Abstract

The Iowa gambling task (IGT) was designed to assess clinically relevant decision-making impairment, yet some studies find high rates of failure in otherwise healthy control groups. The current study examined variables potentially related to IGT failure, including negative affect, intellect, personality, and executive functioning, in a well-screened sample of healthy young adults. In addition, cerebral oxygenation (near-infrared spectroscopy) was assessed. Results indicated that those who failed the IGT had lower estimated intellect, made more commission errors on the 2-back task, and showed less bilateral dorsolateral prefrontal cortex oxygenation, relative to those who passed. Overall findings are consistent with prior literature suggesting that frontal lobe functioning is related to successful IGT performance and that executive functioning and working memory skills are important components of IGT performance, even in those without clinical disorders or evidence of “real world” dysfunction.

Introduction

The Iowa gambling task (IGT) is a measure of risky decision making that, according to its clinical manual, is designed to support diagnosis of brain dysfunction and to assess clinically relevant decision-making impairment (Bechara, 2007). Patients with neurological damage to the ventromedial prefrontal cortex show deficits on the IGT, as do patients with dorsolateral prefrontal cortex damage, particularly in the right hemisphere (Bechara, 2007; Buelow & Suhr, 2009). The IGT has also been used to detect impaired risky decision making in a variety of clinical populations where there is evidence for real-world decision-making and/or executive function difficulties, including substance abuse, pathological gambling, obsessive compulsive disorder, schizophrenia, and psychopathy (Bechara, 2007; Buelow & Suhr, 2009).

With the recent publication of clinical norms for the IGT (Bechara, 2007), it is likely that clinicians will use the IGT in neuropsychological evaluation of their patients, and their interpretation of the findings will likely be driven by these clinical data. However, what does “failure” based on these recently published norms really mean? Early work on the IGT demonstrated that a significant minority (∼30%) of “normal” controls perform poorly (Bechara et al., 2001; Bechara & Damasio, 2002). Reasons for IGT failure in healthy individuals are important to consider if one plans to use the IGT as a measure of clinically relevant risky decision making, to make diagnostic decisions, or to make conclusions about the presence of brain damage/dysfunction in an individual being assessed. There are several possible contributors to IGT failure in a “normal” population. One possible contributor is the presence of personality characteristics associated with risky behavior. Bechara, Tranel, and Damasio (2000) suggested that nonclinical individuals who fail the IGT may have a personality style in which they minimize their attention to future consequences, whether positive or negative. Buelow and Suhr (2009) reviewed the literature supporting this contention and found that riskier IGT performance is related to higher levels of sensation seeking, disinhibition, and impulsivity in nonclinical samples. However, these personality characteristics have been less consistently associated with IGT performance in clinical samples; for example, pathological gamblers (Brand et al., 2005) and binge drinkers (Goudriann, Grekin, & Sher, 2007).

Another potential contributor to IGT failure is negative affect at the time of task completion, which has been related to riskier decision making in a multitude of risky decision paradigms (seeBuelow & Suhr, 2009, for a review). Must and colleagues (2006) found that patients with clinically diagnosed major depressive disorder were impaired on the IGT. In addition, Suhr and Tsanadis (2007) showed that riskier IGT performance was related to higher negative mood, as measured by the Positive and Negative Affect Scale, in a nonclinical population.

It is also possible that intellectual level is related to IGT performance. Most studies use education as a proxy measure of intellectual ability, but findings have been decidedly mixed. Although one study showed higher education was related to worse IGT performance (Evans, Kemish, & Turnbull, 2004), some showed that lower education was related to worse performance (Davis et al., 2008; Denburg, Recknor, Bechara, & Tranel, 2006; Dom, De Wilde, Hulstijn, van den Brink, & Sabbe, 2006; Fein, McGillivray, & Finn, 2007), and others found no relationship between education and IGT performance (Barry & Petry, 2008; Bechara, 2007; Boeka & Lokken, 2006; Cavedini, Riboldi, Keller, D'Annucci, & Bellodi, 2002; Jollant et al., 2007; Lawrence et al., 2006; Verdejo-Garcia, Bechara, Recknoir, & Perez-Garcia, 2006). Studies that have actually assessed or estimated intellectual level rather than using education level as a proxy measure have shown a relationship between higher IQ and better IGT performance (Barry & Petry, 2008; Monterosso, Ehrman, Napier, O'Brien, & Childress, 2001), although there are exceptions (Brand, Recknor, Grabenhorst, & Bechara, 2007).

Another potential contributor to IGT impairment in healthy individuals is the presence of relative weaknesses in executive functioning skills. Of note, studies with clinical populations have tended to find no relationship between IGT performance and performance on the Wisconsin Card Sorting Test (Bechara et al., 2001; Grant, Contoreggi, & London, 2000; Ritter, Meador-Woodruff, & Dalack, 2004; Rotherham-Fuller, Shoptaw, Berman, & London, 2004), although patients with frontal lesions who fail the IGT also perform poorly on working memory tasks (Bechara, Damasio, Tranel, & Anderson, 1998; Manes et al., 2002). Patients with amygdala lesions who fail the IGT also have been shown to perform poorly on a modified card sorting task, the Stroop Color and Word Test, verbal fluency tasks, and the Trail Making Test (Brand, Grabenhorst, Starcke, Vandekerckhove, & Markowitsch, 2007). Few data examining the relationship of executive functioning to IGT performance are available from nonclinical populations. Overman and colleagues (2004) found that IGT scores were not related to Wisconsin Card Sorting Test performance in a nonclinical sample. However, Brand, Recknor, and colleagues (2007) examined a larger and more diverse nonclinical sample and found that IGT performance was related to both perseverative and nonperseverative errors on the Wisconsin Card Sorting Test. This relationship was even stronger when analyses were focused on IGT performance in later quintiles, which appears to be an IGT variable more strongly associated with risky decision making (Brand, Recknor, et al., 2007).

The relationship of IGT performance to frontal lobe dysfunction was first supported by lesion studies (Damasio, 1994), but later gained inconsistent support via functional neuroimaging findings in other clinical populations (Bolla et al., 2003; Tanabe et al., 2007; Tucker et al., 2004). Although there are few studies, functional neuroimaging findings in nonclinical samples also suggest that IGT performance is associated with activation of frontal networks. In an early PET study with healthy controls (Ernst et al., 2002), a predominantly right-sided pattern of activation was observed during IGT performance relative to baseline, including in the orbitofrontal cortex, dorsolateral prefrontal cortex, anterior cingulate, inferior parietal lobe, thalamus, and anterior insula. In addition, correlation analyses showed better IGT performance was associated with more activation in right ventrolateral prefrontal cortex, anterior insula, and head of caudate. In an event-related fMRI study (Fukui, Murai, Fukuyama, Hayashi, & Hanakawa, 2005), only healthy controls who could consciously identify which IGT decks were advantageous and which were not at the end of the task were included in their analyses. Neural activity during selections from advantageous (“nonrisky”) decks was compared with neural selections during disadvantageous (“risky”) decks. Anticipation of plays from risky decks was related to increased activation in the superior anterior cingulate and medial frontal gyrus. Furthermore, medial frontal gyrus activation was correlated with net scores on risky plays but not on nonrisky plays. However, by virtue of participant selection methods, it is likely that their sample did not include people significantly impaired on the IGT, and thus these findings do not necessarily speak to the nature of brain activation related to IGT failure in otherwise normal individuals. In another nonclinical sample, healthy men showed a general activation of bilateral ventromedial prefrontal cortex during completion of a modified IGT task relative to baseline, and activation of the precentral gyrus was positively correlated with a higher net score (Lawrence et al., 2006).

In summary, although there are personality, mood, cognitive, and neuroimaging correlates of IGT performance in nonclinical populations, it is unclear from existing findings whether these factors would be associated with clinically impaired IGT performance in otherwise healthy individuals. The present study sought to further examine variables potentially related to impaired IGT performance in a healthy sample of well-screened young adults. The first hypothesis was that individuals who failed the IGT, based on clinical norms, would have more sensation-seeking personality characteristics than those who passed the IGT. The second hypothesis was that individuals who failed the IGT would perform worse on measures of executive functioning and working memory relative to those who passed the IGT. The third hypothesis was that individuals who failed the IGT would show less cortical activation during completion of neuropsychological tasks, particularly in right dorsolateral prefrontal cortex, a pattern consistent with findings in healthy controls and in some clinical samples. In testing these hypotheses, we sought to control for variables associated with IGT performance in prior research, including affect at the time of the task and estimated intellect.

Materials and Methods

Participants

Participants were 58 carefully screened young adults (23 men, 53 right-handed, >90% Caucasian, age range 18–23 years old) recruited from a large undergraduate research pool in the Department of Psychology at a medium-sized Midwestern university who participated in a much larger study of polysubstance abuse (Hammers & Suhr, in press). Participants in the present study included the 23 impulsive controls who were matched to the polysubstance users for analysis in that study, plus an additional 35 controls who had been run through the study protocol but were not used in that study's analyses. The study was approved by the university's Institutional Review Board. Participants in the present analyses all denied a history of loss of consciousness for >1 hr or post-traumatic amnesia for >15 min, diagnosis of Learning Disorder or Attention Deficit/Hyperactivity Disorder, other neurological history (seizures, brain tumor), and current non-substance-related psychological diagnosis or treatment. In addition, all participants were carefully screened with regard to substance use. No participants scored as “at risk” on the Substance Abuse Subtle Screening Inventory-3, and all denied use of any drugs except alcohol and marijuana. In the total sample, 40 participants indicated that they never smoked marijuana and 18 reported only “rarely” using marijuana (one to two times during their lifetime). Only two individuals in the sample smoked cigarettes. On average, participants reported that they drank three alcoholic drinks/week (with a range from 0 to 8 drinks/week). Participants reported that, on average, they started drinking at 16.9 years of age. Finally, all participants had blood-alcohol levels of less than 0.02% at the time of testing, based on acute screening.

Measures

Screening Measures

The “Drinking and Drug Habits Questionnaire” is a modified version of the Daily Drinking Questionnaire (Collins, Parks, & Marlatt, 1985) and was used to assess alcohol and substance use. Individuals for the present study could not endorse the use of any substances other than alcohol and marijuana and could only endorse “rare” marijuana use. The Daily Drinking Questionnaire has been used extensively in research as a measure of problematic drinking in both community and undergraduate populations (Collins et al., 1985; Larimer, Turner, Mallett, & Geisner, 2004; Lewis & Neighbors, 2004; McNally, Palfai, & Kahler, 2005; Murphy, Corriea, Colby, & Vuchinich, 2005; Testa & Livingston, 2000). The “Substance Abuse Subtle Screening Inventory-3” (Miller & Lazowski, 1999) was used to assess for substance dependence disorder risk based on gender. The total score on the instrument generates a categorical “at risk” or “not at risk” judgment; no participants in the study could score as “at risk” for substance abuse. Evidence from both community and undergraduate samples suggests that the measure has adequate psychometric properties (Clements, 2002; Laux, Perera-Diltz, Smirnoff, & Salyers, 2005; Miller & Lazowski, 1999). The “Alcohawk Precision” (Q3 Innovations, LLC, 2005) brand digital alcohol detector was used to measure the quantity of alcohol consumption at the time of the study. The device provides a digital reading of the participant's approximate blood-alcohol concentration, accurate within 0.01%, with a range 0.000%–0.040% blood-alcohol concentration.

Personality and affect measures

The Behavioral Activation Scale (BAS) from the “Behavioral Inhibition/BAS” (Carver & White, 1994) was used to assess impulsive sensation seeking and reward driven personality characteristics. There are three subscales of this measure: Reward Responsiveness, Reward Drive, and Fun Seeking. Prior data suggest that the Fun-Seeking Scale best taps the impulsive sensation-seeking construct most consistently associated with IGT impairment (Suhr & Tsanadis, 2007) and was the dependent variable of interest for hypothesis testing in the present study. The Fun-Seeking subscale has been shown to be internally reliable (Jorm et al., 1999), temporally stable (Carver & White, 1994; Heubeck, Wilkinson, & Cologon, 1998), and, of the three subscales, most strongly related to extraversion, hypomania, disinhibition, novelty seeking, risk taking, and substance use/abuse behavior (Carver, 2004; Carver & White, 1994; Heubeck et al., 1998; Smillie, Jackson, & Dalgleish, 2006).

The “Positive and Negative Affect Schedule” (Watson, Clark, & Tellegen, 1988) was used to assess participants' state mood. The dependent variable for the present study was state negative mood. This subscale has shown good internal consistency and convergent validity with other measures of negative affect (Watson et al., 1988).

Estimation of intellect

The “Wechsler Abbreviated Scale of Intelligence Matrix Reasoning subtest” (Wechsler, 1999) was used as an estimate of nonverbal intellect. The dependent variable for the present study was the age-corrected T-score. Data suggest the Matrix Reasoning subtest correlated .66 with WAIS-III Full Scale IQ in the standardization sample (Wechsler, 1999). Participants also provided self-reported high-school GPA and ACT scores for further estimation of their general cognitive abilities.

Neuropsychological measures

The computerized “IGT” (Bechara, 2007) was administered. For the purposes of the present study, which was focused on IGT impairment, an overall net score falling at or below T = 39 was considered impaired, as per recommendations in the clinical manual (Bechara, 2007). The “Wisconsin Card Sorting Test” computerized 64-card version (Heaton, 2005) was administered as one test of executive functioning. For the current study, the dependent measures of interest included age-corrected T-scores for number of perseverative errors and number of nonperseverative errors. A computerized “N-back test” assessed working memory. The present version of the task was developed specifically for this study and modeled after a standard paradigm (Bliss & Hamalanian, 2005). For each condition, participants read a series of instructions and then completed a practice trial. Following the practice, participants completed two separate conditions. In each condition, 64 letters were presented for 500 ms each, with an interstimulus interval of 2500 ms. During the 0-back condition, participants were instructed to remember a target letter (i.e., X) and click a mouse button each time they saw the target letter appearing in the sequence of letters. In the 2-back condition, participants were instructed to press the mouse any time they saw a letter that was identical to one presented two-trials back. For both trials, the participants' errors of omission and commission were recorded, as well as the reaction time for each selection. The dependent variables for the present study were omission and commission errors in the 2-back condition.

Cerebral oxygenation measure

Two-channel near-infrared spectroscopy (NIRS) was used to record cerebral oxygenation levels during the neuropsychological measures. The INVOS 5100 Cerebral Oximeter (Somanetics Corporation, Troy, MI, USA) utilizes two alternately illuminated light-emitting diodes (LEDs) at wavelengths of light centered at 730 and 810 nm, in the form of two single-use, disposable sensors; each sensor incorporates the LEDs and two light-collecting optodes at fixed distances from the emitter, 3 and 4 cm, respectively (Thavasothy, Broadhead, Elwell, Peters, & Smith, 2002). The sensors were created to be specifically positioned on the forehead above the eyebrows, approximately 5–6 cm equally distant from the midline (Kwee & Nakada, 2003; Thavasothy et al., 2002). As such, blood flow through the first few millimeters of the dorsal prefrontal cortex, approximately 1 cm laterally to either site of the photodetector, should be measured. However, as head size varies, the exact localization of the tissue being assessed cannot be precisely determined. The INVOS 5100 Cerebral Oximeter assumes a strict ratio of 3:1 for venus:arterial blood, based on the studies of cerebral circulation (Henson, Cartwright, & Chlebowski, 1997) and therefore uses this ratio when calculating values for regional blood oxygenation saturation (rSO2). Light absorption data were collected 15 times per second and a new average rSO2 value was calculated every 3–4 s. The NIRS-dependent variables for the current study were right and left frontal lobe change scores during performance of each task (average resting rSO2 values during a 5-min pre-task baseline subtracted from average rSO2 values during task performance), summed across tasks separately for left and right hemispheres. In addition, we examined left and right hemisphere activation for just the IGT. NIRS has been used in several cognitive activation studies with a variety of clinical populations, and sample data suggest concordance between functional MRI and NIRS findings in a variety of cognitive tasks (seeIrani, Platek, Bunce, Ruocco, & Chute, 2007, for a review).

Procedure

Following informed consent, participants provided demographic information and completed breathalyzer testing. They then completed substance use and personality questionnaires and were administered the Matrix Reasoning test. Then participants were connected to the NIRS. The two self-adhesive sensors were placed above the participant's eyebrows and a headband was placed over the electrodes to block out excess light and ensure a comfortable fit. Participants were instructed to sit quietly for a 5-min baseline. Then each neuropsychological test was administered, with the IGT first, followed by two experimental tasks not of interest in the present analyses, the WCST, and finally the N-back task. There was a rest of 5 min between each task. Following completion of the cognitive tasks, participants were disconnected from the NIRS, debriefed, and compensated for their participation in the form of extra credit points in an introductory psychology course.

Data Analyses

First participants were divided into two groups, based on clinical failure on IGT, as defined above. Descriptive data on IGT performance in both groups were provided to better characterize the overall performance of both groups. The two groups were then compared on dependent variables using t-tests or, where appropriate, the analysis of covariance to test the study hypotheses.

Results

Failures on the IGT

Of the 58 participants, one was excluded because the analysis of their response patterns showed that they picked almost exclusively from one deck during the 100 trials, rendering their response pattern uninterpretable. This individual was excluded from all other analyses. The average IGT net total T-score across the entire sample was 45.5 (SD = 9.6, range 25–62). Among the 57 participants, 15 (26%) failed the IGT according to age- and education-based norms (Bechara, 2007). Examination of the T-scores for advantageous–disadvantageous deck plays separated by quintile for the 15 individuals who failed the IGT showed that all played in a risky fashion across the entire task.

Given that the published norms for the IGT also include patterns of performance and norms per quintile, we also examined whether performance patterns among the remaining participants (n = 42), who passed the IGT based on the total net score, would suggest IGT impairment in any way (e.g., engaging in more risky plays but only at the end of the task). There were 12 participants who failed at least one quintile of the IGT, according to the published norms. For 6 of these 12, the failure occurred on either the first or the second quintile, consistent with learning under ambiguity, and was thus not suggestive of a risky decision-making style (Brand, Grabenhorst, et al., 2007). For four other participants, the fourth or fifth quintile was the only quintile failed, and the final two participants failed both the second and the last quintile. All other “passers” passed every quintile of the IGT.

Those who failed the IGT were not significantly different from those who passed in age, t(55) = −1.60, p = .11, educational level, t(55) = −0.61, p = .55, reported high school GPA, t(55) = −0.12, p = .91, reported composite ACT score, t(53) = −0.05, p = .96, or gender (43% of men in the pass group, 27% in the fail group), χ2(1) = 1.22, p = .27. The two groups were also not different in negative mood, t(55) = 0.50, p = .62. However, those who failed the IGT performed significantly lower than those who passed in the Matrix Reasoning test, t(55) = −2.37, p = .02 (Table 1). Therefore, only estimated intellect will be controlled for in tests of study hypotheses.

Table 1.

Group Comparisons on demographic, mood, intellect, personality, and neuropsychological variables

Variable Failures (N = 15) (mean [SD], range) Passers (N = 42) (mean [SD], range) Effect size (Cohen's d
Age 18.5 (0.5), 18–20 19.1 (1.3), 18–23 .61 
Education 12.3 (0.6), 11–15 12.5 (0.8), 11–15 .28 
Self-reported high school GPA 3.5 (0.4), 2.9–4.0 3.5 (0.4), 2.8–4.0 .00 
Self-reported composite ACT score 24.0 (3.6), 18–30 24.1 (3.7), 19–32 −.03 
PANAS Negative Mood 13.8 (3.7), 10–21 13.4 (2.6), 10–21 .13 
Matrix Reasoning T-score* 49.3 (5.3), 37–57 53.7 (6.5), 39–65 −.74 
Fun Seeking 11.4 (2.3), 6–14 10.3 (2.3), 6–16 .48 
WCST Perseverative Errors T-score 52.8 (13.8), 20–65 52.5 (8.3), 31–65 .03 
WCST Nonperseverative Errors T-score 50.3 (12.1), 28–69 50.4 (11.1), 20–69 .01 
2-back commission errors** 2.0 (2.2), 0–7 0.5 (0.9), 0–4 .89 
2-back omission errors 0.7 (0.9), 0–3 0.6 (0.8), 0.4 .12 
Variable Failures (N = 15) (mean [SD], range) Passers (N = 42) (mean [SD], range) Effect size (Cohen's d
Age 18.5 (0.5), 18–20 19.1 (1.3), 18–23 .61 
Education 12.3 (0.6), 11–15 12.5 (0.8), 11–15 .28 
Self-reported high school GPA 3.5 (0.4), 2.9–4.0 3.5 (0.4), 2.8–4.0 .00 
Self-reported composite ACT score 24.0 (3.6), 18–30 24.1 (3.7), 19–32 −.03 
PANAS Negative Mood 13.8 (3.7), 10–21 13.4 (2.6), 10–21 .13 
Matrix Reasoning T-score* 49.3 (5.3), 37–57 53.7 (6.5), 39–65 −.74 
Fun Seeking 11.4 (2.3), 6–14 10.3 (2.3), 6–16 .48 
WCST Perseverative Errors T-score 52.8 (13.8), 20–65 52.5 (8.3), 31–65 .03 
WCST Nonperseverative Errors T-score 50.3 (12.1), 28–69 50.4 (11.1), 20–69 .01 
2-back commission errors** 2.0 (2.2), 0–7 0.5 (0.9), 0–4 .89 
2-back omission errors 0.7 (0.9), 0–3 0.6 (0.8), 0.4 .12 

Notes: PANAS = Positive and Negative Affect Scale; WCST = Wisconsin Card Sorting Test.

*p < .05.

**p < .001.

The first hypothesis was that the group who failed the IGT would have a more impulsive sensation-seeking personality than the group who passed. In contrast to expectations, those who failed the IGT did not show significantly higher Fun-Seeking characteristics than those who passed, although the results showed a trend in that direction, t(56) = 1.65, p = .11, and the effect size (.48) was moderate (Table 1).

The second hypothesis was that the group who failed the IGT would perform worse on measures of executive functioning skills. Analysis of covariance models (with estimated intellect as a covariate) showed that the two groups were not different in WCST perseverative errors, F(1,52) = 0.16, p = .69, or nonperseverative errors, F(1,52) = 0.19, p = .67. However, the analysis of covariance showed that those who failed the IGT made significantly more commission errors on the 2-back task compared with those who passed, F(1,52) = 8.84, p = .004. The two groups were not different in 2-back omission errors, F(1,52) = 0.20, p = .64 (Table 1).

The third hypothesis was that the group who failed the IGT would show less prefrontal cortex oxygenation than the group who passed. As predicted, those who failed the IGT showed significantly less dorsolateral prefrontal cortex oxygenation on the right (collapsed across all tasks), relative to those who passed, t(53) = −3.23, p = .002. Those who failed also showed less dorsolateral prefrontal cortex oxygenation on the left, t(53) = −2.60, p = .01. Looking at activation specifically during the IGT, the group who failed the IGT showed significantly less right frontal cerebral activation, t = −1.98, p = .05, but were not different in left frontal cerebral activation, t = −1.27, p = .21 (Table 2).

Table 2.

Group comparisons on near-infrared spectroscopy (NIRS) change scores

Variable Failures (N = 14) (mean [SD], range) Passers (N = 42) (mean [SD], range) Effect size (Cohen's d
Left frontal composite** −0.3 (1.1), −2.3 to 1.2 0.4 (1.0), −2.2 to 1.9 −.67 
Left frontal IGT −0.4 (1.9), −3.3 to 3.2 0.3 (1.7), −3.5 to 3.8 .06 
Right frontal composite*** 0.1 (1.1), −2.1 to 1.8 1.0 (0.9), −0.7 to 3.0 .90 
Right frontal IGT* 0.3 (1.2), −1.5 to 2.3 1.4 (1.8), −2.9 to 5.7 .72 
Variable Failures (N = 14) (mean [SD], range) Passers (N = 42) (mean [SD], range) Effect size (Cohen's d
Left frontal composite** −0.3 (1.1), −2.3 to 1.2 0.4 (1.0), −2.2 to 1.9 −.67 
Left frontal IGT −0.4 (1.9), −3.3 to 3.2 0.3 (1.7), −3.5 to 3.8 .06 
Right frontal composite*** 0.1 (1.1), −2.1 to 1.8 1.0 (0.9), −0.7 to 3.0 .90 
Right frontal IGT* 0.3 (1.2), −1.5 to 2.3 1.4 (1.8), −2.9 to 5.7 .72 

Note: IGT = Iowa gambling task.

*p < .05.

**p < .01.

***p < .005.

To explore whether the NIRS differences were mediated by intellect, we tested analysis of covariance models (with Matrix Reasoning as a covariate) for right and left hemisphere NIRS scores. For left NIRS composite, Matrix Reasoning was not a significant covariate, F(1,50) = 1.03, p = .32, and group differences in left activation remained significant, F(1,50) = 3.82, p = .05. For right NIRS composite, Matrix Reasoning was a significant covariate, F(1,50) = 4.57, p = .04. Even when controlling for Matrix Reasoning scores, differences on right NIRS remained significant, F(1,50) = 5.19, p = .03. To explore whether the NIRS differences were mediated by working memory, we tested analysis of covariance models (with 2-back commission errors as a covariate) for right and left hemisphere NIRS scores. For the left NIRS analysis, the covariate did not quite reach significance, F(1,49) = 3.75, p = .06, but the difference on left hemisphere activation dropped from significance, F(1,49) = 1.65, p = .21. For the right NIRS analysis, the covariate did not quite reach significance, F(1,49) = 3.34, p = .07, and the group difference on right hemisphere activation barely dropped out of significance, F(1,49) = 3.8, p = .06.

Discussion

As hypothesized, there were neuropsychological and cerebral oxygenation differences between individuals who clinically failed the IGT relative to those who passed. These results are all the more striking given that this was a nonclinical sample carefully screened for neurological and substance abuse-related factors, which have been linked to IGT failure in prior research.

Those who failed the IGT showed lower general intellect, as estimated by WASI Matrix Reasoning subtest. Our findings are in contrast to those of Brand, Recknor, and colleagues (2007), who found that Matrix Reasoning was not associated with IGT performance; however, their sample had a broader age range and was more educated overall than the present sample. Although the Matrix Reasoning subtest is highly correlated with general intellect and thus is a good estimate of overall intellectual level, it could be argued that the task is also a nonverbal problem solving and reasoning task (Wechsler, 1999), and thus group differences may be related not to overall intellectual differences but to relative executive function difficulties in those who failed the IGT. However, it should be noted that the groups did not differ in self-reported GPA or ACT scores, which served as additional indicators of general cognitive ability. Given the importance of education and intellect to clinical interpretation of this complex task, future studies examining the relationship of general intellect to IGT performance should include more comprehensive assessment of intellectual skills and not rely on education level or only one test to estimate it. Furthermore, the present study, as with most existing studies, includes individuals of only a limited intellectual/educational range. For example, the range of Matrix Reasoning scores in the “passers” fell from 1 SD (T = 39) below to 2.5 SD (T = 65) above standard norms for the task, whereas the scores in the “failures” fell from a bit more than 1 SD (T = 37) below to less than 2 SD (T = 57) above standard norms. Future studies should examine normal control samples that include individuals at lower levels of intellect and/or educational levels, as this is important for clinical interpretation of IGT performance in nonclinical populations.

Inconsistent with hypotheses, those who failed the IGT did not have a significantly more impulsive sensation-seeking personality style than those who passed, although this was predominantly an issue of statistical power, as our data showed a medium effect size difference on Fun Seeking. Furthermore, clinical variables often associated with risky personality characteristics (substance use) were controlled for in the present design, possibly restricting our range relative to a more broadly defined nonclinical sample. Similarly, the two groups were not different in negative affect, although the range of affect in this nonclinical sample was restricted. Given prior studies suggesting that both personality and negative affect are important correlates of IGT performance; however, future studies should continue to assess and consider the contribution of negative mood and impulsive sensation seeking when interpreting IGT findings. Other individual difference variables possibly related to IGT performance that could be explored in future studies have been suggested in recent literature. For example, Werner, Duschek, and Schandry (2009) found that healthy adults who showed less awareness of their own heart activity performed in a more risky fashion on the IGT than those who showed high awareness of their own heart activity. Even though this was a nonclinical sample without impairment in real-world functioning, such findings would be consistent with the Somatic Marker hypothesis (Damasio, 1994). In addition, Caroselli, Hiscock, Scheibel, and Ingram (2006) suggested that, rather than paying attention to overall accumulation of total winnings and losses over time, young healthy adults may simply be paying attention to the frequency of positive outcomes received on any given deck, which leads them to play in an overall disadvantageous style. This is a slightly different type of “short sightedness” than that suggested by Bechara and colleagues (2000) when explaining the “myopia” of individuals with frontal lobe dysfunction and suggests that patterns of performance on the IGT are just as important for clinical interpretation as an overall composite impaired score.

Support for executive functioning hypotheses was mixed. Contrary to predictions, those who failed the IGT did not perform worse on the WCST, but they did make significantly more commission errors on the 2-back trials of the N-back task. Although 2-back performance was correlated with intellect, as estimated by Matrix Reasoning, the group difference in 2-back performance was not accountable for by intellect. Overall, this finding is consistent with prior literature, which has shown working memory deficits in association with poor IGT performance in those with frontal lesions (Bechara et al., 1998; Manes et al., 2002), and studies in nonclinical samples that have experimentally manipulated working memory load demands, which tends to lead to more risky performance on IGT and similar gambling tasks (Dretsch & Tipples, 2008; Hinson, Jameson, & Whitney, 2002). Another possible explanation for our findings, given that the 2-back errors were commission rather than omission errors, is that the present data reflect more behavioral impulsivity in those who failed the IGT (Casbon, Curtin, Lang, & Patrick, 2003). With regard to the WCST findings, the present results are not due to limited power; observed effect sizes were extremely small. The present findings are more similar to those of Overman and colleagues (2004), whose sample was also similar to that of the present study. Brand, Recknor and colleagues (2007) did find a relationship of WCST performance with IGT performance, but in a sample with a much broader age range than the present study. Future studies should continue to include other measures of executive functioning in studies of the IGT, using samples with broader ranges of age and intellect, in order to better understand what other cognitive skills are most associated with IGT failure.

Finally, hypotheses regarding NIRS findings were supported. Those who failed the IGT showed less cerebral oxygenation in right dorsolateral prefrontal cortex, as well as left dorsolateral prefrontal cortex, during the performance on the neuropsychological tests, relative to those who passed. These differences were not accounted for by group differences in intellect. With regard to working memory/behavioral impulsivity, although inclusion of 2-back commission errors as a covariate reduced the significance of group differences on the NIRS, this may have been more an issue of statistical power than mediation, because the covariate was not significant. Overall, these findings are consistent with neuroimaging findings in normal control samples, where better IGT performance is associated with more frontal lobe activation, and with some of the dorsolateral prefrontal cortex findings in clinical samples. A limitation of our functional imaging data was the use of a 2-channel NIRS; although it has been shown to reliably detect dorsolateral prefrontal cortex activation (Kwee & Nakada, 2003) and is not able to detect noncortical or other cortical areas of the frontal lobe. However, our findings are also broadly consistent with NIRS findings (some using 2-channel NIRS) in normal controls completing other executive functioning tasks, including response conflict tasks (Kikuchi et al., 2007; Schroeter, Zysset, Kupka, Kruggel, & von Cramon, 2002), the WCST (Fallgatter & Strik, 2000), and N-back tasks (Herrmann, Ehlis, & Fallgatter, 2004). A limitation of all functional neuroimaging, including NIRS, is that it is unclear what more or less activation in any given brain region may mean. In fact, some recent event-related fMRI data with the IGT suggest that, in a cognitively complex task like the IGT, one can see different activation patterns in response to playing in a risky manner, as a result of learning over repeated trials, or in response to receiving a win or loss on a play. For example, Lawrence and colleagues (2006) found that choosing from disadvantageous relative to advantageous decks was related to higher activation in the medial frontal gyrus, lateral orbitofrontal cortex, and the insula, but activation in the lateral orbitofrontal cortex changed over time, perhaps suggesting that, as the IGT decks become less ambiguous, one may see different patterns of cerebral activation. In addition, they found that activation in striatothalamic regions was related to gains rather than losses. Future studies using more sophisticated imaging and modified IGT procedures will help to explain the conflicting findings in existing imaging literature on the IGT.

With regard to clinical implications for our findings, it is striking that this relatively well-screened nonclinical sample showed a 26% failure rate on the IGT, using the recently published age and education corrected clinical normative data. This rate of failure raises concerns about the use of the suggested T-score to detect clinical impairment on this task, at least for young adults. As our sample consisted of young adults who successfully completed high school and were in their first years of university, it is unlikely that those who failed were demonstrating significant real world executive functioning or risky decision-making impairments, although one might speculate that perhaps those who failed, who played in a disadvantageous fashion across the entire IGT, would be less likely to successfully finish college. Although the IGT manual does caution users that the IGT, as any neuropsychological test, should be interpreted in the context of a thorough assessment, data such as that found in the present study call for even more cautionary notes about interpreting “impaired” IGT performance without evidence of real-world decision-making impairments or any other evidence of brain dysfunction. Thus, our results serve as another reminder for those who use such tests in the clinic or in research settings that tests results alone are not equal to clinical assessment.

Conflict of Interest

None declared.

References

Barry
D.
Petry
N. M.
Predictors of decision-making on the Iowa gambling task: Independent effects of lifetime history of substance use disorders and performance on the Trail Making Test
Brain and Cognition
 , 
2008
, vol. 
66
 (pg. 
243
-
252
)
Bechara
A.
Iowa gambling task professional manual
 , 
2007
Boca Raton, FL
Psychological Assessment Resources, Inc
Bechara
A.
Damasio
H.
Decision-making and addiction (part I): Impaired activation of somatic states in substance dependent individuals when pondering decisions with negative future consequences
Neuropsychologia
 , 
2002
, vol. 
40
 (pg. 
1675
-
1689
)
Bechara
A.
Damasio
H.
Tranel
D.
Anderson
S.
Dissociation of working memory from decision making within the human prefrontal cortex
Journal of Neuroscience
 , 
1998
, vol. 
18
 (pg. 
428
-
437
)
[PubMed]
Bechara
A.
Dolan
S.
Denburg
N.
Hindes
A.
Anderson
S.
Nathan
P.
Decision-making deficits, linked to a dysfunctional ventromedial prefrontal cortex, revealed in alcohol and stimulant users
Neuropsychologia
 , 
2001
, vol. 
39
 (pg. 
376
-
389
)
Bechara
A.
Tranel
D.
Damasio
H.
Characterization of the decision-making deficit of patients with ventromedial prefrontal cortex lesions
Brain
 , 
2000
, vol. 
123
 (pg. 
2189
-
2202
)
Bliss
I.
Hamalanian
H.
Different working memory capacity in normal young adults for visual and tactile letter recognition task
Scandinavian Journal of Psychology
 , 
2005
, vol. 
46
 (pg. 
247
-
251
)
Boeka
A. G.
Lokken
K. L.
The Iowa gambling task as a measure of decision making in women with bulimia nervosa
Journal of the International Neuropsychological Society
 , 
2006
, vol. 
12
 (pg. 
741
-
745
)
Bolla
K.
Eldreth
D.
London
E.
Kiehl
K.
Mouratidis
M.
Contoreggi
C.
, et al.  . 
Orbitofrontal cortex dysfunction in abstinent cocaine abusers performing a decision-making task
NeuroImage
 , 
2003
, vol. 
19
 (pg. 
1085
-
1094
)
Brand
M.
Grabenhorst
F.
Starcke
K.
Vandekerckhove
M. M. P.
Markowitsch
H. J.
Role of the amygdala in decisions under ambiguity and decisions under risk: Evidence from patients with Urbach-Wieths disease
Neuropsychologia
 , 
2007
, vol. 
45
 (pg. 
1305
-
1317
)
Brand
M.
Kalbe
E.
Labudda
K.
Fujiwara
E.
Kessler
J.
Markowitsch
H. J.
Decision making impairments in patients with pathological gambling
Psychiatry Research
 , 
2005
, vol. 
133
 (pg. 
91
-
99
)
Brand
M.
Recknor
E.
Grabenhorst
R.
Bechara
A.
Decisions under ambiguity and decisions under risk: Correlations with executive functions and comparisons of two different gambling tasks with implicit and explicit rules
Journal of Clinical and Experimental Neuropsychology
 , 
2007
, vol. 
29
 (pg. 
86
-
99
)
Buelow
M. T.
Suhr
J. A.
Construct validity of the Iowa gambling task
Neuropsychology Review
 , 
2009
, vol. 
19
 (pg. 
102
-
114
)
Carver
C.
White
T.
Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales
Journal of Personality and Social Psychology
 , 
1994
, vol. 
2
 (pg. 
319
-
333
)
Carver
C. S.
Negative affects deriving from the Behavioral Activation System
Emotion
 , 
2004
, vol. 
4
 (pg. 
3
-
22
)
Casbon
T. S.
Curtin
J. J.
Lang
A. R.
Patrick
C. J.
Deleterious effects of alcohol intoxication: Diminished cognitive control and its behavioral consequences
Journal of Abnormal Psychology
 , 
2003
, vol. 
112
 (pg. 
476
-
487
)
Caroselli
J. S.
Hiscock
M.
Scheibel
R. S.
Ingram
F.
The simulated gambling paradigm applied to young adults: An examination of university students' performance
Applied Neuropsychology
 , 
2006
, vol. 
13
 (pg. 
203
-
212
)
Cavedini
P.
Riboldi
G.
Keller
R.
D'Annucci
A.
Bellodi
L.
Frontal lobe dysfunction in pathological gambling patients
Biological Psychiatry
 , 
2002
, vol. 
51
 (pg. 
334
-
341
)
Clements
R.
Psychometric properties of the Substance Abuse Subtle Screening Inventory-3
Journal of Substance Abuse Treatment
 , 
2002
, vol. 
23
 (pg. 
419
-
423
)
Collins
R.
Parks
G.
Marlatt
G.
Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol
Journal of Consulting and Clinical Psychology
 , 
1985
, vol. 
53
 (pg. 
189
-
200
)
Damasio
A. R.
Decartes' error: Emotion, reason, and the human brain
 , 
1994
New York
Grosset/Putnam
Davis
C.
Fox
J.
Patte
K.
Curtis
C.
Strimas
R.
Reid
C.
, et al.  . 
Education level moderates learning on two versions of the Iowa gambling task
Journal of the International Neuropsychological Society
 , 
2008
, vol. 
14
 (pg. 
1063
-
1068
)
Denburg
N. L.
Recknor
E. C.
Bechara
A.
Tranel
D.
Psychophysiological anticipation of positive outcomes promotes advantageous decision-making in normal older persons
International Journal of Psychophysiology
 , 
2006
, vol. 
61
 (pg. 
19
-
25
)
Dom
G.
De Wilde
B.
Hulstijn
W.
van den Brink
W.
Sabbe
B.
Decision-making deficits in alcohol-dependent patients with and without comorbid personality disorder
Alcoholism: Clinical and Experimental Research
 , 
2006
, vol. 
30
 (pg. 
1670
-
1677
)
Dretsch
M. N.
Tipples
J.
Working memory involved in predicting future outcomes based on past experiences
Brain and Cognition
 , 
2008
, vol. 
66
 (pg. 
83
-
90
)
Ernst
M.
Bolla
K.
Mouratidis
M.
Contoreggi
C.
Matochik
J. A.
Kurian
V.
, et al.  . 
Decision-making in a risk-taking task: A PET study
Neuropsychopharmacology
 , 
2002
, vol. 
26
 (pg. 
682
-
691
)
Evans
C.
Kemish
K.
Turnbull
O.
Paradoxical effects of education on the Iowa gambling task
Brain and Cognition
 , 
2004
, vol. 
54
 (pg. 
240
-
244
)
Fallgatter
A. J.
Strik
W. K.
Reduced frontal functional asymmetry in schizophrenia during a cued continuous performance test assessed with near-infrared spectroscopy
Schizophrenia Bulletin
 , 
2000
, vol. 
26
 (pg. 
913
-
919
)
Fein
G.
McGillivray
S.
Finn
P.
Older adults make less advantageous decisions than younger adults: Cognitive and psychological correlates
Journal of the International Neuropsychological Society
 , 
2007
, vol. 
13
 (pg. 
480
-
489
)
Fukui
H.
Murai
T.
Fukuyama
H.
Hayashi
R.
Hanakawa
T.
Functional activity related to risk anticipation during performance of the Iowa gambling task
NeuroImage
 , 
2005
, vol. 
24
 (pg. 
253
-
259
)
Goudriann
A. E.
Grekin
E. R.
Sher
K. J.
Decision making and binge drinking: A longitudinal study
Alcoholism: Clinical and Experimental Research
 , 
2007
, vol. 
31
 (pg. 
928
-
938
)
Grant
S.
Contoreggi
C.
London
E. D.
Drug abusers show impaired performance in a laboratory test of decision making
Neuropsychologia
 , 
2000
, vol. 
38
 (pg. 
1180
-
1187
)
Hammers
D. B.
Suhr
J. A.
Neuropsychological, impulsive personality, and cerebral oxygenation correlates of undergraduate polysubstance use
Journal of Clinical and Experimental Neuropsychology
  
(in press)
Heaton
R.
PAR Staff
WCST-64: Computer version 2- research edition
 , 
2005
Lutz, FL
Psychological Assessment Resources, Inc
Henson
L.
Cartwright
C.
Chlebowski
S.
Estimation of jugular venous O2 saturation from cerebral oximetry or arterial O2 saturation during isocapnic hypoxia
Anesthesiology
 , 
1997
, vol. 
87
 pg. 
A402
 
Herrmann
M. J.
Ehlis
A. C.
Fallgatter
A. J.
Bilaterally reduced frontal activation during a verbal fluency task in depressed patients as measured by near-infrared spectroscopy
Journal of Neuropsychiatry and Clinical Neurosciences
 , 
2004
, vol. 
16
 (pg. 
170
-
175
)
Heubeck
B. G.
Wilkinson
R. B.
Cologon
J.
A second look at Carver and White's (1994) BIS/BAS scales
Personality and Individual Differences
 , 
1998
, vol. 
25
 
4
(pg. 
785
-
800
)
Hinson
J.
Jameson
T.
Whitney
P.
Somatic markers, working memory, and decision making
Cognitive, Affective, and Behavioural Neuroscience
 , 
2002
, vol. 
2
 (pg. 
341
-
353
)
Irani
F.
Platek
S.
Bunce
S.
Ruocco
A.
Chute
D.
Functional near infrared spectroscopy (fNIRS): An emerging neuroimaging technology with important applications for the study of brain disorders
The Clinical Neuropsychologist
 , 
2007
, vol. 
21
 (pg. 
9
-
37
)
Jollant
F.
Buresi
C.
Guillaume
S.
Jaussent
I.
Bellivier
R.
Leboyer
M.
, et al.  . 
The influence of four serotonin-related genes on decision-making in suicide attempters
American Journal of Medical Genetics, Part B: Neuropsychiatric Genetics
 , 
2007
, vol. 
144
 (pg. 
615
-
624
)
Jorm
A. F.
Christensen
H.
Henderson
A. S.
Jacomb
P. A.
Korten
A. E.
Rodgers
B.
Using the BIS/BAS Scale to measure behavioural inhibition and behavioural activation: Factor structure, validity, and norms in a large community sample
Personality and Individual Differences
 , 
1999
, vol. 
26
 (pg. 
49
-
58
)
Kikuchi
S.
Iwata
K.
Onishi
Y.
Kubota
F.
Nisijima
K.
Tamai
H.
, et al.  . 
Prefrontal cerebral activity during a simple “rock, paper, scissors” task measured by the noninvasive near-infrared spectroscopy method
Psychiatry Research: Neuroimaging
 , 
2007
, vol. 
156
 (pg. 
199
-
208
)
Kwee
I. L.
Nakada
T.
Dorsolateral prefrontal lobe activation declines significantly with age
Journal of Neurology
 , 
2003
, vol. 
250
 (pg. 
525
-
529
)
Larimer
M.
Turner
A.
Mallett
K.
Geisner
I.
Predicting drinking behavior and alcohol-related problems among fraternity and sorority members: Examining the role of descriptive and injunctive norms
Psychology of Addictive Behaviors
 , 
2004
, vol. 
18
 (pg. 
203
-
212
)
Laux
J.
Perera-Diltz
D.
Smirnoff
J.
Salyers
K.
The SASSI-3 Face Valid Other Drugs scale: A psychometric investigation
Journal of Addictions and Offender Counseling
 , 
2005
, vol. 
26
 (pg. 
15
-
21
)
Lawrence
N. S.
Wooderson
S.
Mataix-Cols
D.
David
R.
Speckens
A.
Phillips
M. L.
Decision making and set shifting impairments are associated with distinct symptoms dimensions in obsessive-compulsive disorder
Neuropsychology
 , 
2006
, vol. 
20
 (pg. 
409
-
419
)
Lewis
M.
Neighbors
C.
Gender-specific misperceptions of college student drinking norms
Psychology of Addictive Behaviors
 , 
2004
, vol. 
18
 (pg. 
334
-
339
)
Manes
F.
Sahakian
B.
Clark
L.
Rogers
R.
Antoun
N.
Aitken
M.
, et al.  . 
Decision-making processes following damage to the prefrontal cortex
Brain
 , 
2002
, vol. 
125
 (pg. 
624
-
639
)
McNally
A.
Palfai
T.
Kahler
C.
Motivational interventions for heavy drinking college students: Examining the role of discrepancy-related psychological processes
Psychology of Addictive Behaviors
 , 
2005
, vol. 
19
 (pg. 
79
-
87
)
Miller
G.
Lazowski
L.
The SASSI manual: Substance abuse measures
 , 
1999
2nd ed.
Springville, IN
SASSI Institute
Monterosso
J.
Ehrman
R.
Napier
K. L.
O'Brien
C. P.
Childress
A. R.
Three decision-making tasks in cocaine-dependent patients: Do they measure the same construct?
Addiction
 , 
2001
, vol. 
96
 (pg. 
1825
-
1837
)
Murphy
J.
Corriea
C.
Colby
S.
Vuchinich
R.
Using behavioral theories of choice to predict drinking outcomes following a brief intervention
Experimental and Clinical Psychopharmacology
 , 
2005
, vol. 
13
 (pg. 
93
-
101
)
Must
A.
Szabo
Z.
Bodi
N.
Szasz
A.
Janka
Z.
Keri
S.
Sensitivity to reward and punishment in the prefrontal cortex in major depression
Journal of Affective Disorders
 , 
2006
, vol. 
90
 (pg. 
209
-
215
)
Overman
W. H.
Frassrand
K.
Ansel
S.
Trawalter
S.
Bies
B.
Redmond
A.
Performance on the IOWA card task by adolescents and adults
Neuropsychologia
 , 
2004
, vol. 
42
 (pg. 
1838
-
1851
)
Q3 Innovations.
Alcohawk precision owner's manual
 , 
2005
 
Retrieved March 11, 2010, from http://www.q3i.com
Ritter
L. M.
Meador-Woodruff
J. H.
Dalack
G. W.
Neurocognitive measures of prefrontal cortical dysfunction in schizophrenia
Schizophrenia Research
 , 
2004
, vol. 
68
 (pg. 
65
-
73
)
Rotherham-Fuller
E.
Shoptaw
S.
Berman
S.
London
E.
Impaired performance in a test of decision-making by opiate-dependent tobacco smokers
Drug and Alcohol Dependence
 , 
2004
, vol. 
73
 (pg. 
79
-
86
)
Schroeter
M. L.
Zysset
S.
Kupka
T.
Kruggel
F.
von Cramon
D. Y.
Near-infrared spectroscopy can detect brain activity during a color-word matching Stroop task in an event-related design
Human Brain Mapping
 , 
2002
, vol. 
17
 (pg. 
61
-
71
)
Smillie
L. D.
Jackson
C.
Dalgleish
L. I.
Conceptual distinctions among Carver and White's (1994) BAS scales: A reward-reactivity versus trait impulsivity perspective
Personality and Individual Differences
 , 
2006
, vol. 
40
 (pg. 
1039
-
1050
)
Somanetics Corporation.
INVOS cerebral oximeter monitoring system models 4100 & 5100 operations manual
  
Retrieved March 11, 2010, from http://www.somanetics.net
Suhr
J.
Tsanadis
J.
Affect and personality correlates of the Iowa gambling task
Personality and Individual Differences
 , 
2007
, vol. 
43
 (pg. 
27
-
36
)
Tanabe
J.
Thompson
L.
Claus
E.
Dalwani
M.
Hutchison
K.
Banich
M. T.
Prefrontal cortex activity is reduced in gambling and nongambling substance abusers during decision-making
Human Brain Mapping
 , 
2007
, vol. 
28
 (pg. 
1276
-
1286
)
Testa
M.
Livingston
J.
Alcohol and sexual aggression: Reciprocal relationships over time in a sample of high-risk women
Journal of Interpersonal Violence
 , 
2000
, vol. 
15
 (pg. 
413
-
427
)
Thavasothy
M.
Broadhead
M.
Elwell
C.
Peters
M.
Smith
M.
A comparison of cerebral oxygenation as measured by the NIRO 300 and the INVOS 5100 Near-Infrared Spectrophotometers
Anaesthesia
 , 
2002
, vol. 
57
 (pg. 
999
-
1006
)
Tucker
K. A.
Potenza
M. N.
Beauvais
J. E.
Browndyke
J. N.
Gottschalk
P. C.
Kosten
T. R.
Perfusion abnormalities and decision making in cocaine dependence
Biological Psychiatry
 , 
2004
, vol. 
56
 (pg. 
527
-
530
)
Verdejo-Garcia
A.
Bechara
A.
Recknoir
E.
Perez-Garcia
M.
Executive dysfunction in substance dependent individuals during drug use and abstinence: An examination of the behavioral, cognitive, and emotional correlates of addiction
Journal of the International Neuropsychological Society
 , 
2006
, vol. 
12
 (pg. 
405
-
415
)
Watson
D.
Clark
L.
Tellegen
A.
Development and validation of brief measures of positive and negative affect: The PANAS scales
Journal of Personality and Social Psychology
 , 
1988
, vol. 
54
 
6
(pg. 
1063
-
1070
)
Wechsler
D.
Wechsler Abbreviated Scale of Intelligence (WASI)
 , 
1999
San Antonio, TX
The Psychological Corporation
Werner
N. S.
Duschek
S.
Schandry
R.
Relationships between affective states and decision-making
International Journal of Psychophysiology
 , 
2009
, vol. 
74
 (pg. 
259
-
265
)