- Email: [email protected]
Impaired inhibitory control of behavior in chronic cocaine users
Drug and Alcohol Dependence 66 (2002) 265– 273 www.elsevier.com/locate/drugalcdep
Impaired inhibitory control of behavior in chronic cocaine users Mark T. Fillmore a,*, Craig R. Rush a,b,c b
a Department of Psychology, Uni6ersity of Kentucky, Lexington, KY 40506 -0044, USA Department of Beha6ioral Science, Uni6ersity of Kentucky, Lexington, KY 40539 -0086, USA c Department of Psychiatry, Uni6ersity of Kentucky, Lexington, KY 40539 -0086, USA
Abstract This study examined the ability to inhibit and execute behavioral responses in adult cocaine users and in an aged-matched sample with no history of cocaine use. Subjects (n= 22) were identified as cocaine users by testing positive for the presence of cocaine or benzoylecgonine in urine-analysis and by self-reported cocaine use. Control subjects (n= 22) tested negative in urine-analysis and reported no past cocaine use. Response inhibition and response execution were measured by a stop-signal paradigm using a choice reaction time task that engaged subjects in responding to go-signals when stop-signals occasionally informed them to inhibit the response. Cocaine users displayed significantly poorer ability to inhibit their behavioral responses than did controls. Specifically, cocaine users required more time to inhibit responses to stop-signals and displayed a lower probability of inhibiting their responses. Cocaine users did not differ from controls in their ability to execute responses as measured by their speed and accuracy of responses to go-signals. These findings are important because they identify a specific deficit involving behavioral inhibition that could contribute to cocaine abuse, and explain its association with other disorders of self-regulation, such as ADHD. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cocaine abuse; Response inhibition; Human
1. Introduction The adverse health effects of chronic cocaine abuse have become well-documented in recent years. In particular, there is growing evidence that long-term habitual cocaine users display neuropsychological impairments and have neuroanatomical abnormalities (Caine, 1998; Jentsch and Taylor, 1999). Several studies have found that long-term cocaine use is associated with deficits in global measures of performance based on several neuropsychological tasks, including tests of attention, memory (Ardila et al., 1991), intellectual functioning (O’Malley et al., 1992), learning, problem solving, and perceptual motor speed (Beatty et al., 1995) (for a review, see Strickland and Stein, 1995). The neuroanatomical basis for these neuropsychological deficits is supported by evidence from neuroimaging studies which show impairments of frontal lobe functions that are thought to be important in the control * Corresponding author. Tel.: +1-859-257-4728; fax: + 1-859-3231979. E-mail address: [email protected] (M.T. Fillmore).
and regulation of behavior of these individuals (Volkow et al., 1996). A principal function of this brain region is to control behavior via inhibitory processes that normally serve to regulate behavior by suppressing or terminating pre-potent (i.e. environmentally-triggered) responses (for reviews see Caine, 1998; Jentsch and Taylor, 1999). Some investigators suggest that repeated dopaminergic activation of the neural circuits by chronic cocaine use eventually impairs inhibitory functions, leading to a loss of control over behavioral impulses, including cocaine self-administration (Lyvers, 2000; Volkow et al., 1996). Several lines of research have pointed to an association between long-term cocaine use and impairments of inhibitory processes (Ardila et al., 1991; Biggins et al., 1997; Horner et al., 1996; Volkow et al., 1996). Studies of animals and humans provide converging evidence that repeated cocaine use produces persistent cognitive deficits that often involve inhibitory processes of attention and behavior. For example, studies of animals show long-term deficits in sensory inhibition following repeated cocaine administration (Boutros et al., 1994, 1997). Studies of humans show that individuals with
0376-8716/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 6 - 8 7 1 6 ( 0 1 ) 0 0 2 0 6 - X
266
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
histories of cocaine abuse show patterns of premature responding (Bauer, 2001), perseverative behavior, and an inability to adapt behavior to environmental changes (Lane et al., 1998). In addition, cocaine users are more likely to report symptoms of attention deficit/ hyperactivity disorder (ADHD) and other behavioral self-regulation disorders than are controls with no history of cocaine abuse (Horner et al., 1996; Levin et al., 1998). Despite mounting evidence that cocaine abuse is associated with an impaired ability to inhibit behavior, no research has directly tested cocaine users for specific deficits in their ability to inhibit behavior. To date, most studies of these individuals have employed neuropsychological measures of their global intellectual functioning that do not isolate specific cognitive abilities or processes. However, in recent years several behavioral tasks have been developed to identify the specific inhibitory processes that underlie disorders of self-control. In particular, the development of a cognitive ‘stop-signal paradigm’ has generated considerable research on inhibitory processes and the assessment of specific deficits in the ability to inhibit behavioral responses (Logan and Cowan, 1984; Logan et al., 1984; Schachar et al., 1995). The paradigm is based on a cognitive model of control which asserts that the ability to inhibit an action is determined by the outcome of competitive activating and inhibiting processes elicited by cues to activate or inhibit a response. The time in which each competing process is completed determines the behavioral outcome. If the inhibiting processes are completed first, the response is withheld, but if the activating processes finish first, the response is executed. The model is tested by a stop-signal task which is a dual task that elicits conflicting go- and stop-processes in individuals by engaging them in responding to go-signals, but occasionally requiring them to inhibit the response when a stop-signal occurs. The stop-signal task directly measures the subject’s ability to inhibit (i.e. suppress) a pre-potent behavioral response in the presence of conflicting go- and stop-signals. The task is unique because of its direct assessment of the ability to inhibit a pre-potent action. The paradigm differs from ‘reactive’ inhibition models that measure inhibition of a behavior that results from interference owing to some other competing behavioral response. Such models include: stoop interference (MacLeod, 1991), negative priming (Tipper, 1985), and inhibition of return (Posner and Cohen, 1984). Compared with these measures, the stop-signal model is thought to provide a more direct measure of inhibitory control (Quay, 1997). There is evidence that stop-signal tasks are sensitive to impairments of behavioral inhibition that underlie self-control. The deficit is implicated in many behavioral disorders, including anti-social personality, obses-
sive-compulsive disorder, ADHD, as well as substance abuse disorders, including alcohol dependence (Barkely, 1997; Tannock, 1998). Studies using the stop-signal paradigm have found that behavioral inhibition was reduced in individuals with self-control disorders, such as Oppositional/Defiant Disorder and ADHD (Oosterlaan and Sergeant, 1996; Schachar et al., 1995). The paradigm also has been used to monitor the efficacy of clinical treatments for poor inhibitory self-control (Tannock et al., 1989). The present study employed the stop-signal paradigm to determine if the inhibitory processes of behavior in chronic cocaine users might also be deficient. The study compared the response inhibition of a group of adult cocaine users with a control group of non-users. Given the evidence for neuroanatomical abnormalities and neuropsychological impairments among cocaine users (Ardila et al., 1991; Volkow et al., 1996), we predicted that cocaine abusing individuals would display a specific deficit in the ability to inhibit behavioral pre-potent responses as measured in the stop-signal paradigm.
2. Method
2.1. Subjects Twenty-two chronic cocaine using adults (seven female and 15 male) and 22 control participants (10 female and 12 male) with no history of chronic cocaine use participated in the study. The racial make-up of the cocaine using group was 86% African American and 14% Caucasian, and the make-up of the control group was 36% African American and 64% Caucasian. At the time of the study, 81% of the control group and 27% of the cocaine using group reported current employment. The sample was drawn from adults in the Lexington, KY metropolitan area, who volunteered to participate in studies of behavioral pharmacology. Participants were recruited via notices posted on community bulletin boards and by word of mouth. The study was approved by the University of Kentucky Medical Institutional Review Board. All potential volunteers had to be between the ages of 18 and 50 years, with no self-reported major Axis I psychiatric disorder, head trauma, or other CNS injury. Potential volunteers had to have: a minimum of grade eight education, demonstrated reading ability, and no uncorrected vision or auditory problems. The cocaine using and control groups were matched in terms of age range, range of education, geographic location, and racial make-up. Of these characteristics, it was especially important to ensure a similar age range across groups. Although basic cognitive inhibitory processes are not related to general intellectual functioning, such as IQ performance, there is evidence that inhibitory
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
control declines throughout the life span (Logan, 1994; Schachar and Logan, 1990). Thus it was important to confine our sample to an age range that would avoid extreme age differences (e.g. young adults vs the elderly). The sample was obtained from a pool of approximately 70 potential volunteers who responded to the advertisement for studies conducted in our lab over the period of 1 year. Within this pool of potential volunteers, the most common reason for excluding an individual from the study was failure to meet age or education criteria. Volunteers were classified as cocaine users if they met all of the following criteria: (1) a score of at least four on the Drug Abuse Screening Test (DAST) (Skinner, 1982); (2) self-report of past week cocaine use; (3) positive test for the presence of cocaine or benzoylecgonine in urine-analysis (ONTRAK Abusscreens, Roche Diagnostic Systems, Nutley, NJ); and (4) self-reported history of habitual cocaine use for a minimum of 6 months. Most (91%) of these subjects smoked cocaine in the form of crack and 9% reported intranasal use of powdered cocaine. None of these volunteers were currently in, or seeking, treatment for their substance use. The sample reported an average frequency of cocaine use of 3.8 days in the past week (SD= 1.9), and 16.0 days in the past month (SD= 10.1). The sample reported smoking a mean daily amount of 2.1 rocks of crack cocaine per day (SD= 2.0), and spending an average of $22.0 per day on cocaine (SD= 27.0). The majority of the sample also reported using nicotine (86%) and alcohol (82%). The mean total weekly alcohol consumption for this sample was 44.0 standard drinks (SD = 46.4). Over one half of the sample also reported some past month use of marijuana (59%). Two subjects reported using a benzodiazepine or a hallucinogen in the past month. No other drug use, including prescription medication, was reported by the sample. Volunteers were classified as controls if they met all of the following criteria: (1) a score of less than four on the DAST; (2) no self-reported current cocaine use; (3) a negative test for the presence of cocaine and benzoylecgonine in urine-analysis; and (4) no self-reported history of past habitual cocaine use. Habitual use was defined as at least four times within a month during a subject’s lifetime. These control subjects also reported nicotine use (36%) and alcohol use (68%). The mean total weekly alcohol consumption for this sample was 22.3 standard drinks (SD=38.9). Some controls also reported past month use of marijuana (18%). No other drug use, including prescription medication, was reported.
2.2. Apparatus Response inhibition and response execution were measured by a stop-signal paradigm using a choice
267
reaction time (RT) task that engages participants in responding to go-signals when stop-signals occasionally inform them to inhibit the response. The task was performed on a PC computer and was operated by Micro Experimental Laboratory Software (Psychology Software Tools Inc., Pittsburgh, PA). The ‘go’ stimuli for the choice response were four 1.5 cm letters (A, B, C, and D), presented one at a time in the center of a computer monitor. Letters were preceded by a 500 ms preparation interval in which a fixation point (c) appeared in the center of the computer screen. A letter was displayed for 500 ms, and the computer screen was blank for a 2.5 s inter-stimulus interval before the next letter was displayed. This provided a 3 s period in which the subject could respond to the letter. A subject responded to a letter by pressing one of two adjacent keys on the computer keyboard, using the index and middle fingers of the preferred hand. One key (the period key) was pressed to indicate that one of two letters (A or C) appeared, and the adjacent key (the forward slash key) was pressed to indicate that one of the other two letters (B or D) appeared. A single stop-signal test presented each of the four letter stimuli equally often, for a total of 176 letter presentations. A stop-signal occurred on 27% of the 176 presentations (i.e. 48 times) during a test. The stop-signal was a 500 ms 900 Hz tone generated by the computer at a comfortable listening level. Participants were required to withhold (i.e. inhibit) their response when a stop-signal was presented. Stop-signals were presented 12 times, at each of four delays (i.e. 50, 150, 250 and 350 ms after the onset of a letter). These delays were chosen based on previous research using this task in our lab that showed that the probability of inhibiting decreases in an orderly, linear fashion as the stop-signal delays increase from 50 to 350 ms (Fillmore et al., 2001; Fillmore and Vogel-Sprott, 1999, 2000). The order of letters, stop-signals, and delays was random. A test was completed in approximately 10 min. Response inhibition was determined by the number of times that an individual inhibited responses to stopsignal trials (out of 48), and by the estimated mean latency to inhibit a response (Section 2.4). Response execution to go-signals was measured by the mean RT to the go-signals during a test (i.e. the average time from the onset of letter presentation until a computer key press). Shorter RTs (i.e. faster responding) indicated greater response execution. Choice response errors to go-signals were also recorded. The inhibition and execution measures are highly reliable across trials (alpha coefficients \ 0.90) and stable over sessions (test-retest reliabilities \ 0.85) in both drug using and non-using populations (Fillmore and Rush, 2001; Fillmore and Vogel-Sprott, 1999; Mulvihill et al., 1997).
268
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
2.3. Procedure All participants were tested individually. All sessions were scheduled from Monday to Friday between 8:30 and 16:30 h. Potential volunteers responded to the posted advertisements by phoning the lab. The advertisements stated that adult volunteers were needed to perform behavioral tests and to report on their drug and alcohol use. A research assistant conducted a phone interview and an appointment was then made for the volunteer to attend a familiarization/assessment session at the laboratory. Subjects were instructed not to drink alcohol or take any drug during their scheduled lab days. During all sessions, participants provided a breath sample using an Intoxilyzer Model 400 (CMI Inc. Owensborough, KY) to ensure a zero blood alcohol concentration (BAC). A urine sample also was obtained to test for the presence of cocaine/benzoylecgonine, as well as recent use of benzodiazepines, barbiturates, tetrahydrocannabinol, amphetamine, and opiates. With the exception of cocaine/benzoylecgonine, no volunteers tested positive for the presence of any drug at the time of study. The familiarization/assessment session required approximately 2 h to complete and its purpose was to acquaint the volunteers to the laboratory and to obtain background information relevant to the study and group classification. Upon arriving for the familiarization/assessment session, volunteers were greeted by the research assistant, who explained the general details of the study protocol and obtained the volunteers’ written, signed informed consent. Participants then completed a packet of questionnaires regarding their current and past state of physical and mental health, current and past drug use, as well as demographic measures, including age and education level. The packet included: MiniMental Status Examination (MMSE) (Folstein et al., 1975), DAST (Skinner, 1982), Michigan Alcoholism Screening Test (MAST) (Selzer, 1971), Beck Depression Inventory (BDI) (Beck et al., 1961), Brief Psychiatric Rating Scale (Overall and Gorham, 1962), a general health questionnaire, and detailed drug use questionnaires. The packet also included an abridged form of the Wender Utah Rating Scale (WURS) which is a 25-item symptom rating scale that provides a retrospective measure of childhood symptoms characteristic of ADHD (Ward et al., 1993). After the packet was completed, volunteers were paid $20 for attending, and an appointment was scheduled for the test session. Attempts were made to schedule this session within 1 week of the familiarization/assessment session. The subjects performed the stop-signal task on the test session. Volunteers met the research assistant who escorted them to a quiet testing room that contained the computer that controlled the stop-signal task. The assistant explained the task requirements. Subjects were
instructed to respond to the letters (i.e. go-signals) as quickly and as accurately as possible and to try to withhold their response if they heard a tone sound. Participants then performed one familiarization (practice) test on the task. The research assistant remained in the room to observe the subjects and verify that they understood the requirements. After the familiarization, the research assistant answered any questions that the subjects’ had about the task requirements. Participants then performed the stop-signal test. The research assistant was blind with respect to the purpose of the experiment, and to the measures obtained by the task. The task data were automatically stored on computer disk, and were not available to the assistant. No performance feedback was provided to subjects. After the test, participants were debriefed and paid $20 for their participation in the session.
2.4. Dependent measures of stop-signal performance 2.4.1. Measures of response inhibition The ability to inhibit a response can be impaired in two important ways (Logan, 1994; Tannock et al., 1995): (1) the initiation of the inhibitory process might fail to occur occasionally, thus resulting in less stop-signal inhibitions; (2) The completion of the inhibitory process might be slowed, so that it takes longer to inhibit a response to a stop-signal. The stop-signal paradigm provides a method for assessing these two criterion measures of response inhibition. 2.4.1.1. Probability of inhibition (P-inhibition). The proportion of successfully inhibited responses on the 48 stop-signal trials for a test provided a general measure of the ability to inhibit responses for a subject. Smaller proportions indicated a lower probability of inhibiting a response to a stop-signal (poorer inhibitory control of behavior). 2.4.1.2. Stop-signal reaction time (SSRT). Response inhibition also depends on the time required to inhibit a response. Deficits in inhibitory control might arise because more time is required to inhibit a response. The estimated mean time (in ms) required for a participant to inhibit responses to stop-signals (SSRT) was also measured in the present research. The method for calculating SSRT was based on the subject’s probability of inhibiting responses to stop-signals and the distribution of RTs to go-signal trials. The calculations are explained in detail elsewhere (Logan, 1994), but the general logic of the method is outlined here. SSRT is the time needed to inhibit the pre-potent response once the stop-signal occurs. The estimate is based on a method of determining how long the stop-signal can be delayed after the go-signal before the subject can no longer inhibit the response. As a general observation, the time
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
required to inhibit is less than the time required to respond (Logan and Cowan, 1984). This is evident by the fact that individuals can successfully inhibit a response to a stop-signal that does not occur until some time after the go-signal has been presented. SSRTs are generally related to the number of successfully inhibited responses on a test, with shorter SSRTs associated with a greater number of inhibited responses. However, the SSRT also provides latency information concerning the speed of the inhibitory processes. Longer SSRTs, resulting from a general slowing of the inhibitory processes, indicate weak inhibitory control over behavior.
2.4.2. Measures of response execution The stop-signal task also provides measures of response execution. Responses displayed during go-signal trials (i.e. when no stop-signal occurs) yielded two important criterion measures of response execution: go-signal RT and go-signal accuracy. 2.4.2.1. Go-signal reaction time (RTgo). The strength of response execution to go-signals was measured by a participant’s mean RT (ms) to the 128 go-signal trials presented during a test (i.e. the average time from the onset of go-signals until a computer key press). This produced a mean RTgo score for a participant for each test. All analyses involving RTgo scores were performed on observed mean RTgo scores, and on the mean RTgo scores obtained after removing outlier trial RTs that were 2.5 SDs from the mean RTgo score on a test. On average, less than 2% of trials on a test had outlier RT values. Analyses based on actual RTgo scores and trimmed RTgo scores yielded identical conclusions. The Results section reports analyses based on trimmed RTgo scores. 2.4.2.2. Go-signal response accuracy. Choice response errors on go-signal trials were also recorded. These errors occur when a participant presses the incorrect key in response to a letter presentation during a go-signal trial. The errors generally reflect failures of attention or information processing. Choice response errors during a test are rare and typically occur on less than 5% of all go-signal responses (Fillmore and VogelSprott, 1999, 2000; Logan, 1994).
3. Results
3.1. Demographic and background characteristics Table 1 summarizes the demographic and drug use data for the cocaine using and control group and reports the results of t-tests that compared the groups. The table shows no significant difference in the groups’ age, education level, mental status (MMSE scores), or
269
Table 1 Mean demographic and background characteristics of subjects. Standard deviation (SD) is presented in parentheses. Independent t-tests compared cocaine users with controls Sample
Controls
Age BMI Education Income
40.5 (5.6) 28.8 (6.4) 12.4 (1.3) 14987.8 (8783.9) DAST 1.5 (1.1) MAST 6.6 (8.4) WURS 17.9 (13.2) BDI 6.3 (6.6) MMSE 27.7 (2.2) Weekly drinks 22.3 (38.9) Daily cigarettes 5.2 (10.0) Daily caffeine 523.7 (708.1)
Cocaine users
Significant
40.1 25.1 12.0 9050.0
(7.7) (3.9) (1.5) (5317.5)
ns ** ns *
(5.9) (12.0) (20.3) (8.3) (1.9) (46.4) (9.2) (38.6)
** ** ns ns ns ns ** ns
13.2 18.9 23.7 6.4 28.9 44.0 14.2 312.6
ns, Non-significant difference; BMI, body-mass index score; BDI, Beck Depression Inventory; MMSE, Mini-Mental Status Examination (maximum score = 30); DAST, Drug Abuse Screening Test; MAST, Michigan Alcoholism Screening Test; WURS, Wender Utah Rating Scale for childhood ADHD; income, annual income in dollars; weekly drinks, weekly number of standard alcoholic drinks consumed; daily cigarettes, number of cigarettes smoked per day; daily caffeine, mgs of caffeine consumed per day. * PB0.05. ** PB0.01.
WURS and BDI scores. Cocaine users had significantly higher DAST and MAST scores than control subjects. Cocaine users also reported greater daily use of cigarettes. With respect to weekly alcohol consumption, cocaine users reported a higher mean number of drinks than controls, but this difference was not statistically significant.
3.2. Stop-signal task performance Table 2 reports the stop-signal task measures for each group. The table presents mean scores for the two measures of response inhibition (P-inhibition and SSRT) and for the two measures of response execution Table 2 Mean stop-signal task performance measures of subjects. Standard deviation is in parentheses. Independent t-tests compared cocaine users with controls Sample
Controls
Cocaine users
Significant
Response inhibition measures P-inhibitions 0.69 (0.04) SSRT (ms) 286.4 (11.5)
0.56 (0.05) 349.6 (23.2)
* *
Response execution measures RTgo (ms) 590.6 (23.5) Choice errors 9.8 (1.4)
579.0 (23.8) 8.9 (1.3)
ns ns
ns, Non-significant difference. * PB0.05; **PB0.01.
270
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
each delay. The ANOVA revealed significant main effects of group (F1,42 = 4.3, P= 0.043), and stop-signal delay (F3,126 = 80.9, PB 0.001), but no significant group× stop-signal delay interaction, (F3,126 = 1.0, P= 0.410).
3.2.1.2. Stop-signal RT. Table 2 also shows that cocaine users displayed significantly longer SSRTs than did controls (t42 = 2.4, P= 0.020). A 2-group ANCOVA also obtained a significant difference in SSRT after controlling for subjects’ weekly alcohol use, (F1,41 =7.4, P= 0.009). 3.2.2. Response execution measures Fig. 1. Mean probability of inhibiting a response (P-inhibition) to 12 stop-signals at each of four stop-signal delays (50, 150, 250, 350 ms) for cocaine users and controls. Vertical capped lines indicate standard error of the mean (group n= 22).
(RTgo and choice errors). The table also reports the results of t-tests that compared the two groups on stop-signal task measures.
3.2.1. Response inhibition measures 3.2.1.1. Probability of inhibition (P-inhibition). Table 2 shows that cocaine users displayed a lower probability of inhibiting responses to stop-signals (i.e. P-inhibition scores) compared with controls (t42 =2.1, P =0.043). Given that cocaine users tended to have higher weekly levels of alcohol consumption, an analysis of covariance (ANCOVA) also tested the group difference in P-inhibition after removing any variance due to individual differences in weekly alcohol use. The 2-group analysis of covariance (ANCOVA) also revealed a significant group difference in P-inhibition scores (F1,41 = 5.2, P= 0.028). Thus, the lower probability of inhibiting displayed by cocaine users was not related to their weekly alcohol use. To further explore the group difference in P-inhibition scores, subjects’ scores were plotted as a function of stop-signal delay. Research shows that the probability of inhibiting a response diminishes as the stop-signal delay increases because less time is available to inhibit the pre-potent response (Logan, 1994). Fig. 1 shows the mean probability of inhibiting a response at each of the four stop-signal delays for the two groups. As expected, the probability of inhibiting diminished as the stop-signal delay increased. This pattern was evident in both groups. Moreover, Fig. 1 shows that cocaine users displayed a consistently lower P-inhibition than controls at all stop-signal delays. These observations were confirmed by a 2 (group)×4 (stop-signal delay) mixed analysis of variance of subjects’ P-inhibition scores at
3.2.2.1. RT to go-signals (RTgo). Table 2 presents the results of an independent t-test that compared the RTgo scores of the two groups. No significant difference was obtained. Fig. 2 plots the mean RTgo scores for each group along with the groups’ mean SSRT scores. The figure shows that the mean RTgo score of controls and cocaine users was very similar. By contrast, the SSRT score of cocaine users was longer than that displayed by controls. Together, the two measures of reaction time show that RT differences between controls and cocaine users were specific to the inhibition of responses rather than to their execution. Fig. 2 also provides information regarding the general difference in time required to inhibit a response versus execute the response. The figure shows that the mean SSRT of controls is approximately one half of their mean RTgo (286.4 vs 590.6 ms). This RT difference is consistent with parametric studies of stop-signal performance which found that inhibiting a response requires
Fig. 2. Mean SSRT (response inhibition) and mean RTgo (response execution) for cocaine users and controls. Vertical capped lines atop bars indicate standard error of the mean (group n = 22).
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
about one half of the time needed to execute a response (Logan, 1994).
3.2.2.2. Choice errors. Table 2 presents the results of an independent t-test that compared the number of choice response errors displayed by the two groups. No significant difference was obtained. The mean number of errors was similar in the two groups. Based on 128 go-signal trials these means represented a high accuracy rate (above 90%).
4. Discussion This study used a stop-signal paradigm to compare the inhibitory control of cocaine users to an agedmatched sample of individuals with no history of habitual cocaine use. The study showed that cocaine users displayed a poorer ability to inhibit a response than did controls. Compared with controls, cocaine users displayed a lower probability of inhibiting responses, and required more time to inhibit responses as estimated by their stop-signal reaction time (i.e. SSRT). By contrast, cocaine users did not differ from controls in their ability to execute responses, as measured by their speed and accuracy of responding to go-signals. The reaction time and response accuracy scores in both groups were comparable to those obtained in other studies of healthy volunteers that have used the stop-signal task (Fillmore et al., 2001; Fillmore and Vogel-Sprott, 1999, 2000). The study also showed that response inhibition diminished as a function of increasing the stop-signal delay. Fig. 1 shows the negative slope functions that relate the probability of inhibiting a response to the stop-signal delays. The slopes are consistent with previous studies showing that stop-signal latencies affect the probability of response inhibition (Fillmore and VogelSprott, 1999; Logan and Cowan, 1984). Moreover, the presence of these slopes provides verification that subjects in both groups understood the task requirements and followed instructions. In studies of other populations, such as young children, task performance can be affected by inattention and by random response strategies, owing to a lack of motivation or interest on the part of the subject (Schachar et al., 1995; Tannock et al., 1995). Such response styles are detected by the slope function. Randomly inhibiting and executing responses generates a flat slope function because inhibitions are equally likely to occur at all stop-signal delays under such a strategy. This was not observed in the present study. Rather, the negative slopes demonstrate that response inhibition in both groups was under some degree of stimulus control of the stop signals. The lower probability of response inhibition among cocaine users could not be attributed to faster response
271
execution because both groups displayed similar mean reaction times to go signals. The lower probability might be due to a general slowing of the inhibitory process, so that more time is required to inhibit a response. This time requirement was estimated by the SSRT measure which showed that cocaine users displayed longer estimated latencies to inhibit responses than did controls. The mean SSRT of cocaine users was approximately 350 ms, whereas controls displayed a mean SSRT of 286 ms. SSRT estimates of healthy adults typically range between 200 and 300 ms (Logan, 1994), and estimates above 300 ms are usually observed in special populations, such as children with ADHD (Schachar et al., 1995; Tannock et al., 1995). Although statistically significant, performance differences in the order of milliseconds might raise concern regarding their relevance to more macro-level behaviors, such as drug abuse. However, many fundamental cognitive and perceptual processes, such as inhibitory influences, are considered to operate in a ‘bottom-up’ fashion to exert increasing influence at each stage of higher-order attentional and cognitive functions (McClelland and Rumelhart, 1981). Thus, although cocaine users exhibited a SSRT slowing of approximately 65 ms compared with controls, such a deficiency might exert considerable influence on higher-order cognitive functions, such as impulse-control. Moreover, the similarity between cocaine users and individuals with ADHD in terms of prolonged SSRT estimates could point to a common deficit in both populations. Disorders characterized by deficits of self-control and attention are prevalent among cocaine users (O’Malley et al., 1992). Sensation seeking, impulsivity, antisocial personality, and childhood disorders of self-control, such as conduct disorder and ADHD, are all considered to pose a risk for the development of cocaine use and other drug use (Horner et al., 1996; Ratey et al., 1992). From a clinical perspective, cocaine use is seen as part of a general disinhibitory psychopathology that involves a constellation of traits (Strickland and Stein, 1995). The common phenotypic characteristic among these traits is a deficit in behavioral inhibition. Deficits in inhibitory control could also directly contribute to instances of heavy cocaine and other drug use by compromising the ability to stop repeated drug administrations in the situation. The present findings are new and raise many questions about how such cognitive deficits in inhibitory control could relate to drug abuse. The poorer inhibitory control displayed by cocaine users also could represent a direct effect of recent cocaine administration. These individuals reported using the drug on a weekly basis and tested positive for cocaine in urineanalysis during their initial laboratory assessment session. However, studies of other stimulants, such as, methylphenidate, amphetamine, and caffeine, have reported that these drugs increase (rather than decrease)
272
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
response inhibition on the stop-signal task (Feola et al., 2000; Fillmore and Vogel-Sprott, 1999; Tannock et al., 1995). As yet, the direct effects of acute cocaine administration on response inhibition await to be explored. Nonetheless, the findings obtained in the present research are important because they identify a specific behavioral deficit that could contribute to cocaine use, and explain its association with other disorders of self-regulation, such as ADHD. However, its etiological role in cocaine use remains unclear, and it is uncertain as to whether cocaine use contributes to inhibitory deficits, or if they are part of a pre-existing constellation of ADHD-like symptoms that operate to increase risk for cocaine and other drug use.
Acknowledgements This research was supported by Grants DA14079 and DA10325 from the National Institute on Drug Abuse. Offprint requests should be sent to Mark T. Fillmore, Ph.D., Department of Psychology, University of Kentucky, Lexington, KY 40506-0044. Email: [email protected]
References Ardila, A., Rosselli, M., Strumwasser, S., 1991. Neuropsychological deficits in chronic cocaine abusers. Int. J. Neurosci. 57, 73 – 79. Barkely, R.A., 1997. Behavioral inhibitions, sustained attention, and executive functions: constructing a unifying theory of ADHD. Psychol. Bull. 121, 65 –94. Bauer, L.O., 2001. Antisocial personality disorder and cocaine dependence: their effects on behavioral and electoencephalographic measures of time estimation. Drug Alcohol Depend. 63, 87 – 95. Beatty, W.W., Katzung, V.M., Moreland, V.J., Nixon, S.J., 1995. Neuropsychological performance of recently abstinent alcoholics and cocaine abusers. Drug Alcohol Depend. 37, 247 –253. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E., Erbaugh, J.K., 1961. An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561 – 571. Biggins, C.A., MacKay, S., Clark, W., Fein, G., 1997. Event-related potential evidence for frontal cortex effects of chronic cocaine dependence. Soc. Biol. Psychiatry 42, 472 –485. Boutros, N.N., Uretsky, N., Bernston, G., Bornstein, R., 1994. Effects of cocaine on sensory inhibition in rats: preliminary data. Soc. Biol. Psychiatry 36, 242 –248. Boutros, N.N., Uretsky, N., Lui, J.J., Millana, R.B., 1997. Effects of repeated cocaine administration on sensory inhibition in rats: preliminary data. Soc. Biol. Psychiatry 41, 461 –466. Caine, S.B., 1998. Neuroanatomical basis of the reinforcing stimulus effects of cocaine. In: Higgins, S.T., Katz, J.L. (Eds.), Cocaine Abuse: Behavior, Pharmacology and Clinical Applications. Academic Press, New York, pp. 21 – 50. Feola, T.W., de Wit, H., Richards, J.B., 2000. Effects of d-amphetamine and alcohol on a measurof behavioral inhibition in rats. Behav. Neurosci. 114, 838 –848. Fillmore, M.T., Rush, C.R., 2001. The effects of oral cocaine on inhibitory control in humans. Manuscript in preparation.
Fillmore, M.T., Rush, C.R., Kelly, T.H., Hays, L., 2001. Triazolam impairs inhibitory control of behavior in humans. Exp. Clin. Psychopharmacol. 9, 363 – 371. Fillmore, M.T., Vogel-Sprott, M., 1999. An alcohol model of impaired inhibitory control and its treatment in humans. Exp. Clin. Psychopharmacol. 7, 49 – 55. Fillmore, M.T., Vogel-Sprott, M., 2000. Response inhibition under alcohol: effects of cognitive and motivational conflict. J. Stud. Alcohol 61, 239 – 246. Folstein, M., Folstein, S., Hugh, P., 1975. Mini-mental state — a practical method for grading the cognitive state of patients for the clinician. J. Psychiat. Res. 12, 189 – 198. Horner, B., Scheibe, K., Stine, S., 1996. Cocaine abuse and attentiondeficit hyperactivity disorder: implications of adult symptomatology. Psych. Addict. Behav. 10, 55 – 60. Jentsch, J.D., Taylor, J.R., 1999. Impulsivity resulting from frontostriatal dysfunction in drug abuse: implication for the control of behavior by reward-related stimuli. Psychopharmacology 146, 373 – 390. Lane, S.D., Cherek, D.R., Dougherty, D.M., Moeller, F.G., 1998. Laboratory measurement of adaptive behavior change in humans with a history of substance dependence. Drug Alcohol Depend. 51, 239 – 252. Levin, F.R, Evans, S.M., Kleber, H.D., 1998. Prevalence of adult attention-deficit hyperactivity disorder among cocaine abusers seeking treatment. Drug Alcohol Depend. 52, 15 – 25. Logan, G.D., 1994. On the ability to inhibit thought and action: a users’ guide to the stop-signal paradigm. In: Dagenbach, D., Carr, T.H. (Eds.), Inhibitory Processes in Attention, Memory, and Language. Academic Press, San Diego, CA, pp. 189 – 239. Logan, G.D., Cowan, W.B., 1984. On the ability to inhibit thought and action: a theory of an act of control. Psychol. Rev. 91, 295 – 327. Logan, G.D., Cowan, W.B., Davis, K.A., 1984. On the ability to inhibit simple and choice reaction time responses: a model and a method. J. Exp. Psychol: Hum. Percept. Perform. 10, 276 –291. Lyvers, M., 2000. ‘Loss of control’ in alcoholism and drug addiction: a neuroscientific interpretation. Exp. Clin. Psychopharmacol. 8, 225 – 249. MacLeod, C.M., 1991. Half a century of research on the Stroop effect: an integrative review. Psychol. Bull. 109, 163 – 203. McClelland, J.L., Rumelhart, D.E., 1981. An interactive activational model of context effects in letter perception: part 1. An account of basic findings. Psych. Rev. 88, 375 – 407. Mulvihill, L.E., Skilling, T.A., Vogel-Sprott, M., 1997. Alcohol and the ability to inhibit behavior in men and women. J. Stud. Alcohol 58, 600 – 605. O’Malley, S., Adamse, M., Heaton, R.K., Gawin, F.H., 1992. Neuropsychological impairment in chronic cocaine abusers. Am. J. Drug Alcohol Abuse 18, 131 – 144. Oosterlaan, J., Sergeant, J.A., 1996. Inhibition in ADHD, aggressive, and anxious children: a biologically based model of child psychopathology. J. Abnorm. Child Psychol. 24, 19 – 37. Overall, J.E., Gorham, D.R., 1962. Brief psychiatric rating scale. Psychol. Res. 10, 79 – 812. Posner, M.I., Cohen, Y., 1984. Components of visual orienting. In: Bouma, H., Bouwhuis, D. (Eds.), Attention and Performance X. Lawrance Erlbaum, Hillsdale, NJ. Quay, H.C., 1997. Inhibition and attention deficit hyperactivity disorder. J. Abnorm. Child Psychol. 25, 7 – 13. Ratey, J.J., Greenberg, M.S., Bemporad, J.R., Lindem, K.J., 1992. Unrecognized attention-deficit hyperactivity disorder in adults presenting for outpatient psychotherapy. J. Child. Adoles. Psychopharmacol. 2, 267 – 275. Schachar, R.J., Logan, G.D., 1990. Impulsivity and inhibitory control in normal development and childhood psychopathology. Develop. Psychol. 26, 710 – 720.
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273 Schachar, R., Tannock, R., Marriott, M., 1995. Deficient inhibitory control in attention deficit hyperactivity disorder. J. Abnorm. Child Psychol. 23, 411 – 437. Selzer, M.L., 1971. The Michigan alcoholism screening test: the quest for a new diagnostic instrument. Am. J. Psychiatry 127, 1653 – 1658. Skinner, H.A., 1982. The drug abuse screening test. Addict. Behav. 7, 363 – 371. Strickland, T.L., Stein, R., 1995. Cocaine-induced cerebrovascular impairment: challenges to neuropsychological assessment. Neuropsychol. Rev. 5, 69 –79. Tannock, R., Schachar, R., Logan, G., 1995. Methylphenidate and cognitive flexibility: dissociated dose effects in hyperactive children. J. Abnorm. Child Psychol. 23, 235 –267.
273
Tannock, R., 1998. Attention deficit hyperactivity disorder: Advances in cognitive, neurobiological, and genetic research. J. Child Psychol. Psychiatry. 39, 65 – 99. Tipper, S.P., 1985. The negative priming effect: inhibitory effects of ignored primes. Quart. J. Exp. Psychol. 37A, 571 – 590. Volkow, N.D., Ding, Y., Fowler, J.S., Wang, G., 1996. Cocaine addiction: hypothesis derived from imaging studies with PET. J. Addict. Dis. 15, 55 – 71. Ward, M.F., Wender, P.H., Reimherr, F.W., 1993. The Wender Utah rating scale: an aid in the retrospective diagnosis of childhood attention deficit hyperactivity disorder. Am. J. Psychiatry 150, 885 – 890.
Impaired inhibitory control of behavior in chronic cocaine users Mark T. Fillmore a,*, Craig R. Rush a,b,c b
a Department of Psychology, Uni6ersity of Kentucky, Lexington, KY 40506 -0044, USA Department of Beha6ioral Science, Uni6ersity of Kentucky, Lexington, KY 40539 -0086, USA c Department of Psychiatry, Uni6ersity of Kentucky, Lexington, KY 40539 -0086, USA
Abstract This study examined the ability to inhibit and execute behavioral responses in adult cocaine users and in an aged-matched sample with no history of cocaine use. Subjects (n= 22) were identified as cocaine users by testing positive for the presence of cocaine or benzoylecgonine in urine-analysis and by self-reported cocaine use. Control subjects (n= 22) tested negative in urine-analysis and reported no past cocaine use. Response inhibition and response execution were measured by a stop-signal paradigm using a choice reaction time task that engaged subjects in responding to go-signals when stop-signals occasionally informed them to inhibit the response. Cocaine users displayed significantly poorer ability to inhibit their behavioral responses than did controls. Specifically, cocaine users required more time to inhibit responses to stop-signals and displayed a lower probability of inhibiting their responses. Cocaine users did not differ from controls in their ability to execute responses as measured by their speed and accuracy of responses to go-signals. These findings are important because they identify a specific deficit involving behavioral inhibition that could contribute to cocaine abuse, and explain its association with other disorders of self-regulation, such as ADHD. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cocaine abuse; Response inhibition; Human
1. Introduction The adverse health effects of chronic cocaine abuse have become well-documented in recent years. In particular, there is growing evidence that long-term habitual cocaine users display neuropsychological impairments and have neuroanatomical abnormalities (Caine, 1998; Jentsch and Taylor, 1999). Several studies have found that long-term cocaine use is associated with deficits in global measures of performance based on several neuropsychological tasks, including tests of attention, memory (Ardila et al., 1991), intellectual functioning (O’Malley et al., 1992), learning, problem solving, and perceptual motor speed (Beatty et al., 1995) (for a review, see Strickland and Stein, 1995). The neuroanatomical basis for these neuropsychological deficits is supported by evidence from neuroimaging studies which show impairments of frontal lobe functions that are thought to be important in the control * Corresponding author. Tel.: +1-859-257-4728; fax: + 1-859-3231979. E-mail address: [email protected] (M.T. Fillmore).
and regulation of behavior of these individuals (Volkow et al., 1996). A principal function of this brain region is to control behavior via inhibitory processes that normally serve to regulate behavior by suppressing or terminating pre-potent (i.e. environmentally-triggered) responses (for reviews see Caine, 1998; Jentsch and Taylor, 1999). Some investigators suggest that repeated dopaminergic activation of the neural circuits by chronic cocaine use eventually impairs inhibitory functions, leading to a loss of control over behavioral impulses, including cocaine self-administration (Lyvers, 2000; Volkow et al., 1996). Several lines of research have pointed to an association between long-term cocaine use and impairments of inhibitory processes (Ardila et al., 1991; Biggins et al., 1997; Horner et al., 1996; Volkow et al., 1996). Studies of animals and humans provide converging evidence that repeated cocaine use produces persistent cognitive deficits that often involve inhibitory processes of attention and behavior. For example, studies of animals show long-term deficits in sensory inhibition following repeated cocaine administration (Boutros et al., 1994, 1997). Studies of humans show that individuals with
0376-8716/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 6 - 8 7 1 6 ( 0 1 ) 0 0 2 0 6 - X
266
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
histories of cocaine abuse show patterns of premature responding (Bauer, 2001), perseverative behavior, and an inability to adapt behavior to environmental changes (Lane et al., 1998). In addition, cocaine users are more likely to report symptoms of attention deficit/ hyperactivity disorder (ADHD) and other behavioral self-regulation disorders than are controls with no history of cocaine abuse (Horner et al., 1996; Levin et al., 1998). Despite mounting evidence that cocaine abuse is associated with an impaired ability to inhibit behavior, no research has directly tested cocaine users for specific deficits in their ability to inhibit behavior. To date, most studies of these individuals have employed neuropsychological measures of their global intellectual functioning that do not isolate specific cognitive abilities or processes. However, in recent years several behavioral tasks have been developed to identify the specific inhibitory processes that underlie disorders of self-control. In particular, the development of a cognitive ‘stop-signal paradigm’ has generated considerable research on inhibitory processes and the assessment of specific deficits in the ability to inhibit behavioral responses (Logan and Cowan, 1984; Logan et al., 1984; Schachar et al., 1995). The paradigm is based on a cognitive model of control which asserts that the ability to inhibit an action is determined by the outcome of competitive activating and inhibiting processes elicited by cues to activate or inhibit a response. The time in which each competing process is completed determines the behavioral outcome. If the inhibiting processes are completed first, the response is withheld, but if the activating processes finish first, the response is executed. The model is tested by a stop-signal task which is a dual task that elicits conflicting go- and stop-processes in individuals by engaging them in responding to go-signals, but occasionally requiring them to inhibit the response when a stop-signal occurs. The stop-signal task directly measures the subject’s ability to inhibit (i.e. suppress) a pre-potent behavioral response in the presence of conflicting go- and stop-signals. The task is unique because of its direct assessment of the ability to inhibit a pre-potent action. The paradigm differs from ‘reactive’ inhibition models that measure inhibition of a behavior that results from interference owing to some other competing behavioral response. Such models include: stoop interference (MacLeod, 1991), negative priming (Tipper, 1985), and inhibition of return (Posner and Cohen, 1984). Compared with these measures, the stop-signal model is thought to provide a more direct measure of inhibitory control (Quay, 1997). There is evidence that stop-signal tasks are sensitive to impairments of behavioral inhibition that underlie self-control. The deficit is implicated in many behavioral disorders, including anti-social personality, obses-
sive-compulsive disorder, ADHD, as well as substance abuse disorders, including alcohol dependence (Barkely, 1997; Tannock, 1998). Studies using the stop-signal paradigm have found that behavioral inhibition was reduced in individuals with self-control disorders, such as Oppositional/Defiant Disorder and ADHD (Oosterlaan and Sergeant, 1996; Schachar et al., 1995). The paradigm also has been used to monitor the efficacy of clinical treatments for poor inhibitory self-control (Tannock et al., 1989). The present study employed the stop-signal paradigm to determine if the inhibitory processes of behavior in chronic cocaine users might also be deficient. The study compared the response inhibition of a group of adult cocaine users with a control group of non-users. Given the evidence for neuroanatomical abnormalities and neuropsychological impairments among cocaine users (Ardila et al., 1991; Volkow et al., 1996), we predicted that cocaine abusing individuals would display a specific deficit in the ability to inhibit behavioral pre-potent responses as measured in the stop-signal paradigm.
2. Method
2.1. Subjects Twenty-two chronic cocaine using adults (seven female and 15 male) and 22 control participants (10 female and 12 male) with no history of chronic cocaine use participated in the study. The racial make-up of the cocaine using group was 86% African American and 14% Caucasian, and the make-up of the control group was 36% African American and 64% Caucasian. At the time of the study, 81% of the control group and 27% of the cocaine using group reported current employment. The sample was drawn from adults in the Lexington, KY metropolitan area, who volunteered to participate in studies of behavioral pharmacology. Participants were recruited via notices posted on community bulletin boards and by word of mouth. The study was approved by the University of Kentucky Medical Institutional Review Board. All potential volunteers had to be between the ages of 18 and 50 years, with no self-reported major Axis I psychiatric disorder, head trauma, or other CNS injury. Potential volunteers had to have: a minimum of grade eight education, demonstrated reading ability, and no uncorrected vision or auditory problems. The cocaine using and control groups were matched in terms of age range, range of education, geographic location, and racial make-up. Of these characteristics, it was especially important to ensure a similar age range across groups. Although basic cognitive inhibitory processes are not related to general intellectual functioning, such as IQ performance, there is evidence that inhibitory
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
control declines throughout the life span (Logan, 1994; Schachar and Logan, 1990). Thus it was important to confine our sample to an age range that would avoid extreme age differences (e.g. young adults vs the elderly). The sample was obtained from a pool of approximately 70 potential volunteers who responded to the advertisement for studies conducted in our lab over the period of 1 year. Within this pool of potential volunteers, the most common reason for excluding an individual from the study was failure to meet age or education criteria. Volunteers were classified as cocaine users if they met all of the following criteria: (1) a score of at least four on the Drug Abuse Screening Test (DAST) (Skinner, 1982); (2) self-report of past week cocaine use; (3) positive test for the presence of cocaine or benzoylecgonine in urine-analysis (ONTRAK Abusscreens, Roche Diagnostic Systems, Nutley, NJ); and (4) self-reported history of habitual cocaine use for a minimum of 6 months. Most (91%) of these subjects smoked cocaine in the form of crack and 9% reported intranasal use of powdered cocaine. None of these volunteers were currently in, or seeking, treatment for their substance use. The sample reported an average frequency of cocaine use of 3.8 days in the past week (SD= 1.9), and 16.0 days in the past month (SD= 10.1). The sample reported smoking a mean daily amount of 2.1 rocks of crack cocaine per day (SD= 2.0), and spending an average of $22.0 per day on cocaine (SD= 27.0). The majority of the sample also reported using nicotine (86%) and alcohol (82%). The mean total weekly alcohol consumption for this sample was 44.0 standard drinks (SD = 46.4). Over one half of the sample also reported some past month use of marijuana (59%). Two subjects reported using a benzodiazepine or a hallucinogen in the past month. No other drug use, including prescription medication, was reported by the sample. Volunteers were classified as controls if they met all of the following criteria: (1) a score of less than four on the DAST; (2) no self-reported current cocaine use; (3) a negative test for the presence of cocaine and benzoylecgonine in urine-analysis; and (4) no self-reported history of past habitual cocaine use. Habitual use was defined as at least four times within a month during a subject’s lifetime. These control subjects also reported nicotine use (36%) and alcohol use (68%). The mean total weekly alcohol consumption for this sample was 22.3 standard drinks (SD=38.9). Some controls also reported past month use of marijuana (18%). No other drug use, including prescription medication, was reported.
2.2. Apparatus Response inhibition and response execution were measured by a stop-signal paradigm using a choice
267
reaction time (RT) task that engages participants in responding to go-signals when stop-signals occasionally inform them to inhibit the response. The task was performed on a PC computer and was operated by Micro Experimental Laboratory Software (Psychology Software Tools Inc., Pittsburgh, PA). The ‘go’ stimuli for the choice response were four 1.5 cm letters (A, B, C, and D), presented one at a time in the center of a computer monitor. Letters were preceded by a 500 ms preparation interval in which a fixation point (c) appeared in the center of the computer screen. A letter was displayed for 500 ms, and the computer screen was blank for a 2.5 s inter-stimulus interval before the next letter was displayed. This provided a 3 s period in which the subject could respond to the letter. A subject responded to a letter by pressing one of two adjacent keys on the computer keyboard, using the index and middle fingers of the preferred hand. One key (the period key) was pressed to indicate that one of two letters (A or C) appeared, and the adjacent key (the forward slash key) was pressed to indicate that one of the other two letters (B or D) appeared. A single stop-signal test presented each of the four letter stimuli equally often, for a total of 176 letter presentations. A stop-signal occurred on 27% of the 176 presentations (i.e. 48 times) during a test. The stop-signal was a 500 ms 900 Hz tone generated by the computer at a comfortable listening level. Participants were required to withhold (i.e. inhibit) their response when a stop-signal was presented. Stop-signals were presented 12 times, at each of four delays (i.e. 50, 150, 250 and 350 ms after the onset of a letter). These delays were chosen based on previous research using this task in our lab that showed that the probability of inhibiting decreases in an orderly, linear fashion as the stop-signal delays increase from 50 to 350 ms (Fillmore et al., 2001; Fillmore and Vogel-Sprott, 1999, 2000). The order of letters, stop-signals, and delays was random. A test was completed in approximately 10 min. Response inhibition was determined by the number of times that an individual inhibited responses to stopsignal trials (out of 48), and by the estimated mean latency to inhibit a response (Section 2.4). Response execution to go-signals was measured by the mean RT to the go-signals during a test (i.e. the average time from the onset of letter presentation until a computer key press). Shorter RTs (i.e. faster responding) indicated greater response execution. Choice response errors to go-signals were also recorded. The inhibition and execution measures are highly reliable across trials (alpha coefficients \ 0.90) and stable over sessions (test-retest reliabilities \ 0.85) in both drug using and non-using populations (Fillmore and Rush, 2001; Fillmore and Vogel-Sprott, 1999; Mulvihill et al., 1997).
268
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
2.3. Procedure All participants were tested individually. All sessions were scheduled from Monday to Friday between 8:30 and 16:30 h. Potential volunteers responded to the posted advertisements by phoning the lab. The advertisements stated that adult volunteers were needed to perform behavioral tests and to report on their drug and alcohol use. A research assistant conducted a phone interview and an appointment was then made for the volunteer to attend a familiarization/assessment session at the laboratory. Subjects were instructed not to drink alcohol or take any drug during their scheduled lab days. During all sessions, participants provided a breath sample using an Intoxilyzer Model 400 (CMI Inc. Owensborough, KY) to ensure a zero blood alcohol concentration (BAC). A urine sample also was obtained to test for the presence of cocaine/benzoylecgonine, as well as recent use of benzodiazepines, barbiturates, tetrahydrocannabinol, amphetamine, and opiates. With the exception of cocaine/benzoylecgonine, no volunteers tested positive for the presence of any drug at the time of study. The familiarization/assessment session required approximately 2 h to complete and its purpose was to acquaint the volunteers to the laboratory and to obtain background information relevant to the study and group classification. Upon arriving for the familiarization/assessment session, volunteers were greeted by the research assistant, who explained the general details of the study protocol and obtained the volunteers’ written, signed informed consent. Participants then completed a packet of questionnaires regarding their current and past state of physical and mental health, current and past drug use, as well as demographic measures, including age and education level. The packet included: MiniMental Status Examination (MMSE) (Folstein et al., 1975), DAST (Skinner, 1982), Michigan Alcoholism Screening Test (MAST) (Selzer, 1971), Beck Depression Inventory (BDI) (Beck et al., 1961), Brief Psychiatric Rating Scale (Overall and Gorham, 1962), a general health questionnaire, and detailed drug use questionnaires. The packet also included an abridged form of the Wender Utah Rating Scale (WURS) which is a 25-item symptom rating scale that provides a retrospective measure of childhood symptoms characteristic of ADHD (Ward et al., 1993). After the packet was completed, volunteers were paid $20 for attending, and an appointment was scheduled for the test session. Attempts were made to schedule this session within 1 week of the familiarization/assessment session. The subjects performed the stop-signal task on the test session. Volunteers met the research assistant who escorted them to a quiet testing room that contained the computer that controlled the stop-signal task. The assistant explained the task requirements. Subjects were
instructed to respond to the letters (i.e. go-signals) as quickly and as accurately as possible and to try to withhold their response if they heard a tone sound. Participants then performed one familiarization (practice) test on the task. The research assistant remained in the room to observe the subjects and verify that they understood the requirements. After the familiarization, the research assistant answered any questions that the subjects’ had about the task requirements. Participants then performed the stop-signal test. The research assistant was blind with respect to the purpose of the experiment, and to the measures obtained by the task. The task data were automatically stored on computer disk, and were not available to the assistant. No performance feedback was provided to subjects. After the test, participants were debriefed and paid $20 for their participation in the session.
2.4. Dependent measures of stop-signal performance 2.4.1. Measures of response inhibition The ability to inhibit a response can be impaired in two important ways (Logan, 1994; Tannock et al., 1995): (1) the initiation of the inhibitory process might fail to occur occasionally, thus resulting in less stop-signal inhibitions; (2) The completion of the inhibitory process might be slowed, so that it takes longer to inhibit a response to a stop-signal. The stop-signal paradigm provides a method for assessing these two criterion measures of response inhibition. 2.4.1.1. Probability of inhibition (P-inhibition). The proportion of successfully inhibited responses on the 48 stop-signal trials for a test provided a general measure of the ability to inhibit responses for a subject. Smaller proportions indicated a lower probability of inhibiting a response to a stop-signal (poorer inhibitory control of behavior). 2.4.1.2. Stop-signal reaction time (SSRT). Response inhibition also depends on the time required to inhibit a response. Deficits in inhibitory control might arise because more time is required to inhibit a response. The estimated mean time (in ms) required for a participant to inhibit responses to stop-signals (SSRT) was also measured in the present research. The method for calculating SSRT was based on the subject’s probability of inhibiting responses to stop-signals and the distribution of RTs to go-signal trials. The calculations are explained in detail elsewhere (Logan, 1994), but the general logic of the method is outlined here. SSRT is the time needed to inhibit the pre-potent response once the stop-signal occurs. The estimate is based on a method of determining how long the stop-signal can be delayed after the go-signal before the subject can no longer inhibit the response. As a general observation, the time
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
required to inhibit is less than the time required to respond (Logan and Cowan, 1984). This is evident by the fact that individuals can successfully inhibit a response to a stop-signal that does not occur until some time after the go-signal has been presented. SSRTs are generally related to the number of successfully inhibited responses on a test, with shorter SSRTs associated with a greater number of inhibited responses. However, the SSRT also provides latency information concerning the speed of the inhibitory processes. Longer SSRTs, resulting from a general slowing of the inhibitory processes, indicate weak inhibitory control over behavior.
2.4.2. Measures of response execution The stop-signal task also provides measures of response execution. Responses displayed during go-signal trials (i.e. when no stop-signal occurs) yielded two important criterion measures of response execution: go-signal RT and go-signal accuracy. 2.4.2.1. Go-signal reaction time (RTgo). The strength of response execution to go-signals was measured by a participant’s mean RT (ms) to the 128 go-signal trials presented during a test (i.e. the average time from the onset of go-signals until a computer key press). This produced a mean RTgo score for a participant for each test. All analyses involving RTgo scores were performed on observed mean RTgo scores, and on the mean RTgo scores obtained after removing outlier trial RTs that were 2.5 SDs from the mean RTgo score on a test. On average, less than 2% of trials on a test had outlier RT values. Analyses based on actual RTgo scores and trimmed RTgo scores yielded identical conclusions. The Results section reports analyses based on trimmed RTgo scores. 2.4.2.2. Go-signal response accuracy. Choice response errors on go-signal trials were also recorded. These errors occur when a participant presses the incorrect key in response to a letter presentation during a go-signal trial. The errors generally reflect failures of attention or information processing. Choice response errors during a test are rare and typically occur on less than 5% of all go-signal responses (Fillmore and VogelSprott, 1999, 2000; Logan, 1994).
3. Results
3.1. Demographic and background characteristics Table 1 summarizes the demographic and drug use data for the cocaine using and control group and reports the results of t-tests that compared the groups. The table shows no significant difference in the groups’ age, education level, mental status (MMSE scores), or
269
Table 1 Mean demographic and background characteristics of subjects. Standard deviation (SD) is presented in parentheses. Independent t-tests compared cocaine users with controls Sample
Controls
Age BMI Education Income
40.5 (5.6) 28.8 (6.4) 12.4 (1.3) 14987.8 (8783.9) DAST 1.5 (1.1) MAST 6.6 (8.4) WURS 17.9 (13.2) BDI 6.3 (6.6) MMSE 27.7 (2.2) Weekly drinks 22.3 (38.9) Daily cigarettes 5.2 (10.0) Daily caffeine 523.7 (708.1)
Cocaine users
Significant
40.1 25.1 12.0 9050.0
(7.7) (3.9) (1.5) (5317.5)
ns ** ns *
(5.9) (12.0) (20.3) (8.3) (1.9) (46.4) (9.2) (38.6)
** ** ns ns ns ns ** ns
13.2 18.9 23.7 6.4 28.9 44.0 14.2 312.6
ns, Non-significant difference; BMI, body-mass index score; BDI, Beck Depression Inventory; MMSE, Mini-Mental Status Examination (maximum score = 30); DAST, Drug Abuse Screening Test; MAST, Michigan Alcoholism Screening Test; WURS, Wender Utah Rating Scale for childhood ADHD; income, annual income in dollars; weekly drinks, weekly number of standard alcoholic drinks consumed; daily cigarettes, number of cigarettes smoked per day; daily caffeine, mgs of caffeine consumed per day. * PB0.05. ** PB0.01.
WURS and BDI scores. Cocaine users had significantly higher DAST and MAST scores than control subjects. Cocaine users also reported greater daily use of cigarettes. With respect to weekly alcohol consumption, cocaine users reported a higher mean number of drinks than controls, but this difference was not statistically significant.
3.2. Stop-signal task performance Table 2 reports the stop-signal task measures for each group. The table presents mean scores for the two measures of response inhibition (P-inhibition and SSRT) and for the two measures of response execution Table 2 Mean stop-signal task performance measures of subjects. Standard deviation is in parentheses. Independent t-tests compared cocaine users with controls Sample
Controls
Cocaine users
Significant
Response inhibition measures P-inhibitions 0.69 (0.04) SSRT (ms) 286.4 (11.5)
0.56 (0.05) 349.6 (23.2)
* *
Response execution measures RTgo (ms) 590.6 (23.5) Choice errors 9.8 (1.4)
579.0 (23.8) 8.9 (1.3)
ns ns
ns, Non-significant difference. * PB0.05; **PB0.01.
270
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
each delay. The ANOVA revealed significant main effects of group (F1,42 = 4.3, P= 0.043), and stop-signal delay (F3,126 = 80.9, PB 0.001), but no significant group× stop-signal delay interaction, (F3,126 = 1.0, P= 0.410).
3.2.1.2. Stop-signal RT. Table 2 also shows that cocaine users displayed significantly longer SSRTs than did controls (t42 = 2.4, P= 0.020). A 2-group ANCOVA also obtained a significant difference in SSRT after controlling for subjects’ weekly alcohol use, (F1,41 =7.4, P= 0.009). 3.2.2. Response execution measures Fig. 1. Mean probability of inhibiting a response (P-inhibition) to 12 stop-signals at each of four stop-signal delays (50, 150, 250, 350 ms) for cocaine users and controls. Vertical capped lines indicate standard error of the mean (group n= 22).
(RTgo and choice errors). The table also reports the results of t-tests that compared the two groups on stop-signal task measures.
3.2.1. Response inhibition measures 3.2.1.1. Probability of inhibition (P-inhibition). Table 2 shows that cocaine users displayed a lower probability of inhibiting responses to stop-signals (i.e. P-inhibition scores) compared with controls (t42 =2.1, P =0.043). Given that cocaine users tended to have higher weekly levels of alcohol consumption, an analysis of covariance (ANCOVA) also tested the group difference in P-inhibition after removing any variance due to individual differences in weekly alcohol use. The 2-group analysis of covariance (ANCOVA) also revealed a significant group difference in P-inhibition scores (F1,41 = 5.2, P= 0.028). Thus, the lower probability of inhibiting displayed by cocaine users was not related to their weekly alcohol use. To further explore the group difference in P-inhibition scores, subjects’ scores were plotted as a function of stop-signal delay. Research shows that the probability of inhibiting a response diminishes as the stop-signal delay increases because less time is available to inhibit the pre-potent response (Logan, 1994). Fig. 1 shows the mean probability of inhibiting a response at each of the four stop-signal delays for the two groups. As expected, the probability of inhibiting diminished as the stop-signal delay increased. This pattern was evident in both groups. Moreover, Fig. 1 shows that cocaine users displayed a consistently lower P-inhibition than controls at all stop-signal delays. These observations were confirmed by a 2 (group)×4 (stop-signal delay) mixed analysis of variance of subjects’ P-inhibition scores at
3.2.2.1. RT to go-signals (RTgo). Table 2 presents the results of an independent t-test that compared the RTgo scores of the two groups. No significant difference was obtained. Fig. 2 plots the mean RTgo scores for each group along with the groups’ mean SSRT scores. The figure shows that the mean RTgo score of controls and cocaine users was very similar. By contrast, the SSRT score of cocaine users was longer than that displayed by controls. Together, the two measures of reaction time show that RT differences between controls and cocaine users were specific to the inhibition of responses rather than to their execution. Fig. 2 also provides information regarding the general difference in time required to inhibit a response versus execute the response. The figure shows that the mean SSRT of controls is approximately one half of their mean RTgo (286.4 vs 590.6 ms). This RT difference is consistent with parametric studies of stop-signal performance which found that inhibiting a response requires
Fig. 2. Mean SSRT (response inhibition) and mean RTgo (response execution) for cocaine users and controls. Vertical capped lines atop bars indicate standard error of the mean (group n = 22).
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
about one half of the time needed to execute a response (Logan, 1994).
3.2.2.2. Choice errors. Table 2 presents the results of an independent t-test that compared the number of choice response errors displayed by the two groups. No significant difference was obtained. The mean number of errors was similar in the two groups. Based on 128 go-signal trials these means represented a high accuracy rate (above 90%).
4. Discussion This study used a stop-signal paradigm to compare the inhibitory control of cocaine users to an agedmatched sample of individuals with no history of habitual cocaine use. The study showed that cocaine users displayed a poorer ability to inhibit a response than did controls. Compared with controls, cocaine users displayed a lower probability of inhibiting responses, and required more time to inhibit responses as estimated by their stop-signal reaction time (i.e. SSRT). By contrast, cocaine users did not differ from controls in their ability to execute responses, as measured by their speed and accuracy of responding to go-signals. The reaction time and response accuracy scores in both groups were comparable to those obtained in other studies of healthy volunteers that have used the stop-signal task (Fillmore et al., 2001; Fillmore and Vogel-Sprott, 1999, 2000). The study also showed that response inhibition diminished as a function of increasing the stop-signal delay. Fig. 1 shows the negative slope functions that relate the probability of inhibiting a response to the stop-signal delays. The slopes are consistent with previous studies showing that stop-signal latencies affect the probability of response inhibition (Fillmore and VogelSprott, 1999; Logan and Cowan, 1984). Moreover, the presence of these slopes provides verification that subjects in both groups understood the task requirements and followed instructions. In studies of other populations, such as young children, task performance can be affected by inattention and by random response strategies, owing to a lack of motivation or interest on the part of the subject (Schachar et al., 1995; Tannock et al., 1995). Such response styles are detected by the slope function. Randomly inhibiting and executing responses generates a flat slope function because inhibitions are equally likely to occur at all stop-signal delays under such a strategy. This was not observed in the present study. Rather, the negative slopes demonstrate that response inhibition in both groups was under some degree of stimulus control of the stop signals. The lower probability of response inhibition among cocaine users could not be attributed to faster response
271
execution because both groups displayed similar mean reaction times to go signals. The lower probability might be due to a general slowing of the inhibitory process, so that more time is required to inhibit a response. This time requirement was estimated by the SSRT measure which showed that cocaine users displayed longer estimated latencies to inhibit responses than did controls. The mean SSRT of cocaine users was approximately 350 ms, whereas controls displayed a mean SSRT of 286 ms. SSRT estimates of healthy adults typically range between 200 and 300 ms (Logan, 1994), and estimates above 300 ms are usually observed in special populations, such as children with ADHD (Schachar et al., 1995; Tannock et al., 1995). Although statistically significant, performance differences in the order of milliseconds might raise concern regarding their relevance to more macro-level behaviors, such as drug abuse. However, many fundamental cognitive and perceptual processes, such as inhibitory influences, are considered to operate in a ‘bottom-up’ fashion to exert increasing influence at each stage of higher-order attentional and cognitive functions (McClelland and Rumelhart, 1981). Thus, although cocaine users exhibited a SSRT slowing of approximately 65 ms compared with controls, such a deficiency might exert considerable influence on higher-order cognitive functions, such as impulse-control. Moreover, the similarity between cocaine users and individuals with ADHD in terms of prolonged SSRT estimates could point to a common deficit in both populations. Disorders characterized by deficits of self-control and attention are prevalent among cocaine users (O’Malley et al., 1992). Sensation seeking, impulsivity, antisocial personality, and childhood disorders of self-control, such as conduct disorder and ADHD, are all considered to pose a risk for the development of cocaine use and other drug use (Horner et al., 1996; Ratey et al., 1992). From a clinical perspective, cocaine use is seen as part of a general disinhibitory psychopathology that involves a constellation of traits (Strickland and Stein, 1995). The common phenotypic characteristic among these traits is a deficit in behavioral inhibition. Deficits in inhibitory control could also directly contribute to instances of heavy cocaine and other drug use by compromising the ability to stop repeated drug administrations in the situation. The present findings are new and raise many questions about how such cognitive deficits in inhibitory control could relate to drug abuse. The poorer inhibitory control displayed by cocaine users also could represent a direct effect of recent cocaine administration. These individuals reported using the drug on a weekly basis and tested positive for cocaine in urineanalysis during their initial laboratory assessment session. However, studies of other stimulants, such as, methylphenidate, amphetamine, and caffeine, have reported that these drugs increase (rather than decrease)
272
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273
response inhibition on the stop-signal task (Feola et al., 2000; Fillmore and Vogel-Sprott, 1999; Tannock et al., 1995). As yet, the direct effects of acute cocaine administration on response inhibition await to be explored. Nonetheless, the findings obtained in the present research are important because they identify a specific behavioral deficit that could contribute to cocaine use, and explain its association with other disorders of self-regulation, such as ADHD. However, its etiological role in cocaine use remains unclear, and it is uncertain as to whether cocaine use contributes to inhibitory deficits, or if they are part of a pre-existing constellation of ADHD-like symptoms that operate to increase risk for cocaine and other drug use.
Acknowledgements This research was supported by Grants DA14079 and DA10325 from the National Institute on Drug Abuse. Offprint requests should be sent to Mark T. Fillmore, Ph.D., Department of Psychology, University of Kentucky, Lexington, KY 40506-0044. Email: [email protected]
References Ardila, A., Rosselli, M., Strumwasser, S., 1991. Neuropsychological deficits in chronic cocaine abusers. Int. J. Neurosci. 57, 73 – 79. Barkely, R.A., 1997. Behavioral inhibitions, sustained attention, and executive functions: constructing a unifying theory of ADHD. Psychol. Bull. 121, 65 –94. Bauer, L.O., 2001. Antisocial personality disorder and cocaine dependence: their effects on behavioral and electoencephalographic measures of time estimation. Drug Alcohol Depend. 63, 87 – 95. Beatty, W.W., Katzung, V.M., Moreland, V.J., Nixon, S.J., 1995. Neuropsychological performance of recently abstinent alcoholics and cocaine abusers. Drug Alcohol Depend. 37, 247 –253. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E., Erbaugh, J.K., 1961. An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561 – 571. Biggins, C.A., MacKay, S., Clark, W., Fein, G., 1997. Event-related potential evidence for frontal cortex effects of chronic cocaine dependence. Soc. Biol. Psychiatry 42, 472 –485. Boutros, N.N., Uretsky, N., Bernston, G., Bornstein, R., 1994. Effects of cocaine on sensory inhibition in rats: preliminary data. Soc. Biol. Psychiatry 36, 242 –248. Boutros, N.N., Uretsky, N., Lui, J.J., Millana, R.B., 1997. Effects of repeated cocaine administration on sensory inhibition in rats: preliminary data. Soc. Biol. Psychiatry 41, 461 –466. Caine, S.B., 1998. Neuroanatomical basis of the reinforcing stimulus effects of cocaine. In: Higgins, S.T., Katz, J.L. (Eds.), Cocaine Abuse: Behavior, Pharmacology and Clinical Applications. Academic Press, New York, pp. 21 – 50. Feola, T.W., de Wit, H., Richards, J.B., 2000. Effects of d-amphetamine and alcohol on a measurof behavioral inhibition in rats. Behav. Neurosci. 114, 838 –848. Fillmore, M.T., Rush, C.R., 2001. The effects of oral cocaine on inhibitory control in humans. Manuscript in preparation.
Fillmore, M.T., Rush, C.R., Kelly, T.H., Hays, L., 2001. Triazolam impairs inhibitory control of behavior in humans. Exp. Clin. Psychopharmacol. 9, 363 – 371. Fillmore, M.T., Vogel-Sprott, M., 1999. An alcohol model of impaired inhibitory control and its treatment in humans. Exp. Clin. Psychopharmacol. 7, 49 – 55. Fillmore, M.T., Vogel-Sprott, M., 2000. Response inhibition under alcohol: effects of cognitive and motivational conflict. J. Stud. Alcohol 61, 239 – 246. Folstein, M., Folstein, S., Hugh, P., 1975. Mini-mental state — a practical method for grading the cognitive state of patients for the clinician. J. Psychiat. Res. 12, 189 – 198. Horner, B., Scheibe, K., Stine, S., 1996. Cocaine abuse and attentiondeficit hyperactivity disorder: implications of adult symptomatology. Psych. Addict. Behav. 10, 55 – 60. Jentsch, J.D., Taylor, J.R., 1999. Impulsivity resulting from frontostriatal dysfunction in drug abuse: implication for the control of behavior by reward-related stimuli. Psychopharmacology 146, 373 – 390. Lane, S.D., Cherek, D.R., Dougherty, D.M., Moeller, F.G., 1998. Laboratory measurement of adaptive behavior change in humans with a history of substance dependence. Drug Alcohol Depend. 51, 239 – 252. Levin, F.R, Evans, S.M., Kleber, H.D., 1998. Prevalence of adult attention-deficit hyperactivity disorder among cocaine abusers seeking treatment. Drug Alcohol Depend. 52, 15 – 25. Logan, G.D., 1994. On the ability to inhibit thought and action: a users’ guide to the stop-signal paradigm. In: Dagenbach, D., Carr, T.H. (Eds.), Inhibitory Processes in Attention, Memory, and Language. Academic Press, San Diego, CA, pp. 189 – 239. Logan, G.D., Cowan, W.B., 1984. On the ability to inhibit thought and action: a theory of an act of control. Psychol. Rev. 91, 295 – 327. Logan, G.D., Cowan, W.B., Davis, K.A., 1984. On the ability to inhibit simple and choice reaction time responses: a model and a method. J. Exp. Psychol: Hum. Percept. Perform. 10, 276 –291. Lyvers, M., 2000. ‘Loss of control’ in alcoholism and drug addiction: a neuroscientific interpretation. Exp. Clin. Psychopharmacol. 8, 225 – 249. MacLeod, C.M., 1991. Half a century of research on the Stroop effect: an integrative review. Psychol. Bull. 109, 163 – 203. McClelland, J.L., Rumelhart, D.E., 1981. An interactive activational model of context effects in letter perception: part 1. An account of basic findings. Psych. Rev. 88, 375 – 407. Mulvihill, L.E., Skilling, T.A., Vogel-Sprott, M., 1997. Alcohol and the ability to inhibit behavior in men and women. J. Stud. Alcohol 58, 600 – 605. O’Malley, S., Adamse, M., Heaton, R.K., Gawin, F.H., 1992. Neuropsychological impairment in chronic cocaine abusers. Am. J. Drug Alcohol Abuse 18, 131 – 144. Oosterlaan, J., Sergeant, J.A., 1996. Inhibition in ADHD, aggressive, and anxious children: a biologically based model of child psychopathology. J. Abnorm. Child Psychol. 24, 19 – 37. Overall, J.E., Gorham, D.R., 1962. Brief psychiatric rating scale. Psychol. Res. 10, 79 – 812. Posner, M.I., Cohen, Y., 1984. Components of visual orienting. In: Bouma, H., Bouwhuis, D. (Eds.), Attention and Performance X. Lawrance Erlbaum, Hillsdale, NJ. Quay, H.C., 1997. Inhibition and attention deficit hyperactivity disorder. J. Abnorm. Child Psychol. 25, 7 – 13. Ratey, J.J., Greenberg, M.S., Bemporad, J.R., Lindem, K.J., 1992. Unrecognized attention-deficit hyperactivity disorder in adults presenting for outpatient psychotherapy. J. Child. Adoles. Psychopharmacol. 2, 267 – 275. Schachar, R.J., Logan, G.D., 1990. Impulsivity and inhibitory control in normal development and childhood psychopathology. Develop. Psychol. 26, 710 – 720.
M.T. Fillmore, C.R. Rush / Drug and Alcohol Dependence 66 (2002) 265–273 Schachar, R., Tannock, R., Marriott, M., 1995. Deficient inhibitory control in attention deficit hyperactivity disorder. J. Abnorm. Child Psychol. 23, 411 – 437. Selzer, M.L., 1971. The Michigan alcoholism screening test: the quest for a new diagnostic instrument. Am. J. Psychiatry 127, 1653 – 1658. Skinner, H.A., 1982. The drug abuse screening test. Addict. Behav. 7, 363 – 371. Strickland, T.L., Stein, R., 1995. Cocaine-induced cerebrovascular impairment: challenges to neuropsychological assessment. Neuropsychol. Rev. 5, 69 –79. Tannock, R., Schachar, R., Logan, G., 1995. Methylphenidate and cognitive flexibility: dissociated dose effects in hyperactive children. J. Abnorm. Child Psychol. 23, 235 –267.
273
Tannock, R., 1998. Attention deficit hyperactivity disorder: Advances in cognitive, neurobiological, and genetic research. J. Child Psychol. Psychiatry. 39, 65 – 99. Tipper, S.P., 1985. The negative priming effect: inhibitory effects of ignored primes. Quart. J. Exp. Psychol. 37A, 571 – 590. Volkow, N.D., Ding, Y., Fowler, J.S., Wang, G., 1996. Cocaine addiction: hypothesis derived from imaging studies with PET. J. Addict. Dis. 15, 55 – 71. Ward, M.F., Wender, P.H., Reimherr, F.W., 1993. The Wender Utah rating scale: an aid in the retrospective diagnosis of childhood attention deficit hyperactivity disorder. Am. J. Psychiatry 150, 885 – 890.