- Email: [email protected]
Smoking as a cue for subjective and behavioral responses to a stressor
Journal oi Substance Abuse, 3, 29-38 (1991)
Smoking as a Cue for Subjective and Behavioral Responses to a Stressor Kenneth A. Perkins Leonard H. Epstein I. Richard ]ennings western Psychiatric Institute & Clinic University of Pittsburgh School of Medicine
Smoking in conjunction with a repeated stressor may come to serve as a cue, or discriminative stimulus, for responses to the stressor. This study examined the effects of omitting a previously predictable smoking cue on subjective and behavioral responses to a stressor. Male smokers were instructed to smoke briefly on command immediately prior to ("smoke-pre", n = 10) or several minutes after ("smoke-post", n = 10) each of four acquisition trials of a mental arithmetic task. During two subsequent trials, both groups first engaged in the task without the smoking cue (Test Trial 1), a novel experience for smoke-pre, and then engaged in the task preceded by the smoking cue (Test Trial 2), a novel experience for smoke-post. Omission of the smoking cue during Test Trial 1 produced significantly poorer task performance and greater subjective stress in smoke-pre. Although not significant, insertion of the smoking cue during Test Trial 2 tended to produce poorer performance in smoke-post. Thus, a consistent pattern of smoking in association with a repeated stressor may subsequently lead to impaired responding to the stressor if smoking is omitted.
Smoking is increased during stress (e.g., Rose, Ananda, & Jarvik, 1983), perhaps because of smoking's purported direct influences on reducing negative affect (Epstein, Dickson, McKenzie, & Russell, 1984; Pomerleau & Pomerleau, 1984) or on enhancing task performance (Wesnes & Warburton, 1983). In addition to the direct effects of smoking in ameliorating stress, a consistent pattern of smoking while engaged in repeated episodes of a stressor could lead to smoking becoming a cue, or discriminative stimulus, for behavioral and subjective responses to the stressor. Smokers may acutely increase smoking in anticipation of, as well as during, an expected stressor. For example, Rose et al. (1983) found increased smoking during preparation for an impromptu videotaped speech compared with quiet rest. This predictable association of smoking with a stressor could facilitate coping, possibly by increasing smokers' This research was supported in part by grant DA05807 from the National Institute on Drug Abuse. We would like to thank Bonita Marks and Thomas Dehski for their help in completing the study. Correspondence and requests for reprints should be sent to Kenneth A. Perkins, Western Psychiatric Institute and Clinic, 3811 O'Hara St., Pittsburgh, PA 15213.
29
30
K.A. Perkins, L.H. Epstein, and I.K. lennings
perceived control over the stressor when smoking compared to when not smoking (Epstein & Perkins 1988; Pomerleau & Rosecrans, 1989). On the other hand, removal of this previously predictable smoking cue while exposed to stress may disrupt the perception of control and impair coping. Such an effect could help explain why smoking cessation may lead to poorer coping with stress (Hughes, Higgins, & Hatsukami, 1990) and why stress may increase craving and relapse after cessation (Cummings, Jaen, & Giovino, 1985; Shiftman et al., 1986). Potential cue or discriminative stimulus effects of smoking on stress responding in humans has received little attention, although general effects of smoking and smoking withdrawal on physiological responses to noxious stimuli have been noted. Smoking acutely increases speed of EEG adaptation to aversive stimuli (Friedman, Horvath, & Meares, 1974), whereas smoking abstinence slows electrodermal and cardiovascular adaptation to such stimuli (Knott, 1984; Puddey, Vandongen, Beilin, & English, 1984). Likewise, animal research has demonstrated that avoidance task performance is impaired when saline is substituted for nicotine following repeated exposure to the task in the presence of nicotine (Hall & Morrison, 1973; Morrison, 1074). Termination of nicotine treatment disrupts responding on other tasks previously performed in the presence of nicotine (Carroll, Lac, Asencio, & Keenan, 1089; Corrigall, Herling, & Coen, 1989). However, although physiological responses to smoking stimuli have been examined (e.g., Niaura et al., 1988) and conditioning of other responses to smoke intake has been explored (Payne, Etscheidt, & Corrigan, 1990), almost no research has focused on smoking as a cue for responses to other environmental stimuli. In a recent laboratory study, we found that removal of an expected smoking cue prior to a repeated nonstressful cognitive task increased electrodermal responding to the task (Epstein, Perkins, Jennings, & Pastor, 1990). This finding may b e comparable to an animal study which showed that omission of a previously reliable pharmacological cue (alcohol) preceding intake of morphine heightened physiological responses (i.e., disrupted tolerance) to morphine (Poulos, Hunt, & Cappell, 1988). Similarly, in another study, omission of a previously consistent benzodiazepine cue preceding shock caused a reemergence of previously extinguished fear responding to the shock, whereas this omission had no effect on extinction in animals whose past exposure to benzodiazepine never served as a cue for shock (Bouton, Kenney, & Rosengard, 1990). To our knowledge, there has been no specific examination of the effects of omitting smoking during a repeated stressor previously experienced while smoking. This study examined the effects of acute omission of a smoking cue on behavioral and subjective responses to a repeated stress task previously signalled by the cue. A control group was presented equal, but temporally separate, exposure to the task and to smoking so that smoking could not act as a cue for the task. It was expected that the novel omission of smoking prior to the task would disrupt responses in the experimental group. In an additional, exploratory exposure to the task, we examined whether novel insertion of the smoking cue prior to the task would alter responses of the
Smoking as a Cue for Stress Responding
31
control group, which previously experienced the task without the preceding cue. METHODS Subjects Subjects were 20 healthy men aged 18-30 years (M + S.E.M. = 23.6 + 1.1) who had smoked 15-40 cigarettes per day (M = 22.9 _+ 2.0) for at least 6 months (M = 5.1 _+ 1.3 years). Subjects were paid $8 plus an average of $5 as a performance-contingent incentive (see Stress task). Informed consent was obtained from subjects prior to their participation. Subjective Stress Measure Subjective stress was assessed with the Stress scale of the Stress-Arousal Checklist (SACL), a self-report measure employed in a variety of studies to assess acute fluctuations in subjective stress and arousal (Mackay, 1980; Mackay, Cox, Burrows, & Lazzarini, 1978). T h e scale contains 19 descriptive adjectives directly or inversely related to subjective stress (e.g., "uptight", "relaxed") and four response choices (definitely feel, feel slightly, cannot decide, definitely do not feel). Scoring involves counting up the number of items endorsed in the stress direction (e.g., definitely feel or feel slightly for uptight; cannot decide or definitely do not feel for relaxed). Potential range o f scores is 0-19. Stress Task T h e stressor involved responding orally with the answers to mental arithmetic (MA) problems presented via cassette tape recorder. Each problem consisted o f subtracting 1 two-digit number from a larger two-digit number. Subjects had 6 seconds to respond after presentation of each problem before the next problem was presented. Subjects' responses were recorded by an experimenter present in the room. T h e r e were 12 problems during each of six task trials, each trial lasting approximately 90 s. Trials were presented every 10 min (see Procedure). N u m b e r of problems incorrect was the measure o f performance. In addition, a monetary incentive was provided such that subjects could receive up to $8 beyond their participation payment depending on their task performance. T h e y were told that $.10 would be removed from this $8 bonus for every incorrect answer they gave during task trials. A very similar task was previously shown to elicit significant stress responses (Perkins, Dubbert, Martin, Faulstich, & Harris, 1986). Procedure Subjects were instructed to abstain from smoking after midnight prior to the 90-rain morning session. All sessions were conducted between 9:00 a.m.
32
K.A. Perkins, LH. Epstein, and J.R. Jennings
and 12:00 noon. Subjects were also told to bring their current brand of cigarettes (M nicotine yield = 0.93 + .10 mg). T h r o u g h o u t the session, subjects reclined in comfortable armchairs within a sound-proof experiment room. Following 10 min of quiet rest (baseline), subjects completed the first stress rating. Then, subjects were provided with instructions on the MA stress task and incentive conditions and presented with one of the two smoking manipulations according to their randomized assignment to two groups (n = 10 each). T h e temporal relationship between smoking and the MA stress task for each group is shown in Figure 1. One group ("smoke-pre") was instructed to inhale briefly (2 s) on command from their cigarette four times, once every 15 s, during the 60 s prior to each of the first four trials o f the task ("acquisition trials"). T h e other group ("smokepost") was instructed to inhale in exactly the same manner but not until 2 min after each of the four acquisition trials. This pattern of smoking either immediately before, or a few minutes after the task, was held constant across these first four trials so that smoking could come to serve as a reliable cue for the task in the smoke-pre group but not for the smoke-post group. Immediately following each task trial, subjects completed the SACL and were given feedback on the number of problems incorrectly answered and amount o f money taken away. Subjects in the smoke-post group then smoked on command, as previously described. For both groups, the remainder of each 10-min period was spent resting quietly prior to the next task trial. After these four acquisition trials, both groups were exposed to the same manipulation during Trial 5 (Test Trial 1), in which subjects engaged in the MA task without the preceding smoking cue (see Figure 1), and then smoked on command 2 min later. For the smoke-post group, this procedure was the same as that during Trials 1-4. For the smoke-pre group, however, this involved the novel omission of the smoking cue prior to the task. It was hypothesized that smoke-pre would show impaired performance and heightened stress during this trial compared with smoke-post. For exploratory purposes, a subsequent Trial 6 (Test Trial 2) occurred in which both groups smoked prior to the task, a reinstatement of the smoking cue for smoke-pre, but a novel insertion o f smoking preceding the task for smoke-post. It was expected that smoke-post might show impaired performance and heightened stress during this trial compared with smoke-pre, suggesting that any unexpected change preceding the task, and not just novel omission of the smoking cue, could disrupt responses. Statistical A n a l y s e s
T testg were used to detect any group differences in smoking characteristics, age, and resting baseline stress rating in order to confirm success of randomization. Task performance (number of problems incorrect out of 12) and subjective stress during the acquisition trials (Trials 1-4) were analyzed by analysis o f variance (ANOVA) to determine any differences in responding due to the initial smoking manipulation (i.e., smoke-pre versus smoke-post).
Smoking as a Cue for Stress Responding
33 o
I-O0 LLI El::
I-t,,LI
w Z
[vi~i:l~l I:::Wik',&;I P"4k'qP".4k"~'~
0:I:I_-{ : :IIVI:I '
'<
w I..-
ik~v,4k,'v, dk,Tir41
o-.
o
Or)
I--
I-. Or) W n"
w
._I
-u-)
W
o
0--
O334 HlVH
oo
m
0334
Z
m m
3•
Or)
w I-I-1-O9 i,I I.-
m o,ZNe-Ne--
5
-o
o
~.~o~ A
O9(.0
033--,
o
o
o 111 nn
I
O9 0 i'i
I
LUll. ',.~ ::~
L.UO_ ~.~
O0
O0
o0(9
o0~
34
K.A. Perkins, L.H. Epstein, and I.R. Jennings
Subjective and performance responses during the subsequent Test Trials 1 and 2 were each analyzed by two-variable, mixed repeated-measures analysis of covariance (ANCOVA), with groups and trials (2) as the effects and mean response during the Acquisition Trials 1-4 as the covariate. A Groups X Trials interaction from the ANCOVA was followed up by comparisons between groups at each trial using Fisher's least significant difference t-test procedure (Huitema, 1980)..Values are presented as M + S.E.M. RESULTS
T h e r e were no significant differences between groups in age, smoking rate, smoking history, and nicotine content o f cigarette brand (all ts < 1). Similarly, there was no significant difference in resting baseline stress rating (1.8 + 0.5 versus 4.0 + 1.2 for smoke-pre and smoke-post, respectively; t ( 1 9 ) = 1.66, p > .10). A N O V A results on responding during the four acquisition trials also revealed no significant group or trials effects for stress ratings, 5.1 + 1.1 versus 6.1 _+ 1.1 for smoke-pre and smoke-post, respectively; F(I, 18) < 1, or MA task performance, 5.6 + 1.1 versus 4.8 + 0.9, respectively, F(1, 18) < 1. Thus, differential exposure to smoking as a cue for the task had no independent effect on responses prior to the test trials. Test Trial Task Performance
ANCOVA o f MA task performance during Test Trials 1 and 2 revealed a significant Group X Trials interaction, F(1, 17) = 7.29, p < .02, but there was no main effect o f group or trials (Fs < I). As shown in Figure 2, task performance was significantly poorer in the smoke-pre group compared with the smoke-post group during Test Trial I, t(19) = 2.09, p < .05, in which both groups were exposed to the task without the smoking cue (novel omission o f smoking cue for smoke-pre). During Test Trial 2, in which both groups were exposed to the smoking cue prior to the task (novel insertion of smoking cue for smoke-post, return of smoking cue for smoke-pre), the difference between groups did not reach significance, t(19) = 1.69, p > .10. However, between Test Trials 1 and 2, task performance significantly worsened in the smoke-post group: increase in errors o f 1.4 _+ .8 from Test Trial 1, t(19) = 2.30, p < .05. Notably, performance was poorer for smoke-we during Test Trial 1 and for smoke-post during Test Trial 2 than their performance on any other trial during the session. Test Trial Subjective Stress
Similar results were observed for the ANCOVA of subjective stress, as the Group X Trials interaction was nearly significant F(1, 17) = 4.21, p = .055, but there were no main effects of group, F(1, 17) = 1.61, p z> .I0, or trials, F < 1. As also shown in Figure 2, follow-up comparisons showed that stress was significantly higher during Test Trial I for the smoke-pre group compared
Smoking as a Cue for Stress Responding
35 S.~IK-PRE
F~
S.klK-POST
7L 8 5
_ TEST TI
TEST T2
TEST TI
TEST T2
_
J
~0 9 8 7 6
4 3 2 I
Figure 2. Covariate-adjusted mean (+- S.E.M.) mental arithmetic task performance (number of errors out of 12) and subjective stress for smoke-pre and smoke-post groups during'Test Trial 1 (omission of smoking cue) and Test Trail 2 (insertion of smoking cue). ** p < .01; * p < .05. with smoke-post, t(19) = 4.16, p < .01. There was no significant difference in stress during Test Trial 2, t(19) = 1.27, n.s. DISCUSSION These results indicated that removal of a previously predictable smoking cue preceding a repeated stressor may affect responses to that stressor. Specifically, exposure to a stress task in the absence of a smoking cue led to poorer performance and greater subjective stress in subjects who had previously performed the task only following the smoking cue (smoke-pre), compared with those who had equal recent exposure to the task and to smoking but without smoking acting as a cue for the task (smoke-post).
36
K.A. Perkins, L.H. Epstein, and J.R. lennings
Importantly, exposure to smoking before (i.e., as a cue) versus after the task did not influence task performance or subjective stress during the four acquisition trials. It was only when this predictable smoking cue was unexpectedly omitted that performance and stress were affected. Previous research has shown that physiological and subjective responses to a repeated noxious stimulus may be heightened if that stimulus is changed across presentations compared to when it is held constant (Epstein, 1973). Sirlailarly, adaptation to a repeated stimulus is slower when characteristics of the stimulus are added, altered, or omitted over time (Poulos et al., 1988; Siddle & Hirschhorn, 1986). T h e results of this study may be related to those findings. Although the smoking cue occurred immediately prior to (rather than concurrent with) the stress task, the temporal contiguity of smoking with stress may have served to create an association between smoking and the task. Thus, for the smoke-pre group, the novel omission of smoking during Test Trial I may have been equivalent to altering the characteristics of the task by removing one o f its implicit components (i.e., smoking). It is also conceivable that novel insertion o f smoking prior to a task may alter the characteristics o f the task by adding a new component. When such a change was instituted in this study during Test Trial 2 for the smoke-post group, performance significantly worsened compared with Test Trial 1, although there was no significant change in subjective stress. However, lack o f an additional control group (i.e., a group never receiving the preceding smoking cue on this or any previous trial) limits the usefulness of this particular comparison, perhaps leaving unanswered the question of whether novel insertion o f smoking influences task performance and stress. Although speculative, it is possible that cessation-induced disruption of a long-term, predictable pattern of smoking in conjunction with naturally occurring stressors could similarly impair coping with the stressors and help to explain stress-related craving and relapse (Shiffman et al., 1986). This notion is somewhat different from, but not incompatible with, other conditioning explanations o f drug craving. For example, Siegel, Krank, and Hinson (1988) suggested that reliable pairing of an environmental context (conditioned stimulus) with drug intake subsequently led to the elicitation of drug-opposite compensatory responses in the presence of that context, which then produced drug craving. However, our results suggest that drug intake (i.e., smoking) itself can come to serve as a reliable discriminative stimulus for responses to a task, which, if subsequently attempted in the absence of drug intake, may lead to poorer performance and increased distress, thus encouraging a resumption o f drug use in conjunction with the task. Generalization of these results to different behavioral and subjective responses to other stressors is needed. Although significant results were observed with the behavioral and subjective measures used here, the absolute differences between groups were rather small. Greater temporal separation between exposures to the stressor, use of more naturalistic stressors, and an increase in the number of pairings of smoking with the stressor may reveal greater effects of the removal o f smoking on task performance and subjective measures. Furthermore, similarity of these results with those of our previous study
Smoking as a Cue for Stress Responding
37
involving responses to a nonstress task (Epstein et al., 1990) indicate that the disruptive effects of novel omission of smoking may generalize across many types of tasks, perhaps more closely resembling state-dependent learning (Peeke & Peeke, 1984). Finally, intake of other substances in conjunction with repeated stressors (e.g., alcohol) may also have similar effects when unexpectedly omitted, as suggested by animal studies noted earlier. It is not clear if the critical element of the smoking cue was intake of nicotine or the behavior of puffing on a cigarette. Subjects took only four brief puffs prior to each trial and consumed only two cigarettes over the course of the entire session. Thus, amount o f nicotine intake per trial was probably small, suggesting that smoking behavior may have been the most salient cue component. In addition, our previous study (Epstein et al., 1990) found that, whereas removal of a smoking cue (and nicotine) affected electrodermal response to a repeated nonstress task previously preceded by smoking, presentation of a smoking cue without nicotine (i.e., smoking a nonnicotine cigarette) did not affect responding. On the other hand, other research has shown puff-by-puff changes in cognitive task performance (Revell, 1988), presumably due to transient increases in brain levels of nicotine, and there is a great deal of evidence that nicotine can serve as a potent discriminative stimulus for responding in animals (Morrison & Stephenson, 1969; Rosecrans, 1989). Future research employing nonnicotine cigarettes and/or an alternative method of nicotine delivery (e.g., Perkins, Epstein et al., 1986) may shed light on the importance of the behavioral versus pharmacological aspects of smoking in its function as a cue for stress responding. REFERENCES Bouton, M.E., Kenney, F.A., & Rosengard, C. (1990). State-dependent fear extinction with two benzodiazepine tranquilizers. Behavioral Neuroscience, 104, 44-55. Carroll, M.E., Lac, S.T., Asencio, M., & Keenan, R.M. (1989). Nicotine dependence in rats. Life Sciences, 45, 1381-1388. Corrigall, W.A.,'tterling, S., & Coen, K.M. (1989). Evidence for a behavioral deficit during withdrawal from chronic nicotine treatment. Pharmacology, Biochemistry, & Behavior, 33, 559-562. Cummings, K.M., Jaen, C.R., & Giovino, G. (1985). Circumstances surrounding relapse in a group of recent exsmokers. Preventive Medicine, 14, 195-202. Epstein, L.H., Dickson, B.E., McKenzie, S., & Russell, P.O. (1984). The effect of smoking on perception of muscle tension. Psychopharmacology, 83, 107-I 13. Epstein, L.H., & Perkins, K.A. (1988). Smoking, stress, and coronary heart disease. Journal of Consulting and Clinical Ps~'cholo~', 56, 342-349. Epstein, L.H., Perkins, K.A., Jennings, J.R., & Pastor, S. (1990). The effects of smoking context on habituation to a repeated cognitive task. Ps)'chopharmacolog'y, 100, 366-371. Epstein, S. (1973). Expectancy and magnitude of reaction to a noxious UCS. Psychophysiolog-y, 10, 100-107. Friedman, J., Horvath, T., & Meares, R. (1974). Tobacco smoking and a "'stimulus barrier." Nature, 248, 455-456. Hall, G.H., & Morrison, C.F. (1973). New evidence for a relationship between tobacco smoking, nicotine dependence, and stress. Nature, 243, 199-201. Hughes, J.R., Higgins, S.T., & Hatsukami, D. (1990). Effects of abstinence from tobacco. In L.T. Kozlowski, H.M. Annis, H.D. Cappell, F.B. Glaser, M.S. Goodstadt, Y. Israel, H.
38
K.A. Perkins, L.H. Epstein, and J.R. lennings
Kalant, E.M. Sellers, & E.R. Vingilis (Eds.), Research advances in alcohol and drug problems, (Vol. 10, pp. 317-398). New York: Plenum. Huitema, B. (1980). Analysis of covariance and alternath,es. New York: Wiley-lnterscience. Knott, V.J. (1984). Electrodermal activity during aversive stimulation: Sex differences in smokers and nonsmokers. Addictive Behaviors, 9, 195-199. Mackay, C.J. (1980). The measurement of mood and psychophysiological activity using self-report techniques. In I. Martin, & P.H. Venables (Eds.), Techniquesin psychoph~'siology(pp. 501-562). New York: Wiley. Mackay, C.J., Cox, T., Burrows, G.C., & Lazzarini, A.J. (1978). An inventory for the measurement of self reported stress and arousal. British Journal of Social and Clinical Psychology, 17, 283-284. Morrison, C.F. (1974). Effects of nicotine and its withdrawal on the performance of rats on signalled and unsignalled avoidance schedules. Ps.ychopharnmcologia, 38, 25-35. Morrison, C.F., & Stephenson, J.A. (1969). Nicotine injections as the conditioned stimulus in discrimination learning. Ps~'chopharmacologia, 15, 351-360. Niaura, R.S., Rohsenow, D.J., Binkoff, J.A., Monti, P.M., Pedraza, M., & Abrams, D.B. (1988). Relevance of cue reactivity to understanding alcohol and smoking relapse. Journal of Abnormal Ps)'chology, 97, 133-152. Payne, T.J., Etscheidt, M., & Corrigan, S.A. (1990). Conditioning arbitrary stimuli to cigarette smoke intake: A preliminary study. Journal of Substance Abuse, 2, 113-119. Peeke, S.C., & Peeke, H.V.S. (1984). Attention, memory, and cigarette smoking. Psychopharmacolog3", 84, 205-214. Perkins, K.A., Dubbert, P.M., Martin, J.E., Faulstich, M., & Harris, J.K. (1986). Cardiovascular reactivity to psychological stress in aerobically trained versus untrained mild hypertensives and normotensives. Health Ps~'chology, 5, 407-421. Perkins, K.A., Epstein, L.H., Stiller, R., Jennings, J.R., Christiansen, C., & McCarthy, T. (1986). An aerosol spray alternative to cigarette smoking in the study of the behavioral and physiological effects of nicotine. Behavior Research Methods, Instruments, and Computers, 18, 420-426. Pomerleau, O.F., & Pomerleau, C.S. (1984). Neuroregulators and the reinforcement of smoking: Towards a biobehavioral explanation. Neuroscience and Biobehavioral Rea,iews, 8, 503-513. Pomerleau, O.F., & Rosecrans, J. (1989). Neuroregulatory effects of nicotine. Psychoneuroepdocrinology, 14, 407-423. Poulos, C.S., Hunt, T., & Cappell, H. (1988). Tolerance to morphine analgesia is reduced by the novel addition or omission of an alcohol cue. Psychopharmacology, 94, 412-416. Puddey, I.B., Vandongen, R., Beilin, L.J., & English, D. (1984). Hemodynamic and neuroendocrine consequences of stopping smoking--a controlled study. Clipical and Experime~,tal Pharmacology & Ph)~iolo~', 11, 423-426. Re~'ell, A.D. (1988). Smoking and performance--a puff-by-puff analysis. Psychopharmacology, 96, 563-565. Rose, J.E., Ananda, S., & jarvik, M.E. (1983). Cigarette smoking during anxiety-provoking and monotonous tasks. Addicth,e Behaviors, 8, 353-359. Rosecrans, J.A. (1989). Nicotine as a discriminative stimulus: A neurobiobehavioral approach to studying central cholinergic mechanisms. Journal of Substance Abuse, 1, 287-300. Shiffman, S.M., Shumaker, S.A., Abrams, D., Cohen, S., Garvey, J., Grunberg, N., & Swan, G. (1986). Models of smoking relapse. Health Psschology, 5 (Suppl. 1), 13-27. Siddle, D.A.T., & Hirschhorn, T. (1986). Effects of stimulus omission and stimulus novelty on dishabituation of the skin conductance response. Psychophysiology, 23, 309-314. Siegel, S., Krank, M.D., & Hinson, R.E. (1988). Anticipation of pharmacological and nonpharmacological events: Classical conditioning and addictive behavior. In S. Peele (Ed.), Visions of addiction: Major contemporaryperspectiveson addiction and alcoholism (pp. 85-116). Lexington MA: D.C. Heath. Wesnes, K., & Warburton, D.M. (1983). Smoking, nicotine, and human performance. Pharmacology & Therapeutics, 21, 189-208.
Smoking as a Cue for Subjective and Behavioral Responses to a Stressor Kenneth A. Perkins Leonard H. Epstein I. Richard ]ennings western Psychiatric Institute & Clinic University of Pittsburgh School of Medicine
Smoking in conjunction with a repeated stressor may come to serve as a cue, or discriminative stimulus, for responses to the stressor. This study examined the effects of omitting a previously predictable smoking cue on subjective and behavioral responses to a stressor. Male smokers were instructed to smoke briefly on command immediately prior to ("smoke-pre", n = 10) or several minutes after ("smoke-post", n = 10) each of four acquisition trials of a mental arithmetic task. During two subsequent trials, both groups first engaged in the task without the smoking cue (Test Trial 1), a novel experience for smoke-pre, and then engaged in the task preceded by the smoking cue (Test Trial 2), a novel experience for smoke-post. Omission of the smoking cue during Test Trial 1 produced significantly poorer task performance and greater subjective stress in smoke-pre. Although not significant, insertion of the smoking cue during Test Trial 2 tended to produce poorer performance in smoke-post. Thus, a consistent pattern of smoking in association with a repeated stressor may subsequently lead to impaired responding to the stressor if smoking is omitted.
Smoking is increased during stress (e.g., Rose, Ananda, & Jarvik, 1983), perhaps because of smoking's purported direct influences on reducing negative affect (Epstein, Dickson, McKenzie, & Russell, 1984; Pomerleau & Pomerleau, 1984) or on enhancing task performance (Wesnes & Warburton, 1983). In addition to the direct effects of smoking in ameliorating stress, a consistent pattern of smoking while engaged in repeated episodes of a stressor could lead to smoking becoming a cue, or discriminative stimulus, for behavioral and subjective responses to the stressor. Smokers may acutely increase smoking in anticipation of, as well as during, an expected stressor. For example, Rose et al. (1983) found increased smoking during preparation for an impromptu videotaped speech compared with quiet rest. This predictable association of smoking with a stressor could facilitate coping, possibly by increasing smokers' This research was supported in part by grant DA05807 from the National Institute on Drug Abuse. We would like to thank Bonita Marks and Thomas Dehski for their help in completing the study. Correspondence and requests for reprints should be sent to Kenneth A. Perkins, Western Psychiatric Institute and Clinic, 3811 O'Hara St., Pittsburgh, PA 15213.
29
30
K.A. Perkins, L.H. Epstein, and I.K. lennings
perceived control over the stressor when smoking compared to when not smoking (Epstein & Perkins 1988; Pomerleau & Rosecrans, 1989). On the other hand, removal of this previously predictable smoking cue while exposed to stress may disrupt the perception of control and impair coping. Such an effect could help explain why smoking cessation may lead to poorer coping with stress (Hughes, Higgins, & Hatsukami, 1990) and why stress may increase craving and relapse after cessation (Cummings, Jaen, & Giovino, 1985; Shiftman et al., 1986). Potential cue or discriminative stimulus effects of smoking on stress responding in humans has received little attention, although general effects of smoking and smoking withdrawal on physiological responses to noxious stimuli have been noted. Smoking acutely increases speed of EEG adaptation to aversive stimuli (Friedman, Horvath, & Meares, 1974), whereas smoking abstinence slows electrodermal and cardiovascular adaptation to such stimuli (Knott, 1984; Puddey, Vandongen, Beilin, & English, 1984). Likewise, animal research has demonstrated that avoidance task performance is impaired when saline is substituted for nicotine following repeated exposure to the task in the presence of nicotine (Hall & Morrison, 1973; Morrison, 1074). Termination of nicotine treatment disrupts responding on other tasks previously performed in the presence of nicotine (Carroll, Lac, Asencio, & Keenan, 1089; Corrigall, Herling, & Coen, 1989). However, although physiological responses to smoking stimuli have been examined (e.g., Niaura et al., 1988) and conditioning of other responses to smoke intake has been explored (Payne, Etscheidt, & Corrigan, 1990), almost no research has focused on smoking as a cue for responses to other environmental stimuli. In a recent laboratory study, we found that removal of an expected smoking cue prior to a repeated nonstressful cognitive task increased electrodermal responding to the task (Epstein, Perkins, Jennings, & Pastor, 1990). This finding may b e comparable to an animal study which showed that omission of a previously reliable pharmacological cue (alcohol) preceding intake of morphine heightened physiological responses (i.e., disrupted tolerance) to morphine (Poulos, Hunt, & Cappell, 1988). Similarly, in another study, omission of a previously consistent benzodiazepine cue preceding shock caused a reemergence of previously extinguished fear responding to the shock, whereas this omission had no effect on extinction in animals whose past exposure to benzodiazepine never served as a cue for shock (Bouton, Kenney, & Rosengard, 1990). To our knowledge, there has been no specific examination of the effects of omitting smoking during a repeated stressor previously experienced while smoking. This study examined the effects of acute omission of a smoking cue on behavioral and subjective responses to a repeated stress task previously signalled by the cue. A control group was presented equal, but temporally separate, exposure to the task and to smoking so that smoking could not act as a cue for the task. It was expected that the novel omission of smoking prior to the task would disrupt responses in the experimental group. In an additional, exploratory exposure to the task, we examined whether novel insertion of the smoking cue prior to the task would alter responses of the
Smoking as a Cue for Stress Responding
31
control group, which previously experienced the task without the preceding cue. METHODS Subjects Subjects were 20 healthy men aged 18-30 years (M + S.E.M. = 23.6 + 1.1) who had smoked 15-40 cigarettes per day (M = 22.9 _+ 2.0) for at least 6 months (M = 5.1 _+ 1.3 years). Subjects were paid $8 plus an average of $5 as a performance-contingent incentive (see Stress task). Informed consent was obtained from subjects prior to their participation. Subjective Stress Measure Subjective stress was assessed with the Stress scale of the Stress-Arousal Checklist (SACL), a self-report measure employed in a variety of studies to assess acute fluctuations in subjective stress and arousal (Mackay, 1980; Mackay, Cox, Burrows, & Lazzarini, 1978). T h e scale contains 19 descriptive adjectives directly or inversely related to subjective stress (e.g., "uptight", "relaxed") and four response choices (definitely feel, feel slightly, cannot decide, definitely do not feel). Scoring involves counting up the number of items endorsed in the stress direction (e.g., definitely feel or feel slightly for uptight; cannot decide or definitely do not feel for relaxed). Potential range o f scores is 0-19. Stress Task T h e stressor involved responding orally with the answers to mental arithmetic (MA) problems presented via cassette tape recorder. Each problem consisted o f subtracting 1 two-digit number from a larger two-digit number. Subjects had 6 seconds to respond after presentation of each problem before the next problem was presented. Subjects' responses were recorded by an experimenter present in the room. T h e r e were 12 problems during each of six task trials, each trial lasting approximately 90 s. Trials were presented every 10 min (see Procedure). N u m b e r of problems incorrect was the measure o f performance. In addition, a monetary incentive was provided such that subjects could receive up to $8 beyond their participation payment depending on their task performance. T h e y were told that $.10 would be removed from this $8 bonus for every incorrect answer they gave during task trials. A very similar task was previously shown to elicit significant stress responses (Perkins, Dubbert, Martin, Faulstich, & Harris, 1986). Procedure Subjects were instructed to abstain from smoking after midnight prior to the 90-rain morning session. All sessions were conducted between 9:00 a.m.
32
K.A. Perkins, LH. Epstein, and J.R. Jennings
and 12:00 noon. Subjects were also told to bring their current brand of cigarettes (M nicotine yield = 0.93 + .10 mg). T h r o u g h o u t the session, subjects reclined in comfortable armchairs within a sound-proof experiment room. Following 10 min of quiet rest (baseline), subjects completed the first stress rating. Then, subjects were provided with instructions on the MA stress task and incentive conditions and presented with one of the two smoking manipulations according to their randomized assignment to two groups (n = 10 each). T h e temporal relationship between smoking and the MA stress task for each group is shown in Figure 1. One group ("smoke-pre") was instructed to inhale briefly (2 s) on command from their cigarette four times, once every 15 s, during the 60 s prior to each of the first four trials o f the task ("acquisition trials"). T h e other group ("smokepost") was instructed to inhale in exactly the same manner but not until 2 min after each of the four acquisition trials. This pattern of smoking either immediately before, or a few minutes after the task, was held constant across these first four trials so that smoking could come to serve as a reliable cue for the task in the smoke-pre group but not for the smoke-post group. Immediately following each task trial, subjects completed the SACL and were given feedback on the number of problems incorrectly answered and amount o f money taken away. Subjects in the smoke-post group then smoked on command, as previously described. For both groups, the remainder of each 10-min period was spent resting quietly prior to the next task trial. After these four acquisition trials, both groups were exposed to the same manipulation during Trial 5 (Test Trial 1), in which subjects engaged in the MA task without the preceding smoking cue (see Figure 1), and then smoked on command 2 min later. For the smoke-post group, this procedure was the same as that during Trials 1-4. For the smoke-pre group, however, this involved the novel omission of the smoking cue prior to the task. It was hypothesized that smoke-pre would show impaired performance and heightened stress during this trial compared with smoke-post. For exploratory purposes, a subsequent Trial 6 (Test Trial 2) occurred in which both groups smoked prior to the task, a reinstatement of the smoking cue for smoke-pre, but a novel insertion o f smoking preceding the task for smoke-post. It was expected that smoke-post might show impaired performance and heightened stress during this trial compared with smoke-pre, suggesting that any unexpected change preceding the task, and not just novel omission of the smoking cue, could disrupt responses. Statistical A n a l y s e s
T testg were used to detect any group differences in smoking characteristics, age, and resting baseline stress rating in order to confirm success of randomization. Task performance (number of problems incorrect out of 12) and subjective stress during the acquisition trials (Trials 1-4) were analyzed by analysis o f variance (ANOVA) to determine any differences in responding due to the initial smoking manipulation (i.e., smoke-pre versus smoke-post).
Smoking as a Cue for Stress Responding
33 o
I-O0 LLI El::
I-t,,LI
w Z
[vi~i:l~l I:::Wik',&;I P"4k'qP".4k"~'~
0:I:I_-{ : :IIVI:I '
'<
w I..-
ik~v,4k,'v, dk,Tir41
o-.
o
Or)
I--
I-. Or) W n"
w
._I
-u-)
W
o
0--
O334 HlVH
oo
m
0334
Z
m m
3•
Or)
w I-I-1-O9 i,I I.-
m o,ZNe-Ne--
5
-o
o
~.~o~ A
O9(.0
033--,
o
o
o 111 nn
I
O9 0 i'i
I
LUll. ',.~ ::~
L.UO_ ~.~
O0
O0
o0(9
o0~
34
K.A. Perkins, L.H. Epstein, and I.R. Jennings
Subjective and performance responses during the subsequent Test Trials 1 and 2 were each analyzed by two-variable, mixed repeated-measures analysis of covariance (ANCOVA), with groups and trials (2) as the effects and mean response during the Acquisition Trials 1-4 as the covariate. A Groups X Trials interaction from the ANCOVA was followed up by comparisons between groups at each trial using Fisher's least significant difference t-test procedure (Huitema, 1980)..Values are presented as M + S.E.M. RESULTS
T h e r e were no significant differences between groups in age, smoking rate, smoking history, and nicotine content o f cigarette brand (all ts < 1). Similarly, there was no significant difference in resting baseline stress rating (1.8 + 0.5 versus 4.0 + 1.2 for smoke-pre and smoke-post, respectively; t ( 1 9 ) = 1.66, p > .10). A N O V A results on responding during the four acquisition trials also revealed no significant group or trials effects for stress ratings, 5.1 + 1.1 versus 6.1 _+ 1.1 for smoke-pre and smoke-post, respectively; F(I, 18) < 1, or MA task performance, 5.6 + 1.1 versus 4.8 + 0.9, respectively, F(1, 18) < 1. Thus, differential exposure to smoking as a cue for the task had no independent effect on responses prior to the test trials. Test Trial Task Performance
ANCOVA o f MA task performance during Test Trials 1 and 2 revealed a significant Group X Trials interaction, F(1, 17) = 7.29, p < .02, but there was no main effect o f group or trials (Fs < I). As shown in Figure 2, task performance was significantly poorer in the smoke-pre group compared with the smoke-post group during Test Trial I, t(19) = 2.09, p < .05, in which both groups were exposed to the task without the smoking cue (novel omission o f smoking cue for smoke-pre). During Test Trial 2, in which both groups were exposed to the smoking cue prior to the task (novel insertion of smoking cue for smoke-post, return of smoking cue for smoke-pre), the difference between groups did not reach significance, t(19) = 1.69, p > .10. However, between Test Trials 1 and 2, task performance significantly worsened in the smoke-post group: increase in errors o f 1.4 _+ .8 from Test Trial 1, t(19) = 2.30, p < .05. Notably, performance was poorer for smoke-we during Test Trial 1 and for smoke-post during Test Trial 2 than their performance on any other trial during the session. Test Trial Subjective Stress
Similar results were observed for the ANCOVA of subjective stress, as the Group X Trials interaction was nearly significant F(1, 17) = 4.21, p = .055, but there were no main effects of group, F(1, 17) = 1.61, p z> .I0, or trials, F < 1. As also shown in Figure 2, follow-up comparisons showed that stress was significantly higher during Test Trial I for the smoke-pre group compared
Smoking as a Cue for Stress Responding
35 S.~IK-PRE
F~
S.klK-POST
7L 8 5
_ TEST TI
TEST T2
TEST TI
TEST T2
_
J
~0 9 8 7 6
4 3 2 I
Figure 2. Covariate-adjusted mean (+- S.E.M.) mental arithmetic task performance (number of errors out of 12) and subjective stress for smoke-pre and smoke-post groups during'Test Trial 1 (omission of smoking cue) and Test Trail 2 (insertion of smoking cue). ** p < .01; * p < .05. with smoke-post, t(19) = 4.16, p < .01. There was no significant difference in stress during Test Trial 2, t(19) = 1.27, n.s. DISCUSSION These results indicated that removal of a previously predictable smoking cue preceding a repeated stressor may affect responses to that stressor. Specifically, exposure to a stress task in the absence of a smoking cue led to poorer performance and greater subjective stress in subjects who had previously performed the task only following the smoking cue (smoke-pre), compared with those who had equal recent exposure to the task and to smoking but without smoking acting as a cue for the task (smoke-post).
36
K.A. Perkins, L.H. Epstein, and J.R. lennings
Importantly, exposure to smoking before (i.e., as a cue) versus after the task did not influence task performance or subjective stress during the four acquisition trials. It was only when this predictable smoking cue was unexpectedly omitted that performance and stress were affected. Previous research has shown that physiological and subjective responses to a repeated noxious stimulus may be heightened if that stimulus is changed across presentations compared to when it is held constant (Epstein, 1973). Sirlailarly, adaptation to a repeated stimulus is slower when characteristics of the stimulus are added, altered, or omitted over time (Poulos et al., 1988; Siddle & Hirschhorn, 1986). T h e results of this study may be related to those findings. Although the smoking cue occurred immediately prior to (rather than concurrent with) the stress task, the temporal contiguity of smoking with stress may have served to create an association between smoking and the task. Thus, for the smoke-pre group, the novel omission of smoking during Test Trial I may have been equivalent to altering the characteristics of the task by removing one o f its implicit components (i.e., smoking). It is also conceivable that novel insertion o f smoking prior to a task may alter the characteristics o f the task by adding a new component. When such a change was instituted in this study during Test Trial 2 for the smoke-post group, performance significantly worsened compared with Test Trial 1, although there was no significant change in subjective stress. However, lack o f an additional control group (i.e., a group never receiving the preceding smoking cue on this or any previous trial) limits the usefulness of this particular comparison, perhaps leaving unanswered the question of whether novel insertion o f smoking influences task performance and stress. Although speculative, it is possible that cessation-induced disruption of a long-term, predictable pattern of smoking in conjunction with naturally occurring stressors could similarly impair coping with the stressors and help to explain stress-related craving and relapse (Shiffman et al., 1986). This notion is somewhat different from, but not incompatible with, other conditioning explanations o f drug craving. For example, Siegel, Krank, and Hinson (1988) suggested that reliable pairing of an environmental context (conditioned stimulus) with drug intake subsequently led to the elicitation of drug-opposite compensatory responses in the presence of that context, which then produced drug craving. However, our results suggest that drug intake (i.e., smoking) itself can come to serve as a reliable discriminative stimulus for responses to a task, which, if subsequently attempted in the absence of drug intake, may lead to poorer performance and increased distress, thus encouraging a resumption o f drug use in conjunction with the task. Generalization of these results to different behavioral and subjective responses to other stressors is needed. Although significant results were observed with the behavioral and subjective measures used here, the absolute differences between groups were rather small. Greater temporal separation between exposures to the stressor, use of more naturalistic stressors, and an increase in the number of pairings of smoking with the stressor may reveal greater effects of the removal o f smoking on task performance and subjective measures. Furthermore, similarity of these results with those of our previous study
Smoking as a Cue for Stress Responding
37
involving responses to a nonstress task (Epstein et al., 1990) indicate that the disruptive effects of novel omission of smoking may generalize across many types of tasks, perhaps more closely resembling state-dependent learning (Peeke & Peeke, 1984). Finally, intake of other substances in conjunction with repeated stressors (e.g., alcohol) may also have similar effects when unexpectedly omitted, as suggested by animal studies noted earlier. It is not clear if the critical element of the smoking cue was intake of nicotine or the behavior of puffing on a cigarette. Subjects took only four brief puffs prior to each trial and consumed only two cigarettes over the course of the entire session. Thus, amount o f nicotine intake per trial was probably small, suggesting that smoking behavior may have been the most salient cue component. In addition, our previous study (Epstein et al., 1990) found that, whereas removal of a smoking cue (and nicotine) affected electrodermal response to a repeated nonstress task previously preceded by smoking, presentation of a smoking cue without nicotine (i.e., smoking a nonnicotine cigarette) did not affect responding. On the other hand, other research has shown puff-by-puff changes in cognitive task performance (Revell, 1988), presumably due to transient increases in brain levels of nicotine, and there is a great deal of evidence that nicotine can serve as a potent discriminative stimulus for responding in animals (Morrison & Stephenson, 1969; Rosecrans, 1989). Future research employing nonnicotine cigarettes and/or an alternative method of nicotine delivery (e.g., Perkins, Epstein et al., 1986) may shed light on the importance of the behavioral versus pharmacological aspects of smoking in its function as a cue for stress responding. REFERENCES Bouton, M.E., Kenney, F.A., & Rosengard, C. (1990). State-dependent fear extinction with two benzodiazepine tranquilizers. Behavioral Neuroscience, 104, 44-55. Carroll, M.E., Lac, S.T., Asencio, M., & Keenan, R.M. (1989). Nicotine dependence in rats. Life Sciences, 45, 1381-1388. Corrigall, W.A.,'tterling, S., & Coen, K.M. (1989). Evidence for a behavioral deficit during withdrawal from chronic nicotine treatment. Pharmacology, Biochemistry, & Behavior, 33, 559-562. Cummings, K.M., Jaen, C.R., & Giovino, G. (1985). Circumstances surrounding relapse in a group of recent exsmokers. Preventive Medicine, 14, 195-202. Epstein, L.H., Dickson, B.E., McKenzie, S., & Russell, P.O. (1984). The effect of smoking on perception of muscle tension. Psychopharmacology, 83, 107-I 13. Epstein, L.H., & Perkins, K.A. (1988). Smoking, stress, and coronary heart disease. Journal of Consulting and Clinical Ps~'cholo~', 56, 342-349. Epstein, L.H., Perkins, K.A., Jennings, J.R., & Pastor, S. (1990). The effects of smoking context on habituation to a repeated cognitive task. Ps)'chopharmacolog'y, 100, 366-371. Epstein, S. (1973). Expectancy and magnitude of reaction to a noxious UCS. Psychophysiolog-y, 10, 100-107. Friedman, J., Horvath, T., & Meares, R. (1974). Tobacco smoking and a "'stimulus barrier." Nature, 248, 455-456. Hall, G.H., & Morrison, C.F. (1973). New evidence for a relationship between tobacco smoking, nicotine dependence, and stress. Nature, 243, 199-201. Hughes, J.R., Higgins, S.T., & Hatsukami, D. (1990). Effects of abstinence from tobacco. In L.T. Kozlowski, H.M. Annis, H.D. Cappell, F.B. Glaser, M.S. Goodstadt, Y. Israel, H.
38
K.A. Perkins, L.H. Epstein, and J.R. lennings
Kalant, E.M. Sellers, & E.R. Vingilis (Eds.), Research advances in alcohol and drug problems, (Vol. 10, pp. 317-398). New York: Plenum. Huitema, B. (1980). Analysis of covariance and alternath,es. New York: Wiley-lnterscience. Knott, V.J. (1984). Electrodermal activity during aversive stimulation: Sex differences in smokers and nonsmokers. Addictive Behaviors, 9, 195-199. Mackay, C.J. (1980). The measurement of mood and psychophysiological activity using self-report techniques. In I. Martin, & P.H. Venables (Eds.), Techniquesin psychoph~'siology(pp. 501-562). New York: Wiley. Mackay, C.J., Cox, T., Burrows, G.C., & Lazzarini, A.J. (1978). An inventory for the measurement of self reported stress and arousal. British Journal of Social and Clinical Psychology, 17, 283-284. Morrison, C.F. (1974). Effects of nicotine and its withdrawal on the performance of rats on signalled and unsignalled avoidance schedules. Ps.ychopharnmcologia, 38, 25-35. Morrison, C.F., & Stephenson, J.A. (1969). Nicotine injections as the conditioned stimulus in discrimination learning. Ps~'chopharmacologia, 15, 351-360. Niaura, R.S., Rohsenow, D.J., Binkoff, J.A., Monti, P.M., Pedraza, M., & Abrams, D.B. (1988). Relevance of cue reactivity to understanding alcohol and smoking relapse. Journal of Abnormal Ps)'chology, 97, 133-152. Payne, T.J., Etscheidt, M., & Corrigan, S.A. (1990). Conditioning arbitrary stimuli to cigarette smoke intake: A preliminary study. Journal of Substance Abuse, 2, 113-119. Peeke, S.C., & Peeke, H.V.S. (1984). Attention, memory, and cigarette smoking. Psychopharmacolog3", 84, 205-214. Perkins, K.A., Dubbert, P.M., Martin, J.E., Faulstich, M., & Harris, J.K. (1986). Cardiovascular reactivity to psychological stress in aerobically trained versus untrained mild hypertensives and normotensives. Health Ps~'chology, 5, 407-421. Perkins, K.A., Epstein, L.H., Stiller, R., Jennings, J.R., Christiansen, C., & McCarthy, T. (1986). An aerosol spray alternative to cigarette smoking in the study of the behavioral and physiological effects of nicotine. Behavior Research Methods, Instruments, and Computers, 18, 420-426. Pomerleau, O.F., & Pomerleau, C.S. (1984). Neuroregulators and the reinforcement of smoking: Towards a biobehavioral explanation. Neuroscience and Biobehavioral Rea,iews, 8, 503-513. Pomerleau, O.F., & Rosecrans, J. (1989). Neuroregulatory effects of nicotine. Psychoneuroepdocrinology, 14, 407-423. Poulos, C.S., Hunt, T., & Cappell, H. (1988). Tolerance to morphine analgesia is reduced by the novel addition or omission of an alcohol cue. Psychopharmacology, 94, 412-416. Puddey, I.B., Vandongen, R., Beilin, L.J., & English, D. (1984). Hemodynamic and neuroendocrine consequences of stopping smoking--a controlled study. Clipical and Experime~,tal Pharmacology & Ph)~iolo~', 11, 423-426. Re~'ell, A.D. (1988). Smoking and performance--a puff-by-puff analysis. Psychopharmacology, 96, 563-565. Rose, J.E., Ananda, S., & jarvik, M.E. (1983). Cigarette smoking during anxiety-provoking and monotonous tasks. Addicth,e Behaviors, 8, 353-359. Rosecrans, J.A. (1989). Nicotine as a discriminative stimulus: A neurobiobehavioral approach to studying central cholinergic mechanisms. Journal of Substance Abuse, 1, 287-300. Shiffman, S.M., Shumaker, S.A., Abrams, D., Cohen, S., Garvey, J., Grunberg, N., & Swan, G. (1986). Models of smoking relapse. Health Psschology, 5 (Suppl. 1), 13-27. Siddle, D.A.T., & Hirschhorn, T. (1986). Effects of stimulus omission and stimulus novelty on dishabituation of the skin conductance response. Psychophysiology, 23, 309-314. Siegel, S., Krank, M.D., & Hinson, R.E. (1988). Anticipation of pharmacological and nonpharmacological events: Classical conditioning and addictive behavior. In S. Peele (Ed.), Visions of addiction: Major contemporaryperspectiveson addiction and alcoholism (pp. 85-116). Lexington MA: D.C. Heath. Wesnes, K., & Warburton, D.M. (1983). Smoking, nicotine, and human performance. Pharmacology & Therapeutics, 21, 189-208.