352-360.

33 Blackburn, J.R., Phillips, A.G. and Fibiger, H.C., Dopkne andpreparatory behavior: III. Effects of metoclopramidr d thiori-dazine, B&al: Nertrosci., 103 (1989) 903-906.

34 Blackburn, J-R., Phillips, A.G., Jakubovic, A. and Fibigtx-, H.C.,Dopamine and preparatory behavior: II. A neurochcmii:il anal)’sis, Behalt. Ntvrrosci., 103 (1989) 15 -23.

35 Borison, R.L., Havdala, H.S. and Diamontl, B.1.. Chronicphcnylethylamine stcrcotypy in rats: a nt’w aninxtl nwJt’1 forschizophrenia?, Life Sci., 21 (1977) 117- 122.

36 Bradberry, C.W., Gruen, K.J., Brrridg~, C.W. and Roth. R.H.,individual differences in lwh;wior;~I nxxsurtx C~)rr~littic~ns withnucleus accumbcns dopaminc rwxwrcrt by nkwdk~l!*sis. Pjrar-mad. Biochtw~. Bt~hur:, 30 (1991) 877482.

37 Braff, D.L. and Gcycr, MA., Sensorirnotor gitting ;irlc the

283

neurobiology of schi7ophrcnin: human and animal model stud-ies. In SC. Schultz and C.A. Tamminga (Eds.), Schizophrenia:&+rrfjfic ~IX~,~XW, Oxford IJniversity Press, New York, 1989, pp.124- 1%.

38 Brown. E.E. and Fibiger, 11X’., Cocaine-induced conditionedlocomotion: absence of associated increase in dopamine release,Akwrost-item, 48 ( 1992) 62 l-629.

39 Brown, E.E., Robertson, G.S. and Fibiger, H.C., Evidence forconditional neuronal activation following exposure to a cocaine-paired environment: role of forebrain limbic structures, J. Nen-rosci., 12 (1992) 4112-4121.

40 Brownc, R.G. and Segal, D.S., Metabolic and experimentalfactors in the behavioral response to repeated amphetamine,Pharmacol. Riochum. Behart. ) 6 ( 1977) 545-552.

41 Cabanac, M., Physiological role of pleasure, Scknce, 173 (1971)1103- 1107.

42 Cabanac, M., Sensory pleasure, Q. ReLv. Biol., 54 (1979) l-29.43 Cabanac, M. and Lafrance, L., Postingestive alliesthesia: the rat

tells the same story, Physiol. Behal*., 47 (1990) 539-543.44 Cador, M., Taylor, J.R. and Robbins, T.W., Potentiation of the

effects of reward-related stimuli by dopaminergic-dependentmechanisms in the nucleus accumbens, Psychopharmucology,104 (1991) 377-335.

45 Camp, D.M. and Robinson, T.E., Susceptibility to sensitization.I. Sex differences in the enduring effects of chronic D-

amphetamine treatment on locomotion, stereotyped behaviorand brain monoamines, Behw. Brain Res., 30 (1988) 55-68.

46 Camp, D.M. and Robinson, T.E., Susceptibility to sensitization.II. The influence of gonadal hormones on enduring changes inbrain monoamines and behavior produced by the repeated ad-ministration of D-amphetamine or restraint stress, Behao. BrainRes., 30 (1988) 69-88.

47 Carlson, J.N. and Glick, S.D., Cerebral lateralization as a sourceof interindividual differences in behavior, Experientiu, 45 (1989)788-798.

48 Castaiieda, E., Becker, J.B. and Robinson, T.E., The long-termeffects of repeated amphetamine treatment in vivo on am-phetamine, KC1 and electrical stimulation evoked striataldopamine release in vitro, Life Sci., 42 (1988) 2447-2456.

: 49 Childress, A.R., Ehrman, R., McLellan, A.T., MacRae, J., Na-‘ tale, M. and O’Brien, C.P., Negative mood states trigger condi-

tioned drug craving and conditioned withdrawal in opiate abusepatients, J. Subst. Abuse Treat, in press.

50 Childress, A.R., McLellan, A.T., Ehrman, R. and O’Brien, C.P.,Classically conditioned responses in opioid and cocaine depen-dence: a role in relapse?, NIDA Res . Monogr . , 84 (1988) 25-43 ., 51 Church, W.H., Justice Jr., J.B. and Neill, D.B., Detecting behav-iorally relevant changes in extracellular dopamine with micro-dialysis, Brain Res., 412 (1987) 397-399.

52 Clarke, P.B.S., Mesolimbic dopamine activation - the key tonicotine reinforcement? In Ciba Foundation Symposium 152:The Biology of Nicotine Dependence, Wiley, Chichester, 1990, pp.153-168.

53 Clifton, P.G., Rusk, I.N. and Cooper, S.J., Stimulation andinhibition of food intake by the selective dopamine D2 agonist,N-0437: a meal pattern analysis, Pharmacol. Biochem. Behav., 33(1989) 21-26.

54 Cole, B.J., Cador, M., Stinus, L., Rivier, C., Rivier, J., Vale, W.,t 1 Le Moal, M. and Koob; G.F., Critical role of the hypothalamic

pituitary adrenal axis in amphetamine-induced sensitization ofbehavior, Life Sci., 47 (1990) 1715-1720.

55 Cole, B.J., Cador, M., Stinus, L., Rivier, J., Vale, W., Koob, G.F.and Le Meal, M., Central administration of a CRF antagonistblocks the development of stress-induced behavioral sensitiza-tion, Brain Res., 512 (1990) 343-346.

56 Cole, J.O. and Chiarello, R.J., The benzodiazepines as drugs ofabuse, L Psychiat. Res., 24 (1990) 135-144.

57 Cools, A.R., Differential role of mineralocorticoid and glucocor-ticoid receptors in the genesis of dexamphetamine-induced sen-sitization of mesolimbic, alpha 1 adrenergic receptors in theventral striatum, Neuroscience, 43 (1991) 419-428.

58 Crabbe, J.C., Johnson, N.A., Gray, D.K., Kosobud, A. and

Young, E.R., Biphasic effects of ethanol on open-field activity:sensitivity and tolerance in C57W./6N and DBA/2N mice, J.Camp. Physiol Psychol., 96 ( 1982) 440-45 1.

59 Cummings, J.L. and Cunningham, K., Obsessive-compulsive dis-order in Huntington’s disease, &u/. Psychiclfry, 31 (1992) 263-270.

60 Cunningham, C.L., Nichus, D.R., Malott, D.H. and Prather,L.K., Genetic differences in the rewarding and activating effectsof morphine and ethanol, Ps))cllo,lrharrrla~~o~o~y, 107 (1992) 385-393.

61 Cunningham, C.L. and Noble, D., Conditioned activation in-duced by ethanol: role in sensitization and conditioned placepreference, Pharmacol . Biochem. Behal*. , 4 3 (1992) 307-313 .

62 Cunningham, S.T. and Kelley, A.E., Evidence for opiate-dopa-mine cross-sensitization in the nucleus accumbens: studies ofconditioned reward, Brain Res. Bldi., 29 (1992) 675-680.

63 Dackis, C.A. and Gold, M.S., Bromocriptine as a treatment ofcocaine abuse, The Lancet, May (1985) 1151-l 152.

64 Dackis, C.A. and Gold, M.S., New concepts in cocaine addic-tion: the dopamine depletion hypothesis, Neurosci. Biobehao.Ret!. , 9 (1985) 469-477.

65 Damsma, G., Day, J. and Fibiger, H.C., Lack of tolerance tonicotine-induced dopamine release in the nucleus accumbens,Eur. J. Pharmacol., 168 (1989) 363-368.

66 Damsma, G., Pfaus, J.G., Wenkstern, D., Phillips, A.G. andFibiger, H.C., Sexual behavior increases dopamine transmissionin the nucleus accumbens and striatum of male rats: comparisonwith novelty and locomotion, Behav. Neurosci., 106 (1992) 181-191.

67 Davis, J.D., The microstructure of ingestive behavior, Ann. NYAcad. Sci., 575 (1989) 106-119.

68 Deminiere, J.M., Piazza, P.V.., Guegan, G., Abrous, N., Maccari,S., Le Moal, M. and Simon, H., Increased locomotor response tonovelty and propensity to intravenous amphetamine self-admin-istration in adult offspring of stressed mothers, Brain Res., 586(1992) 135-139.

.

69 Deroche, V., Piazza, P.V., Casolini, P., Maccari, S., Le Moal, M.and Simon, H., Stress-induced sensitization to amphetamine andmorphine psychomotor effects depend on stress-induced corti-costerone secretion, Brain Res., 598 (1992) 343-348.

70 Deroche, V., Piazza, P.V., Maccari, S., Le Moal, M. and Simon,H., Repeated corticosterone administration sensitizes the loco-motor response to amphetamine, Brain Res., 584 (1992) 309-313.

71 Deutch, A.Y. and Cameron, D.S., Pharmacological characteriza-tion of dopamine systems in the nucleus accumbens core andshell, Neuroscience, 46 (1992) 49-56.

72 Dews, P., Remarks. In T. Thompson and K.R. Unna (Eds.),Predicting Dependence Liability of Stimulant and DepressantDrugs, University Park Press, Baltimore, 1977, pp. 75-79.

73 Di Chiara, G. and Imperato, A., Drugs abused by humanspreferentially increase synaptic dopamine concentrations in themesolimbic system of freely moving rats, Proc. Natl. Acad. Sci.USA, 85 (1988) 5274-5278.

74 Di Lullo, S.L. and Martin-Iverson, M.T., Evidence for presynap-tic dopamine mechanisms underlying amphetamine-conditionedlocomotion, Brain Res., 578 (1992) 161-167.

75 Doyle, T.J., Berridge, KC. and Gosnell, B.A., Morphine en-hances tas te p leasure , Pharmacol . Biochem. Behao. , i n p r e s s .

76 Drew, K.L. and Glick, S.D., Classical conditioning of am-phetamine-induced lateralized and nonlateralized activity in rats,Psychopharmacology , 92 (1987) 52-57 .

77 Drew, K.L. and Glick, SD., Role of D-l and D-2 receptorstimulation in sensitization to amphetamine-induced circlingbehavior and in expression and extinction of the Pavlovianconditioned response, Psychopharmacology, 101 (1990) 465-471.

78 Edmonds, D.E. and Gallistel, C.R., Reward versus performancein self-stimulation: electrode-specific effects of alpha-methyl-p-tyrosine on reward in the rat, J. Comp. Physiot. Psychol., 91(1977) 962-974.

79 Edwards, G., Nomenclature and classification of drug- andalcohol-related problems: a WHO memorandum, Bull. WorldHealth Org., 59 (1981) 225-242.

284

80 Ehrman, R., Ternes, J., O’Brien, C.P. and McLellan, A.T.,Conditioned tolerance in human opiate addicts, Psychopharma-cology, 108 (1992) 218-224.

81 Emmett-Oglesby, M.W. and Lane, J.D., Tolerance to the rein-forcing effects of cocaine, Behav. Pharmacol., 3 (1992) 193-200.

82 Ettenberg, A., Dopamine, neuroleptics and reinforced behavior,Neurosci. Biobehav. Rev., 13 (1989) 105-I 11.

83 Ettenberg, A., Haloperidol prevents the reinstatement of am-phetamine-rewarded runway responding in rats, Pharmacol.Biochem. Behav . , 36 (1990) 635-638 .

84 Everitt, B.J., Sexual motivation: a neural and behavioural analy-sis of the mechanisms underlying appetitive and copulatoryresponses of male rats, Neurosci. Biobehav. Rev., 14 (1990)217-232.

85 Everitt, B.L. and Robbins, T.W., Amygdala-ventral striatal in-teractions and reward-related processes. In J.P. Aggleton (Ed.),The Amygdala: Neurobiological Aspects o f Emotion, Memory, andMental Dysfunction, Wiley-Liss, New York, 1992, pp. 401-429.

86 Falk, J.L., Dews, P.B. and Schuster, C.R., Commonalities in theenvironmental control of behavior. In PK. Levison, D.R. Ger-stein and D.R. Maloff (Eds.), Commonalities in Substance Abuseand Habitual Behavior, D.C. Heath and Co., Lexington, MA,1983, pp. 47-110.

87 Falk, J.L. and Feingold, D.A., Environmental and cultural fac-tors in the behavioral action of drugs. In H.Y. Meltzer (Ed.),Psychopharmacology: The Third Generation of Progress, RavenPress, New York, 1987, pp. 1503-1510.

88 Falk, J.L., Ma, F. and Lau, C.E., Chronic oral cocaine self-ad-ministration: pharmacokinetics and effects on spontaneous anddiscriminative motor functions, J. Pharmacol. Exp. Ther., 257(1991) 457-465.

89 Fibiger, H.C. and Phillips, A.G., Increased intracranial self-stimulation in rats after long-term administration of de-sipramine, Science, 214 (1981) 683-685.

90 Fibiger, H.C. and Phillips, A.G., Reward, motivation, cognition:psychobiology of mesotelencephalic dopamine systems. InHandbook o f Physiology, Vol . W, Intrinsic ReguIa tory Systems o fthe Bruin, American Physiology Society, Bethesda, MD, 1986,pp. 647-675.

91 Fink, J.S. and Reis, D.J., Genetic variations in midbraindopamine cell number: parallel with differences in responses todopaminergic agonists and in naturalistic behaviors mediated bycentral dopaminergic systems, Brain Res., 222 (1981) 335-349.

92 Finlay, J.M., Damsma, G. and Fibiger, H.C., Benzodiazepine-in-duced decreases in extracellular concentrations of dopamine inthe nucleus accumbens after acute and repeated administration,Psychopharmacology, 106 (1992) 202-208.

93 Finlay, J.M., Jakubovic, A., Phillips, A.G. and Fibiger, H.C.,Fentanyl-induced conditional place preference: lack of associ-ated condi t ional neurochemical events , Psychopharmacology , 96(1988) 534-540.

94 Fischman, M.W., Relationship between self-reported drug ef-fects and their reinforcing effects: studies with stimulant drugs,NIDA Res. Monogr., 92 (1989) 21 l-230.

95 Fischman, M.W. and Foltin, R.W., Self-administration of co-cainrl by humans: a laboratory perspective. In G.R. Bock and J.Whelan (Eds.), Cocaine: Scientific and Social Dimensions. CIBAFoundation Symposium No. 166, Wiley , Chiches ter , UK, 1992,pp. 165-180.

96 Fischman, M.W., Foltin, R.W., Nestadt, G. and Pearlson, G.D.,Effects of desipramine maintenance on cocaine self-administra-tion by humans, J. Pharmacol. Exp. Tiler., 253 (1990) 760-770.

97 Fischman, M.W., Schuster, C.R., Javaid, J., Hatano, Y. andDavis, J., Acute tolerance development to the cardiovascularand subjective effects of cocaine, J. Pharmacol. Exp. Ther., 235(1985) 677-682.

98 Faitin, R.W. and Fischman, M.W., Assessment of :tbuse liabilityof stimulant drugs in humans: a methodological survey, DrugAImho &perrd., 28 (1991) 3-48.

99 Fwriezos, G. and Wise, R.A., Pimozide-induced extinction ofintracranial self-stimulation: response pat terns rule out motor orperformance deficits, Rruiri Rex, 103 (1976) 377-350.

100 Fujiwara, Y., Kazahaya, Y., Nakashima, M., Sato, M. and ot.suki, S., Behavioral sensitization to methamphetamine in therat: an ontogenic study, Psychopharmacology, 91 (1987) 316-319,

101 Fung, Y.K., Effects of chronic nicotine pretreatment on (+bamphetamine and nicotine-induced synthesis and release of(3H]dopamine from 13H]tyrosine in rat nucleus accumbens, J.Pharm. Pharmacol . , 41 (1989) 66-68 .

102 Gaiardi, M., Bartoletti, M., Bacchi, A., Gubellini, C., Costa, M.and Babbini, M., Role of repeated exposure to morphine indetermining its affective properties: Place and taste conditioningstudies in rats, Psychopharmacology, 103 (1991) 183-186.

103 Gallistel, C.R., Boytim, M., Gomita, Y. and IUebanoff, L., Doespimozide block the reinforcing effect of brain stimulation?,Pharmacol. B&hem. Behav., 17 (19821769-781.

104 Gallistel, C.R., Shizgal, P. and Yeomans, J.S., A portrait of thesubstrate for self-stimulation, Psychol. Rev., 88 (1981) 22-273.

105 Gawin, F.H. and Ellinwood, E.J., Cocaine and other stimulants,Actions, abuse and treatment, N. Engl. J. Med., 318 (1988)1173-1182.

106 Geary, N. and Smith, G.P., Pimozide decreases the positivereinforcing effect of sham fed sucrose in the rat. Pharrnacol.B&hem. Behav., 22 (1985) 787-790.

107 Glick, S.D. and Hinds, P.A., Sex differences in sensitization tococaine-induced rotation, Eur. J. Pharmacol., 99 (1984) 119-121.

108 Glick, S.D. and Hinds, P.A., Differences in amphetamine andmorphine sensitivity in lateralized and non-lateralized rats: loco-motor activity and drug self-administration, Eur. J. Pharmacol.,118 (1985) 239-244.

109 Glick, S.D., Merski, C., Steindorf, S., Wang, S., Keller, R.W.and Carlson, J.N., Neurochemical predisposition to self-adminis-ter morphine in rats, Brain Res., 578 (1992) 215-220.

110 Glick, S.D., Shapiro, R.M., Drew, K.L., Hinds, P.A. and Carl-son, J.N., Differences in spontaneous and amphetamine-inducedrotational behavior and in sensitization to amphetamine, amongSprague-Dawley derived rats from different sourct’s, Physiol.Behac., 38 (1986) 67-70.

111 Glickman, S.E. and Schiff, B.B., A biological theory of rein-forcement, Psychol. Rev., 74 (1967) 81-109.

112 Gnegy, M.E., Hewlett, G.H.K., Yee, S.L. and W&h, M.J.,Alterations in calmodulin content and localization in areas ofrat brain after repeated intermittent amphetamine. Brain Res.,562 (1991) 6-12.

113 Gold, L.H., Swerdlow, N.R. and Koob, G.F., Tht: role ofmesolimbic dopamine in conditioned locomotion prmbed byamphetamine, Behav. Neurosci . , 102 (1988) 544-552 .

114 Goldstein, K.R., Knapp, D.J., Saiff, E.I., Pohorecky. L.A. andBenjamin, D., Sensitization to ethanol demonstrated in place-preference and locomotor activation, Sot Neurosci. Xxstr., 18(1992) 107.

115 Gratton, A., Wise, R.A. and Kiyatkin, E., Chronoampzrometricmeasurements of dopamine levels in the rat nucleus acc’umbensduring cocaine self-administration, Sot Neurosci. Abstr.. 1 S (1992)1076.

116 Greenberg, B.D. and Segal, D.S., Acute and chronic b&avioralinteractions between phencyclidine (PCP) and amphctarnine:evidence for a dopaminergic role in some PCP-induce11 behav-iors, Phurmucol. Biochem. B&al*. , 23 (1985) 99-105.

117 Greenberg, B.D. and Segal, D.S., Evidence for multiple opiatereceptor involvement in different phencyclidine-induced uncon-di t ioned behaviors in rats, P~cltoplrnrrllcrcolog)?, 88 (19% ) -14-53.

118 Grill, H.J. and Norgren, R., Chronically decerebriw ratsdemonstrate satiation but not bait shyness, Science, 201 (1978)267-269.

119 Grill, H.J. and Norgren, R., The taste reactivity test. I. niimeticresponses to gustatory stimuli in neurologically noriml rats,Brain Res., 143 (1978) 263-279.

120 Haertzen, C.A., Kocher, T.R. and Miyasato, K., Reinforcementsfrom the first drug experience can predict I:lter drug habitsand/or addiction: results with coffee, c&rcttes, alcohol. twbi-turates, minor and major tranquilizers, stimulants, nwijuma,hallucinogens, heroin, opiates and cocaine, Drl(,l: HwhoI De-pend., 11 (1983) 147-165.

285

121 Hamamura, T., Akiyama, K., Akimoto, K., Kashihara, K., Oku-murit, K., Ujikc, Ii. and Otsuki, S., Co-;tdn~inistr;ltion of either aselective Dl or D2 dopaminc antagonist with methamphctnmincprevents methanlphctnnline-induced behavioral sensitization andneurochemical change, studied by in vivo intracerebral dialysis,Brain Rc*.s., S-16 ( 199 1) W-46.

122 Ii;trris, G.C. and Williams, LT., Sensitization of locus cerulcusneurons during withdrawal from chronic stimulants and antide-pressants, .I. Hturr~rt~~l. Eq). T/w-., 261 (1992) 476-483.

123 Harsing, L.G., Sershen, I I. and Lajtha, A., Dopamine effluxfrom striatum after chronic nicotine: evidence for autoreceptordesensitization, J. Neurochem., 59 (1992) 48-54.

124 Henry, D.J. and White, F.J., Repeated cocaine administrationcauses persistent enhancement of Dl dopamine receptor sensi-tivity within the rat nucleus accumbens, J. Pharmacol. Exp.TIrer., 258 (199 1) 882-890.

125 Herman, J.P., Guillonneau, D., Dantzer, R., Scat&, B., Se-merdjian-Rouquier, L. and Le Meal, M., Differential effects ofinescapable footshocks and of stimuli previously paired withinescapable footshocks on dopamine turnover in cortical andlimbic areas of the rat, Life Sci., 30 (1982) 2207-2214.

126 Hernandez, L. and Hoebel, B.G., Food reward and cocaineincrease extracellular dopamine in the nucleus accumbens asmeasured by microdialysis, Life Sci., 42 (1988) 1705-1712.

127 Hill, R.T., Facilitation of conditioned reinforcement as a mecha-nism of psychomotor stimulation. In E. Costa and S. Garattini(Eds.), Amphetamine and Related Compounds, Raven Press,New York, 1970, pp. 781-795.

125 Hinson, R.E. and Poulos, C.X., Sensitization to the behavioraleffects of cocaine: modification by Pavlovian conditioning, Phar-macol. B&hem. Behau., 15 (1981) 559-562.

129 Hiroi, N. and White, N.M., Conditioned stereotypy: behavioralspecification of the UCS and pharmacological investigation ofthe neural change, Pharmacol. B&hem. Behav., 32 (1989) 249-258.

130 Hiroi, N. and White, N.M., The reserpine-sensitive dopaminepool mediates ( + )-amphetamine-conditioned reward in theplace preference paradigm, Brain Res., 510 (1990) 33-42.

131 Hoebel, B.G., Neuroscience and motivation: pathways and pep-tides that define motivational systems. In R.C. Atkinson, R.J.Herrnstein, G. Lindtay and R.D. Lance (Eds.), Steven’s Hand-book of Experimental Psychology, Vol. I, Perception and Motica-tion, John Wiley, New York, 1988, pp. 547-597.

132 Hooks, M.S., Colvin, A.C., Juncos, J.L. and Justice Jr., J.B.,Individual differences in basal and cocaine-stimulated extracel-lualr dopamine in the nucleus accumbens using quantitativemicrodialysis, Brain Res., 587 (1992) 306-312.

133 Hooks, M.S., Jones, G.H., Liem, B.J. and Justice Jr., J.B.,Sensitization and individual differences to IP amphetamine,cocaine or caffeine following repeated intracranial amphetamineinfusions, Pharmacol. B&hem. Behav., 43 (1992) 815-823.

134 Hooks, M.S., Jones, G.H., Neill, D.B. and Justice Jr., J.B.,Individual differences in amphetamine sensitization: dose-de-pendent effects, Pharmacol . B&hem. Behav . , 41(1992) 2 0 3 - 2 1 0 .

135 Hooks, M.S., Jones, G.H., Smith, A.D., Neill, D.B. and JusticeJr., J.B., Individual differences in locomotor activity and sensiti-za t ion , Pharmacol . B&hem. B e h a v . , 3 8 (1991) 467-470 .

136 Horger, B.A., Giles, M.K. and Schenk, S., Preexposure to am-phetamine and nicotine predisposes rats to self-administer a lowdose of cocaine, Psychopharmacology, 107 (1992) 271-276.

137 Horger, B.A., Shelton, K. and Schenk, S., Preexposure sensitizesrats to the rewarding effects of cocaine, Pharm. B&hem. Behav.,370990) 707-711.

138 1 forger, B.A., Wellman, P.J., Morien, A., Davies, B.T. andSchenk, S., Caffeine exposure sensitizes rats to the reinforcingeffects of cocaine, Neuroreport, 2 (1991) 53-56.

139 Horvitz, J.C. and Ettenberg, A., Conditioned incentive proper-ties of a food-paired conditioned stimulus remain intact duringdopamine receptor blockade, Behav. Neurosci., 105 (1991) 536-541.

140 Hu, X.-T., Brooderson, R.J. and White, F.J., Repeated stimula-tion of Dl dopamine receptors causes time-dependent alter-

ations in the sensitivity of both Dl and D2 dopamine receptorswithin the rat striatum, Neuroscience, 50 (1992) 137-147.

141 Hunt, H.F., Some effects of drugs on classical (type s) condition-.ing, Ann. NY Acad. Sci., 65 (1956) 258-267.

142 Hunt, W.A. and Lands, W.E., A role for behavioral sensitizationin uncontrolled ethanol intake, Alco/zol, 9 (1992) 327-328.

143 Hurd, Y.L., Weiss, F., Koob, G.F., And, N.E. and Ungerstedt,U., Cocaine reinforcement and extracellular dopamine overflowin rat nucleus accumbens: an in vivo microdialysis study, BrainRes., 498 (1989) 199-203.

144 Ichikawa, J., Changes in behavior and central monoaminergicsystems in the rat after repeated methamphetamine pretreat-ment: presynaptic regulatory mechanism, Yakubutsu SeishinKodo, 8 (1988) 3 8 9 - 4 0 3 .

145 Ingram, D.K. and Corfman, T.P., An overview of neurobiologi-cal comparisons in mouse strains, Neurosci. Biobehw. Rev., 4(1980) 421-435.

146 Invernizzi, R., Pozzi, L. and Samanin, R., Release of dopamineis reduced by diazepam more in the nucleus accumbens than inthe caudate nucleus of conscious rats, Neuropharmacolugy, 30(1991) 575-578.

147 Iwamoto, E.T., Comparison of the pharmacologic effects ofN-allylnormetazocine and phencyclidine: sensitization, cross-sensitization and opioid antagonist activity, Psychopharmacol-ogy, 89 (1986) 221-229.

148 Jaffe, J.H., Drug addiction and drug abuse. In A.G. Gilman,T.W. Rall , A.S. Nies and P. Taylor (Eds.), The PharmacologicalBasis of Therapeutics, Pergamon Press, New York, 1990, pp.522-573.

149 Jaffe, J.H., Current concepts of addiction. In C.P. O’Brien andJ.H. Jaffe (Eds.), Addictive States, Raven Press, New York,1992, pp. l-21.

150 Jaffe, J.H., Cascella, N.G., Kumor, KM. and Sherer, M.A.,Cocaine-induced cocaine craving, Psychopharmacology, 97 (1989)59-64.

151 Jones, B.E. and Prada, J.A., Drug-seeking behavior duringmethadone maintenance, Psychopharmacologia, 41 (1975) 7-10.

152 Jones, G.H., Marsden, C.A. and Robbins, T.W., Increased sensi-tivity to amphetamine and reward-related stimuli following so-cial isolation in rats: possible disruption of dopamine-dependentmechanisms of the nucleus accumbens, Psychopharmacology,102 (1990) 364-372.

.

153 Joyce, E.M. and Iversen, S.D., The effect of morphine appliedlocally to mesencephalic dopamine cell bodies on spontaneousmotor activity in the rat, Neurosci. Lett., 14 (1979) 207-212.

154 Kalivas, P.W., Sensitization to repeated enkephalin administra-tion into the ventral tegmental area of the rat. II. Involvementof the mesolimbic dopamine system, J. Pharmacol. Exp. Ther.,235(1985)544-550.

155 Kalivas, P.W. and Duffy, P., Effects of daily cocaine and mor-phine treatment on somatodendritic and terminal field dopaminerelease, J. Neurochem., 50 (1988) 1498-1504.

156 Kalivas, P.W. and Duffy, P., Similar effects of daily cocaine andstress on mesocorticolimbic dopamine neurotransmission in therat, Biol. Psychiatry, 25 (1989) 913-928.

157 Kalivas, P.W. and Duffy, P., Effect of acute and daily cocainetreatment on extracellular dopamine in the nucleus accumbens,Synapse, 5 (1990) 48-58.

158 Kalivas, P.W. and Duffy, P., Time course of extracellulardopamine and behavioral sensitization to cocaine: I. Dopamineaxon terminals, J. Neurosci., 13 (1993) 266-275.

159 Kalivas, P.W., Duffy, P., Abhold, R. and Dilts, R.P., Sensitiza-tion of mesolimbic dopamine neurons by neuropeptides andstress. In P.W. Kalivas and C.D. Barnes (Eds.), Sensitization inthe Nervous System, Telford Press, Caldwell, NJ, 1988, pp.119-143.

160 Kalivas, P.W. and Stewart, J., Dopamine transmission in theinitiation and expression of drug- and stress-induced sensitiza-tion of motor activity, Brain Res. Rev., 16 (1991) 223-244.

161 Karler, R., Calder, L.D., Chaudhry, I.A. and Turkanis, S.A.,Blockade of “reverse tolerance” to cocaine and amphetamine byMK-801, Life Sci., 45 (1989) 599-606.

286

162 Karler, R., Calder, L.D. and Turkanis, S.A., Changes in CNSsensitivity to cannabinoids with repeated treatment: toleranceand auxoesthesia, NIDA Res. Monogr., 54 (1984) 312-322.

163 Karler, R., Calder, L.D. and Turkanis, S.A., DNQX blockade ofamphetamine behavioral sensitization, Brain Res., 552 (1991)295-300.

164 Karler, R., Chaudhry, LA., Calder, L.D. and Turkanis, S.A.,Amphetamine behavioral sensitization and the excitatory aminoacids, Brain Res., 537 (1990) 76-82.

165 Katz, J.L. and Goldberg, S.R., Preclinical assessment of abuseliability of drugs, Agents and Actions, 23 (1988) 18-26.

166 Kazahaya, Y., Akimoto, K. and Otsuki, S., Subchronic metham-phetamine treatment enhances methamphetamine- or cocaine-induced dopamine efflux in vivo., Biol. Psychiarry, 25 (1989)903-912.

167 Keller, R.W., Maisonneuve, I.M., Carlson, J.N. and Glick, S.D.,Within-subject sensitization of striatal dopamine release after asingle injection of cocaine: an in vivo microdialysis study,Synapse, 11 (1992) 28-34.

168 Kelley, A.E. and Delfs, J.M., Dopamine and conditioned rein-forcement. I. Differential effects of amphetamine microinjec-tions into striatal subregions, Psychopharmacology, 103 (1991)187-196.

169 Kelley, A.E. and Delfs, J.M., Dopamine and conditioned rein-forcement. II. Contrasting effects of amphetamine microinjec-tion into the nucleus accumbens with peptide microinjectioninto the ventral tegmental area, Psychopharmacology, 103 (1991)197-203.

170 Khantzian, E.J., The self-medication hypothesis of addictivedisorders: focus on heroin and cocaine dependence, Am. I.Psychiatry, 142 (1985) 1259-1264.

171 Kita, T., Okamoto, M. and Nakashima, T., Nicotine-inducedsensitization to ambulatory stimulant effect produced by dailyadministration into the ventral tegmental area and the nucleusaccumbens in rats, Life Sci., 50 (1992) 583-590.

172 Kiyatkin, E., Wise, R.A. and Gratton, A., Chronamperometicmeasurements of dopamine levels in the rat nucleus accumbens,Sot. Neurosci. Abstr., 18 (1992) 374.

173 Knoblich, G., Curtis, D., Faustman, W.O., Zarcone, V., Stewart,S., Mefford, I. and King, R., Increased CSF HVA with cravingin long-term abstinent cocaine abusers, Biol. Psychiatry, 32(1992) 96-100.

174 Kokkinidis, L., Zacharko, R.M. and Anisman, H., Amphetaminewithdrawal: a behavioral evaluation, Life Sci., 38 (1986) 1617-1623.

175 Kolta, M.G., Scalzo, F.M., Ali, S.F. and Holson, R.R., Ontogenyof the enhanced behavioral response to amphetamine in am-phetamine-pretreated rats, Psychopharmacology, 100 (1990)377-382.

176 Kolta, M.G., Shreve, P., De Souza, V. and Uretsky, N.J., Timecourse of the development of the enhanced behavioral andbiochemical responses to amphetamine after pretreatment withamphetamine, Neuropharmhcology, 24 (1985) 823-829 .

177 Kolta, M.G., Shreve, P. and Uretsky, N.J., Effect of meth-ylphenidate pretreatment on the behavioral and biochemicalresponses to amphetamine, Eur . 1. Pharmacol . , 117 (1985) 279-2 8 2 .

178 Kolta, M.G., Shreve, P. and Uretsky, N.J., Effect of pretreat-ment with amphetamine on the interaction between am-phetamine and dopamine neurons in the nucleus accumbens,Neuropharmacology , 28 ( 1989) 9- 14.

179 Koob, G.F., Dopamine, addiction and reward, Sem. Neurosci., 4(1992) 139-148.

180 Koob, G.F. and Bloom, F.E., Cellular and molecular mecha-nisms of drug dependence, Science, 242 (1988) 715-723.

181 Koob, G.F. and Goeders, N.E., Neuroanatomical substrates ofdrug self-administration. In J.M. Liebman and S.J. Cooper(Eds.), The lVcrrrophar~zlacolo~ica1 Basis of Reward, Oxford Uni-versity Press, New York, 1989, pp. 214-263.

182 Koob, G.F., Stinus, I.,., Le Moal, M. and Bloom, F.E., Opponentprocess theory of motivation: neurobiological evidence from

studies of opiate dependence, Neurosci. Biobehac. Rev., 13 (1989)135-140.

183 Kosobud, A.E., Harris, A.E. and Chapin, J.K., Actions of drugsof abuse on reward-related activity in neurons of the ventraltegmental area and prefrontal cortex in the rat, Sot. Neur~~~~Abstr., 18 (1992) 900.

184 Kozlowski, L.T. and Wilkinson, D.A., Use and misuse of theconcept of craving by alcohol, tobacco and drug researchers, ~~~J. Addict., 82 ( 1987) 31-36.

185 Krieckhaus, E.E. and Wolf, G., Acquisition of sodium by rats:interactions of innate mechanisms and latent learning, J. COAX,Physiol. Psychol., 65 (1968) 197-201.

186 Ksir, C., Hakan, R,, Hall, D.P. and Kellar, K.J., Exposure tonicotine enhances the behavioral stimulant effect of nicotineand increases binding of [3H]acetylcholine to nicotinic receptors, Neuropharmacology , 24 (1985) 527-531.

187 Kuczenski, R. and Leith, NJ., Chronic amphetamine: isdopamine a link in or mediator of the development of toleranceor reverse tolerance ?, Pharmucol. Biochem. Behuc., 15 (1981)405-413.

388 Kuczenski, R. and Segal, D.S., Psychomotor stim tjlant-inducedsensitization: behavioral and neurochemical correlates. In P.W.Kalivas and C.D. Barnes (Eds.), Sensitization iu the NervousSystem, Telford Press, Caldwell, NJ, 1988, pp. 175-205.

189 Kuroki, T., Tsutsumi, T., Hirano, M., Matsumoto, T., Tate-bayashi, Y., Nishiyama, K., Uchimura, H., Shiraishi, A., Naka-hara, T. and Nakamura, K., Behavioral sensitization to beta-phenylethylamine (PEA): enduring modifications of specificdopaminergic neuron systems in the ra t , Psychopharmacology,102 (1990) 5-10.

190 La Hoste, G.J., Mormede, P., Rivet, J.M. and L-e Moal, M.,Differential sensitization to amphetamine and stress responsivityas a function of inherent lateral& Brain Res.. 453 (1988)381-384.

191 Lal, H., Miksic, S., Drawbaugh, R., Numan, R. and Smith, N.,Alleviation of narcotic withdrawal syndrome by conditional stim-uli, Parllorq J. Biol. Sci., 11 (1976) 251-262.

192 Lamb, R.J., Preston, K.L., Schindler, C., Meisch, R.A., Davis,F., Katz, J.L., Henningfield, J.E. and Goldberg, S.R.. The rein-forcing and subjective effects of morphine in post-addicts: adose-response study, J. Pharmacol. Exp. Ther., 259 (1991) 1165-1173.

193 Lasagna, L., von Felsinger, J.M. and Beecher, H.K.. Drug-in-duced mood changes in man. 1. Observations on hc&hy sub-jects, chronically ill patients and “postaddicts”, JAM4, 157(1955) 1006-1020.

194 Leith, N.J. and Kuczenski, R., Two dissociable conqwnents ofbehavioral sensitization following repeated amphetamine admin-istration, Psychopharmacology, 76 (1982) 310-5.

195 Lett, B.T., Repeated exposures intensify rather th;jn diminishthe rewarding effects of amphetamine, morphine and cocaine,Psychopharmacology , 98 (1989) 3 5 7 - 3 6 2 .

196 Leyton, M. and Stewart, J., Preexposure to foot-shock wnsitizesthe locomotor response to subsequent systemic morphine andintra-nucleus accumbtns amphetamine, I%crrr?rczcol. IJiochem.Behaly., 37 (1990) 303-310.

197 Lin-Chu, G., Robinson, T.E. and Becker, J.B., Sensiti.T:ltion ofrotational behavior produced by a single exposure to cocaine,Pharmacol. Biochcm Behat?. , 22 (1985) 901-903.

198 Ljungberg, T., Apicella, P. and Schultz, W., Responses of mon-key midbrain dopaminc neurons during delayed alterrut i:m per-formance, Brain Res., 567 (199 1) 337-341.

199 Ljungberg, T., Apicclln, P. and Schultz, W., Responses of mon-key dopamine neurons during learning of bch:lvioral rc x’t ions,J. Ncurophysiol. , 67 ( 1992) 135 - 1 b3.

200 MacLennan, A.J. :rntl Mnier, S.F., Coping md the swss-in-duced potentiation of stimulant stcrcotypy in the rat, Suh’e,219 (1983) 1091-1093.

201 Magos, L., Pcrsistencc of the cffcct of ampht’taminc on stereo-typed ;jct ivity OII rats, IGrr. J. Pharrn(u-01.. 6 (l%% 200-201.

202 Maj, J. and Wcdzony, K., Repeated trt%ment with imipr:tmine

LO.3 Maj. J., Wrdzony, K. and Klimek, V., Desipramine given repeat-edly enhances behavioural effects of dopamine and d-amphetamine injected into the nucleus accumbens, &rr. J. !%ur-??lClCOi., 140 (1987) 179- 185.

204 blarkou, A. and Koob, G.F., Postcocaine anhedonia: an animalmodel of cocaine withdrawal, Nelcropsychopllarrzlncolo~, 4(1991) 17-X.

20s Martin, G.M., Bcchara, A. and van der Kooy, D., The perccp-tion of emotion: parallel neural processing of the affective anddiscriminative properties of opiates, Psychobiology, 19 (1991)147- 152.

206 Martin, S.D., Yeragani, V.K., Lodhi, R. and Galloway, M.P.,Clinical ratings and plasma HVA during cocaine abstinence,

0 Biol. Psychititry, 26 (1989) 356-362.207 Martin-Iverson, M.T., Ortmann, R. and Fibiger, H.C., Place

preference conditioning with methylphenidate and nomifensine,Bruin Res., 332 ( 1985) 59-67.

208 Masur, J. and Boerngen, R., The excitatory component ofethanol in mice: a chronic study, Pharmacol. Biochem. Behav.,13 (1980) 777-780.

209 Masur, J., Oliveira de Souza, M.L. and Zwicker, A.P., Theexcitatory effect of ethanol: absence in rats, no tolerance andincreased sensitivity in mice, Pharmacol. Biochem. Behav., 24(1986) 1225-1228.

2 10 Mazurski, E.J. and Beninger, R.J., Environment-specific condi-tioning and sensitization with (+ )-amphetamine, Pharmacol.Biochem. Behal 1.) 27 ( 1987) 6 l-65.

211 McAuliffe, W.E., A second look at first effects: the subjectiveeffects of opiates on nonaddicts, J. Drug Issues, 5 (1975) 369-399.

212 McAuliffe, W.E., A test of Wikler’s theory of relapse: thefrequency of relapse due to conditioned withdrawal sickness,Int. J. Addict, 17 (1982) 19-33.

213 McAuliffe, W.E. and Gordon, R.A., A test of Lindesmith’stheory of addiction: the frequency of euphoria among long-termaddicts, Am. J. Social., 79 (1974) 795-840.

214 Meisel, R.L., Camp, D.M. and Robinson, T.E., A microdialysisstudy of ventral striatal dopamine during sexual behavior infemale Syrian hamsters, Behav. Brain Res . , i n p r e s s .

215 Meyer, R.E., Conditioning phenomena and the problem ofrelapse in opioid addicts and alcoholics, NIDA Res. Monogr., 84(1988) 161-179.

216 Miller, G.A., Galanter, E. and Pribram, K.H., Plans and theStructure of BehaGor, Holt, Rinehart & Winston, New York,1965.

217 Miller, J.D., Sanghera, M.K. and German, DC., Mesencephalicdopaminergic unit activity in the behaviorally conditioned rat,Life Sci., 29 (1981) 1255-1263.

21X Mitchell, J.B. and Stewart, J., Facilitation of sexual behaviors inthe male rat associated with intra-VTA injections of opiates,Pharmacol. B&hem. Behav., 35 (1990) 643-650.

219 Mithani, S., Martin-Iverson, M.T., Phillips, A.G. and Fibiger,H.C., The effects of haloperidol on amphetamine- and meth-ylphenidate-induced conditioned place preferences and locomo-tor activity, Psychopharmucology, 90 (1986) 247-252.

220 Mittleman, G., Castafieda, E., Robinson, T.E. and Valenstein,E.S., The propensity for nonregulatory ingestive behavior isrelated to differences in dopamine systems: behavioral andbiochemical evidence, Behav. Neurosci., 100 (1986) 213-220.

221 Mittleman, G. and Valenstein, E.S., Individual differences innon-regulatory ingestive behavior and catecholamine systems,Brain Res., 348 (1985) 112-117.

222 Morency, M.A. and Beninger, R.J., Dopaminergic substrates ofcocaine-induced place conditioning, Brain Res., 399 (1986) 33.41.

223 Morley, J.E., Levine, A.S. and Rowland, N.E., Stress inducedeating, Life Sci., 32 (1983) 2169-2182.

224 Mueller, K., Kunko, P.M., Whiteside, D. and Haskett, C., Timecourse of amphetamine-induced locomotor s tereotypyopen field, Psychopharmacology, 99 ( 1989) 501-507.

in an

287

225 Nabeshima, T., Fukaya, H., Yamaguchi, K., Ishikawa, K., Fu-rukawa, H. and Kameyama, T., Development of tolerance andsupersensitivity to phencyclidine in rats after repeated adminis-tration of phencyclidinc, /Grr. J. flrcrrmacol., 135 (1987) 23-33.

226 Nakayama, M., Koyama, T. and Yamashita, I., Long-lastingdecrease in dopamine uptake sites following repeated adminis-tration of methamphetamine in the rat striatum, Bruin Res., 601(1993) 209-212.

227 Nestler, E.J., Moltxular mechanisms of drug addiction, J. Neu-rosci., 12 (1992) 2439-2450.

228 Nisbett, R.E. and Wilson, T.D., Telling more than we can know:verbal reports on mental processes, Psychol. Rev., 84 (1977)231-259.

229 O’Brien, C.P., Childress, A.R., McLellan, A.T. and Ehrman, R.,A learning model of addiction. In C.P. O’Brien and J.H. Jaffe(Eds.), Addictive States, Raven Press, New York, 1992, pp.157- 177.

230 Parker, L.A., Taste reactivity responses elicited by reinforcingdrugs: a dose-response analysis, Behav. Neurosci., 105 (1991)955-964.

231 Patrick, S.L., Thompson, T.L., Walker, J.M. and Patrick, R.L.,Concomitant sensitization of amphetamine-induced behavioralstimulation and in vivo dopamine release from the rat caudatenucleus, Bruin Res., 538 (1991) 343-346.

232 Paulson, P.E., Camp, D.M. and Robinson, T.E., The time courseof transient behavioral depression and persistent behavioralsensitization in relation to regional brain monoamine concentra-tions during amphetamine withdrawal in rats, Psychopharmacol-ogy, 103 (1991) 480-492.

233 Paulson, P.E. and Robinson, T.E., Sensitization to systemicamphetamine produces an enhanced locomotor reoponse to asubsequent intra-accumbens amphetamine in rats, Psychophar-macology, 104 (1991) 140-141.

234 PCrez-Cruet, J., Conditioning of striatal dopamine metabolismwith methadone, morphine or bulbocapnine as an uncondi-tioned stimulus, Pal*lov J. Biol. Sci., 11 (1976) 237-250.

235 Pirez-de la Mora, M., tipez-Quiroz, D., Mkndez-France, J.and Drucker-Colin, R., Chronic administration of amphetamineincreases glutamic acid decarboxylase activity in the rat substan-tia nigra, Neurosci. Lett., 109 (1990) 315-320.

236 Peris, J. and Zahniser, N.R., One injection of cocaine producesa long-lasting increase in [ “HI-dopamine release, Pharmacol.Biochem. Behav . , 27 (1987) 533-535 .

237 Peris, J. and Zahniser, N.R., Persistent augmented dopaminerelease after acute cocaine requires dopamine receptor activa-tion, Pharmacol. B&hem. Behav., 32 (1989) 71-76.

238 Pettit, H.O., Pan, H.T., Parsons, L.H. and Justice Jr., J.B.,Extracellular concentrations of cocaine and dopamine are en-hanced during chronic cocaine administration, J. Neurochem.,55 (1990) 798-804.

.

239 Phillips, A.G., Pfaus, J.G. and Blaha, C.D., Dopamine andmotivated behavior: insights provided by in vivo analyses. In P.Willner and J. Scheel-Kruger (Eds.), The Mesolimbic DopamineSystem: from Motivation to Action, John Wiley, New York, 1991,pp. 199-224.

240 Phillips, T.J. and Crabbe, J.C., Behavioral studies of geneticdifferences in alcohol action. In J.C. Crabbe and R.A. Harris(Eds.), The Genetic Basis of Alcohol and Drug Actions, PlenumPress, New York, 1991, pp. 25-104.

241 Piazza, P.V., Deminigre, J.-M., Maccari, S., Le Moal, M.,Morm&de, P. and Simon, H., Individual vulnerability to drugself-administration: action of corticosterone on dopaminergicsystems as a possible pathophysiological mechanism. In P. Will-ner and J. Scheel-Kriiger (Eds.), The Mesolimbic DopamineSystem: From Motivation to Action, John Wiley, New York,1991, pp. 473-495.

242 Piazza, P.V., DeminiZre, J.M., Le Moal, M. and Simon, H.,Factors that predict individual vulnerability to amphetamineself-administration, Science, 245 (1989) 1511-1513.

243 Piazza, P.V., Deminiire, J.M., Le Moal, M. and Simon, H.,Stress- and pharmacologically induced behavioral sensitization

2 8 8

increases vulnerability to acquisition of amphetamine self-ad-ministration, Bruin Res., 514 (1990) 22-26.

244 Piazza, P.V., Deroche, V., Roug&Pont, F., Deminiire, J.M.,Maccari, S., Le Moal, M. and Simon, H., Individual differencesin the sensitivity to corticosterone’s reinforcing effects and incorticosterone-induced dopamine release may be a biologicalbasis for sensation-seeking, Sot. Neurosci. Abstr., 18 (1992) 1076.

245 Pinel, J.P.J. and Mana, M.J., Adaptive interactions of rats withdangerous inanimate objects: support for a cognitive theory ofdefensive behavior. In R.J. Blanchard, P.F. Brain, D.C. Blan-chard and S. Parmigiani (Eds.), Ethoexperimental Approaches tothe Study of Behavior , Kluwer Academic Publ ishers , Dordrecht ,The Netherlands, 1989, pp. 137-150.

246 Pleim, E.T., Matochik, J.A., Barfield, R.J. and Auerbach, S.B.,Correlation of dopamine release in the nucleus accumbens withmasculine sexual behavior in rats, Brain Res., 524 (1990) 160-163.

247 Poncelet, M., Dangoumau, L., Soubrie, P. and Simon, P., Effectsof neuroleptic drugs, clonidine and lithium on the expression ofconditioned behavioral excitation in rats, Psychopharmacology,92 (1987) 393-397.

248 Post, R., Intermittent versus continuous stimulation: effect oftime interval on the development of sensitization or tolerance,Life Sci., 26 (1980) 1275-1282.

249 Post, R.M., Cocaine psychoses: a continuum model, Am. J.Psychiatry, 132 (1975) 225-231.

250 Post, R.M. and Contel, N.R., Human and animal studies ofcocaine: implications for development of behavioral pathology.In I. Creese (Ed.), Stimulants: Neurochemical, Behal.ioral andClinical Perspectives, Raven Press, New York, 1983, pp. 169-203.

251 Post, R.M., Lockfeld, A., Squillace, KM. and Contel, N.R.,Drug-environment interaction: context dependency of cocaine-induced behavioral sensitization, Life Sci., 28 (1981) 755-760.

252 Post, R.M., Weiss, S.R., Fontana, D. and Pert, A., Conditionedsensitization to the psychomotor stimulant cocaine. In P,W.Kalivas and H.H. Samson (Eds.), Annals of the New YorkAcademy of Sciences: The Neurobiology of Drug and AlcoholAddiction, The New York Academy of Sciences, New York,1992, pp. 386-399.

253 Powell, J., Bradley, B. and Gray, J., Classical conditioning andcognitive determinants of subjective craving for opiates: an

. investigation of their relative contributions, Br. J. Addict., 87(1992) 1133-1144.

254 Randrup, A. and Munkvad, I., Stereotyped activities producedby amphetamine in several animal species and man, Psychophar-macologia, 11 (1967) 300-310.

255 Rapoport, J.L., The biology of obsessions and compulsions, Sci.Am., 260 (1989) 82-89.

256 Rapoport, J.L., Obsessive compulsive disorder and basal gangliadysfunction, Psychol. Med., 20 (1990) 465-469.

257 Rebec, G.V., Changes in brain and behavior produced by am-phetamine: a perspective based on microdialysis, voltammeryand single-unit electrophysiology in freely moving animals. InR.R. Watson (Ed.), Biochemistry and Physiology of Substance

Abuse, CRC Press, Boca Raton, FL, 1991, pp. 93-l 15.258 Rebec, G.V. and Segal, D.S., Enhanced responsiveness to intra-

ventricular infusion of amphetamine following its repeated sys-temic administration, Psychophurmacology, 62 (1979) 101-102.

259 Rescorla, R.A., Hierarchical associative relations in Pavlovianconditioning and instrumental training, Curr. Direct Psychol.Sci., 1 (1992) 66-70.

260 Risinger, F.O., Dickinson, S.D. and Cunningham, C.L.,Ilaloperidol reduces ethanol-induced motor activity stimulationbu t no t cond i t ioned p lace preference, Psychopharmacology, 1 0 7(1992) 453-456.

261 Rivet , J .M., St inus, L. , Le Moal, M. and Mormede, P., Behav-ioral sensitization to amphetamine is dependent on corticos-teroid receptor activation, Brain Res., 498 (1989) 149-153.

262 Robbins, T.W., Cador, M., Taylor, J.R. and Everitt, B.J., Lim-bit-striatal interactions in reward-related processes, Neurosci.Biobehn I*. Rcr 7. , 13 (1989) 155-162.

263 Robbins, T.W. and Everitt, B.J., Functions of dopaminc in thedorsal and ventral striatum, SC/X Newosci., 4 (1992) 119-127.

264 Robbins, T.W., Watson, B.A., Gaskin, M. and Ennis, C., ~~~trasting interactions of pipradrol, d-amphetamine, cocaine, aCaine analogues, apomorphine and other drugs with conditionerreinforcement, Psychopharmacology, 80 (1983) 113-119.

265 Roberts-Lewis, J .M., Roseboom, P.H,, Iwaniec, L.M. and Gnegy, M.E., Differential down-regulation of Dl-stimulated adenylate cyclase activity in rat forebrain after in vivo amphetaminetreatments, J, Neurosci., 6 (1986) 2245-2251.

266 Robinson, J.H. and Pritchard, W.S., The meaning of addiction:rep ly to West , Psychopharmacology , 108 (1992) 41 l - 4 1 6 .

267 Robinson, T.E., Behavioral sensitization: characterization (enduring changes in rotational behavior produced by interm;tent injections of amphetamine in male and female rats, ~5chopharmacology , 84 ( 1984) 466-475.

268 Robinson, T.E., Stimulant drugs and stress: factors influencin,individual differences in the susceptibility to sensitization. InP.W. Kalivas and C. Barnes (Eds.), Sensitization of the NervousSystem, Telford Press, Caldwell, N.J., 1988, pp. 145-173.

269 Robinson, T.E., The neurobiology of amphetamine psychosis:evidence from studies with an animal model. In T. Nakazawa(Ed .), Taniguchi Symposia on Brain Sciences , Vol . 14 , BiologicalBasis of Schizophrenia, Japan Scientific Societies Press, Tokyo,1991, pp. 185-201.

270 Robinson, T.E., Persistent sensitizing effects of drugs on braindopamine systems and behavior: implications for addiction andrelapse. In S.G. Korenman and J.D. Barchas (Eds.). The Biologi-cal Basis of Substance Abuse, Oxford University Press, NewYork, in press.

271 Robinson, T.E. and Becker, J.B., Behavioral sensitization isaccompanied by an enhancement in amphetamine-stimulateddopamine release from striatal tissue in vitro, Eur. J. Pharmacol.,85 (1982) 253-254.

272 Robinson, T.E. and Becker, J.B., Enduring changes in brain andbehavior produced by chronic amphetamine administration: areview and evaluation of animal models of amphet‘imine psy-chosis, Brain Res. Ret*., 396 (1986) 157-198.

273 Robinson, T.E., Becker, J.B. and Presty, SK., Long-term facili-tation of amphetamine-induced rotational behavior and striataldopamine release produced by a single exposure to am-phetamine: sex differences, Brain Res., 253 (1982) 2X-241.

274 Robinson, T.E., Jurson, P.A., Bennett, J.A. and Bentgen, KM.,Persistent sensitization of dopamine neurotransmis&Jn in ven-tral striatum (nucleus accumbens) produced by past experiencewith ( + )-amphetamine: a microdialysis study in frw$ movingrats, Br. Res., 462 (1988) 21 l-222.

275 Romo, R. and Schultz, W., Dopamine neurons of the monkeymidbrain: contingencies of responses to active touch duringself-initiated arm movements, J. Neurophysiol., 63 ( WO) 5%606.

276 Roseboom, P.H. and Gnegy. M.E., Acute in vivo amphetamineproduces a homolgous desensitization of dopamine receptor-coupled adenylate cyclase activities and decreases agw%t bind-ing to the Dl site, Mol. Phwmacof., 34 (1989) 138~1%.

277 Rossetti, Z.L., D’Aquila, P.S., Hmaidan, Y., Gessa, G.L. andSerra, G., Repeated treatment with imipramine potenriates CO-Caine-induced dopamine release and motor stimulation. Eur. J.Phurmucol., 201 (1991) 243-245.

278 Rozin, P. and Fallon, A.E., A perspective on disgust. Psychol.Rel’., 94 (1987) 23-41.

279 Sato, M., Acute exacerbation of methamphetamine psychosisand lasting dopaminergic supr:rscnsitivity - a clinical survey,P~lc’~2i)I)l)urt)1ucol. Bull., 22 (19%) 75 I-756.

280 Sato, M., Chen, C.C., Akiy:ulla, K. and Otsuki, S., Acute) wwr-bation of paranoid psychotic state after long-term abstiwnce inpatients with previous methamphetrln~ine psychosis, ~‘wI. &Y-chiutr)), 18 (1983) 429-340.

281 Schcnk, S., Lacclle, G., German, K. and Amit, Z., (‘(Icaineself-administration in r;its illtluenced by en\*ironment;tl condi-tions: implications for the t’tiology of drug abuse, Ah-osci.ht., 81 (1987) 227-231.

282 Schcnk, S., V;rl;\dcz, A., McNamar;ti, C. and l-lorgcr-. B.A.,Blockxic of sensitizing cfft’cts of amphct;unillc prwqww-e on

289

cocninc self-atirnirlistr;lti(~n by the NMDA ant;lgonist MK-SOI,Sot. Mwosci. .4lwW., 1 S ( WL) 1237.

283 Schiff, S.R., ~onditioried dopaniirlergic activity, Biol. Psychiatry,17 (Pm 135- 15-1.

283 Schiot-ring, E., Amphetamine induced stlcctive stimulation ofcertain hehaviour itans with concurrent inhibition of others inan open-field test with rats, RchaC’orrr, 39 (1971) l-17.

285 Schiorring, E., An open field study of stereotyped locomotoractivity in amphetamine-trcatcj rats, f.sychclykartrlacolo~~, 66(1979) 281-287.

286 Schiorring. E.. Psychopathology induced by “speed drugs”,Pharmacol. Biochcm. Bchar *. , 1 ( 198 1) 109- 122.

287 Schultz, W., Responses of midbrain dopamine neurons to behav-ioral trigger stimuli in the monkey, J. Ncurophysiol., 56 (1986)1339-1361.

(Ed.), P.~yc.llo~~~J~lrf7lac~~l~~gy in the Practice of MdicirJe, Applc-ton-Century-Crofts, New York, 1977.

306 Siegel, S., Drug anticipation and drug toltxmce. In M. Lader(Ed.), ~?/Jc Psyc.llo~?tlcrrrrJacology of Addiction, Oxford UniversityPress, New York, 1988, pp. 73-97.

307 Skinner, B.F., Are theories of learning necessary?, Psychol. ReL:,57 (1950) 193-216.

308 Smith, G.P. and Schneider, L.H., Relationships betweenmesolimbic dopamine function and eating behavior, Ann. NYAcad. Sci., 537 (1988) 254-261.

288 Schultz. W., Activity of dopamine neurons in the behavingprimate, Sum. Neurosci., 4 (1992) 129-138.

289 Schultz. W. and Romo. R., Dopamine neurons of the monkeymidbrain: contingencies of responses to stimuli eliciting immedi-ate behavioral reactions, J. Neurophysiol., 63 ( 1990) 607-624.

290 Schuster, C., Drug-seeking behaviour: implications for theoriesof drug dependence. In G. Edwards and M. Lader (Eds.), TheNature o f Drug Dependence, Oxford, New York, 1990, pp. 17 l-193.

309 Smith, I.E., Co, C., Freeman, M.E. and Lane, J.D., Brainneurotransmitter turnover correlated with morphine-seeking be-havior of rats, Pharmacol. Biochcm. BehaLl., 16 (1982) 509-519.

310 Smith, J.E. and Dworkin, S.I., Behavioral contingencies deter-mine changes in drug-induced transmitter turnover, Drug Dec.Res., 20 (1990) 337-348. .

311 Sijderpalm, B., Svensson, L.FHulthe, P., Johannessen, K. andEngel, J.A., Evidence for a role for dopamine in the diazepamlocomotor st imulating effect , Psychopharmacology, 104 (1991)97-102.

291 Segal, D.S., Behavioral and neurochemical correlates of re-peated d-amphetamine administration, A~u. B&hem. Psy-chopharmacol . , 1 3 (197% 2 4 7 - 2 6 2 .

292 Segal, D.S., Geyer, M.A. and Schuckit, M.A., Stimulant-inducedpsychosis: an evaluation of animal models, Essays Neurochem.Neuropharmacol., 5 ( 198 1) 9% 129.

293 Segal, D.S., Kelly, P.H., Koob, G. and Roberts, D.C.S., Nonstri-atal dopamine mechanisms in the response to repeated d-amphetamine administration. In E. Usdin, I. Kopin and J.Barchas (Eds.), Catecholamines: Basic and Clinical Frontiers , Vol.2, Pergamon Press, New York, 1979, pp. 1672-1674.

294 Segal, D.S. and Kuczenski, R., Individual differences in respon-siveness to single and repeated amphetamine administration:behavioral characteristics and neurochemical correlates, J. Phar-macol. fip. Ther., 242 (1987) 917-926.

295 Segal, D.S. and Kuczenski, R., In vivo microdialysis reveals adiminished amphetamine-induced DA response correspondingto behavioral sensitization produced by repeated amphetaminepretreatment, Brain Res., 571 (1992) 330-337.

296 Segal, D.S. and Kuczenski, R., Repeated cocaine administrationinduces behavioral sensitization and corresponding decreasedextracellular dopamine responses in caudate and accumbens,Brain Res., 577 (1992) 351-355.

312 Solomon, R.L., Acquired motivation and affective opponent-processes. In J . Madden (Ed.) , Neurobiology o f Learning, Emo-tion and Affect, Raven Press, New York, 1991, pp. 307-348.

313 Sorensen, S.M., Hattox, S., Johnson, S.W., Bickford, P., Mur-phy, R. and Freedman, R., Norepinephrine-dependent and in-dependent mechanisms of persistent effects of amphetamine inrat cerebellum, Life Sci., 36 (1985) 2383-2389.

314 Sorensen, S.M., Johnson, S.W. and Freedman, R., Persistenteffects of amphetamine on cerebellar Purkinje neurons follow-ing chronic administration, Brain Res., 247 (1982) 365-371.

315 Sorg, B.A. and Kalivas, P.W., Effects of cocaine and footshockstress on extracellular dopamine levels in the ventral striatum,Brain Rex, 559 (1991) 29-36.

297 Segal, D.S. and Schuckit, M.A., Animal models of stimulant-in-duced psychosis . In I . Creese (Ed.) , Stimulants: Neurochemical,Behacioral and Clinical Perspectives, Raven Press, New York,1983, pp. 131-167.

316 Spanos, L.J. and Yamamoto, B-K., Acute and subchronic effectsof methylenedioxy-methamphetamine [( + /-)MDMA] on loco-motion and serotonin syndrome behavior in the rat, Pharmacol.B&hem. Behal!., 32 (1989) 835-840 [published erratum appearsin Pharmacol. Biochem. Behav. 1989, 34:679].

317 Spyraki, C., Fibiger, H.C. and Phillips, A.G., Cocaine-inducedplace preference conditioning: lack of effects of neurolepticsand 6-hydroxydopamine lesions, Brain Res., 253 (1982) 195-203.

318 Squire, L.R. and Zola-Morgan, S., The media1 temporal lobememory system, Science, 253 (1991) 1380-1386.

319 Steiner, J.E., Human facial expressions in response to taste andsmell stimulation, Adc. Child Dec. Beharq., 13 (1979) 257-295.

320 Steketee, J.D., Sorg, B.A. and Kalivas, P.W., The role of thenucleus accumbens in sensitization to drugs of abuse, Prog.Neuropsychopharmacol . Biol . Psychiatry, 1 6 (1992) 2 3 7 - 2 4 6 .

321 Stevens, J.R., An anatomy of schizophrenia?, Arch. Gen. Psychi-atry, 29 (1973) 177-189.

298 Sherer, M.A., Intravenous cocaine: psychiatric effects, biologicalmechanisms, Biol. Psychiatry, 24 (1988) 865-885.

299 Shippenberg, T.S., Emmett-Oglesby, M.W., Ayesta, F.J. andHerz, A., Tolerance and selective cross-tolerance to the motiva-tional effects of opioids, Psychopharmacology, 96 (1988) 1 lo- 115.

300 Short, P.H. and Shuster, L., Changes in brain norepinephrineassociated with sensitization to d-amphetamine, Psychopharma-cology, 48 (1976) 59-67.

322 Stewart, J., Reinstatement of heroin and cocaine self-adminis-tration behavior in the rat by intracerebral application of mor-phine in the ventral tegmental area, Pharmacol. Biochem. Be-hao., 20 (1984) 917-923.

323 Stewart, J., Conditioned stimulus control of the expression ofsensitization of the behavioral activating effects of opiate andstimulant drugs. In I. Gormezano and E.A. Wasserman (Eds.),Learning and Memory: The Behaliorul and Biological Substrutes,Erlbaum, Hillsdale, NJ, 1992, pp. 129-151.

301 Shuster, L., Hudson, J., Anton, M. and Righi, D., Sensitizationof mice to methylphenidate, Psychopharmacology, 77 (1982)31-36.

302 Shuster, L., Webster, G.W. and Yu, G., Increased runningresponse to morphine in morphine-pretreated mice, J. Pharma-col. Exp. Ther., 192 (1975) 64-67.

324 Stewart, J., Tolerance and sensitization to the behavioral effectsof drugs. In P. Nencini (Ed.), Controllo Farmacologico de1 C’om-portamento, Unione Tipografico, Torino, Italy, 1992, pp. 161-183.

303 Shuster, L., Webster, G.W. and Yu, G., Perinatal narcoticaddiction in mice: sensitization to morphine stimulation, Addict.Dis., 2 (1975) 277-292.

325 Stewart, J. and De Wit, H., Reinstatement of drug-taking behav-ior as a method of assessing incentive motivational properties ofdrugs. In M.A. Bozarth (Ed.), Methods of Assessing the Reinforc-ing Properties of Abused Drugs, Springer-Verlag, New York,1987, pp. 211-227.

304 Shuster, L., Yu, G. and Bates, A., Sensitization to cocaine 326 Stewart, J., de Wit, H. and Eikelboom, R., Role of uncondi-stimulation in mice, Psychopharmacology, 52 (1977) 185-190. tioned and conditioned drug effects in the self-administration of

305 Siegel, S., Learning and psychopharmacology. In M.E. Jarvik opiates and stimulants, Psychol. Rev., 91 (1984) 251-268.

290

327 Stewart, .J. and Vezina, P., A comparison of the effects ofintra-accumbens injections of amphetamine and morphine onreinstatement of heroin intravenous self-administration behav-ior, Brain Res., 457 (1988) 287-294.

328 Stewart, J. and Vezina, P., Extinction procedures abolish condi-tioned stimulus control but spare sensitized responding to am-phetamine, Behal*. Pharmacol., 2 (1991) 65-71.

329 Stewart, J. and Wise, R.A., Reinstatement of heroin self-admin-istration habits: morphine prompts and naltrexone discouragesrenewed responding after extinction, Psychopharmacology, 108(1992) 79-84.

330 Strecker, R.E., Steinfels, G.F. and Jacobs, B.L., Dopaminergicunit activity in freely moving cats: lack of relationship to feeding,satiety and glucose injections, Brain Res., 260 ( 1983) 317-321.

331 Swerdlow, N.R., Braff, D.L., Masten, V.L. and Geyer, M.A.,Schizophrenic-like sensorimotor gating abnormalities in rats fol-lowing dopamine infusion into the nucleus accumbens, Psy-chopharmacology, 1 0 1 (1990) 4 1 4 - 4 2 0 .

332 Swerdlow, N.R. and Koob, G.F., Dopamine, schizophrenia, ma-nia and depression: toward a unified hypothesis of cortico-striato-pallido-thalamic function, Behar: Brain Sci., 10 (1987)197-245.

333 Taylor, J.R. and Robbins, T.W., Enhanced behavioural controlby conditioned reinforcers following microinjections of d-amphetamine into the nucleus accumbens, Psychopharmacology,84 (1984) 405-412.

334 Taylor, J.R. and Robbins, T.W., 6-Hydroxydopamine lesions ofthe nucleus accumbens, but not of the caudate nucleus, attenu-ate enhanced responding with reward-related stimuli producedby intra-accumbens d-amphetamine, Psychopharmacology, 90( 1986) 390-397.

335 Ternes, J .W., Ehrman, R.N. and O’Brien, C.P. , Nondependentmonkeys self-administer hydromorphone, Behal*. Neurosci., 99(1985) 583-588.

336 Thierry, A.M., Tassin, J.P., Blanc, G. and Glowinski, J., Selec-tive activation of the mesocortical DA system by stress, Nature,263 ( 1976) 242-243.

337 Tiffany, S.T., A cognitive model of drug urges and drug-usebehavior: role of automatic and nonautomatic processes, Psy-chol. Rev., 97 (1990) 147-168.

338 Tilson, H.A. and Rech, R.A., Conditioned drug effects andabsence of tolerance to d-amphetamine induced motor activity,Pharmacol. Biochem. Behal*., 1 (1973) 149-153.

339 Toates, F., Motkational Systems, Cambridge University Press,Cambridge, 1986.

340 Treit, D. and Berridge, KC., A comparison of benzodiazepine,serotonin and dopamine agents in the taste-reactivity paradigm,Pharmacol. Biochem. Behar q., 37 ( 1990) 45 l-456.

341 Trulson, M.E. and Preussler, D.W., Dopamine-containing ven-tral tegmental area neurons in freely moving cats: activity duringthe sleep-waking cycle and effects of stress, Exp. Neurol., 83( 1984) 367-377.

342 Ujike, H., Tsuchida, H., Kanzaki, A., Akiyama, K. and Otsuki,S., Competitive and non-competitive N-methyl-D-aspartate an-tagonists fail to prevent the induction of methamphetamine-in-duced sensitization, Lz”e Sci., 50 (1992) 1673-1681.

343 Vaccarino, F.J., Schiff, B.B. and Glickman, S.E., Biological viewof reinforcement. In S.B. Klein and R.R. Mowrer (Eds.), Corz-tempory Learning Theories, Erlbaum, Hillsdale, NJ, 1989, pp.111-142.

344 Vezina, P. and Stewart, J., Conditioning and place-specific sen-sitization of increases in activity induced by morphine in theVTA, Pharmncol. Riochem. Bcha~ , 20 ( 1984) 925-934.

345 Vezina, P. and Stewart, J., The cffcct of dopamine receptorblockade on the development of sensitization to the locomotoractivating effects of amphetamine and morphine, Brain Res., 499(1989) 108-120.

346 Vezina, P. and Stewart, J., Amphetamine administered to theventral tegmcntal area but not to the nucleus accumbens scnsi-tizes rats to systemic morphine: I;lck of conditioned effects,Brain Rex, 516 (1990) 99- 106.

347 Vezina, P. and Zurakowski, B., Prior exposure to jntra-WAamphetamine sensitizes the nucleus accumbens dopamjnergicresponse to systemic amphetamiqe as measured by invivo micro-dialysis, Sot. Neurosci. Abstr., 18 (1992) 720.

348 Walter, S. and Kuschinsky, K., Conditioning of mor@ljne-inwduced locomotor activity and stereotyped behaviour in rats, J,Neural Transm. Gen. Sect., 78 (19891 231-247.

349 Watson, S.J., Trujillo, K.A., Herman, J.P. and Akil, H., Neu-roanatomical and neurochemical substrates of drug-seeking be-havior: overview and future directions. In A. Goldstein (Ed.),Molecular and Cellular Aspects of the Drug Addictions, Springer-Verlag, New York, 1989, pp. 29-91.

350 Weiskrantz, L., Warrington, E.K., Sanders, M.D. and Marshall,J., Visual capacity in the hemianopic field following a restrictedoccipital ablation, Brain, 97 (1974) 709-728.

351 Weiss, F., Markou, A., Lorang, M.T. and Koob, G.F., Basalextracellular dopamine levels in the nucleus accumbens aredecreased during cocaine withdrawal after unlimited-accessself-administration, Brain Res., 593 (1992) 314-3 18.

352 Weiss, S.R., Post, R.M., Pert, A., Woodward, R. and Murman,D., Context-dependent cocaine sensitization: differential effectof haloperidol on development versus expression. Pharmacol.Biochem. Behat *. , 34 ( 1989) 655 -66 1.

353 Whishaw, I.Q. and Kornelsen, R.A., Two types of action re-vealed by ibotenic acid nucleus accumbens lesions: dissociationof food carrying and hoarding and the role of primary incentivemotivation, Behat?. Brain Res., in press.

354 White, F.J. and Wolf, M.E., Psychomotor stimulants. In J. Pratt(Ed.), The Biological Bases of Drug Tolerance and Dependence,Academic Press, New York, 1991, pp. 153-197.

355 White, N., Sklar, L. and Amit, Z., The reinforcing action ofmorphine and its paradoxical side effect, Psychoplrurmacology,52 (1977) 63-66.

356 White, N.M. and Milner, P.M., The psychobiology of rein-forcers, Annu. Rel*. Psychol., 43 ( 19921 443-47 1.

357 Wikler, A., Recent progress in research on the neurtophysiologi-cal basis of morphine addiction, Am. 1. Psychiatry. 105 (1948)329-338.

358 Wilcox, R.A., Robinson, T.E. and Becker, J.B., Enduring en-hancement in amphetamine-stimulated striatal dopamine re-lease in vitro produced by prior exposure to amphetamine orstress in vivo, Eur. J. Pharmacol., 124 (1986) 375-376.

359 Wise, R.A., Catecholamine theories of reward: a critical review,Brain Res., 152 (1978) 215-247.

360 Wise, R.A., Neuroleptics and operant behavior: the ;Inhedoniahypothesis, Behav. Brain Sci., 5 (1982) 39-87.

361 Wise, R.A., The anhedonia hypothesis: Mark III, Bc+ur*. BrainSci., 8 (1985) 178-186.

362 Wise, R.A., The role of reward pathways in the development ofdrug dependence, Pharmacol. Ther., 35 (1987) 227~2M.

363 Wise, R.A., The neurobiology of craving: implicatic~ll~ for theunderstanding and treatment of addiction. J. Abnortrr. PsychoI*,97 (1988) 118-132.

364 Wise, R.A. and Bozarth, M.A., Brain mechanism of drugreward and euphoria, Psychiatry Mcd., 3 (19SS) 445-4~0.

365 Wise: R.A. and Bozarth, M.A.. A psychomotor stinlulilnt theoryof addiction, Psychol. Relq.. 94 (1987) 369-392.

366 Wise, R.A. and Hoffman, D.C., Localization of drug rewardnlech;inisrns by intracranial injections, ~~II<I/ISP. 10 (1‘1’12) 247-263.

367 Wise, R.A., Murray, A. and Bozarth, M.A.. Bromocriptine self-administr;~tion and bromocriptine-reinstatement 01’ cocaine-trained ;md heroin-trained lever pressing in rats. i?sydzipharma-c*ch‘~y, 100 ( 1090) 355 -360.

368 Wise, R.A. and Rompre, P.-P., Brain dopaminc and reward,Annrr. Rw. P.s~x~hc~l.. 40 (1989) 191-225.

369 Wolf, ME. ;tnd Khansa, M.R., Repented ;1drninistr;(tion ofMK-SO 1 products sensitization to its own locomotor st irnuiantcffccts but blocks sensitization to amph~t:lmine, Brain Kcs., 562(1901) 163- 168 .

370 Wolf, ME., White, F.J., Nasser, R., Brooderson, R.J. and

291

Kh;ux\, M.R., Differential development of autoreceptor sub-sensitivity and cnh:\nccd dopamine rclcase during amphctamincsensi tiz;t t ion, J. I%trrmtrcol. Ekp. Tiw., 20-l ( 199 3) 24% 1255.L

371 Woods, J.H., Katz, J.L. and Wingcr, G., Use and abuse ofbenzodiazepines. .I. Am. Med. Assoc.. 260 ( IWO 3476-3480.

372 Woods, J.H. and Schuster, C.R., Reinforcement properties ofmorphine, cocaine and SPA as a function of unit dose, Irtf. J.Addict., 3 ( 1968) 23 l-237.

373 Woolverton, W.L., Cervo, L. and Johanson, C.E., Effects ofrepeated methamphetamine administration on metham-phetamine self-administration in rhesus monkeys, Phtzrmtrcol.Biochtwk B&w., 2 1 ( 1984) 737-74 1.

374 Menakis, S. and Sclafani, A., The dopamincrgic mediation of asweet reward in normal and VMH hyperphngic rats, Phurmucol.Biochcm. B&l 3. , 16 (1982) 293-302.

375 Yamada, S., Kojima, Ii., Yokoo, f I., Tsutsumi, T., Takamuki, K.,Anraku, S., Nishi, S. and Inanaga, K., Enhancement of dopamincrelease from striatal slices of rats that were subchronicallytreated with methamphetamine, Biof. Psychicttry, 24 (1988) 399-408.

376 Young, A.M., Joseph, M.H. and Gray, J.A., Increased dopaminerelease in vivo in nucleus accumbens and caudate nucleus of therat during drinking: a microdialysis study, Neuroscience, 48 (1992)871476.

377 Young, A.M.J., Gray, J.A. and Joseph, MN., Dopamine func-tion in the nucleus accumbens in conditioning studied by micro-dialysis, Abstructs of the 7th Internatiorlul Cktecholamine Sympo-sium in Amsterdam, (1992) 360.