Addictive Behaviors, Vol. 15, pp. 137-146, Printed in the USA. All rights reserved.
REACTIVITY
1990 Copyright
TO ALCOHOL CUES AND INDUCED ALCOHOLICS
MARK D. LIFT, NED L. COONEY,* University
of Connecticut
and RONALD
0306-4603&O $3.00 + .OO e 1990 Pergamon Press plc
MOODS IN
M. KADDEN
Health Center
LARRY GAUPP VA Medical Center, Newington,
Connecticut
Abstract - It has been theorized that respondent conditioning processes in part underlie desire for alcohol and thus contribute to relapse after alcoholism treatment. One implication of this theory is that the relevant conditioned responses could be eliminated by respondent extinction, in which the alcoholic patient is exposed to alcohol-related stimuli while being prevented from consuming alcohol. However, exteroceptive cues such as the sight and smell of alcoholic beverages are not always sufficient to elicit desire for alcohol. In view of this, it has been suggested that interoceptive cues, such as mood states, may also play a role in eliciting desire for alcohol. To test this, eight alcoholic subjects were induced to experience negative or neutral moods on four separate days, and then exposed to the sight and smell of their favorite alcoholic drink, and to a neutral stimulus (seltzer water), in a within-subjects design. Results from this work indicate that: (a) negative moods can be. reliably induced in the laboratory as confirmed by subjects’ reports; (b) exposure to alcohol cues had no effect on desire for alcohol while subjects were in a relaxed, neutral mood state; (c) the presence of negative mood states alone appeared to be sufficient to elicit desire for alcohol in some subjects, regardless of whether alcohol or water was presented. These data argue that negative mood states may cue desire for alcohol independent of other cues. The data also suggest that reactivity to alcohol cues may be substantially reduced by relaxation.
In recent years a great deal of interest has focused on the idea that classically conditioned responses underlie desire for alcohol in alcoholics and increase the likelihood of alcohol consumption (Poulos, Hinson, & Siegel, 1981). This conceptualization is especially compelling because it implies that a program of desensitization, or extinction of cue-related responses, could be a useful adjunct to standard alcoholism treatment because of its potential for the prevention of relapse. Both clinical and laboratory studies of alcoholics’ reactions to alcohol cues support the idea that some conditioned subjective and physiological responses occur in the presence of alcohol cues, and that this phenomenon may be related to relapse. For example, Kennedy (1971) and Monti, Abrams, Bit&off, Rohsenow, and Zwick (1987), in laboratory cue exposure studies, found that those alcoholics most physiologically reactive to alcohol-related stimuli were the most likely to relapse soon after treatment. Some investigators (e.g., Hodgson, Rankin, & Stockwell, 1979; Kaplan, Meyer, & Stroebel, 1983; Laberg & Ellertsen, 1987) have examined reactivity to alcohol cues following priming doses of alcohol. After a priming dose, alcoholic subjects were more likely to report increased desire to drink in the presence of alcohol cues than when not exposed to cues. Kaplan et al. (1985) conducted a cue exposure study in which no alcohol consumption
*Now at VA Medical Center, New Haven, CT. Support for this article was provided by NIAAA grants P50-AA-035 10 and 2T32-AA07290. The authors would like to thank Kenneth SanGiacomo and Deborah Busher for their help in data collection and analysis. Requests for reprints should be sent to Mark D. Litt, Department of Behavioral Sciences and Community Health, University of Connecticut Health Center, Farmington, CT 06032. 137
138
MARK D. LITT et al.
was allowed. Alcoholic subjects showed greater heart rate and skin conductance level during alcohol cue exposure than did nonalcoholic subjects. However, a reanalysis of their data reveals that only 56% of alcoholic subjects responded to the sight and smell of alcohol with increased desire to drink. Similar rates of subjective responding when no alcohol consumption is allowed (i.e., about 50% of subjects showing reactivity) have been reported elsewhere (e.g., Laberg & Ellertsen, 1987; Monti et al., 1987). Thus, although alcohol cues do appear to elicit desire for alcohol in some, this is not true for all alcoholics. It appears that alcohol cues alone may not be sufficiently salient to elicit desire for many alcoholics in the absence of an opportunity to consume. If extinction-based treatments are to be clinically useful for abstinent alcoholics, it will be necessary to find other cues that elicit a desire to drink. Much of the research to date on cue reactivity in alcoholics has focused on exteroceptive cues, namely the sight and smell of alcoholic beverages. It has also been suggested, however, that interoceptive cues, particularly mood states, can also elicit desire for alcohol. This may occur as a result of the classical conditioning that takes place if a person repeatedly drinks while in a particular mood, if a particular mood state accompanies withdrawal on a number of occasions, or if a negative mood increases the salience of the immediate effects of alcohol. Poulos et al. (1981) suggest that, under such conditions, affective states become part of a conditioned stimulus complex that may subsequently elicit conditioned craving or withdrawal responses. There is some evidence that mood states, particularly negative ones, may cue desire for alcohol. Marlatt and Gordon (1980), for instance, interviewed relapsed alcoholics and found that a majority of relapses occurred under conditions of interpersonal stress or negative mood state. Additionally, Childress, McLellan, Natale, and O’Brien (1987) found that when negative mood states were hypnotically induced in opiate addicts, two-thirds of their subjects reported “craving” or withdrawal symptoms regardless of whether drug cues were present. The purpose of the present study was to examine whether negative mood states can indeed serve as interoceptive cues for alcohol desire, either by themselves or as part of a stimulus complex that includes exteroceptive alcohol cues (i.e., the sight and smell of a preferred alcoholic beverage). Inpatient alcoholics were induced to experience either negative or neutral mood states, and were then presented either with their preferred alcoholic beverage or with a neutral stimulus (seltzer water), in a within-subjects design. It was hypothesized that subjects would exhibit subjective desire for alcohol and physiological arousal when exposed to alcohol cues, and that these responses would be augmented when the subject was in a negative mood state. METHOD
Subjects
Eight male veterans were recruited from an inpatient alcoholism treatment program at a Veterans Administration Medical Center. Subjects were recruited if their primary (DSMIII-R) diagnosis was alcohol dependence, and if they met each of the following criteria: (a) no concurrent diagnosis of other substance abuse or dependence, (b) no indication of suicidality, homicidality, thought disorder, or organic brain syndrome, (c) no current use of disulfiram, anti-depressant, anti-hypertensive, anxiolytic, or vasoactive medications, and (d) no medical conditions that would prohibit participation. The subjects, seven white and one black, ranged from 23 to 62 years of age (M = 42, SD = 13.2). All subjects had been abstinent from alcohol for a minimum of seven days prior to the beginning of the study. Prior to admission the subjects had been drinking, on average, the equivalent of six or more standard drinks per day on 15 of the preceding 30 days. All subjects had tried to stop drinking in the past, with five of the subjects having received one or more formal treatments for alcoholism.
Alcohol cue reactivity
Session
Timeline
Mood
Stimulus
Induction
Presentation
Water
Negative
Attach sensors.
139
Resting
’
base’ine
L
’
-
Alcohol
/’
Recovery
Acclimation -
0
10
/I Neutral
15
\ Alcohol
35
Time
-
water
40
/
45
50
in Minutes
Fig. 1. Session timeline showing the order of procedures for a single session. Four such sessions were completed for each subject, with mood-induction and order of stimulus being alternated on different days.
Procedure Patients found eligible on the basis of the preceding criteria were told that the study was concerned with the effects of alcohol cues and mood states on the experience of craving for alcohol. It was explained that participation would involve one 45min session with a research assistant who would test the subject’s ability to relax, followed by four l-hour experimental sessions on subsequent days. Patients who agreed to participate were asked to read and sign a statement of informed consent. Of 15 eligible patients contacted, seven declined participation for lack of interest or scheduling problems. Following the signing of the consent form, the experimenter inquired as to the subject’s preferred alcoholic beverage and administered the short form of the Inventory of Drinking Situations (Annis, 1984). This inventory lists 42 situations and requests subjects to rate how often they drink heavily in each one, on a scale from 1 (never) to 4 (almost always). Each subject’s pattern of responding on the inventory was used to identify a target negative mood state frequently associated with heavy drinking. The experimenter then elicited from each subject a specific episode in the past in which he felt the target mood strongly (e.g., “It looks like you often drink heavily when you feel nervous or under pressure. Can you tell me about a time recently when you felt very nervous or really under pressure?“). This episode (mood scene) was recorded in detail for use later to induce the target negative mood during the appropriate experimental session. The experimenter took care to select a mood scene that did not include the use or presence of alcoholic beverages. Subjects also met individually with a research assistant for a 45min period during which they completed several instruments assessing severity of alcohol dependence and hypnotic susceptibility. These included the Rand Alcohol Dependence Scale (Polich, Armor, & Braiker, 1981), tapping alcohol dependence symptoms experienced by the subject in the previous 30 days, and the Stanford Hypnotic Clinical Scale (SHCS; Hilgard & Hilgard, 1975), a measure of hypnotic susceptibility. The four experimental sessions were conducted on separate days between 4:00 p.m. and 5:00 p.m. All sessions were conducted in a small laboratory room measuring approximately 10 ft x 15 ft. The room was dimly lit to enhance the hypnotic induction procedure. The
140
MARK D. LITT et al.
subject was seated in a large, comfortable chair and a blood pressure cuff was placed on his non-dominant arm. The cuff was attached to a Critikon Dynamap 845XT vital signs monitor that provided periodic measurements of systolic and diastolic blood pressure. The experimenter sat at a desk to the right of the subject. A small table separated the subject’s chair from the desk. Subjects were randomly assigned to one of four orders of presentation of mood induction (negative or neutral) and stimuli (alcoholic beverage first or seltzer water first) over the four sessions. In this way every subject experienced each combination of induced mood and stimulus presentation, and the effect of sequencing was controlled for. Figure 1 depicts the sequence of events for a session. Each session began with a 6-min acclimation period, during which three blood pressure readings were taken and the experimenter acquainted the subject with the Within-Session Rating Form (WSRF) which would be used to collect data on subjective mood state and desire for alcohol. The WSRF consisted of several visual analog scales. Four of these scales asked the subject to rate the degree to which he was happy, angry, tense, or depressed at that moment by placing an “X” at the point on a 1OOmm line, from “Not at all” to “Extremely,” that corresponded to how he felt. These items were included to check on the mood induction procedure. Another scale asked subjects to indicate their desire for an alcoholic drink. A final scale tapped subjects’ estimates of the difficulty they would have in resisting an alcoholic drink at that moment. The acclimation period was followed by a 5-min resting baseline period during which blood pressure readings were taken every two min. At the end of the baseline period subjects completed the WSRF. Depending upon the sequence to which he had been assigned, the subject then underwent a neutral or negative mood induction. A hypnotic mood induction procedure was used because of its ability to produce strong emotions capable of lasting through a variety of tasks over several minutes (Bower, 1981). This involved a IO-min initial hypnotic induction emphasizing relaxation and comfort. This was followed by suggestions to (a) imagine traveling cross-country by bus, not feeling any particular way at all, just neutral, (the neutral mood scene), or (b) imagine in detail the negative mood scene elicited by the experimenter prior to the beginning of the study. The subject was then asked to rate aloud, on a scale from 0 to 10, the degree to which he was experiencing the target negative mood or the neutral mood, depending on the experimental condition. After reaching a sufficient absorption in the desired mood (a self-rating of 6 or more on the O-10 scale), the subject was asked to open his eyes slowly and watch as the beverage (seltzer water or preferred alcoholic drink in its labeled bottle) was poured into a glass and placed on the table to the subject’s right. He was then asked to hold and sniff the beverage for three mitt, during which two blood pressure readings were taken. At the end of this presentation, with the beverage still in view, the subject again completed the WSRF. After completing the WSRF, the subject was again asked to close his eyes and imagine the appropriate scene (i.e., either riding on a bus or the negative mood scene). As before, a self-rating of absorption in the mood was requested and the mood scene was described until a rating of at least 6 out of 10 was reported. The second beverage was then presented in the same manner as the first, with a 3-min exposure during which two blood pressure measurements were taken, followed by completion of the WSRF. The subject was then asked to close his eyes and to return to a relaxed state. During this time three “recovery” blood pressure readings were taken. After the subject reported feeling relaxed and comfortable, a final WSRF was completed. Succeeding sessions proceeded in the same manner, with different moods (negative or neutral) induced on alternate days, and the order of presentation of beverages alternated.
Alcohol cue reactivity
0
loo-
141
Subject
EE3 Baseline
go--
Response Value
80-70 -60-so-40-30 -20-lO--
0 Sl S2 S3 S4 S5 S6 S7 58 Water Negative
SlS2SSS4SSS6S7S8 Alcohol Mood Condition
100 90 80 z
70 60
z % p F
50 40 30
10 0 SlS2S3S4SSS6S758 Water
SlS2SSSISSS6S7S8 Alcohol Neutral
Mood Condition
Fig. 2. Target mood ratings made along a 100 mm visual analog scale for each combination induction and stimulus. Data from all subjects (Sl-S8) are presented separately.
of mood
RESULTS
Dependent variables for this study were reported desire for alcohol, reported difficulty resisting alcohol, and systolic blood pressure. Although diastolic blood pressure was also collected, the systolic blood pressure is reported because of findings by Niaura, Wilson, and Westrick (1988) showing that only systolic blood pressure was reactive to stress. Data from each subject’s first two experimental sessions were dropped due to the large degree of variability in subjective and physiological responses often observed as subjects adapted to the novel surroundings and the alcohol presentation. In this regard Newlin and Pretorius (1987)
142
MARKD.LITTetal
loo90 -80--
0
Subject
tZJ
Baseline
Response Value
70-60 -50-40 -30 -20-lO--
0
SlS2S3S4S5S6S7S8 Water Negative
SlS2SSS4S5S6S7SB Alcohol Mood Condition
1OOT
90
t
Y
80 --
c ‘iT
70--
n 0
60 -50 --
!! .-
40--
g
30--
0
20 -10 0 SlS2S3SIS5S6S7S8 Water Neutral
SlS2S3S4S5S6S7SS Alcohol Mood Condition
Fig. 3. Ratings of Desire to Drink made along a 100 mm visual analog scale for each combination of mood induction and stimulus. Data from all subjects (Sl-S8) are presented separately.
have reported that cue exposure data collected in an unfamiliar environment may be contaminated by responses attributable to the novelty of the situation. Keeping only the data from the last two sessions allowed examination of responses by subjects well-acquainted with the procedure while still preserving the counterbalancing across subjects of the order of mood induction and stimulus presentation. Prior to analyzing for effects of stimulus or mood, subjects’ baseline ratings of mood, desire for alcohol, and difficulty resisting alcohol, as well as baseline systolic blood pressure readings, were compared using Wilcoxon signed-rank matched pairs tests. These
Alcohol cue reactivity
analyses revealed no significant differences in baseline scores between implying that conditions were comparable across these two sessions.
143
sessions
3 and 4,
Manipulation check Figure 2 shows each subject’s ratings of target mood, as taken from the WSRF, following each beverage presentation during each type of mood induction. Also shown are baseline target mood ratings. Difference scores (subject’s induced mood response minus baseline mood response) were calculated for each subject for each combination of mood induction and stimulus presentation. Because of the large amount of variability in distributions of these data, a Friedman non-parametric 2-way analysis of variance by ranks (Siegel, 1956) was used to detect systematic differences among the within-subjects cells. The omnibus Friedman test indicated significant differences among the conditions (x2, = 15.08, df = 3, p < .Ol). Wilcoxon matched pairs signed-ranks tests (Siegel, 1956) with adjusted p values were used to detect differences attributable to mood induction, stimulus, or an interactionof mood and stimulus. These tests indicated a significant main effect for mood induction (T = 1, p < .Ol) with higher negative mood ratings in the negative mood conditions, a significant effect for stimulus (7’ = 3.5, p < .025), with a greater negative mood response in the presence of water, and no significant interaction effect. The Spearman rank-order correlation between ratings of negative mood after mood induction and scores on the SHCS was rs = .69 @ < .05). These results indicate that the negative mood induction technique used was successful in producing the desired moods, and that this was in part a function of hypnotic susceptibility. The Spearman correlation between ratings of negative mood and scores on the Rand dependence scale was r, = - .04. Alcohol cue and mood efects Omnibus Friedman 2-way ANOVAs with post-hoc Wilcoxon tests were used in the same manner as above to analyze subjects’ responses (expressed as difference scores) along the parameters of desire for alcohol, difficulty resisting alcohol, and systolic blood pressure. Figure 3 shows subjects’ ratings of desire for alcohol for each combination of mood induction and stimulus, along with baseline values. The Friedman ANOVA, corrected for ties among rankings (Siegel, 1956), indicated significant differences in desire ratings among the conditions (x2, = 8.3 1, df= 3, p < .05). Wilcoxon post-tests showed a significant main effect for mood induction (T = 4, p < .025), with greater desire for alcohol in the negative mood conditions, no significant main effect for stimulus, and no significant interaction. Spearman rank-order correlations also indicated that ratings of desire for alcohol during negative mood induction were highly correlated with target mood rating (rs = .71, p < .03). Rand dependence scores did not correlate significantly with alcohol desire ratings in any condition, indicating that desire for alcohol during negative mood induction was not a function of severity of alcohol dependence. Subjects’ ratings of difficulty resisting alcohol were analyzed in the same manner as above. An omnibus Friedman ANOVA on difference scores (subjects’ ratings of difficulty resisting alcohol in each condition minus baseline rating) failed to detect significant differences among the conditions. Friedman ANOVAs performed on difference scores on the blood pressure measure also failed to find significant differences among the conditions. DISCUSSION
The idea that an extinction procedure can be used as one element of a treatment program to reduce the probability of relapse after treatment is extremely appealing. If an extinction procedure is to be generally useful, however, stimuli must be found that will reliably elicit desire for alcohol. Previous research using exteroceptive cues, such as alcoholic beverages,
144
MARK D. LI’lT et al
has shown that these cues do not always elicit desire for alcohol in alcoholic subjects. The purpose of the present study was to explore the possibility that an interoceptive cue, negative mood state, may play a role in eliciting desire in alcoholic subjects. The present results suggest, first, that the mood induction procedure used here appears to have been effective for a majority of subjects. In general, it appears that negative moods can indeed be induced in the laboratory, and that these moods can be maintained through the course of an alcohol cue exposure session. In the present study, however, the ability to achieve a negative mood was at least in part a function of hypnotic susceptibility, which may thus limit the general applicability of a hypnotic induction procedure in extinction-based treatment paradigms. The most interesting, and potentially important, finding of the present study was the indication that negative mood states can elicit desire for alcohol independent of the presence of exteroceptive cues such as the sight or smell of alcoholic beverages. This result is similar to that found by Childress et al. (1987) with opiate addicts. In contrast to findings in previous research (e.g., Kaplan et al., 1985), there was no effect on any parameter of desire when subjects were exposed to alcohol beverage cues in the absence of a negative mood state. Although responses to the alcohol stimulus during the neutral mood induction condition were not expected to be as great as those during the negative mood induction, some response was expected. One possible reason for this lack of cue reactivity in the neutral mood condition is that the relaxation used in the mood inductions inhibited subjective and blood pressure responses to alcohol cues. This hypothesis is supported by Loberg (1987), who found that relaxation induction led to decreased craving responses in alcoholics exposed to alcohol cues. In short, we suspect that the relaxation in the neutral mood conditions inhibited desire reactions. In the negative mood conditions reactions to alcohol were obscured by the mood effects. In other words, the effects of alcohol cues and negative mood state may not be additive. This, however, is not a very satisfying explanation for this apparent lack of alcohol cue reactivity, and these effects will need to be reexamined in further studies. The suggestion that relaxation may inhibit desire for alcohol deserves further study. It implies that future mood induction paradigms may need to be altered to minimize relaxation that might interfere with alcohol cue reactivity and thus slow the extinction process in a cue-exposure treatment. It further suggests, however, that systematic relaxation training may in itself be a useful adjunct to relapse prevention training, perhaps to be used in times of mood disturbance. It is notable that, although ratings of desire for alcohol varied significantly across conditions, ratings of difficulty resisting alcohol did not. Additionally, ratings of alcohol desire and difficulty resisting alcohol were not well correlated in session 3 (Spearman rank order correlations were rs = .28 in session 3 and .68 in session 4). One explanation for this may be that the subjects, all inpatients in an alcohol treatment program and trying to maintain abstinence, may have been willing to acknowledge desire for alcohol, but may not have been willing to admit a difficulty on their part in being able to resist alcohol. This explanation was supported by the fact that ratings of difficulty resisting alcohol were uniformly lower than ratings of desire. Systolic blood pressure responses were unaffected by mood induction, stimulus presentation, or the interaction of mood and stimulus. In addition, blood pressure readings were not well correlated with either subjective measure of distress. Thus, we can draw no conclusions as to subjects’ physiological reactivity to mood state or to alcoholic beverages. Future studies should employ additional physiological measures to shed light on the process of “craving” or desire for alcohol. Rankin, Hodgson, and Stockwell (1979), for example, found body
Alcohol cue reactivity
145
temperature to be a correlate of alcohol desire. Likewise, in their work with opiate addicts, Childress et al. (1987) found skin temperature to be reactive to opiate cue exposure. It would be useful to determine if mood states alone can have equivalent effects on skin temperature in alcoholics. It is unclear why target mood ratings should have been higher during presentation of seltzer water than during presentation of alcohol. During debriefing, however, it was found that two of the most reactive subjects reported feeling extremely distressed after the negative mood induction, but their distress decreased upon presentation of the alcoholic beverage. This raises the possibility that, at least for some subjects, consumption of alcohol may not be necessary to relieve negative mood states; the mere presence of alcohol may in itself be a conditioned stimulus for relief of distress, possibly by eliciting a conditioned appetitive response (Stewart, DeWit, & Eikelboom, 1984). This idea deserves further study. A possible limitation of the present study is that the results obtained may have been a function of demand characteristics. That is, subjects may have expected to be more reactive in the negative mood induction conditions and therefore may have reported greater desire for alcohol. If demand characteristics were playing a role, however, we might also expect higher ratings of alcohol desire in those conditions in which alcohol was presented as the beverage stimulus. The fact that there were no beverage effects in this study tends to diminish the argument that demand characteristics were responsible for the results. To conclude, the present data are in some ways quite consistent with what many alcoholic patients tell us about relapse; namely, that they frequently leave treatment programs feeling fine until something happens that leads them to feel distressed, and then they experience a desire for alcohol that leads them to seek a drink. Perhaps, at least for a subset of alcoholics, interoceptive cues like mood states may be important in leading to the first drink of a relapse. Thereafter, exteroceptive cues provided by the sight, smell and taste of alcohol, may play a role in maintaining further drinking behavior. This model is, in fact, consistent with results reported by Laberg and Ellertsen (1987), who found little alcohol cue reactivity in alcoholics unless a priming dose of alcohol was first consumed. Future research, along the lines explored here, and with a much larger number of subjects, is required to clarify the extent of variability in cue reactivity and the roles of interoceptive and exteroceptive cues in the process of relapse. This research may, in turn, provide direction for more effective relapse prevention strategies.
REFERENCES Annis, H.M. (1984). Invenro~ of drinking situations. Toronto: Addiction Research Foundation, Bower, G.H. (1981). Mood and memory. American Psychologist, 36, 129-148. Childress, A.R., McLellan, A.T., Natale, M., & O’Brien, C.P. (1987). Mood states can elicit conditioned withdrawal and craving in opiate abuse patients. Problems of drug dependence, 1986. NIDA Research Monograph, Vol. 76 (DHHS No. (ADM) 87-1508) (PP. 137-144). Hilgard, E.R.. & Hilgard, J.R. (1975). Hypnosis in the reief of pain. Los Altos. CA: William Kaufman. Hodgson, R., Rankin, H., & Stockwell, T. (1979). Alcohol dependence and the priming effect. Behaviour Research and Therapy, 17. 379-387. Kaplan, R.F., Cooney, N.L., Baker, L.H., Gillespie, R.A., Meyer, R.E., & Pomerleau, O.F. (1985). Reactivity to alcohol-related cues: Physiological and subjective responses in alcoholics and non-problem drinkers, Journal of Studies on Alcohol, 46, 267-272. Kaplan, R.F., Meyer, R.E., & Stroebel, C.F. (1983). Alcohol dependence and responsivity to an ethanol stimulus as predictors of alcohol consumption. British Journal of Addiction, 78, 259-267. Kennedy, D.A. (1971). Pupilometric analysis as an aid in the assessment of maturation, impact of treatment, and prognosis of chronic alcoholics. Unpublished dissertation, University of Utah, Laberg, J.C., & Ellertsen, B. (1987). Psychophysiological indicators of craving in alcoholics: Effects of cue exposure. British Journal of Addiction, 82, 1341-1348. Loberg, T. (1987). Alcohol cue reactivity in alcoholics. Paper presented at the Fourth International Conference on
146
MARK D. LITT et al.
the Treatment of Addictive Behaviors, Bergen, Norway. Marlatt, G.A., & Gordon, J.R. (1980). Determinants of relapse; implications for the maintenance of behavior change. In P. Davidson (Ed.), Behavioral medicine: Changing health lifestyles (Vol 10). New York: Brunner-Mazel. Mont& P., Abrams, D., Binkoff, J., Rohsenow, D., & Zwick, W. (1987). Cue reacrivity in alcoholics. Paper presented at the Fourth International Conference on the Treatment of Addictive Behaviors, Bergen, Norway. Newlin, D.B., & Pretorius, M.B. (1987). Greater alcohol effect in a familiar than a novel environment: A preliminary test of the response competition model. Alcoholism: Clinical and Experimenral Research, 11, 218 (abstract). Niaura, R., Wilson, G.T., & Westrick, E. (1988). Self-awareness, alcohol consumption, and reduced cardiovascular reactivity. Psychosomatic Medicine, 50, 360-380. Polich, J.M., Armor, D.J., & Braiker, H.B. (1981). The course ofakoholism: Four years after freafment. New York: Wiley. Poulos, C.X., Hinson, R.E., & Siegel, S., (1981). The role of Pavlovian processes in drug tolerance and dependence: Implications for treatment. Addicfive Behaviors, 6, 205-212. Rankin, H., Hodgson, R., & Stockwell, T. (1979). The concept of craving and its measurement. Eehaviour Research and Therapy, 17, 389-396. Siegel, S. (1956). Nonparametric srarisfics for the behavioral sciences. New York: McGraw Hill. Stewart, J., DeWit, H., & Eikelboom, R. (1984). Role of unconditioned and conditioned drug effects in the self-administration of opiates and stimulants. Psychological Review, 91, 251-268.
REACTIVITY
1990 Copyright
TO ALCOHOL CUES AND INDUCED ALCOHOLICS
MARK D. LIFT, NED L. COONEY,* University
of Connecticut
and RONALD
0306-4603&O $3.00 + .OO e 1990 Pergamon Press plc
MOODS IN
M. KADDEN
Health Center
LARRY GAUPP VA Medical Center, Newington,
Connecticut
Abstract - It has been theorized that respondent conditioning processes in part underlie desire for alcohol and thus contribute to relapse after alcoholism treatment. One implication of this theory is that the relevant conditioned responses could be eliminated by respondent extinction, in which the alcoholic patient is exposed to alcohol-related stimuli while being prevented from consuming alcohol. However, exteroceptive cues such as the sight and smell of alcoholic beverages are not always sufficient to elicit desire for alcohol. In view of this, it has been suggested that interoceptive cues, such as mood states, may also play a role in eliciting desire for alcohol. To test this, eight alcoholic subjects were induced to experience negative or neutral moods on four separate days, and then exposed to the sight and smell of their favorite alcoholic drink, and to a neutral stimulus (seltzer water), in a within-subjects design. Results from this work indicate that: (a) negative moods can be. reliably induced in the laboratory as confirmed by subjects’ reports; (b) exposure to alcohol cues had no effect on desire for alcohol while subjects were in a relaxed, neutral mood state; (c) the presence of negative mood states alone appeared to be sufficient to elicit desire for alcohol in some subjects, regardless of whether alcohol or water was presented. These data argue that negative mood states may cue desire for alcohol independent of other cues. The data also suggest that reactivity to alcohol cues may be substantially reduced by relaxation.
In recent years a great deal of interest has focused on the idea that classically conditioned responses underlie desire for alcohol in alcoholics and increase the likelihood of alcohol consumption (Poulos, Hinson, & Siegel, 1981). This conceptualization is especially compelling because it implies that a program of desensitization, or extinction of cue-related responses, could be a useful adjunct to standard alcoholism treatment because of its potential for the prevention of relapse. Both clinical and laboratory studies of alcoholics’ reactions to alcohol cues support the idea that some conditioned subjective and physiological responses occur in the presence of alcohol cues, and that this phenomenon may be related to relapse. For example, Kennedy (1971) and Monti, Abrams, Bit&off, Rohsenow, and Zwick (1987), in laboratory cue exposure studies, found that those alcoholics most physiologically reactive to alcohol-related stimuli were the most likely to relapse soon after treatment. Some investigators (e.g., Hodgson, Rankin, & Stockwell, 1979; Kaplan, Meyer, & Stroebel, 1983; Laberg & Ellertsen, 1987) have examined reactivity to alcohol cues following priming doses of alcohol. After a priming dose, alcoholic subjects were more likely to report increased desire to drink in the presence of alcohol cues than when not exposed to cues. Kaplan et al. (1985) conducted a cue exposure study in which no alcohol consumption
*Now at VA Medical Center, New Haven, CT. Support for this article was provided by NIAAA grants P50-AA-035 10 and 2T32-AA07290. The authors would like to thank Kenneth SanGiacomo and Deborah Busher for their help in data collection and analysis. Requests for reprints should be sent to Mark D. Litt, Department of Behavioral Sciences and Community Health, University of Connecticut Health Center, Farmington, CT 06032. 137
138
MARK D. LITT et al.
was allowed. Alcoholic subjects showed greater heart rate and skin conductance level during alcohol cue exposure than did nonalcoholic subjects. However, a reanalysis of their data reveals that only 56% of alcoholic subjects responded to the sight and smell of alcohol with increased desire to drink. Similar rates of subjective responding when no alcohol consumption is allowed (i.e., about 50% of subjects showing reactivity) have been reported elsewhere (e.g., Laberg & Ellertsen, 1987; Monti et al., 1987). Thus, although alcohol cues do appear to elicit desire for alcohol in some, this is not true for all alcoholics. It appears that alcohol cues alone may not be sufficiently salient to elicit desire for many alcoholics in the absence of an opportunity to consume. If extinction-based treatments are to be clinically useful for abstinent alcoholics, it will be necessary to find other cues that elicit a desire to drink. Much of the research to date on cue reactivity in alcoholics has focused on exteroceptive cues, namely the sight and smell of alcoholic beverages. It has also been suggested, however, that interoceptive cues, particularly mood states, can also elicit desire for alcohol. This may occur as a result of the classical conditioning that takes place if a person repeatedly drinks while in a particular mood, if a particular mood state accompanies withdrawal on a number of occasions, or if a negative mood increases the salience of the immediate effects of alcohol. Poulos et al. (1981) suggest that, under such conditions, affective states become part of a conditioned stimulus complex that may subsequently elicit conditioned craving or withdrawal responses. There is some evidence that mood states, particularly negative ones, may cue desire for alcohol. Marlatt and Gordon (1980), for instance, interviewed relapsed alcoholics and found that a majority of relapses occurred under conditions of interpersonal stress or negative mood state. Additionally, Childress, McLellan, Natale, and O’Brien (1987) found that when negative mood states were hypnotically induced in opiate addicts, two-thirds of their subjects reported “craving” or withdrawal symptoms regardless of whether drug cues were present. The purpose of the present study was to examine whether negative mood states can indeed serve as interoceptive cues for alcohol desire, either by themselves or as part of a stimulus complex that includes exteroceptive alcohol cues (i.e., the sight and smell of a preferred alcoholic beverage). Inpatient alcoholics were induced to experience either negative or neutral mood states, and were then presented either with their preferred alcoholic beverage or with a neutral stimulus (seltzer water), in a within-subjects design. It was hypothesized that subjects would exhibit subjective desire for alcohol and physiological arousal when exposed to alcohol cues, and that these responses would be augmented when the subject was in a negative mood state. METHOD
Subjects
Eight male veterans were recruited from an inpatient alcoholism treatment program at a Veterans Administration Medical Center. Subjects were recruited if their primary (DSMIII-R) diagnosis was alcohol dependence, and if they met each of the following criteria: (a) no concurrent diagnosis of other substance abuse or dependence, (b) no indication of suicidality, homicidality, thought disorder, or organic brain syndrome, (c) no current use of disulfiram, anti-depressant, anti-hypertensive, anxiolytic, or vasoactive medications, and (d) no medical conditions that would prohibit participation. The subjects, seven white and one black, ranged from 23 to 62 years of age (M = 42, SD = 13.2). All subjects had been abstinent from alcohol for a minimum of seven days prior to the beginning of the study. Prior to admission the subjects had been drinking, on average, the equivalent of six or more standard drinks per day on 15 of the preceding 30 days. All subjects had tried to stop drinking in the past, with five of the subjects having received one or more formal treatments for alcoholism.
Alcohol cue reactivity
Session
Timeline
Mood
Stimulus
Induction
Presentation
Water
Negative
Attach sensors.
139
Resting
’
base’ine
L
’
-
Alcohol
/’
Recovery
Acclimation -
0
10
/I Neutral
15
\ Alcohol
35
Time
-
water
40
/
45
50
in Minutes
Fig. 1. Session timeline showing the order of procedures for a single session. Four such sessions were completed for each subject, with mood-induction and order of stimulus being alternated on different days.
Procedure Patients found eligible on the basis of the preceding criteria were told that the study was concerned with the effects of alcohol cues and mood states on the experience of craving for alcohol. It was explained that participation would involve one 45min session with a research assistant who would test the subject’s ability to relax, followed by four l-hour experimental sessions on subsequent days. Patients who agreed to participate were asked to read and sign a statement of informed consent. Of 15 eligible patients contacted, seven declined participation for lack of interest or scheduling problems. Following the signing of the consent form, the experimenter inquired as to the subject’s preferred alcoholic beverage and administered the short form of the Inventory of Drinking Situations (Annis, 1984). This inventory lists 42 situations and requests subjects to rate how often they drink heavily in each one, on a scale from 1 (never) to 4 (almost always). Each subject’s pattern of responding on the inventory was used to identify a target negative mood state frequently associated with heavy drinking. The experimenter then elicited from each subject a specific episode in the past in which he felt the target mood strongly (e.g., “It looks like you often drink heavily when you feel nervous or under pressure. Can you tell me about a time recently when you felt very nervous or really under pressure?“). This episode (mood scene) was recorded in detail for use later to induce the target negative mood during the appropriate experimental session. The experimenter took care to select a mood scene that did not include the use or presence of alcoholic beverages. Subjects also met individually with a research assistant for a 45min period during which they completed several instruments assessing severity of alcohol dependence and hypnotic susceptibility. These included the Rand Alcohol Dependence Scale (Polich, Armor, & Braiker, 1981), tapping alcohol dependence symptoms experienced by the subject in the previous 30 days, and the Stanford Hypnotic Clinical Scale (SHCS; Hilgard & Hilgard, 1975), a measure of hypnotic susceptibility. The four experimental sessions were conducted on separate days between 4:00 p.m. and 5:00 p.m. All sessions were conducted in a small laboratory room measuring approximately 10 ft x 15 ft. The room was dimly lit to enhance the hypnotic induction procedure. The
140
MARK D. LITT et al.
subject was seated in a large, comfortable chair and a blood pressure cuff was placed on his non-dominant arm. The cuff was attached to a Critikon Dynamap 845XT vital signs monitor that provided periodic measurements of systolic and diastolic blood pressure. The experimenter sat at a desk to the right of the subject. A small table separated the subject’s chair from the desk. Subjects were randomly assigned to one of four orders of presentation of mood induction (negative or neutral) and stimuli (alcoholic beverage first or seltzer water first) over the four sessions. In this way every subject experienced each combination of induced mood and stimulus presentation, and the effect of sequencing was controlled for. Figure 1 depicts the sequence of events for a session. Each session began with a 6-min acclimation period, during which three blood pressure readings were taken and the experimenter acquainted the subject with the Within-Session Rating Form (WSRF) which would be used to collect data on subjective mood state and desire for alcohol. The WSRF consisted of several visual analog scales. Four of these scales asked the subject to rate the degree to which he was happy, angry, tense, or depressed at that moment by placing an “X” at the point on a 1OOmm line, from “Not at all” to “Extremely,” that corresponded to how he felt. These items were included to check on the mood induction procedure. Another scale asked subjects to indicate their desire for an alcoholic drink. A final scale tapped subjects’ estimates of the difficulty they would have in resisting an alcoholic drink at that moment. The acclimation period was followed by a 5-min resting baseline period during which blood pressure readings were taken every two min. At the end of the baseline period subjects completed the WSRF. Depending upon the sequence to which he had been assigned, the subject then underwent a neutral or negative mood induction. A hypnotic mood induction procedure was used because of its ability to produce strong emotions capable of lasting through a variety of tasks over several minutes (Bower, 1981). This involved a IO-min initial hypnotic induction emphasizing relaxation and comfort. This was followed by suggestions to (a) imagine traveling cross-country by bus, not feeling any particular way at all, just neutral, (the neutral mood scene), or (b) imagine in detail the negative mood scene elicited by the experimenter prior to the beginning of the study. The subject was then asked to rate aloud, on a scale from 0 to 10, the degree to which he was experiencing the target negative mood or the neutral mood, depending on the experimental condition. After reaching a sufficient absorption in the desired mood (a self-rating of 6 or more on the O-10 scale), the subject was asked to open his eyes slowly and watch as the beverage (seltzer water or preferred alcoholic drink in its labeled bottle) was poured into a glass and placed on the table to the subject’s right. He was then asked to hold and sniff the beverage for three mitt, during which two blood pressure readings were taken. At the end of this presentation, with the beverage still in view, the subject again completed the WSRF. After completing the WSRF, the subject was again asked to close his eyes and imagine the appropriate scene (i.e., either riding on a bus or the negative mood scene). As before, a self-rating of absorption in the mood was requested and the mood scene was described until a rating of at least 6 out of 10 was reported. The second beverage was then presented in the same manner as the first, with a 3-min exposure during which two blood pressure measurements were taken, followed by completion of the WSRF. The subject was then asked to close his eyes and to return to a relaxed state. During this time three “recovery” blood pressure readings were taken. After the subject reported feeling relaxed and comfortable, a final WSRF was completed. Succeeding sessions proceeded in the same manner, with different moods (negative or neutral) induced on alternate days, and the order of presentation of beverages alternated.
Alcohol cue reactivity
0
loo-
141
Subject
EE3 Baseline
go--
Response Value
80-70 -60-so-40-30 -20-lO--
0 Sl S2 S3 S4 S5 S6 S7 58 Water Negative
SlS2SSS4SSS6S7S8 Alcohol Mood Condition
100 90 80 z
70 60
z % p F
50 40 30
10 0 SlS2S3S4SSS6S758 Water
SlS2SSSISSS6S7S8 Alcohol Neutral
Mood Condition
Fig. 2. Target mood ratings made along a 100 mm visual analog scale for each combination induction and stimulus. Data from all subjects (Sl-S8) are presented separately.
of mood
RESULTS
Dependent variables for this study were reported desire for alcohol, reported difficulty resisting alcohol, and systolic blood pressure. Although diastolic blood pressure was also collected, the systolic blood pressure is reported because of findings by Niaura, Wilson, and Westrick (1988) showing that only systolic blood pressure was reactive to stress. Data from each subject’s first two experimental sessions were dropped due to the large degree of variability in subjective and physiological responses often observed as subjects adapted to the novel surroundings and the alcohol presentation. In this regard Newlin and Pretorius (1987)
142
MARKD.LITTetal
loo90 -80--
0
Subject
tZJ
Baseline
Response Value
70-60 -50-40 -30 -20-lO--
0
SlS2S3S4S5S6S7S8 Water Negative
SlS2SSS4S5S6S7SB Alcohol Mood Condition
1OOT
90
t
Y
80 --
c ‘iT
70--
n 0
60 -50 --
!! .-
40--
g
30--
0
20 -10 0 SlS2S3SIS5S6S7S8 Water Neutral
SlS2S3S4S5S6S7SS Alcohol Mood Condition
Fig. 3. Ratings of Desire to Drink made along a 100 mm visual analog scale for each combination of mood induction and stimulus. Data from all subjects (Sl-S8) are presented separately.
have reported that cue exposure data collected in an unfamiliar environment may be contaminated by responses attributable to the novelty of the situation. Keeping only the data from the last two sessions allowed examination of responses by subjects well-acquainted with the procedure while still preserving the counterbalancing across subjects of the order of mood induction and stimulus presentation. Prior to analyzing for effects of stimulus or mood, subjects’ baseline ratings of mood, desire for alcohol, and difficulty resisting alcohol, as well as baseline systolic blood pressure readings, were compared using Wilcoxon signed-rank matched pairs tests. These
Alcohol cue reactivity
analyses revealed no significant differences in baseline scores between implying that conditions were comparable across these two sessions.
143
sessions
3 and 4,
Manipulation check Figure 2 shows each subject’s ratings of target mood, as taken from the WSRF, following each beverage presentation during each type of mood induction. Also shown are baseline target mood ratings. Difference scores (subject’s induced mood response minus baseline mood response) were calculated for each subject for each combination of mood induction and stimulus presentation. Because of the large amount of variability in distributions of these data, a Friedman non-parametric 2-way analysis of variance by ranks (Siegel, 1956) was used to detect systematic differences among the within-subjects cells. The omnibus Friedman test indicated significant differences among the conditions (x2, = 15.08, df = 3, p < .Ol). Wilcoxon matched pairs signed-ranks tests (Siegel, 1956) with adjusted p values were used to detect differences attributable to mood induction, stimulus, or an interactionof mood and stimulus. These tests indicated a significant main effect for mood induction (T = 1, p < .Ol) with higher negative mood ratings in the negative mood conditions, a significant effect for stimulus (7’ = 3.5, p < .025), with a greater negative mood response in the presence of water, and no significant interaction effect. The Spearman rank-order correlation between ratings of negative mood after mood induction and scores on the SHCS was rs = .69 @ < .05). These results indicate that the negative mood induction technique used was successful in producing the desired moods, and that this was in part a function of hypnotic susceptibility. The Spearman correlation between ratings of negative mood and scores on the Rand dependence scale was r, = - .04. Alcohol cue and mood efects Omnibus Friedman 2-way ANOVAs with post-hoc Wilcoxon tests were used in the same manner as above to analyze subjects’ responses (expressed as difference scores) along the parameters of desire for alcohol, difficulty resisting alcohol, and systolic blood pressure. Figure 3 shows subjects’ ratings of desire for alcohol for each combination of mood induction and stimulus, along with baseline values. The Friedman ANOVA, corrected for ties among rankings (Siegel, 1956), indicated significant differences in desire ratings among the conditions (x2, = 8.3 1, df= 3, p < .05). Wilcoxon post-tests showed a significant main effect for mood induction (T = 4, p < .025), with greater desire for alcohol in the negative mood conditions, no significant main effect for stimulus, and no significant interaction. Spearman rank-order correlations also indicated that ratings of desire for alcohol during negative mood induction were highly correlated with target mood rating (rs = .71, p < .03). Rand dependence scores did not correlate significantly with alcohol desire ratings in any condition, indicating that desire for alcohol during negative mood induction was not a function of severity of alcohol dependence. Subjects’ ratings of difficulty resisting alcohol were analyzed in the same manner as above. An omnibus Friedman ANOVA on difference scores (subjects’ ratings of difficulty resisting alcohol in each condition minus baseline rating) failed to detect significant differences among the conditions. Friedman ANOVAs performed on difference scores on the blood pressure measure also failed to find significant differences among the conditions. DISCUSSION
The idea that an extinction procedure can be used as one element of a treatment program to reduce the probability of relapse after treatment is extremely appealing. If an extinction procedure is to be generally useful, however, stimuli must be found that will reliably elicit desire for alcohol. Previous research using exteroceptive cues, such as alcoholic beverages,
144
MARK D. LI’lT et al
has shown that these cues do not always elicit desire for alcohol in alcoholic subjects. The purpose of the present study was to explore the possibility that an interoceptive cue, negative mood state, may play a role in eliciting desire in alcoholic subjects. The present results suggest, first, that the mood induction procedure used here appears to have been effective for a majority of subjects. In general, it appears that negative moods can indeed be induced in the laboratory, and that these moods can be maintained through the course of an alcohol cue exposure session. In the present study, however, the ability to achieve a negative mood was at least in part a function of hypnotic susceptibility, which may thus limit the general applicability of a hypnotic induction procedure in extinction-based treatment paradigms. The most interesting, and potentially important, finding of the present study was the indication that negative mood states can elicit desire for alcohol independent of the presence of exteroceptive cues such as the sight or smell of alcoholic beverages. This result is similar to that found by Childress et al. (1987) with opiate addicts. In contrast to findings in previous research (e.g., Kaplan et al., 1985), there was no effect on any parameter of desire when subjects were exposed to alcohol beverage cues in the absence of a negative mood state. Although responses to the alcohol stimulus during the neutral mood induction condition were not expected to be as great as those during the negative mood induction, some response was expected. One possible reason for this lack of cue reactivity in the neutral mood condition is that the relaxation used in the mood inductions inhibited subjective and blood pressure responses to alcohol cues. This hypothesis is supported by Loberg (1987), who found that relaxation induction led to decreased craving responses in alcoholics exposed to alcohol cues. In short, we suspect that the relaxation in the neutral mood conditions inhibited desire reactions. In the negative mood conditions reactions to alcohol were obscured by the mood effects. In other words, the effects of alcohol cues and negative mood state may not be additive. This, however, is not a very satisfying explanation for this apparent lack of alcohol cue reactivity, and these effects will need to be reexamined in further studies. The suggestion that relaxation may inhibit desire for alcohol deserves further study. It implies that future mood induction paradigms may need to be altered to minimize relaxation that might interfere with alcohol cue reactivity and thus slow the extinction process in a cue-exposure treatment. It further suggests, however, that systematic relaxation training may in itself be a useful adjunct to relapse prevention training, perhaps to be used in times of mood disturbance. It is notable that, although ratings of desire for alcohol varied significantly across conditions, ratings of difficulty resisting alcohol did not. Additionally, ratings of alcohol desire and difficulty resisting alcohol were not well correlated in session 3 (Spearman rank order correlations were rs = .28 in session 3 and .68 in session 4). One explanation for this may be that the subjects, all inpatients in an alcohol treatment program and trying to maintain abstinence, may have been willing to acknowledge desire for alcohol, but may not have been willing to admit a difficulty on their part in being able to resist alcohol. This explanation was supported by the fact that ratings of difficulty resisting alcohol were uniformly lower than ratings of desire. Systolic blood pressure responses were unaffected by mood induction, stimulus presentation, or the interaction of mood and stimulus. In addition, blood pressure readings were not well correlated with either subjective measure of distress. Thus, we can draw no conclusions as to subjects’ physiological reactivity to mood state or to alcoholic beverages. Future studies should employ additional physiological measures to shed light on the process of “craving” or desire for alcohol. Rankin, Hodgson, and Stockwell (1979), for example, found body
Alcohol cue reactivity
145
temperature to be a correlate of alcohol desire. Likewise, in their work with opiate addicts, Childress et al. (1987) found skin temperature to be reactive to opiate cue exposure. It would be useful to determine if mood states alone can have equivalent effects on skin temperature in alcoholics. It is unclear why target mood ratings should have been higher during presentation of seltzer water than during presentation of alcohol. During debriefing, however, it was found that two of the most reactive subjects reported feeling extremely distressed after the negative mood induction, but their distress decreased upon presentation of the alcoholic beverage. This raises the possibility that, at least for some subjects, consumption of alcohol may not be necessary to relieve negative mood states; the mere presence of alcohol may in itself be a conditioned stimulus for relief of distress, possibly by eliciting a conditioned appetitive response (Stewart, DeWit, & Eikelboom, 1984). This idea deserves further study. A possible limitation of the present study is that the results obtained may have been a function of demand characteristics. That is, subjects may have expected to be more reactive in the negative mood induction conditions and therefore may have reported greater desire for alcohol. If demand characteristics were playing a role, however, we might also expect higher ratings of alcohol desire in those conditions in which alcohol was presented as the beverage stimulus. The fact that there were no beverage effects in this study tends to diminish the argument that demand characteristics were responsible for the results. To conclude, the present data are in some ways quite consistent with what many alcoholic patients tell us about relapse; namely, that they frequently leave treatment programs feeling fine until something happens that leads them to feel distressed, and then they experience a desire for alcohol that leads them to seek a drink. Perhaps, at least for a subset of alcoholics, interoceptive cues like mood states may be important in leading to the first drink of a relapse. Thereafter, exteroceptive cues provided by the sight, smell and taste of alcohol, may play a role in maintaining further drinking behavior. This model is, in fact, consistent with results reported by Laberg and Ellertsen (1987), who found little alcohol cue reactivity in alcoholics unless a priming dose of alcohol was first consumed. Future research, along the lines explored here, and with a much larger number of subjects, is required to clarify the extent of variability in cue reactivity and the roles of interoceptive and exteroceptive cues in the process of relapse. This research may, in turn, provide direction for more effective relapse prevention strategies.
REFERENCES Annis, H.M. (1984). Invenro~ of drinking situations. Toronto: Addiction Research Foundation, Bower, G.H. (1981). Mood and memory. American Psychologist, 36, 129-148. Childress, A.R., McLellan, A.T., Natale, M., & O’Brien, C.P. (1987). Mood states can elicit conditioned withdrawal and craving in opiate abuse patients. Problems of drug dependence, 1986. NIDA Research Monograph, Vol. 76 (DHHS No. (ADM) 87-1508) (PP. 137-144). Hilgard, E.R.. & Hilgard, J.R. (1975). Hypnosis in the reief of pain. Los Altos. CA: William Kaufman. Hodgson, R., Rankin, H., & Stockwell, T. (1979). Alcohol dependence and the priming effect. Behaviour Research and Therapy, 17. 379-387. Kaplan, R.F., Cooney, N.L., Baker, L.H., Gillespie, R.A., Meyer, R.E., & Pomerleau, O.F. (1985). Reactivity to alcohol-related cues: Physiological and subjective responses in alcoholics and non-problem drinkers, Journal of Studies on Alcohol, 46, 267-272. Kaplan, R.F., Meyer, R.E., & Stroebel, C.F. (1983). Alcohol dependence and responsivity to an ethanol stimulus as predictors of alcohol consumption. British Journal of Addiction, 78, 259-267. Kennedy, D.A. (1971). Pupilometric analysis as an aid in the assessment of maturation, impact of treatment, and prognosis of chronic alcoholics. Unpublished dissertation, University of Utah, Laberg, J.C., & Ellertsen, B. (1987). Psychophysiological indicators of craving in alcoholics: Effects of cue exposure. British Journal of Addiction, 82, 1341-1348. Loberg, T. (1987). Alcohol cue reactivity in alcoholics. Paper presented at the Fourth International Conference on
146
MARK D. LITT et al.
the Treatment of Addictive Behaviors, Bergen, Norway. Marlatt, G.A., & Gordon, J.R. (1980). Determinants of relapse; implications for the maintenance of behavior change. In P. Davidson (Ed.), Behavioral medicine: Changing health lifestyles (Vol 10). New York: Brunner-Mazel. Mont& P., Abrams, D., Binkoff, J., Rohsenow, D., & Zwick, W. (1987). Cue reacrivity in alcoholics. Paper presented at the Fourth International Conference on the Treatment of Addictive Behaviors, Bergen, Norway. Newlin, D.B., & Pretorius, M.B. (1987). Greater alcohol effect in a familiar than a novel environment: A preliminary test of the response competition model. Alcoholism: Clinical and Experimenral Research, 11, 218 (abstract). Niaura, R., Wilson, G.T., & Westrick, E. (1988). Self-awareness, alcohol consumption, and reduced cardiovascular reactivity. Psychosomatic Medicine, 50, 360-380. Polich, J.M., Armor, D.J., & Braiker, H.B. (1981). The course ofakoholism: Four years after freafment. New York: Wiley. Poulos, C.X., Hinson, R.E., & Siegel, S., (1981). The role of Pavlovian processes in drug tolerance and dependence: Implications for treatment. Addicfive Behaviors, 6, 205-212. Rankin, H., Hodgson, R., & Stockwell, T. (1979). The concept of craving and its measurement. Eehaviour Research and Therapy, 17, 389-396. Siegel, S. (1956). Nonparametric srarisfics for the behavioral sciences. New York: McGraw Hill. Stewart, J., DeWit, H., & Eikelboom, R. (1984). Role of unconditioned and conditioned drug effects in the self-administration of opiates and stimulants. Psychological Review, 91, 251-268.
