Assessing individual differences in ethanol preference using a cumulative dosing procedure

Preference for ethanol versus a placebo was assessed in 12 normal volunteers using a cumulative dosing preference test. The test consisted of four sampling sessions followed by three choice sessions. During the sampling sessions subjects received either five cumulating oral doses of ethanol (0.1g/kg per dose) or equal volumes of placebo, at 15-min intervals. Subjective and observer-rated drug effects, psychomotor performance, drug liking ratings, and breath ethanol levels were measured at regular intervals. During choice sessions, subjects first chose which of the two substances (drug or placebo) they wished to take and ingested one unit dose. Then, at 15-min intervals throughout the session, they could ingest up to ten additional unit doses of the same substance (i.e., up to 1.1 g/kg ethanol). On average, the subjects chose the ethanol-containing beverage on 75% of the choice sessions, and they self-administered a mean total dose of 0.8g/kg per session. Subjects varied in the amount of ethanol ingested on choice sessions, and the amount they chose was related to their subjective responses to the drug during sampling. Subjects who chose the most ethanol reported experiencing stimulant-like effects from the ethanol, whereas the subjects who chose ethanol less frequently and ingested lower doses reported primarily sedative-like effects from the drug. The results demonstrate that the cumulative dosing procedure can be used effectively to evaluate drug preferences and dose preferences in normal volunteers.     

Non-specific effect of naltrexone on ethanol consumption in social drinkers

Rationale: Clinical studies have shown that the opioid antagonist naltrexone is effective in the treatment of alcoholism. However, the mechanism by which it produces this effect is not understood. Objective: This study was designed to investigate the effect of acute naltrexone on consumption of ethanol in healthy, non-problem social drinkers. Methods: Subjects (n=24) participated in an eight-session, within-subject, placebo-controlled choice procedure which measured ethanol preference and consumption. The procedure consisted of two blocks of four sessions in which subjects received either naltrexone (50 mg oral) or placebo 1 h before consuming an ethanol or placebo beverage. On the first two sessions of each block, subjects received a color-coded beverage containing ethanol (0.75 g/kg) or placebo, in five equal portions at 15-min intervals. On the next two sessions of each block, subjects chose which beverage they preferred (i.e., placebo or ethanol) and how much they wished to take, in unit doses (placebo or ethanol 0.15 g/kg/dose). The primary behavioral measures were (1) the number of times subjects chose ethanol over placebo, and (2) the number of doses they consumed. Subjects rated their mood states and subjective drug effects at regular intervals during each session. Results: Naltrexone did not alter the frequency of ethanol (versus placebo) choice. Although naltrexone did decrease the total number of ethanol doses subjects took (mean 2.7 doses after naltrexone; 3.4 doses after placebo), it also decreased the number of placebo ”doses” subjects took on sessions when they chose the placebo beverage (mean 1.6 placebo doses after naltrexone; 2.8 doses after placebo). Ethanol produced its prototypic subjective effects (e.g., increased ratings of ”feel drug”, ”like drug” and ”high”), and these effects were not altered by naltrexone. Naltrexone produced mild sedative-like effects, and several subjects reported adverse effects such as nausea. Conclusions: These findings show that naltrexone reduces ethanol consumption in healthy volunteers, as it does in alcoholics. However, this reduction was not specific to alcohol; subjects also consumed less of a non-alcoholic, placebo beverage. These findings suggest that naltrexone may reduce alcohol consumption by a non-specific mechanism. Received: 17 September 1998 / Final version: 14 April 1999     

Individual differences in the priming effect of ethanol in social drinkers

OBJECTIVE: Alcohol-dependent individuals frequently report increased desire for alcohol and exhibit increased alcohol-seeking behaviors following a single drink. The phenomenon, known as priming, has been demonstrated in the laboratory in alcohol-dependent humans and in nonhumans, but the effect is inconsistently observed in nonproblem social drinkers. The current experiment examined this effect in healthy, nonproblem social drinkers across a range of preload doses. METHOD: Using a repeated-measures design, 12 social drinkers were given ethanol (0.2, 0.4 or 0.8 g/kg) or placebo preloads. Various subjective effects measures were obtained at regular intervals. In addition, before and after consuming, the preload subjects performed an operant task on which they made repeated choices for either ethanol or money. RESULTS: Ethanol dose-dependently increased subjective reports of drug liking and desire to take more ethanol. When data from all subjects were examined, ethanol did not affect choices for ethanol on the choice task. However, in subjects who reported the greatest positive mood effects from ethanol (0.8 g/kg), the ethanol preload increased choices for ethanol over money on the choice task. CONCLUSIONS: These findings provide evidence for a priming effect of ethanol in social drinkers as measured by increased subjective desire for drug. The findings also suggest that the priming effects may be strongest in individuals who experience the greatest subjective positive mood effects from ethanol.     

Effect of setting on the reinforcing and subjective effects of ethanol in social drinkers

The reinforcing and subjective effects of two doses of ethanol [0.5 g/kg (LOW) and 0.8 g/kg (HIGH)] were evaluated under two conditions, a social condition (SOC), in which subjects were tested with two or three other subjects, and a socially isolated condition (ISO), in which subjects were tested alone. Forty-one social drinkers participated in a double-blind, seven-session choice procedure. Subjects were randomly assigned to one of four experimental groups: SOC-LOW, SOC-HIGH, ISO-LOW, or ISO-HIGH. On the first four sessions, subjects sampled ethanol (0.5 or 0.8 g/kg) on two occasions and placebo on the other two occasions. On the three remaining sessions, subjects selected and consumed whichever of the two previously sampled substances they preferred. The number of sessions on which they chose ethanol was the primary measure of the reinforcing effects of ethanol. Standardized self-report questionnaires and a psychomotor test were used to measure subjective and objective drug effects. Subjects in the SOC condition chose ethanol over placebo on significantly more sessions than subjects in the ISO condition. Ethanol produced positive subjective effects (e.g., increased ratings of drug liking and euphoria) for subjects in the SOC condition, but for subjects in the ISO condition, it produced apparently negative effects (e.g., increased ratings of dysphoria). These results extend previous reports that the behavioral effects of ethanol depend upon the social condition in which it is consumed.     

Preference for diazepam,but not buspirone,in moderate drinkers

The purpose of the present study was to determine the preference for buspirone, an anxiolytic predicted to have minimal abuse potential, in comparison with diazepam in moderate drinkers. Preference for diazepam and buspirone was assessed in 55 moderate drinkers using a seven-session procedure consisting of four sampling sessions followed by three choice sessions. On each sampling session subjects ingested five capsules, one every 30 min. Color-coded capsules contained placebo on two sessions and drug on two sessions. Each drug capsule contained diazepam (4 mg) for 30 subjects and buspirone (5 mg) for 25 subjects. On choice sessions subjects chose whichever of the two color-coded capsules, i.e., drug or placebo, they wished to take. After ingesting one capsule, every 30 min they had the option of ingesting another capsule of the same color and content, for a maximum of seven capsules over the session (maximum of 28 mg diazepam or 35 mg buspirone). In the diazepam group 70% of subjects chose diazepam over placebo on at least two of the three choice sessions, whereas in the buspirone group only 24% of subjects chose buspirone over placebo on at least two sessions. Both diazepam and buspirone increased measures of sedation. Only diazepam increased ratings of liking and impaired performance, whereas only buspirone decreased ratings of feeling Friendly. These results replicate previous findings indicating that diazepam has reinforcing effects in moderate drinkers. Further, these results demonstrate the pharmacological specificity of this effect by showing that buspirone did not function as a reinforcer under these same conditions.     

Behavioral responses to ethanol in light and moderate social drinkers following naltrexone pretreatment

The opioid antagonist naltrexone has been shown to be effective in the treatment of alcoholism, possibly by dampening the subjective effects of ethanol. However, naltrexone does not consistently attenuate the effects of ethanol in social drinkers in laboratory-based challenge studies. In the present study, 25 healthy volunteers, who were either light drinkers (mean=3 drinks per week) or moderate drinkers (mean=16 drinks per week), participated in six evening sessions. At each session, subjects ingested a capsule containing naltrexone (25 or 50 mg) or placebo, and 1 hr later they consumed a beverage containing ethanol (0.25 g/kg, equivalent to about two standard alcoholic drinks) or placebo. Subjects received all combinations of pretreatments and beverages. They completed self-report mood questionnaires and psychomotor tests at regular intervals. This low dose of ethanol produced modest but significant effects on self-report measures such as ratings of feeling a drug effect and of liking the drug effect. However, naltrexone (25 or 50 mg) pretreatment had no dampening effect on subjects' responses to ethanol. These results indicate that acute doses of naltrexone that are effective when administered chronically to alcoholics do not attenuate the acute effects of a low dose of ethanol in non-problem drinkers.     

Effects of naltrexone pretreatment on the subjective and performance effects of ethanol in social drinkers

Clinical trials suggest that opioid antagonists may be effective in the treatment of alcoholism. For example, two recent clinical trials reported that alcoholics treated with the opioid antagonist naltrexone exhibited higher abstinence rates, decreased craving and a decrease in the amount of alcohol consumed if drinking occurred. The present study examined the hypothesis that naltrexone pretreatment would attenuate the behavioral responses to an acute dose of ethanol in normal, healthy social drinkers. Thirteen healthy male and female social drinkers participated in a six-session, double-blind, placebo-controlled, crossover design study. On each session, subjects ingested a capsule containing naltrexone (25 or 50mg) or placebo and one hour later consumed a beverage containing ethanol (0.5g/kg) or placebo. For three hours after the beverage was consumed, breath alcohol levels were measured and subjects completed standardized subjective effects questionnaires and performance tasks at regular intervals. Ethanol alone produced its prototypic effects, including positive subjective responses such as euphoria and increased ratings of overall liking, as well as increased ratings of confusion. Ethanol also impaired performance on a verbal recall task. Naltrexone alone produced few subjective effects and did not impair psychomotor or verbal recall performance. Contrary to our hypothesis, pretreatment with naltrexone did not alter the positive subjective effects, or any other effects, of ethanol. Further research is needed to determine the influence of factors such as baseline level of ethanol consumption or duration of naltrexone treatment on the interaction between ethanol and the endogenous opioid system.     

Effect of ethanol self-administration on choice behavior: money vs. socializing.

Volunteer chronic alcoholic subjects were exposed to a discrete-trail choice procedure within a residential research setting. Twelve daily trials occurred at 20 min intervals. In each trial a subject chose between 2 mutually exclusive options involving either receipt of money or the opportunity for socializing. The effect of ethanol self-administration was evaluated by requiring randomly over days that a subject consume either 8 drinks of orange juice or 8 drinks of ethanol (89.12 g ethanol total). For all 4 subjects, the mean rate of choosing socialization over money was significantly greater on sessions involving ethanol self-administration than on sessions involving orange juice self-administration.     

Assessing pentobarbital preference in normal volunteers using a cumulative dosing procedure

Preference for pentobarbital was assessed in 12 normal healthy volunteers using a seven-session cumulative dosing choice procedure. On the first four sessions subjects sampled the drug and a placebo, and on the last three sessions they chose the substance they preferred. During each of the sampling sessions they ingested, at 30-min intervals, five capsules containing either pentobarbital (30 mg per dose) or placebo. During the choice sessions subjects first chose which capsules they preferred to take (drug or placebo), and then took from one to seven of these capsules, separated by 30 min between ingestions. Self-report measures of subjective effects were obtained at regular intervals during each session. Subjects chose the pentobarbital-containing capsules on average 52% of choice sessions, and ingested an average total dose of 132 mg. Although the drug produced only modest, sedative-like subjective and behavioral effects and there was little evidence of euphoric effects in the group as a whole, individual differences in drug liking and choice were observed. The results are discussed in terms of variables that affect the reinforcing effects of pentobarbital in normal volunteers, and they are contrasted to previous findings using this procedure with other sedative drugs.     

Ethanol as a hypnotic in insomniacs: self administration and effects on sleep and mood.

The purpose of this study was to assess the effects of low ethanol doses on sleep and mood and to assess its reinforcing effects used as a hypnotic. Twenty healthy adults, aged 21-45 yrs, all moderate social drinkers, were studied: eleven subjects had insomnia and nine were normal sleepers, as documented by clinical polysomnography. On two sampling nights each, ethanol (0.5 g/kg) or placebo was administered before sleep in color-coded cups presented in three doses (0.2, 0.2, and 0.1 g/kg) separated by 15 min. On three subsequent nights subjects chose their preferred presleep beverage (0.2 g/kg ethanol or placebo) based on cup color and were given an opportunity for 3 additional refills (0.2 g/kg each) of the chosen beverage at 15 min intervals, yielding a total possible dose of 0.8 g/kg. Insomniacs chose ethanol 67% of nights and normals 22%. Insomniacs chose significantly more ethanol refills than normals for an average nightly dose of 0.45 g/kg and normals took significantly more placebo refills. On the sampling nights 0.5 g/kg ethanol reduced REM sleep for both groups for the 8-hr sleep period and in insomniacs increased stage 3-4 sleep and reduced stage 1 sleep during the first half of the night to the level seen in the normals. Other sleep variables were not altered in either group or halves of the night. Presleep improvements in the Profile of Mood States tension and concentration factors were also associated with ethanol administration. Thus, acutely, both sleep and mood effects appear to be associated with the reinforcing effects of ethanol as a hypnotic for insomniacs.     

Preference for diazepam in anxious adults

This study investigated whether anxious adults desiring treatment for their anxiety would choose to take 10 mg of diazepam (Valium) or placebo after sampling both substances under double-blind conditions. Subjective effects of the drugs were also assessed, and the relationship between self-reported subjective effects and the number of times subjects chose diazepam or placebo was examined. Fourteen male and female volunteers meeting criteria for generalized anxiety disorder were recruited. They participated in a nine-session choice experiment in which they sampled diazepam 10 mg and a placebo on the first four sessions and chose whichever they preferred on the next five sessions. Only three subjects chose diazepam on all five choice occasions, no subjects chose diazepam on three or four occasions, and 11 subjects chose diazepam on two or fewer occasions. Overall, diazepam produced typical, tranquilizer-like subjective effects. However, subjective responses to diazepam differed in the 0-2-time choosers compared with the 5-time choosers: the 0-2-time choosers showed an increase on the measure of confusion, while the 5-time choosers showed decreases on measures of anxiety and confusion and increases on measures of stimulation.     

Choice of sevoflurane and its subjective and psychomotor effects in light and moderate drinkers

BACKGROUND: Sevoflurane, an inhalant of the volatile anesthetic class, has neurobiological and behavioral effects in common with abused inhalants and ethanol. We sought to determine if choice for subanesthetic doses of sevoflurane, and its subjective and psychomotor effects, would differ as a function of alcohol-drinking status in healthy volunteers. METHODS: The effects of four concentrations of sevoflurane (0, 0.2, 0.4, 0.8% sevoflurane in oxygen) were studied in 16 light drinkers and 16 moderate drinkers. During each of four sessions, subjects sampled a concentration of sevoflurane and 100% O(2) (placebo) for 10 min each. Subjective and psychomotor testing commenced 5 min into each sampling trial. Later, within the session, subjects chose nine times, once every 5 min, among sevoflurane (e.g., "Agent A"), placebo (e.g., "Agent B," 100% O(2)), or neither (and were administered 100% O(2), identified as "drug-free air"). RESULTS: Choice for sevoflurane at the 0.4% concentration was significantly higher in the moderate drinkers than in the light drinkers. A number of subjective effects reported during inhalation of sevoflurane were markedly lower in the moderate-drinking group than in the light-drinking group. However, psychomotor impairment induced by sevoflurane was similar in magnitude in both groups. CONCLUSIONS: Alcohol-drinking status affected sevoflurane choice. The results are consistent with several studies comparing light and heavier drinkers, using other drugs. Although both drinking groups were similarly impaired by sevoflurane, the moderate drinkers reported less of a subjective response than light drinkers, suggestive of cross-tolerance.     

Self-administration of pentobarbital in light and moderate alcohol drinkers.

Preference for a moderate dose of pentobarbital was assessed in light and moderate alcohol drinkers using a double-blind, placebo-controlled laboratory choice procedure. Sixteen light drinkers (less than six alcoholic drinks per week) and 13 moderate drinkers (six or more drinks per week) participated in a seven-session study in which they first sampled capsules containing pentobarbital (150 mg) or placebo and then chose and ingested the capsule they preferred. Subjective and behavioral measures were obtained at regular intervals during each session to characterize the drug's effects. Both groups chose pentobarbital less often than placebo: Mean pentobarbital choice in light drinkers was 20.8% and in moderate drinkers was 38.5%. Pentobarbital choice and drug liking ratings were highest among male moderate drinkers but still did not exceed placebo levels. The drug did not increase scores on standardized measures of drug euphoria, even among the most frequent choosers or the heaviest alcohol consumers. The results extend previous reports showing that individuals without histories of drug abuse, even those who are moderate consumers of alcohol, do not self-administer sedative/anxiolytic drugs or experience their effects as euphorigenic.     

Preferences for ethanol and diazepam in anxious individuals: an evaluation of the self-medication hypothesis

The self-medication hypothesis of addictive disorders postulates that individuals with psychiatric symptoms use drugs to alleviate their symptoms. Although commonly cited to explain the etiology of substance abuse, self-medication has not been experimentally validated. This study evaluated one version of the self-medication hypothesis by formulating it into a testable hypothesis: are highly anxious volunteers more likely to self-administer anxiolytic drugs than non-anxious controls. Anxious (ANX,n=22) and control (CTL,n=23) subjects participated in two double-blind placebo-controlled experiments, one testing ethanol (0.8 g/kg) and the other testing diazepam (20 mg). Subjects sampled and then chose between ethanol and placebo in one experiment, and diazepam and placebo in the other. The main dependent measures were choice of drug over placebo and subjective responses to the drugs. Ethanol decreased self-reported anxiety in ANX subjects, but ANX subjects did not choose ethanol more often than CTL subjects. Diazepam did not measurably reduce anxiety, but ANX subjects nevertheless chose diazepam more often than did CTL subjects. Thus, there were some differences in drug responses between the ANX and CTL subjects, and the study provided limited support for the self-medication hypothesis. However, drug choice was not directly related to anxiolytic drug effects with either ethanol or diazepam. The procedure may be used to test other formulations of the self-medication hypothesis (e.g., examining other psychiatric risk factors).     

Reinforcing and subjective effects of diazepam in nondrug-abusing volunteers

Preference for diazepam was assessed in 18 light and 12 moderate social drinkers using a cumulative dosing procedure. The 7-session procedure consisted of: 1) four sampling sessions, during which participants ingested color-coded capsules containing either diazepam (five 4-mg capsules administered at 30-min intervals; total dose 20 mg) or placebo, and 2) three choice sessions, during which they could ingest up to 7 capsules of their preferred color of capsule, each separated by 30 min. Subjective (mood) and behavioral (performance) measures were obtained throughout the 4-hour sessions. The light social drinkers chose diazepam over placebo on 66% of the choice sessions, and ingested a mean dose per session of about 16 mg. The moderate drinkers chose diazepam on 100% of the choice sessions, and ingested an average dose of 25 mg per session. Diazepam produced sedation in both groups, but in the moderate drinkers it also increased measures of subjective effects suggestive of "euphoria." The results indicate that diazepam can serve as a positive reinforcer under laboratory conditions in nondrug-abusing individuals who are moderate users of alcohol and other drugs. Greater reinforcing efficacy may be indicative of higher risk of abuse. The results illustrate the usefulness of the cumulative dosing procedure to measure both drug preference and dose preference.     

Alcohol pretreatment increases preference for cocaine over monetary reinforcement

Non-dependent cocaine users participated in a two-phase experiment conducted under controlled laboratory conditions. During phase 1, subjects sampled intranasal cocaine (100 mg) and placebo (96 mg lactose +4 mg cocaine) in separate sessions and under double-blind conditions. Sampling sessions were followed by a single choice session in which subjects made a maximum of ten choices between 10 mg unit doses of cocaine or placebo. Only subjects who reliably (70%) chose cocaine over placebo in phase 1 participated in phase 2. During phase 2, subjects participated in a series of nine experimental sessions conducted on different days in which they were pretreated with varying doses of alcohol (placebo, 0.5, and 1.0 g/kg) and made a maximum of ten choices between 10 mg unit doses of cocaine and an alternative reinforcer (i.e., varying amounts of money). Visual-analog ratings of drug effects and cardiac function were monitored across all experimental sessions. Cocaine was reliably chosen over placebo by the majority (9 of 11) of subjects during phase 1, demonstrating that the drug functioned as a reinforcer. During phase 2, alcohol pretreatment significantly increased choice of cocaine over the alternative reinforcer, while increasing monetary value decreased cocaine choice. Ratings on some visual-analog scales (e.g., good effects) paralleled cocaine choice, with alcohol pretreatment increasing ratings and greater monetary value decreasing them. Cardiac output increased above baseline levels across all alcohol and monetary conditions, but maximal effects were observed during sessions involving pretreatment with the active alcohol doses. Overall, these results demonstrate (a) that alcohol can increase preference for cocaine over alternative reinforcers and thereby may thwart efforts to reduce or abstain from cocaine use, (b) that availability of an alternative, non-drug reinforcer can effectively decrease preference for cocaine, and (c) that combined use of alcohol and cocaine increases cardiac risk compared to use of cocaine alone.     

Choice of nitrous oxide and its subjective effects in light and moderate drinkers

BACKGROUND: Alcohol-drinking status has been shown to modulate the reinforcing and subjective effects of a number of drugs. We have previously published two studies on the modulating effects of alcohol-drinking status on choice for, and subjective effects of, nitrous oxide, but the results were equivocal. Using a methodology different from our previous studies, we sought to determine in a more definitive fashion the degree to which the choice of nitrous oxide and its subjective effects were modulated by drinking status. METHODS: Four concentrations of nitrous oxide (0, 20, 30, and 40%) were administered to 16 moderate drinkers (MDs) and 16 light drinkers (LDs) across four 3.5-h sessions. During experimental sessions, subjects first completed two 10-min sampling trials in which one of the nitrous oxide concentrations and placebo (100% oxygen) were inhaled. Subjective and psychomotor tests were given 5min into each sampling trial. During the subsequent choice period, subjects were allowed to choose what they wanted to inhale (drug, placebo, or "drug-free air") on nine contiguous 5-min choice trials. RESULTS: Choice of nitrous oxide was modulated by drinking status: MDs but not LDs chose nitrous oxide significantly more times than placebo, and MDs also chose nitrous oxide significantly more times than did LDs. At each active nitrous oxide concentration, MDs reported more abuse liability-related subjective effects, especially at the 20% and 30% concentrations. CONCLUSIONS: The results of the present study provide more conclusive evidence that choice as well as subjective effects of nitrous oxide is modulated by alcohol-drinking status.     

Caffeine reinforcement: the role of withdrawal

This study examined caffeine’s acute and withdrawal effects in moderate caffeine consumers (mean = 379 mg/day caffeine) to compare the relative contributions each might have to caffeine reinforcement. Subjects were caffeine restricted on the night before each of three sessions, which generally occurred at weekly intervals; these restrictions lasted until the session was completed approximately 19 h later. During the first two sessions, subjects received either placebo or caffeine (each subject’s average daily intake). These two conditions occurred using a double-blind, quasi-random, crossover design. At the end of each session subjects completed the POMS, a caffeine withdrawal questionnaire, and a Multiple-Choice Form on which subjects made a series of discrete choices between receiving the drug again or receiving varying amounts of money. This form also included negative money amounts to assess how much subjects would forfeit to avoid placebo (e.g., withdrawal symptoms after placebo). During the third session, one of the previous choices was randomly selected and the consequence of that choice was implemented. Placebo increased self-reported feelings of “worn out,”“headache,” and “flu-like feelings,” and decreased “alert,”“upset stomach,”“helpful,” and “well-being” relative to caffeine. On the Multiple-Choice Forms, subjects chose to receive caffeine rather than an average of $0.38 and to forfeit $2.51 to avoid receiving placebo again. “Headache” was significantly correlated with amount of money forfeited to avoid placebo. These results suggest that, under these conditions, choice of caffeine is more potently controlled by avoiding withdrawal than it is by the positive effects of caffeine. Received: 21 December 1995 / Final version: 26 October 1996     

Learning alcohol tolerance by mental or physical practice.

Five groups of six male social drinkers learned a psychomotor task (Tracometer) and subsequently attended five sessions to perform the task after drinking. On each of the first four sessions, subjects received 0.62 g/kg alcohol. On session 5, a placebo was administered when alcohol was expected. During treatment sessions 1-3, two groups performed the task with a valuable consequence for drug-compensatory performance: either information (IO) or information plus money (MI). This MI experience was mentally rehearsed by a third group (MR). Two control groups performed the task, either with no outcome (N), or with money for compensatory performance but no information about earnings until the study concluded (MO). Sessions 4 and 5 assessed the effect of the prior treatments when all groups performed the task with money and information. Groups MI, IO and MR displayed comparable and significantly more tolerance and a stronger compensatory response to placebo than control groups MO and N. The evidence indicates that mental or physical practice associating drug-compensatory performance with some valuable outcome enhances tolerance to moderate doses of alcohol.