A. Eugene Le Blanc and Howard Cappell Addiction Research Foundation, Toronto, Canada.
It is clear from even a cursory reading of the vast literature on drug tolerance that approaches to this interesting phenomenon vary widely in basic conception. Whereas some investigators concentrate their efforts upon isolated, albeit fundamental units of biological function, such as neurotransmitters (e.g., 1), others have opted instead for an approach that emphasizes an integrated functional adaptation of entire organisms behaving in a demanding environment (e.g., 2). There is no necessary conflict in these approaches; indeed, any thoughtful observer must see an ultimate connection between these apparently different enterprises. It will become clear that our inclination is to view the organism in the broader context of environmental adaptation; indeed, it is the thesis of this paper that tolerance to drug effects represents a special case of the general adaptive capacity of organisms. Yet this will not be the thesis of apostates who score belief in the events that go on within the venerable "black box"; on the contrary, studies involving manipulations of these events are basic to our conception of tolerance. The strategy will be to draw parallels between adaptations to the behavioral and biologic consequences of drugs and other forms of demand that appear to evoke an adaptive response. Additionally, we will attempt to show how aspects of tolerance to drug effects are subject to environmental control in a way that requires a more complex interpretation of tolerance than changes of receptor sensitivity at the site of a drug's action can alone provide.
The purpose of this paper is to attempt a synthesis. For that reason our strategy will be to look for consistencies despite an awareness that all is not as neat as the analysis may imply. In this regard, it must be admitted from the outset that the parallels to be noted essentially involve the phenomenon of tolerance to the relative exclusion of physical dependence, despite evidence that they are intimately associated developments. Thus it is much more difficult to find compelling examples of postadaptational supersensitivity when dealing with nonpharmacological rather than pharmacological adaptations. With this hopefully disarming admission of fallibility on record, let us now begin to turn to the issues at hand.
WHAT IS TOLERANCE?
Since the word tolerance will be used repeatedly, it behooves us to begin with a definition of its descriptive requirements. A widely acceptable definition includes the following requirements: (a) a drug effect will diminish in magnitude with repeated exposure to a fixed dose; and (b) a drug effect diminished in magnitude by repeated exposure to a fixed dose can be reinstated to its original level by increasing the dose.
Although it is possible to define tolerance precisely, use of the term is complicated by the existence of special cases such as "physiological" tolerance, "behavioral" tolerance, and "functional" tolerance. An added difficulty in speaking of tolerance is that it develops at different rates depending on the effect and drug in question, and does not develop at all where some drug effects are concerned. As far as possible, we will attempt to avoid the pitfalls attendant to the linguistic and empirical complications associated with discourses on tolerance phenomena.
Ultimately, any discussion of tolerance is incomplete to the extent that it skirts the issue of the mechanism or , mechanisms involved. It seems more than a matter of faith that tolerance is accountable in reductionist terms, and indeed such accounts exist (e.g., 1). Yet not all productive conceptions of tolerance are rooted in a reductionist tradition, as will become evident in our discussion of work containing a large element of behavioral analysis. What remains then, is a consideration of the pertinent literature, with particular emphasis on work conducted in the laboratories of the Addiction Research Foundation.
Effects of Cortical Ablations
In looking for parallels between tolerance and general adaptive processes, one potentially useful strategy is to compare the impact on these processes of attempts to interfere with them. One such attempt has involved surgical intervention by ablation of the frontal cortex. A considerable body of research (3) indicates that an intact frontal cortex is necessary in several species to acquire behavior requiring a delay of response or a sensory discrimination in order to receive reward. Thus, frontal cortical ablations may interfere with learning. Moreover, lesions of the frontal cortex have been shown to impair the physiological adaptation to thermally elicited tachycardia that is evident in normal animals (4). If tolerance shares something in common with these other adaptive phenomena, it should be possible to interfere with its acquisition by means of comparable ablations: just such an hypothesis was tested in an experiment by Le Blanc, Matsunaga, and Kalant (5). Rats were first made tolerant to the impairment produced by alcohol on the "moving-belt" test, which provides sensitive measure of motor performance (cf. 6). Tolerance was established over a three week period during which a maximum daily dose of 6 g/kg (by gavage) of ethanol was attained. Tolerance was assessed with periodic test trials in which the impairing effect of 2 g/kg i.p. of ethanol was measured. A period of one month was allowed for recovery before animals were assigned to groups selected for lesions, sham operations, or nonoperated controls. The surgery involved bilateral frontal polar lesions 3 x 5 mm in extent. Following a nine day postoperative recovery period, all animals were once more subjected to the same regime of alcohol exposure as before. Rats that were sham-operated or nonoperated became tolerant to the extent that impairment was reduced by more than 50% over the course of six test trials; lesioned animals, in contrast, displayed virtually no improvement in performance during a similar schedule of exposure to alcohol. A smaller pilot study yielded similar results where lesions to the occipital cortex were concerned. In summary, while much remains to be learned of the basis for this positive result, alcohol tolerance was shown to share a property in common with complex learning and with a fundamental process of physiological adaptation. Other research on amphetamine (7) suggests that the involvement of the frontal cortex in tolerance development is not peculiar to alcohol or to motor impairment.
Inhibition of Protein Synthesis
The evidence on the effects of frontal cortical ablations provided some support for a commonality of process in tolerance development and the acquisition of new responses. To the extent that protein synthesis is involved in learning, another opportunity for pursuing the analogy presents itself. Evidence of a role for protein synthesis in learning is provided in a study by Segal, Squire, and Barondes (8), who showed that cycloheximide, an inhibitor of central protein synthesis, interfered with the retention of discrimination learning in mice. What of the effect of cycloheximide on alcohol tolerence? This question was pursued by Le Blanc et al. (5) using a design and procedures very similar to that adopted in the ablation study; the major difference was that the effects on reacquisition of tolerance after treatment with cycloheximide was at issue. During exposure to the tolerance inducing regime, cycloheximide treatment (0.3 mg/kg) was combined with alcohol gavage in the critical experimental group. On tests of the reacquisition of tolerance to the impairing effects of ethanol, the performance of controls improved over time by more than 50%, whereas rats treated with cycloheximide acquired virtually no tolerance whatsoever. Hence, the evidence for our general hypothesis increases, although it is conceded that this experiment does not rule out mechanisms of interference with tolerance acquisition not involving the inhibition of central protein synthesis (e.g., subclinical convulsions). The effect of cycloheximide on tolerance is not unique to alcohol, since cycloheximide has been found to inhibit the development of tolerance to the analgesic property of morphine in mice (9).
Depletion of Serotonin
A third source of evidence for the tolerance/adaptation parallel derives from recent data on the role of serotonin-in the development of tolerance to ethanol. That depletion of serotonin retards habituation to non-drug stimuli has been amply demonstrated in studies of acoustic startle responding (10,11,12). In all of these studies, the serotonin level was manipulated by prior treatment with p-chlorophenylalanine (pCPA), an established depletor of serotonin in the CNS. The drug studies (13) employed the moving-belt test to measure tolerance. Both alcohol and pentobarbital were investigated, but since the results were quite similar with each drug only the ethanol work will be presented in any detail. The essential strategy of the study was to expose animals depleted of serotonin to a schedule of alcohol administration that would be expected to promote tolerance to the impairing effects of a test dose of 2.2 g/kg of ethanol. Treatment with pCPA (100 mg/kg/day, i.p.) was given daily for ten days before chronic exposure to alcohol (5.0 g/kg/day p.o. 25% V.V.), and continued for twenty-five further days during which experimental animals were also exposed daily to large doses of alcohol by gavage. Tests of impairment on the moving-belt apparatus were interspersed at intervals during the chronic regime of treatment. Control animals exposed to ethanol, but not pCPA, clearly developed tolerance to the test dose; rats exposed to pCPA also displayed some tolerance, but the rate of acquisition was significantly impaired, and the level did not approach that of controls during the course of the experiment.
Much the same assertion can be made where pentobarbital was concerned.
As with the other data presented in support of our general hypothesis, these findings are by no means free of alternative interpretation. Yet they do seem to enhance the credibility of the hypothesis by providing supporting data from another general domain of investigation. Moreover, as with the previous illustrations in the first two sections of this chapter, there is evidence that comparable manipulations retard the development of tolerance to a pharmacologically distinct compound, namely morphine (15).
Topographical Similarity Between Tolerance and Learning
A fourth source of parallels between tolerance and adaptation derives from a topographical similarity in one aspect of tolerance and learning. One property of learning that is demonstrable by appealing to both personal experience and the scientific literature (16), is that responses are reacquired after a period of disuse with much greater facility than they are initially mastered. A similar process can be demonstrated in studies of adaptations of a more fundamental physiological nature; for example, physiological adaptation to thermal stimuli proceeds more rapidly to the extent that an organism has a history of adaptation to those stimuli (4). The parallel was nicely confirmed in a study by Kalant, Le Blanc and Gibbins (17), in which repeated cycles of acquisition of tolerance to alcohol were studied using the moving-belt test. Rats were exposed to as many as four cycles of acquisition with seventeen day drug-free intervals between cycles to permit recovery to baseline levels of impairment. The basic finding was that the same level of tolerance was achieved during each cycle, but that this level was attained in fewer and fewer trials over successive cycles of acquisition. Whereas thirteen to sixteen days of chronic exposure were required for maximal tolerance to be attained during an initial cycle, maximal tolerance was evident with four days of treatment during a fourth cycle of tolerance acquisition.
Summary of "Parallels"
Clearly, we have presented an idealized version of the similarities between tolerance and other forms of adaptive response. Although specifics were not considered, there are alternative interpretations for many of the findings presented in support of the general hypothesis advanced here. Nonetheless, the possibility of genuine parallels seems too strong to be dismissed easily, and the overall weight of the evidence is consistent with the hypothesis. Moreover, the parallels appear to cover a broad spectrum of pharmacological agents to which tolerance can be demonstrated, including morphine, amphetamine, alcohol, and pentobarbital. It is obvious that more evidence will be required to move this thesis beyond the point of credible speculation, but there is good reason to believe that such evidence will be possible to obtain.
INTERACTIONS BETWEEN BEHAVIORAL MANIPULATIONS AND TOLERANCE DEVELOPMENT
The previous sections of this paper dealt with direct parallels between tolerance and other forms of adaptation. Now we turn to a different but not entirely independent issue, namely, the modifiability of tolerance by behavioral interventions.
Tolerance and Reinforcement
One of the earliest, and still one of the most intriguing demonstrations of the behavioral modifiability of tolerance was reported by Schuster et al. (2), who analysed tolerance to amphetamine as a form of functional adaptation to environmental contingencies. Rats were trained to press a lever on two schedules of reinforcement that required temporal control of responding in order to optimize the receipt of reinforcement in the form of food pellets. When the animals were responding stably on these schedules they were repeatedly injected with d-Amphetamine (1.0 mg/kg) prior to the behavioral sessions. Amphetamine had a clear disruptive effect on response rate compared to control sessions in which only saline was administered. The interesting result occurred in examining the extent to which response rate returned to normal over a course of chronic exposure to amphetamine. For some animals, the disruption in responding was such that it eventuated in a reduction in reinforcements obtained during a test. For others, although response rate was clearly affected, the reinforcement schedule was such that there was no loss of reinforcements despite the alteration in response rate. In the case of animals that did not suffer a loss of reinforcement, the response rate did not return to control levels even over a course of chronic drug treatment. However, when there was a loss of reinforcements as a result of the behavioral disruption, there was a return of responding to a degree that restored the receipt of reinforcements to control levels. Moreover, a second experiment demonstrated a failure to develop tolerance when injections of amphetamine improved the animals' responding to avoid shock. This led the authors to postulate a form of adaptation that they described as "behavioral tolerance": "Behavioral tolerance will develop in those aspects of the organism's behavioral repertoire where the action of the drug is such that it disrupts the organism's behavior in meeting the environmental requirement for reinforcements. Conversely, where the actions of the drug enhance or do not affect the organism's behavior in meeting reinforcement requirements we do not expect the development of behavioral tolerance." (2)
This fascinating demonstration has had the effect of substantially altering subsequent conceptions of tolerance for many investigators, for it demonstrated that sheer exposure to a drug was not in itself sufficient to provoke tolerance to one of its demonstrable consequences. Rather, the adaptation was functional in the sense that it occurred only to the extent that it was instrumental in restoring to normal or actually enhancing a particular level of reinforcement.
In a study directly derivative from the work of Schuster et al. (2), Carlton and Wolgin (18) examined the acquisition of tolerance to the anorexigenic effect of amphetamine. By varying the temporal relationship between drug administration and exposure to the behavioral test, they were able to hold constant the pharmacologic stimulus while varying the behavioral contingency. Some of their rats received injections (2.0 or 3.0 mg/kg of d-Amphetamine) twenty minutes prior to a test of milk intake, and some received injections only after the drinking test was completed.. Over repeated test trials those animals receiving injections before the test recovered to normal levels of milk consumption. During the same period, the milk intake of the rats that received amphetamine after the test was unaffected. When tolerance was clearly evident in the "before" condition, the temporal relationship between injection and test was reversed in the "after" condition. Interestingly, milk consumption was depressed to an extent predictable if their animals had never before been exposed to amphetamine; moreover, the rate of acquisition of tolerance was not distinguishable from that of the "before" group, that was, in fact, drug-naive before tolerance development was assessed. Carlton and Wolgin (18) arrived at essentially the same interpretation as did Schuster et al. (2); namely, the development of tolerance depended upon the drug-induced loss of reinforcement for its occurrence. They labelled this phenomenon "contingent tolerance" a term that is conceptually similar to "behavioral tolerance".
When we earlier proposed a parallel between tolerance and adaptation, the semantic loading of the word "adaptation" did not go beyond description of a relationship between repeated stimulation and a recovery of response; in this research on amphetamine, however, the word takes on a new meaning that is tinged with the notion of purposiveness and benefit to the organism. In other words, the implication from this research is that tolerance does not develop to certain drug effects unless it is, in a sense, useful to the organism.
"Behavioral" versus "Physiological" Tolerance
Behavioral tolerance is a discovery of comparatively recent vintage. It is customary to contrast it with the more traditional conception of tolerance that has come to be known as "physiological". The latter term implies a change in the sensitivity of the neurons directly affected by the drug, and should be a consequence of mere exposure; this is to be distinguished from the functional adaptation implied by behavioral tolerance.
One of the earliest systematic studies of this distinction was reported by Chen (19), who used a design not unlike that of Carlton and Wolgin (18). In this instance, however, alcohol was the drug of interest. Chen's behavioral group received injections of alcohol (1.2 g/kg) before each of four trials in a maze task in which approach behavior was maintained by a food reward. A physiologic group received an equivalent injection of alcohol after testing for the first three trials, but before testing on the fourth. As measured by performance on the fourth trial, tolerance was evident in the behavioral but not in the physiologic group. Chen's experiment was criticized on several grounds by Le Blanc et al. (20), but most importantly because the evidence was insufficient to warrant the qualitative distinction between the two varieties of tolerance implied by Chen. Rather, they argued, the difference might simply be one of rate of acquisition; Chen's experimental design did not permit a test of this hypothesis. Le Blanc et al. (20) were able to assess their alternative essentially by extending Chen's procedure to include many more test trials within the context of a generally more complete experimental design. The results were clearcut: although a "behavioral" group developed tolerance much more quickly than a "physiological" group, the performance of the latter eventually improved to the level of the former. It is clear from the data of the study that Chen's conclusions were based on a schedule of testing that was simply terminated prematurely. Le Blanc et al. (20) concluded that if the only difference in these two ostensibly different types of tolerance was one of rate in attaining the same asymptotic levels, there was no need to postulate more than one basic cellular mechanism of tolerance. For this reason, they coined the phrase "behaviorally augmented tolerance" to describe the interaction between behavioral demand and the rate of tolerance development, while negating the suggestion of qualitatively different mechanisms of tolerance. Moreover, other evidence based on findings with opiates (21), barbiturates (22) and chlorpromazine (23) can be adduced in support of this argument, although one study of morphine (24) could be interpreted as showing a difference in asymptote as well as rate of tolerance development. Unfortunately, in any given study it is impossible to determine whether testing for tolerance development was terminated before genuinely asymptotic levels were attained.
The evidence in favor of a unitary mechanism underlying behavioral and physiologic tolerance was recently extended (25). In an initial phase of this experiment rats were exposed to the standard procedure for contrasting behaviorally augmented and physiologic tolerance (20); a major difference was that the procedures were not drawn out to the point of convergence in tolerance as they were in the earlier study. Using the moving-belt test to measure tolerance, Le Blanc et al. (25) confirmed the expectation that tolerance would develop in a group receiving alcohol before exposure to the behavioral task but not afterward. This result was, of course, not novel; what was compelling, however, was the finding that the behavioral augmentation group was tolerant on a second behavioral task (maze performance) on which performance was empirically shown to be entirely independent of the first (i.e., there was no evidence of transfer of training), but the physiologic group, despite pharmacologically equivalent exposure to alcohol, performed no better than controls who were pharmacologically naive. This finding is of crucial importance because it demonstrates that behaviorally augmented tolerance is more than learned compensation for the impairment produced in a particular behavioral test; such an argument is untenable in view of the fact that prior learning in one task was without facilitating consequences in acquiring the other. Rather, a more tenable argument is that behavioral augmentation facilitated the development of a fundamental adaptation at the neuronal level that is common to all manifestations of tolerance to alcohol. Parenthetically, it is worth mentioning that behavioral augmentation of physical dependence was also shown in this work.
In summary, the hypothesis that currently enjoys the strongest support is that there is but one basic process underlying tolerance phenomena where neuronal mechanisms are concerned. Yet this process is significantly influenced by the adaptive or functional requirements of the organism in the face of drug-produced impairment. This interaction cannot be ignored in any general mechanistic theory of tolerance.
TOLERANCE AS CONDITIONING: A RADICAL VIEW
It is generally accepted that behavioral conditioning processes are intimately involved in drug dependence. A role for conditioning has found one of its most persuasive advocates in Wikler (e.g., 26) although his concern has been primarily to account for relapse using the concepts of Pavlovian or classical conditioning. Recently, Siegal (27) showed that Pavlovian conditioning could completely account for tolerance to the analgesic properties of morphine. In Siegal's words: "According to the present conditioning theory, tolerance to the analgesic effects of morphine results because environmental cues regularly paired with the administration of the drug come to elicit a compensatory (conditioned response) hyperalgesia, which algebraically summates with the stable, unconditioned analgesic effects of the narcotic. Thus environmental cues consistently predicting the systemic effects of the drug should be crucial to the development of tolerance since they enable the subject to make timely compensatory (conditioned responses) in anticipation of the analgesic (unconditioned response)." (27).
Siegal began by presenting ample evidence that repeated pairing of environmental conditioned stimuli with certain pharmacological unconditioned stimuli can result in the elicitation of a response that does not mimic the direct effects of the unconditioned stimulus, but rather appears to be a compensatory physiological response. In one test of the conditioning hypothesis, Siegal measured the acquisition of tolerance to morphine (5 mg/kg) using a hot plate test. Three groups of rats received morphine injections under different environmental circumstances during three initial sessions spaced at forty-eight hour intervals. One was injected with morphine and tested on the hot plate when it was in fact hot. A second group received the same treatment except that it was placed on the hot plate surface at room temperature. In a third group the animals received morphine injections but experienced the effects in their home cages and were not exposed to the hot plate. For the critical test, all animals were injected with morphine and placed on the hot plate. The two groups that had had the drug effect paired with the test (i.e., hot plate) cues in the past displayed tolerance in the form of decreased response latencies; in contrast, the animals without such a conditioning history showed no signs of tolerance, even though they had had the same pharmacological history as the other two groups. Two other demonstrations enhanced the credibility of this compensory theory of tolerance. Siegal showed that if a set of environmental stimuli that had previously been paired with morphine injections was later followed by a saline injection it was possible to obtain conditioned hyperalgesia. This is direct evidence for compensatory response theory, since hyperalgesia is the response that should be compensatory to the direct pharmacologic effect of the drug on the response to pain. In a final experiment it was shown that tolerance could be reversed by simply exposing rats to environmental stimuli previously paired with morphine but then followed by saline on a series of trials. A control group was provided to show that a comparable period of time of simple withdrawal from morphine without exposure to the drug-associated environmental cues was insufficient to produce loss of tolerance. Presumably, extinction of the compensatory response underlying the development of tolerance was achieved in the experimental group but not in the controls.
These results are intriguing in that they suggest that tolerance can be an exclusive function of environmental control under circumstances in which the pharmacological stimulus per se is insufficient to promote an adaptation. However, the ultimate importance of the data is still impossible to assess; a high priority would be to determine whether similar results could be obtained with doses of morphine higher than 5 mg/kg. Clearly, this dose is only a small fraction of that to which rats can ultimately develop tolerance.
This brief survey of the literature on tolerance can be crystallized into three broad summary statements: (a) there are direct analogies between tolerance and other forms of adaptation to stimuli; (b) important aspects of tolerance are subject to alteration by nonpharmacological manipulation, although it does not seem necessary to invoke more than a single underlying mechanism to account for this; and (c) tolerance can be shown to be under the complete control of environmental stimuli in circumstances in which a pharmacological stimulus per se fails to provoke tolerance.
At one level, we can summarize these statements by saying that tolerance is a complicated business indeed. But there appears to be something more substantial to be said by way of elaboration. Tolerance seems subject to modification by interventions and processes that must be represented in CNS pathways that are not entirely coincident with the site or sites at which tolerance-producing drugs exert their primary pharmacological effects. It thus seems logical to conclude that there is more to many instances of tolerance than a change in receptor sensitivity at the site of a drug's action. Although the latter may be an important consideration, it cannot be invoked independently of the myriad of CNS events that must be involved in the adaptive interaction of an organism with its environment. The data can lead to no other conclusion.
1. Goldstein, D.B.: Testing the homeostat hypothesis of drug addiction, Biological and Behavioral Approaches to Drug Dependence. Edited by Cappell, H.D. and Le Blanc, A.E., Addiction Research Foundation of Ontario, Toronto, p. 1-11 (1975).
2. Schuster, C.R., Dockens, W.S. and Woods, J.H.: Behavioral variables affecting the development of amphetamine tolerance. J. Pharmacol. Exp. Ther. 9:170-182 (1966).
3. Grossman, S.P.: A Textbook of Physiological Psychology. John Wiley and Sons, Inc., New York (1967).
4. Glaser, E.M.: The Physiological Basis of Habituation. Oxford University Press, London (1966).
5. Le Blanc, A.E., Matsunaga, M. and Kalant, H.: Effects of frontal polar cortical ablation and cycloheximide on ethanol tolerance in rats. Pharmacol. Biochem. Behay. 4: 175-179 (1976).
6. Gibbins, R.J., Kalant, H. and LeBlanc, A.E.: A technique for accurate measurement of small degrees of alcohol intoxication in small animals. J. Pharmacol. Exp. Ther. /59:236-242 (1968).
7. Glick, S.D.: Impaired tolerance to the effects of oral amphetamine intake in rats with frontal cortical ablations. Psychopharmacologia (Berl.) 28:363-371 (1973).
8. Segal, D.S., Squire, L.R. and Barondes, S.H.: Cycleheximide: Its effects on activity are dissociable from its effects on memory. Science 172:82-84 (1971).
9. Loh, H.H., Shen, F.H. and Way, E.L.: Inhibition of morphine tolerance and physical dependence development and brain serotonin synthesis by cycloheximide. Biochem. Pharmacol. 18:2711-2721 (1969).
10. Carlton, P.L. and Advokat, C.: Attenuated habituation due to parachlorophenylalanine. Pharmacol. Biochem. Behay. 1:657-663 (1973).
11. Connor, R.L. at al.: Parachlorophenylalanine and habituation to repetitive auditory startle stimuli in rats. Physiol. Behay. 5:1215-1219 (1970).
12. Swonger, A.K., Chambers, W.F. and Rech, R.H.: The effects of alterations in brain 5HT on habituation of the cortical evoked response and the startle response in rats. Pharmacologist 12:207 (1970).
13. Frankel, D. et al.: Effect of p-Chlorophenylalanine on the acquisition of tolerance to ethanol and pentobarbital. Psychopharmacologia (Berl.) 44:247-252 (1975).
14. Shen, F.H., Loh, H. and Way, E.L.: Brain serotonin turnover in morphine tolerant-dependent mice. J. Pharmacol. Exp. Ther. 175:427-434 (1970).
15. Way, E.L., Loh, H.H. and Shen, F.H.: Morphine tolerance, physical dependence and synthesis of brain 5-hydroxytryptamine. Science 162:1290-1292 (1968).
16. Kimble, G.A.: Hilgard and Marquis' Conditioning and Learning (2nd Edition). Appleton-Century-Crofts, Inc., New York (1961).
17. Kalant, H., Le Blanc, A.E. and Gibbins, R.J.: Pharmacological and behavioral variables in the development of alcohol tolerance. Committee on Problems of Drug Dependence. Nat. Acad. Sci. 2:1283-1296 (1971).
18. Carlton, P.L. and Wolgin, D.L.: Contingent tolerance to the anorexigenic effects of amphetamine. Physiol. Behay. 7:221-223 (1971).
19. Chen, C.S.: A study of the alcohol-tolerance effect and an introduction of a new behavioral technique. Psychopharmacologia (Berl.) 12:433-440 (1968).
20. Le Blanc, A.E., Kalant, H. and Gibbins, R.J.: Behavioral augmentation of tolerance to ethanol in the rat. Psychopharmacologia (Berl.) 30:117-122 (1973).
21. Kayan, S., Woods, L.A. and Mitchell, C.L.: Experience as a factor in the development of tolerance to the analgesic effect of morphine. Europ. J. Pharmacol. 6:333-339 (1969).
22. Wahlstrom, G.: Differences in tolerance to hexobarbital (Enhexymalum NFN) pretreatment during activity or rest. Acta Pharmacol. (Kbh.) 26:92-104 (1968).
23. Irwin, S.: Influence of external factors and arousal mechanisms on the rate of tolerance development. Arch. Int. Pharmacodyn. 142:152-162 (1963).
24. Kayan, S., Ferguson, R.K. and Mitchell, C.L.: An investigation of pharmacologic and behavioral tolerance to morphine in rats. J. Pharmacol. Exp. Ther. 185:300-306 (1973).
25. Le Blanc, A.E., Gibbins, R.J. and Kalant, H.: Generalization of behaviorally augmented tolerance to ethanol, and its relation to physical dependence. Psychopharmacologia (Berl.) 44:241-246 (1975).
26. Wikler, A.: Dynamics of drug dependence: Implications of a conditioning theory for research and treatment. Arch. gen. Psychiat. 28:611-616 (1973).
27. Siegel, S.: Evidence from rats that morphine tolerance is a learned response. J. comp. physiol. Phychol. 89: 498-506 (1975).