Psychopharmacology (1999) 145 : 52–60
© Springer-Verlag 1999
ORIGINAL INVESTIGATION
Jennifer W. Tidey · Stephen T. Higgins Warren K. Bickel · Sandra Steingard
Effects of response requirement and the availability of an alternative reinforcer on cigarette smoking by schizophrenics
Received: 20 April 1998 /Final version : 20 January 1999
Abstract Cigarette smoking and other forms of drug abuse are more prevalent among schizophrenics than the general population. Despite the clinical importance of this problem, there has been relatively little experimental study of schizophrenic drug use. We examined under controlled laboratory conditions the effects of response requirement and the availability of an alternative (monetary) reinforcer on cigarette smoking by schizophrenics. Subjects were six heavy smokers with diagnoses of schizophrenia or schizoaffective disorder. Before each session, subjects provided carbon monoxide samples indicating recent smoking abstinence. During 3-h sessions, subjects obtained opportunities to smoke (2 puffs/ opportunity) under a fixed ratio (FR) schedule of reinforcement, which varied across sessions from FR50 to FR6400. In half of the sessions, subjects also were able to earn a small amount of money ($0.25/ratio completed) under an FR400 schedule. Increasing the response requirement for smoking decreased smoking and increased smoking-maintained responding. The availability of the monetary reinforcer decreased smoking and smoking-maintained responding by approximately half. These results are consistent with those seen previously in community volunteers without major mental illness studied under the same experimental conditions, suggesting that smoking by these two populations is controlled, at least in part, by a common set of determinants.
J.W. Tidey · S.T. Higgins (*) · W.K. Bickel · S. Steingard University of Vermont, Department of Psychiatry, 38 Fletcher Place, Burlington, VT 05401, USA e-mail :
[email protected] Fax: +1-802-656-9628 S.T. Higgins · W.K. Bickel University of Vermont, Department of Psychology, 38 Fletcher Place, Burlington, VT 05401, USA
Key words Schizophrenia · Psychopathology · Fixed-ratio schedule · Cigarette smoking · Nicotine · Alternative reinforcer · Substance abuse · Dual-diagnosis
Introduction Substance abuse is more common among persons with schizophrenia than the general population. Approximately 60–90% of schizophrenics smoke cigarettes, which is more than the general population (25–35%), persons with other psychiatric illnesses (35–55%), or controls matched on socio-economic status, marital status and alcohol use (30–35%; Hughes et al. 1986; Goff et al. 1992; Chiles et al. 1993; de Leon et al. 1995). The prevalence of current alcohol abuse or dependence among schizophrenics (34%) is 3 times that found in the general population and the prevalence of other substance abuse (28%) is 6 times that found in the general population (Regier et al. 1990). The high frequency of substance abuse by schizophrenics has a serious negative impact on the health of these individuals. Cocaine and other forms of substance abuse by schizophrenics can exacerbate positive psychotic symptoms such as thought disorder and decrease medication compliance, resulting in high rates of emergency room admissions and rehospitalization (Negrete et al. 1986; Seibyl et al. 1993; Shaner et al. 1995; Swofford et al. 1996). Schizophrenics are at greater risk than the general population for illnesses related to substance abuse, including HIV infection and cardiovascular disease (Herman et al. 1983; Allenbeck 1989; Cournos et al. 1991). Such health problems are exacerbated when substance-abusing schizophrenics use a great proportion of their disability payments to purchase drugs, leaving less money for food and housing (Drake and Wallach 1989; Shaner et al. 1995). In spite of the clear need for substance abuse treatment programs for dually diagnosed patients, few such
53
programs have been initiated (Leong and Horn 1989; Breckenridge 1990; Drake et al. 1991; Ziedonis and George 1997). In part, the paucity of treatment programs for schizophrenics may result from perceptions that substance abuse by these individuals is intractable (Smith and Grant 1989). Several theories have been advanced to explain the high correlation between substance abuse and schizophrenia (see reviews by LeDuc and Mittleman 1995; Lavin et al. 1996). Most of these theories suggest that biological factors underlie the high rates of substance abuse seen among schizophrenics. For example, one frequently cited proposal suggests that schizophrenics engage in substance abuse in order to self-medicate symptoms of their illness. According to this theory, schizophrenics abuse substances to ameliorate symptoms such as anxiety, distractibility or depression (e.g., Khantzian 1985; Dixon et al. 1991; Glassman 1993). A related theory is that schizophrenics abuse substances to attenuate the sedative or parkinsonian effects of antipsychotic medications (Schneier and Siris 1987; Goff et al. 1992; Sandyk 1993; LeDuc and Mittleman 1995). Other biologically based theories are that substance abuse may precipitate psychotic episodes in people who are genetically vulnerable to schizophrenia, or that the correlation between substance abuse and schizophrenia may result from a shared genetic vulnerability to both conditions (Tien and Anthony 1990; Turner and Tsaung 1990). Such hypotheses imply that variables controlling substance abuse by schizophrenics probably differ from those that control substance abuse by non-schizophrenics. Alternative hypotheses offered to explain the high rates of substance abuse among schizophrenics place greater emphasis on social and environmental variables that may differentially affect schizophrenics versus persons in the general population. For example, schizophrenics generally are not employed full-time and are already socially marginalized, and thus may be less influenced by the social costs and workplace sanctions on these behaviors (Schuster et al. 1979; Higgins 1995; Lavin et al. 1996). Such hypotheses imply that altering the environmental contingencies associated with substance abuse should decrease it. The aim of the present study was to investigate how manipulating environmental contingencies in a laboratory setting affects cigarette smoking by schizophrenic subjects. Cigarette smoking was selected as a convenient form of substance abuse to study because of its ubiquity among schizophrenics and because nicotine dependence shares many of the features of other forms of substance dependence (Henningfield 1984; Henningfield et al. 1991). Two variables were manipulated: response requirement to obtain the opportunity to smoke, and the presence or absence of a competing alternative activity (i.e., the opportunity to earn money). Based on the results of previous work in community volunteers without serious mental illness who
were studied using the same experimental arrangement (Bickel et al. 1995a), we postulated that smoking would decrease as the response requirement for smoking increased and when a competing reinforcer was available.
Materials and methods Subjects Subjects were four male and two female cigarette smokers with diagnoses of schizophrenia or schizoaffective disorder who were recruited from a local outpatient mental health center, and gave written informed consent. Diagnoses were made based on chart review by a board-certified psychiatrist. All participants were taking antipsychotic medications at the time of the study and three also were taking antidepressant medications. All participants were unemployed and receiving disability benefits at the time of the study. To be included in the study, subjects had to provide a breath sample with a carbon monoxide (CO) reading of at least 18 ppm at the initial interview. Subjects also had to score a minimum of 6 on the Fagerstrom Tolerance Questionnaire, a self-report measure of nicotine dependence in which scores can range from 0 to 11 and higher scores indicate greater levels of dependence (Fagerstrom and Schneider 1989). Subjects ranged in age from 20 to 48 years and reported smoking an average of 26.7 ± 4.9 (mean ± SEM) cigarettes per day (range: 10–40) and 21.7 ± 2.8 (range: 10–29) years of daily cigarette smoking. Four subjects usually smoked a cigarette brand that delivered medium levels of nicotine (1.0–1.2 mg); one subject usually smoked a brand that delivered low levels of nicotine (0.7 mg) and one subject usually smoked a brand that delivered high levels of nicotine (1.4 mg). Initial CO values ranged from 18 to 46 ppm and Fagerstrom scores ranged from 6 to 9. Apparatus Breath CO levels were measured using a MiniCO carbon monoxide meter (MSA, Pittsburgh, Pa., USA). During sessions, subjects were seated alone in isolated, ventilated rooms (4’ × 8’ × 8’) which contained one response console and one microcomputer. The response console consisted of three Lindsley plungers (Gerbands No. R6310; Arlington, Mass., USA) centered across the front. Subjects responded by pulling the left and right plungers; responses on the middle plunger had no progammed consequences. Responses on the left and right plungers were associated with different consequences, either cigarette puffs or money. Once responding on a plunger was initiated, responses on the other plunger had no programmed consequences until the response requirement on the initial plunger was completed. Cigarette puffs were taken through a plastic cigarette holder that was fitted to the filtered end of the cigarette. All cigarettes in the experiment were those regularly smoked by the participants and were supplied by the experimenters. The cigarette holder was connected via tubing to a volumetric low pressure transducer (Grass Instruments, Model PT5) which measured the volume of cigarette smoke inhaled. The computer monitor displayed cigarette puff volumes, number of smoking opportunities earned, total amount of money earned during the session (if available), time remaining in the session, and time remaining during inter-trial intervals (described below). Procedure Participants engaged in an initial screening interview, a 1-h practice session, and sixteen 3-h sessions. During the initial interview, participant eligibility was determined and participants provided a
54 breath sample for CO analysis. At the start of each subsequent session, participants were required to provide CO values at or below 50% of this initial value. Participants were instructed that to meet this criterion, it was necessary to abstain from smoking for 5–6 h before each session. When this criterion was not met, the session was rescheduled. Subjects reported to the laboratory at the same time each day, 5 days per week. Thirty minutes before the start of sessions, participants took 1 uniform puff on their preferred brand of cigarette. This presession exposure served to equate the time since last tobacco smoke exposure across participants (Henningfield and Griffiths 1981). During the 30-min interval before the experimental session, subjects rated the extent to which they were currently craving a cigarette on a 10-point scale (“how much are you currently craving cigarettes?” 0 = no craving, 9 = extreme craving). At the end of experimental sessions, subjects rated their cigarette craving along the same scale (“how much did you crave cigarettes during the session today?” 0 = no craving, 9 = extreme craving). In a 1-h practice session, participants were acclimated to the experimental procedures. During this session, 2 puffs on a cigarette were available after the completion of 100 responses [fixed ratio (FR)-100 schedule of reinforcement]. In subsequent sessions, the response requirement for two cigarette puffs and the availability of the monetary reinforcer were randomly varied. The FR values for cigarette smoking were 50, 100, 200, 400, 800, 1600, 3200 and 6400. At the start of each session, participants were informed of that session’s FR value for smoking and whether or not the monetary reinforcer was available. Additionally, small signs indicating the FR values for smoking and money (when available) were posted over the smoking- and money-associated levers. Each response requirement was tested once during each condition (monetary reinforcer available versus not available). Previous studies from our laboratory have shown that single-session exposures to the different response requirements produce reliable results (Bickel et al. 1992,1995a; Degrandpre et al. 1994). The schedule of monetary reinforcement, when available, was always FR400. The amount of money obtained per completed FR was $0.25 for all participants. The puffing procedure was modeled after one described previously (Zacny et al. 1987). Upon completion of the FR for monetary or smoking reinforcement, the computer emitted a tone, incremented a counter indicating to subjects the total amount of money and/ or puffs earned, and initiated a 5-min intertrial interval (ITI). If the completed FR was for smoking reinforcement, the participant then lit a cigarette without inhaling and inserted the cigarette into a plastic cigarette holder that was connected by tubing to the puff volume sensor. Participants were prompted by the computer to take one puff of a specified volume (approximately 60 cc of smoke). As subjects inhaled the cigarette smoke, puff volume was displayed on the computer monitor and a tone signaled when puff volume reached 60 cc. Participants were instructed to stop inhaling when they heard this signal. At 5 s after this tone, participants were prompted by another computer-generated tone to exhale. The second puff was taken 25 s later, using the same procedure. Participants were instructed to extinguish the cigarette after taking the second puff. A series of tones and the words “You may respond now” signaled the end of the 5-min inter-trial interval. A research assistant monitored experimental sessions and ensured compliance with the puffing procedure from outside the rooms via a computer monitor and a one-way mirror. When not responding on the plungers or smoking, participants were allowed to read newspapers or magazines and listen to the radio inside the experimental rooms. Two urine samples were collected from each participant on randomly selected days and were screened for illicit drug use using EMIT (Syva Corp., San Jose, Calif., USA). Statistical analysis Two-way repeated-measures analyses of variance (ANOVA) were used to examine the effects of response requirement and the pres-
ence versus absence of the monetary reinforcer option on smoking and smoking-maintained responding. Total cigarette puffs earned per session and total number of smoking-reinforced responses per session were log-transformed prior to analysis. The log transformations were needed to stabilize the variance in order to satisfy the homogeneity of variance assumption in the ANOVA (Box et al. 1978, p. 234). The standard deviations associated with observations were proportional to their mean. Because log transformations cannot be performed on zero data points, for sessions in which no responses were emitted and no puffs were earned we used the log of 1 for the number of puffs earned and the log of half of the value of the response requirement for that session for the number of responses emitted. This was only necessary for one subject, who failed to respond for smoking at the higher response requirements twice when the monetary reinforcer was not available and three times when the monetary reinforcer was available. Two-way repeated-measures ANOVAs also were used to examine the effects of response requirement and the presence versus absence of the monetary reinforcer on pause times and running rates of responding. Pause time was defined as session duration minus the sum of intertrial intervals and time spent responding for either smoking or the monetary reinforcer (when available). Running rate was defined as total responses per session divided by time spent responding on either plunger. A one-way repeated-measures ANOVA was used to examine the effect of response requirement on money earned and responding for the monetary reinforcer. A three-way repeated-measures ANOVA was used to examine the effects of response requirement, the presence versus absence of the monetary reinforcer option, and time of rating (pre- versus postsession) on the single item cigarette craving measure. For all statistical tests, the criterion for significance was P ≤ 0.05. For behavioral economics analyses, response requirements to obtain 2 cigarette puffs were converted to unit prices (unit price equals response requirement divided by two puffs), yielding the following unit prices: 25, 50, 100, 200, 400, 800, 1600, 3200. The number of cigarette puffs earned and total responses at each unit price were averaged across subjects under each alternative reinforcer condition. The functional relationships between consumption of cigarette puffs (C) and unit price (P) under both the alternative reinforcer available and unavailable conditions were modeled using an equation proposed by Hursh et al. (1988): C = LPbeaP
(1)
or, restated in logarithmic coordinates: log C = log L + b(log P) + aP
(2)
In these equations, L and b are related to initial consumption and slope of the demand curve, respectively, and a is a measure of acceleration, or rate of change in slope. Parameter estimates were obtained through linear regression. Demand curves fit through the averages of subjects’ data accounted for 96% and 93% of the variance under the no alternative reinforcer and reinforcer available conditions, respectively.
Results Cigarette smoking The availability of monetary reinforcement significantly decreased smoking [F(1,5) = 7.39, P < 0.05]. Averaging across all FR values, smoking decreased by 52.2 ± 6.7% when the monetary reinforcer was available (mean ± SEM; Fig. 1, top panel). Varying response requirement also significantly affected smoking [F(7,35) = 30.5, P < 0.01]. Averaging
55
tions, corresponding to rate of change in slope of the demand curves, were found to be significantly different, suggesting that the lines described by these equations are not parallel [F(1,10) = 5.57, P < 0.05]. No other parameters were found to differ between monetary reinforcement conditions. Smoking-maintained responding
Fig. 1 Effects of response requirement on cigarette smoking (top panel ) and smoking-maintained responding (bottom panel) by schizophrenic subjects when an alternative, monetary reinforcer was (closed symbols) or was not (open symbols) available. Points represent means ± SEM for the group of six subjects. Note log-log scales on both graphs
across both monetary reinforcement conditions, smoking decreased from 44.6 ± 2.4 puffs under the FR50 condition to 3.2 ± 0.5 puffs under the FR6400 condition, a 93% decrease (Fig. 1, top panel). There was no significant interaction between monetary reinforcement condition and response requirement. Although the availability of monetary reinforcement decreased smoking by a greater magnitude under the FR50 to FR3200 conditions (45–75%) than under the FR6400 condition (10%), this difference was not statistically significant (F(7,35) = 1.94, P = 0.09). We used behavioral economics analysis to investigate further the functional relationship between consumption of cigarette puffs (C) and unit price of smoking (P). Under the no monetary reinforcement condition, this relationship is described by the equation: log C = 5.1590.281 logP90.000574 P. In the monetary reinforcement available condition, the equation describing this relationship is: log C = 5.3190.526 logP90.00005 P. These equations are significantly different, suggesting that two lines fit these data better than one line [F(3,10) = 16.6, P < 0.001]. The coefficients of unit price (a parameters) of these equa-
The availability of monetary reinforcement significantly decreased smoking-maintained responding [F(1,5) = 7.41, P < 0.05]. Averaging across all FR values, smoking-maintained responding was decreased by 52.0 ± 6.5% when the monetary reinforcer was available (Fig. 1, bottom panel). Varying response requirement also significantly affected smoking-maintained responding [F(7,35) = 17.61, P < 0.01]. Smoking-maintained responding initially increased steeply as response requirement increased, then reached an asymptote at higher FR values (Fig. 1, bottom panel). Averaging across both monetary reinforcement conditions, smoking-maintained responding increased from 1121 ± 57.9 responses/ session under the FR50 condition to 9789 ± 1346.7 responses/ session under the FR6400 condition, a 773% increase. There was no significant interaction between monetary reinforcement condition and response requirement on smoking-maintained responding. To examine whether session duration determined consumption, that is, whether participants responded continuously throughout the session, we analyzed the effects of monetary reinforcement condition and response requirement on pause times and running rates of responding. Monetary reinforcement condition and response requirement significantly affected pause times, and there was a significant interaction between these variables [F(7,35) = 3.08, P < 0.05; F(7,35) = 3.60, P < 0.01; F(1,5) = 13.91, P < 0.05, respectively]. When monetary reinforcement was not available, pause times increased from 25.5 ± 11.8 min under the FR50 condition to 94.2 ± 19.0 min under the FR6400 condition. When monetary reinforcement was available, pause times did not change systematically across response requirement conditions, (e.g., FR50: 10.1 ± 8.4 min, FR6400 : 11.3 ± 7.1 min; Table 1). Neither monetary reinforcement condition nor response requirement affected running rates of responding. Running rates averaged 3.69 ± 0.01 resp /s across monetary reinforcement conditions and response requirements (Table 1). Average trial durations ranged from approximately 15 s when the FR for smoking was 50 responses up to 39 min when the FR for smoking was 6400 responses. Three participants required 30 min or less to complete the FR6400 condition; two participants required approximately 60 min to complete this condition, and one participant
56 Table 1 Pause times (min) and running rates (resp/s) when the alternative reinforcer was and was not available Response requirement 50
100
200
Pause times No alt. reinforcer Alt. reinforcer available
25.5 ± 11.8 10.1 ± 8.4
32.7 ± 15.4 23.1 ± 10.5 9.3 ± 6.6 11.1 ± 6.0
Running ratesb No alt. reinforcer Alt. reinforcer available
3.9 ± 0.4 3.7 ± 0.3
400
800
1600
3200
6400
69.6 ± 29.8 21.1 ± 13.5
72.5 ± 28.4 22.8 ± 11.4
94.2 ± 19.0 11.3 ± 7.1
3.5 ± 0.3 3.9 ± 0.5
3.3 ± 0.5 3.9 ± 0.4
3.1 ± 0.5 3.9 ± 0.4
a
3.4 ± 0.3 3.7 ± 0.4
3.8 ± 0.4 3.8 ± 0.3
22.5 ± 6.9 45.7 ± 18.2 10.7 ± 10.5 4.8 ± 3.9 3.8 ± 0.5 3.7 ± 0.4
3.8 ± 0.4 3.9 ± 0.4
Data represent mean ± SEM for the group of 6 participants a Pause time = total session duration – (time spent responding on either lever + inter-trial intervals) b Running rate = total session responses divided by time spent responding on either lever
did not emit any responses under this condition (data not shown).
an average of $3.71 ± 0.25 per session and made 5942 ± 402 responses/session on the money-associated plunger.
Monetary reinforcement and responding for money Subject ratings Varying response requirement for smoking did not affect the amount of money earned or responding on the money-associated plunger (Fig. 2). When monetary reinforcement was available, participants earned
Neither time of rating (pre- versus post-session) nor monetary reinforcement condition had significant main effects on the single-item cigarette craving measure. Response requirement had a significant main effect on craving [F(7,35) = 3.01, P < 0.05], but that effect interacted with time of rating [pre- versus post-session; F(7,35) = 2.32, P = 0.05]. When participants rated their cigarette cravings at the beginning of sessions, craving scores were similar across all response requirements (e.g., FR50 : 7.1 ± 1.0; FR6400 : 7.2 ± 0.9). Post-session craving scores were low following sessions in which FR values for smoking were low and subjects earned more smoking opportunities, and higher following sessions in which FR values for smoking were high and subjects earned fewer smoking opportunities (e.g., FR50 : 3.9 ± 0.9, FR6400 : 6.4 ± 0.9). Other drug use Urinalysis results indicated that illicit drug use during the study was infrequent. Two participants tested positive for marijuana use on both occasions that specimens were tested and the remaining four participants tested negative for all drugs screened on both occasions.
Discussion
Fig. 2 Effects of response requirement on smoking and money earned (top panel; smoking : squares; money: triangles) and responding maintained by smoking or money (bottom panel ). Points represent means ± SEM for the group of six subjects. Note log-log scales on both graphs
Cigarette smoking by schizophrenic outpatients was reduced by increasing the response requirement for smoking and providing an alternative, monetary reinforcer. Smoking-maintained responding was increased by increasing the response requirement for smoking and was reduced by providing an alternative reinforcer.
57
These results are similar to those reported previously for community volunteers without serious mental illness (Bickel et al. 1995a), suggesting that the functional relationships between response requirement, the presence of an alternative reinforcer, smoking, and smoking-maintained responding are similar for schizophrenic and non-schizophrenic smokers. The results of this experiment are consistent with the widely established finding that increasing the response requirement for a drug reinforcer decreases self-administration and increases drug-maintained responding in human and non-human subjects (see reviews by Griffiths et al. 1980; Bickel and DeGrandpre 1996). For example, in rats or monkeys, self-administration of morphine, ethanol, pentobarbital, phencyclidine and cocaine decreases as response requirement increases (Weeks and Collins 1964; Goldberg et al. 1971; Meisch and Thompson 1973; DeNoble et al. 1982; Lemaire and Meisch 1984; Carroll et al. 1991; Nader and Woolverton 1992; Rodefer et al. 1997). Similarly, in humans, self-administration of ethanol, diazepam, pentobarbital, nicotine and methadone decreases as response requirement increases (Mello et al. 1968; Bigelow and Liebson 1972; Bigelow et al. 1976; Bickel et al. 1991, 1992; Van Etten et al. 1995; Spiga et al. 1996). The relationship between increasing response requirement and overall increase in responding observed in the present study is similarly well-documented across different species and drugs (e.g., Griffiths et al. 1980; Bickel and DeGrandpre 1996). The results of the present study extend the generality of these basic functional relationships between response requirement, drug self-administration and drug-maintained responding to smoking by humans with major mental illness. One potential explanation for the decreases in smoking at high response requirements could be that subjects did not have time to earn additional smoking opportunities. Three subjects required 30 min or less to complete the highest response requirement for smoking (FR6400), two subjects required approximately 60 min, and one subject did not emit any responses under this condition. Based on these trial durations, all participants had enough time to earn additional smoking opportunities. Instead of earning more smoking opportunities at higher FRs, participants chose one of two courses of action, depending on whether the alternative reinforcer was or was not available : (1) when the alternative reinforcer was not available, subjects increased pause times; (2) when the alternative reinforcer was available, pause times were minimal but subjects still could have earned more smoking opportunities by forgoing responding for the monetary reinforcer. Instead, the amount of time allocated to the monetary reinforcer was stable across all response requirements for smoking. Thus, the decreases in smoking and smoking-maintained responding at high response requirements do not appear to result from
participants’ inability to earn more smoking opportunities due to time constraints. A second important finding of this study was that the availability of monetary reinforcement decreased cigarette smoking. This is consistent with prior reports that the presence of a non-drug reinforcer can decrease drug consumption in humans and non-humans. For example, concurrent access to food or a sweetened solution decreases cocaine, phencyclidine or ethanol self-administration in rats and rhesus monkeys (Carroll 1985; Nader and Woolverton 1991; Carroll and Lac 1993; Carroll and Rodefer 1993; Carroll et al. 1995; Comer et al. 1996). Similarly, in laboratory studies with human subjects, concurrent access to a non-drug reinforcer reduces drug self-administration (Vuchinich and Tucker 1988; Zacny et al. 1992; Foltin and Fischman 1994; Hatsukami et al. 1994; Higgins et al. 1994a; Higgins 1997). The finding that alternative reinforcers reduce drug self-administration has important clinical utility, as illustrated by studies showing reductions in cocaine or opiate use by drug users who are given the opportunity to earn vouchers or other privileges contingent on drug abstinence (reviewed by Stitzer and Higgins 1995; Higgins 1997). The results of the present study support the idea that such contingency-management procedures also might be an effective way of reducing smoking or other drug use among schizophrenics. Behavioral economics analysis has been applied to drug self-administration research as a framework for conceptualizing how drug consumption is affected by environmental conditions. According to the principles of behavioral economics, increasing response requirement and making an alternative reinforcer concurrently available increase the price of drug consumption which should lead to a decrease in drug taking, as observed in this study (Samuelson and Nordhaus 1985; Hursh and Bauman 1987). The behavioral economics analysis of the relationship between consumption of cigarette puffs and unit price for smoking determined that two linear equations provided a better fit for the data than one equation, consistent with the analysis of variance finding that the alternative reinforcer condition significantly affected smoking. The results of the two analyses differed somewhat in that, with the analysis of variance test, the interaction between alternative reinforcement condition and response requirement did not achieve statistical significance, whereas with the behavioral economics analysis, the lines described by the two equations were not found to be parallel. This discrepancy appears to have arisen because the presence of the monetary reinforcer had very little effect on smoking or smoking-maintained responding at the highest unit price, and this unit price weighs heavily when calculating acceleration associated with change in slope in Eqn 1. This discrepancy suggests that behavioral economics analyses may be more sensitive to interactions between unit price and the presence of an alternative reinforcer than analysis of variance.
58
As noted above, the present findings are similar to those from a prior study using the same experimental arrangement with non-schizophrenic community volunteers (Bickel et al. 1995a). In that study, cigarette smoking by community volunteers decreased as response requirement increased and when a monetary reinforcer was available. We re-analyzed the results from the Bickel et al. (1995a) study and compared them to the results from the present study. This comparison revealed that the two groups of subjects had similar levels of smoking at low FR values with the alternative reinforcer unavailable, and comparable decreases in cigarette smoking produced by increases in response requirement or the presence of the alternative reinforcer. Thus, we have no evidence from these studies to suggest that smoking by these schizophrenic subjects was controlled any differently by the independent variables than smoking by community volunteers without serious mental illness. Note that the schizophrenic participants in the present study were outpatients who were stabilized on antipsychotic medications. Thus, the possibility exists that smoking by less-stable patients may show other characteristics. Response requirement for smoking did not affect the amount of money earned during sessions in which the monetary response option was available. According to behavioral economics theory, when two reinforcers are concurrently available, they can interact in three ways: when consumption of the reinforcers is inversely related they are referred to as substitutes; when consumption of the reinforcers is directly related they are referred to as complements; and when consumption of the reinforcers is not related they are referred to as independent (Hursh and Bauman 1987; Bickel et al. 1995b). The type of interaction that exists between reinforcers is contextually determined and not an inherent property of the reinforcers themselves (Hursh and Bauman 1987; Bickel et al. 1995b). In the present study, increasing the price (i.e., response requirement) of smoking decreased its consumption but did not increase the consumption of a second reinforcer with a constant price (i.e., the monetary reinforcer), suggesting that smoking and the monetary reinforcer are independent in this context. This finding has also been reported previously for smokers who do not have major mental illness, again supporting the idea that the environmental variables manipulated in these studies had similar effects in schizophrenic and non-schizophrenic smokers (DeGrandpre et al. 1994; Bickel et al. 1995a). Self-reports of cigarette smoking craving increased significantly as response requirement for smoking increased. This effect appeared to be due to increases in post-session craving scores when the FR for smoking was high and participants therefore received fewer opportunities to smoke. The results from this craving scale illustrate that this relatively simple craving instrument is sensitive to smoking deprivation in schizophrenic outpatients.
Finally, the results of this study support the idea that the high rate of smoking by schizophrenics may be due, in part, to environmental factors that promote or condone smoking. Schizophrenics may experience fewer constraints on their smoking than the general population, and have fewer alternative reinforcers available to compete with smoking in the natural environment (Schuster et al. 1979; Schuster 1986; Higgins 1995; Lavin et al. 1996). The substantial decreases in smoking produced by the increased work requirements and by the opportunity to work for money in the present study suggest that the absence of constraints and the paucity of alternative reinforcing activities available to schizophrenics may contribute to the high rates of smoking among these individuals. A prior study by our group (Roll et al. 1998) demonstrated that offering monetary reinforcement contingent on smoking abstinence (e.g., a contingency-management intervention) reduced smoking by schizophrenics, and a study by another group demonstrated the same for cocaine use (Shaner et al. 1997). Further study of environmental arrangements that promote reductions in substance abuse by schizophrenics appears warranted. Acknowledgements The authors thank Mindee Peterson and Suzanne O’Neill for technical assistance and Gary Badger for statistical assistance with this study. This research was supported by NIDA grants DA-08076 (S.T.H.), DA-07242 (J.W.T.), and DA06526 (W.K.B.).
References Allenbeck P (1989) Schizophrenia: a life-shortening disease. Schizophr Bull 15: 81–89 Bickel WK, DeGrandpre RJ (1996) Psychological science speaks to drug policy: the clinical relevance and policy implications of basic behavioral principles. In: Bickel WK, DeGrandpre RJ (eds) Drug policy and human nature. Psychological perspectives on the prevention, management and treatment of illicit drug abuse. Plenum Press, New York, London, pp 31–52 Bickel WK, DeGrandpre RJ, Higgins ST, Hughes JR (1991) Behavioral economics of drug self administration. I. Functional equivalence of response requirement and drug dose. Life Sci 47 :1501–1510 Bickel WK, Hughes JR, DeGrandpre RJ, Higgins ST, Rizzuto P (1992) Behavioral economics of drug self-administration. IV. The effects of response requirement on the consumption of and interaction between concurrently available coffee and cigarettes. Psychopharmacology 107:211–216 Bickel WK, DeGrandpre RJ, Higgins ST, Hughes JR, Badger GJ (1995a) Effects of simulated employment and recreation on cigarette smoking: a behavioral economic analysis. Exp Clin Psychopharmacol 3:467–476 Bickel WK, DeGrandpre RJ, Higgins, ST (1995b) The behavioral economics of concurrent drug reinforcers: a review and reanalysis of drug self-administration research. Psychopharmacology 118 :250–259 Bigelow GE, Liebson IA (1972) Cost factors controlling alcoholic drinking. Psychol Rec 22: 305–314 Bigelow GE, Griffiths RR, Liebson IA (1976) Effects of response requirement upon human sedative self-administration and drug-seeking behavior. Pharmacol Biochem Behav 5 : 681–685
59 Box GEP, Hunter WG, Hunter JS (1978) Statistics for experimenters: an introduction to design, data analysis, and model building. Wiley, New York Breckenridge JS (1990) Smoking by outpatients. Hosp Commun Psychiatry 41:454–455 Carroll ME (1985) Concurrent phencyclidine and saccharin access: presentation of an alternative reinforcer reduces drug intake. J Exp Anal Behav 43: 131–144 Carroll ME, Lac ST (1993) Autoshaping i.v. cocaine self-administration in rats: effects of nondrug alternative reinforcers on acquisition. Psychopharmacology 110:5–12 Carroll ME, Rodefer JS (1993) Income alters choice between drug and an alternative nondrug reinforcer in monkeys. Exp Clin Psychopharmacol 1:110–120 Carroll ME, Carmona GG, May SA (1991) Modifying drug-reinforced behavior by altering the economic conditions of the drug and a nondrug reinforcer. J Exp Anal Behav 56: 361–376 Carroll ME Rodefer JS, Rawleigh JM (1995) Concurrent selfadministration of ethanol and an alternative nondrug reinforcer in monkeys : effects of income (session length) on demand for drug. Psychopharmacology 120:1–9 Chiles JA, Cohen S, Maiuro R, Wright R (1993) Smoking and schizophrenic psychopathology. Am J Addict 2: 315–319 Comer SD, Lac ST, Wyvell CL, Carroll ME (1996) Combined effects of buprenorphine and a nondrug alternative reinforcer on IV cocaine self-administration in rats maintained under FR schedules. Psychopharmacology 125:355–360 Cournos F, Empfield M, Horwath E, McKinnon K, Meyer I, Schrage H, Currie C, Agosin B (1991) HIV seroprevalence among patients admitted to two psychiatric hospitals. Am J Psychiatry 148 :1225–1230 DeGrandpre RJ, Bickel WK, Higgins ST, Hughes JR (1994) A behavioral economic analysis of concurrently available money and cigarettes. J Exp Anal Behav 61:191–201 de Leon J, Dadvand M, Canuso C, White AO, Stanilla JK, Simpson GM (1995) Schizophrenia and smoking: An epidemiological survey in a state hospital. Am J Psychiatry 152: 453–455 DeNoble VJ, Svikis DS, Meisch RA (1982) Orally delivered pentobarbital as a reinforcer for rhesus monkeys with concurrent access to water: effects of concentration, fixed-ratio size, and liquid positions. Pharmacol Biochem Behav 16:113–117 Dixon L, Haas G, Weiden PJ, Sweeney J, Frances AJ (1991) Drug abuse in schizophrenic patients: Clinical correlates and reasons for use. Am J Psychiatry 148: 224–230 Drake RE, Wallach MA (1989) Substance abuse among the chronic mentally ill. Hosp Commun Psychiatry 40: 1041–1046 Drake RE, McLaughlin P, Pepper B, Minkoff K (1991) Dual diagnosis on major mental illness and substance disorder : an overview. New Dir Ment Health Serv 50:3–12 Fagerstrom K-O, Schneider NG (1989) Measuring nicotine dependence: a review of the Fagerstrom Tolerance Questionnaire. J Behav Med 12 :159–182 Foltin RW, Fischman MW (1994) Effects of buprenorphine on the self-administration of cocaine by humans. Behav Pharmacol 5: 79–89 Glassman AH (1993) Cigarette smoking: implications for psychiatric illness. Am J Psychiatry 150:546–553 Goff DC, Henderson DC, Amico E (1992) Cigarette smoking in schizophrenia: relationship to psychopathology and medication side effects. Am J Psychiatry 149: 1189–1194 Goldberg SR, Hoffmeister F, Schlichting UU, Wuttke W (1971) A comparison of pentobarbital and cocaine self-administration in rhesus monkeys: effects of dose and fixed-ratio parameter. J Pharmacol Exp Ther 179:277–283 Griffiths RR, Bigelow GE, Henningfield JE (1980) Similarities in animal and human drug-taking behavior. In: Mello NK (ed) Advances in substance abuse, vol 1. JAI Press, Greenwich, pp 1–90 Hatsukami DK, Thompson TN, Pentel PR, Flygare BK, Carroll ME (1994) Self-administration of smoked cocaine. Exp Clin Psychopharmacol 2 :115–125
Henningfield JE (1984) Behavioral pharmacology of cigarette smoking. In: Thompson T, Dews PB, Barrett JE (eds) Advances in behavioral pharmacology, vol 4. Harcourt Brace Jovanovich, Orlando, pp 131–210 Henningfield JE, Griffiths RR (1981) Cigarette smoking and subjective response: effects of d-amphetamine. Clin Pharmacol Ther 30: 497–505 Henningfield JE, Cohen C, Slade JD (1991) Is nicotine more addictive than cocaine? Br J Addict 86:565–569 Herman HE, Baldwin JA, Christie D (1983) A record-linkage study of mortality and general hospital discharge in patients diagnosed as schizophrenics. Psychol Med 13:581–593 Higgins ST (1995) Comments on combined behavioral and pharmacological treatments of drug abuse. In: Onkey LS, Blaine JD, Boren JJ (eds) NIDA Research Monograph #150, Integrating behavioral therapies with medications in the treatment of drug dependence. NIH Pub. No. 95–3899, US Government Printing Office, Washington, pp 170–179 Higgins ST (1997) The influence of alternative reinforcers on cocaine use and abuse: a brief review. Pharmacol Biochem Behav 57 : 419–427 Higgins ST, Bickel WK, Hughes JR (1994a) Influence of an alternative reinforcer on human cocaine self-administration. Life Sci 55 :179–187 Hughes JR, Hatsukami DK, Mitchell JE, Dahlgren LA (1986) Prevalence of smoking among psychiatric outpatients. Am J Psychiatry 143:993–997 Hursh SR, Bauman RA (1987) The behavioral analysis of demand. In: Green L, Kagel JH (eds) Advances in behavioral economics, vol 1. Ablex Publishing, Norwood, NJ, pp 117–165 Hursh SR, Raslear TG, Shurtleff D, Bauman R, Simmons L (1988) A cost-benefit analysis of demand for food. J Exp Anal Behav 50 :419–440 Khantzian EJ (1985) The self-medication hypothsis of adictive disorders: focus on heroin and cocaine dependence. Am J Psychiatry 142:1259–1264 Lavin MR, Siris SG, Mason SE (1996) What is the clinical importance of cigarette smoking in schizophrenia? Am J Addict 5 :189–208 LeDuc PA, Mittleman G (1995) Schizophrenia and psychostimulant abuse: a review and re-analysis of clinical evidence. Psychopharmacology 121:407–427 Lemaire GA, Meisch RA (1984) Pentobarbital self-administration in rhesus monkeys: drug concentration and fixed-ratio size interactions. J Exp Anal Behav 42: 37–49 Leong GB, Horn M (1989) Economic motivation for smoking cessation in chronic psychotic patients. Acta Psychiatr Scand 80 :657–658 Meisch RA, Thompson T (1973) Ethanol as a reinforcer: effects of fixed-ratio size and food deprivation. Psychopharmacologia 28 : 171–183 Mello NK, McNamee HB, Mendelson JH (1968) Drinking patterns of chronic alcoholics: gambling and motivation for alcohol. In: Cole JO (ed) Clinical research in alcoholism. American Psychiatric Association, Psychiatric Report No. 24, Washington, pp 83–118 Nader MA, Woolverton WL (1991) Effects of increasing the magnitude of an alternative reinforcer on drug choice in a discrete trials choice procedure. Psychopharmacology 105 : 169–174 Nader MA, Woolverton WL (1992) Effects of increasing response requirement on choice between cocaine and food in rhesus monkeys. Psychopharmacology 108: 295–300 Negrete JC, Knapp WP, Douglas DE, Smith WB (1986) Cannabis affects the severity of schizophrenic symptoms: results of a clinical survey. Psychol Med 16: 515–520 Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, Goodwin FK (1990) Comorbidity of mental disorders with alcohol and other drug abuse. JAMA 264:2511–2518
60 Rodefer JS, Mattox AJ, Thompson SS, Carroll ME (1997) Effects of buprenorphine and an alternative nondrug reinforcer, alone and in combination on smoked cocaine selfadministration in monkeys. Drug Alcohol Depend 45: 21–29 Roll JM, Higgins ST, Steingard S, McGinley M (1998) Use of monetary reinforcement to reduce the cigarette smoking of persons with schizophrenia : a feasibility study. Exp Clin Psychopharmacol 6:157–161 Samuelson PA, Nordhaus WD (1985) Economics, 12th edn. McGraw-Hill, New York Sandyk R (1993) Cigarette smoking : effects on cognitive functions and drug-induced parkinsonism in chronic schizophrenia. Int J Neurosci 70:193–197 Schneier FR, Siris SG (1987) A review of psychoactive substance use and abuse in schizophrenia – patterns of drug choice. J Nerv Ment Dis 175:641–652 Schuster CR (1986) Implications of laboratory research for the treatment of drug dependence. In: Goldberg SR, Stolerman IP (eds) Behavioral analysis of drug dependence. Academic Press, Orlando, pp 357–386 Schuster CR, Renault PE, Blaine J (1979) An analysis of the relationship of psychopathology to non-medical drug use. In: Pickens R, Heston LL (eds) Psychiatric factors in drug abuse. Grune and Stratton, New York, pp 1–19 Seibyl JP, Satel SL, Anthony D, Southwick SM, Krystal JH, Charney DS (1993) Effects of cocaine on hospital course in schizophrenia. J Nerv Ment Dis 181: 31–37 Shaner A, Eckman TA, Roberts LJ, Wilkins JN, Tucker DE, Tsuang JW, Mintz J (1995) Disability income, cocaine use, and repeated hospitalization among schizophrenic cocaine abusers – A government-sponsored revolving door? N Engl J Med 333:777–783 Shaner A, Roberts LJ, Eckman TA, Tucker DE, Tsuang JW, Wilkins JN, Mintz J (1997) Monetary reinforcement of abstinence from cocaine among mentally ill patients with cocaine dependence. Psychiatr Serv 48: 807–810
Smith WR, Grant BL (1989) Effects of a smoking ban on a general hospital psychiatric service. Hosp Commun Psychiatry 40 : 497–502 Spiga R, Grabowski J, Silverman PB, Meisch RA (1996) Human methadone self-administration: effects of dose and ratio requirement. Behav Pharmacol 7:130–137 Stitzer ML, Higgins ST (1995) Behavioral treatment of drug and alcohol abuse. In: Bloom FE, Kupfer DJ (eds) Psychopharmacology: the fourth generation of progress. Raven Press, New York, pp 1807–1819 Swofford CD, Kasckow JW, Scheller-Gilkey G, Inderbitzin LB (1996) Substance use: a powerful predictor of relapse in schizophrenia. Schizophr Res 20:145–151 Tien AY, Anthony JC (1990) Epidemiological analysis of alcohol and drug use as risk factors for psychotic experiences. J Nerv Ment Dis 178: 473–480 Turner WM, Tsuang MT (1990) Impact of substance abuse on the course and outcome of schizophrenia. Schizophr Bull 16 : 87–85 Van Etten ML, Higgins ST, Bickel WK (1995) Effects of response cost and unit dose on alcohol self-administration in moderate drinkers. Behav Pharmacol 6: 1–5 Vuchinich RE, Tucker JA (1988) Contributions from behavioral theories of choice as a framework to an analysis of alcohol abuse. J Abnorm Psychol 92:408–416 Weeks J, Collins RJ (1964) Factors affecting voluntary morphine intake in self-maintained addicted rats. Psychopharmacologia 6 :267–279 Zacny JP, Stitzer ML, Brown FJ, Yingling JE, Griffiths RR (1987) Human cigarette smoking: Effects of puffs and inhalation parameters on smoke exposure. J Pharmacol Exp Ther 240 : 554–564 Zacny JP, Divane WT, De Wit H (1992) Assessment of magnitude and availability of a non-drug reinforcer on preference for a drug reinforcer. Hum Psychopharmacol 7: 281–286 Ziedonis DM, George TP (1997) Schizophrenia and nicotine use: report of a pilot smoking cessation program and review of neurobiological and clinical uses. Schizophr Bull 23: 247–254
