Posttraumatic Stress Disorder in the U.S. Warfighter

Abstract

Little research has been done to explore the integrity of emotion-based decision-making performance in individuals with posttraumatic stress disorder (PTSD). In the current study, performance on two decision-making tasks with both positive and negative reinforcement, the standard Iowa Gambling Task (IGT) and the variant Iowa Gambling Task (vIGT), and measures of mood symptoms, were compared between U.S. active-duty soldiers diagnosed with PTSD (n = 23) and soldiers with no PTSD (n = 23). The results revealed that the PTSD group, when compared to controls, reported significantly higher anxiety and depression symptoms. The PTSD group showed similar behavioral performance as controls on the standard IGT but failed to choose advantageously on the vIGT, which has been shown to reflect hypersensitivity to punishment. Medicated participants, being treated with antidepressants, showed significantly better overall performance on the IGT but not on the vIGT compared to nonmedicated participants. The results suggest that soldiers being treated for PTSD have a unique decision-making pattern that may be attributed to difficulty in processing delayed reward when presented with immediate punishment.

Keywords

posttraumatic stress disorder antidepressant Iowa Gambling Task decision-making punishment

Introduction

For many individuals, posttraumatic stress disorder (PTSD) is a debilitating anxiety disorder that results in maladaptive behavioral functioning within personal, social, and occupational domains (Diagnostic and Statistical Manual of Mental Disorders, 4th ed., text rev., American Psychiatric Association, 2000). This poor behavioral functioning often stems from negative alterations in cognition and mood (e.g., markedly diminished interest or participation in significant activities; persistent inability to experience positive emotions) and alterations in arousal and reactivity (e.g., problems with concentration; irritable or aggressive behavior; Spitzer, First, & Wakefield, 2007). These behavioral issues related to PTSD are particularly relevant within the Armed Services as there is a high incidence of PTSD in soldiers returning from wars in Afghanistan and Iraq, also known as Operation Enduring Freedom (OEF) and Operation Iraqi Freedom (OIF). Current estimates suggest that approximately 15% of soldiers returning from deployment develop PTSD (Tanielian & Jacox, 2008).

The literature is replete of findings that PTSD is associated with compromised cognitive processes such as attention and memory (Brandes et al., 2002; Vasterling et al., 2002; Yehuda, Golier, Tischler, Stavitsky, & Harvey, 2005). Much less work has been done exploring how PTSD impacts other important processes such as decision making. The ability to have foresight and make advantageous decisions based on outcome expectancy is necessary for all soldier occupations. For example, decisions are made under various dynamic environmental and psychological conditions for which learned information is constantly being updated to make optimal choices. Many of these choices in the combat environment have real-life, deleterious consequences for which human life, resources, and the mission are potentially compromised.

Decision making depends on several temporally and partially distinct psychological processes and neurobiological substrates (Ernst et al., 2002). Namely, there are concerted interactions between cognitive and emotional neural networks that are predominately regulated by cortical regions such as the dorsolateral prefrontal cortex (DLPFC), ventromedial prefrontal cortex (VMPFC), cingulate cortex, and amygdala (Ernst & Paulus, 2005). The Iowa Gambling Task (IGT; Bechara, Tranel, & Damasio, 2000) is sensitive to the neurocircuitry necessary for emotional-based decision making under conditions of uncertainty whereupon maximizing advantageous choices relies on trial-by-error learning. Seminal evidence of the involvement of specific brain regions in decision making on the IGT is supported by findings from patients with lesions to either the VMPFC (Bechara et al., 2000) or amygdala (Bechara, Damasio, Damasio, & Lee, 1999), who perform significantly worse than controls. Furthermore, evidence from functional brain imaging techniques with healthy individuals further supports the role of the VMPFC in IGT performance in addition to other regions such as the orbitofrontal cortex, insula, secondary somatosensory cortex, and striatum and cingulate gyrus (Lawrence, Jollant, O’Daly, Zelaya, & Phillips, 2008; Li, Lu, D’Argembeau, Ng, & Bechara, 2010; Lin, Chiu, Cheng, & Hsieh, 2008; Windmann et al., 2006).

Copious evidence illustrates that neurobiological and functional changes of the VMPFC are implicated in PTSD (Koenigs & Grafman, 2009; Lanius, Bluhm, Lanius, & Pain, 2006; Villarreal et al., 2002). However, the degree to which pathophysiological changes associated with PTSD may impact decision making has not been thoroughly investigated with the IGT.

The IGT involves asking individuals to select cards from four decks in order to maximize monetary reward. After selecting a card from a specific deck, feedback is displayed informing the participant of the amount of money won and lost. On the standard version of the task, two of the decks incur high immediate rewards but eventually incur larger losses as the task progresses (100 selections are made in total). The remaining (advantageous) decks yield smaller immediate rewards but also smaller losses, leading to net gain if consistently selected throughout the task. In contrast, in the variant IGT (vIGT) the earnings and losses are reversed so that the advantageous decks yield high immediate losses but also higher delayed earning, for a net gain if consistently selected. The disadvantageous decks have smaller immediate losses but also smaller gains that result in a net loss (Bechara, Dolan, & Hindes, 2002).

When the vIGT is used concomitantly with the standard version, it potentially demonstrates sensitivity for differentiating decision-making impairments that are characteristic of hypersensitivity to reward, hyposensitivity to punishment, hypersensitivity to punishment, and insensitivity to both good and bad outcomes. For example, behaviorally, both hypersensitivity to reward and hyposensitivity can be characterize by impaired performance on the standard IGT but not on the vIGT, hypersensitivity to punishment is characteristic of normal performance on the standard IGT but not on the vIGT, and insensitivity to outcomes regardless of valence is characteristic of impaired performance on both versions (Bechara et al., 2002; Crone, Bunge, Latenstein, & van der Molen, 2005).

Insofar, there is only minimal evidence delineating how individuals with PTSD perform on the IGT. In a prior study, Roca, Hart, Kimbrell, and Freeman (2006) administered the standard IGT to 10 veterans diagnosed with PTSD with a co-occurring dissociative disorder and 17 veterans with PTSD and no dissociative disorder. The results indicate there are no significant differences between these PTSD subgroups on the IGT. However, a major limitation of this study is the lack of a control group.

Given that PTSD is an anxiety disorder, one could make different predictions of how individuals with PTSD would perform on either version of the IGT based on studies exploring the contribution of trait-anxiety levels. Interestingly, the evidence using the standard IGT is mixed in that Miu, Heilman, and Houser (2008) provide evidence of an inverse relationship between trait anxiety and IGT performance, whereas Werner, Duschek, and Schandry (2009) show a positive relationship. Although both of these studies provide translational findings, the contribution of trait anxiety was assessed in relatively healthy individuals. Other studies suggest that the mediation of anxiety levels on IGT performance are gender specific (de Visser et al., 2010). In contrast to variability of trait-like mood states, PTSD is a clinical disorder accompanied by high anxiety and depression comorbidity (Hankin, Spiro, Miller, & Kazis, 1999).

Some, but not all anxiety disorders can be linked with emotion-based decision-making impairments. De Rocha, Alvarenga, Malloy-Diniz, and Corrêa (2011) provide findings that patients with obsessive-compulsive disorder (OCD) are impaired on the IGT compared to healthy controls. In contrast, findings by Mueller, Nguyen, Ray, and Borkovec (2010) reveal superior performance on both the IGT and vIGT in individuals diagnosed with generalized anxiety disorder (GAD). Although the evidence concerning anxiety disorders and decision-making performance is mixed, it remains unclear how individuals with PTSD will perform on either the IGT or the vIGT.

Compromised performance on the IGT is not only observed in anxiety disorders other than PTSD but also in mood disorders such as major depressive disorder (MDD; Han et al., 2012). Being that there is a high comorbidity rate of depressive symptomatology affecting 30% to 50% patients with PTSD (Campbell et al., 2007; Hankin et al., 1999), depression may partially mediate emotion-based decision making as assessed by the IGT.

In light of evidence that PTSD is an anxiety disorder with high comorbid depression, and is linked with abnormal VMPFC functioning (Koenigs & Grafman, 2009), continued effort is necessary to explore behavioral performance on a variety of tasks sensitive to both mood symptoms (e.g., depression) and VMPC function. The present study was aimed at investigating the integrity of emotion-based decision making in individuals with PTSD. Our population included active-duty, U.S. Army soldiers with a history of past deployments and a DSM-IV-TR diagnosis of PTSD. Performance was compared to a matched group of active-duty soldiers with similar combat experience but no diagnosis of PTSD. Furthermore, in addition to the standard task, the variant version of the IGT was administered to assess whether sensitivity to immediate punishments would have an effect on advantageous decision making in individuals with PTSD. Individual differences in personality traits (e.g., sensitivity to reward and punishment) and mood symptomatology (e.g., anxiety, depression, PTSD severity) were also assessed.

Material and Methods

Participants

Forty-six U.S. Army soldiers (10 women and 36 men) with prior combat experience voluntarily participated in this cross-sectional study at a military medical treatment center. The PTSD group (n = 23; 91% male) consisted of individuals with a current DSM-IV-TR diagnosis for PTSD from a health care provider (annotated in Armed Forces Longitudinal Technology Application; ALTHA or hard copy of medical record) as assessed via the Clinician-Administered PTSD Checklist (CAPS). Individuals were excluded if there existed a comorbid psychiatric diagnosis of a psychotic disorder (e.g., Schizophrenia), bipolar disorder, or were currently being treated for substance dependency. If a participant reported taking medication with a psychogenic effect that might affect their ability to maintain wakefulness during the study, he or she was informed by the investigators that he or she was not eligible to participate. Participants were recruited from the Behavioral Health Department, PTSD Treatment Clinic, and Community Mental Health Clinic at Fort Bliss, Texas. The control group (n = 23; 65% male) consisted of soldiers with no PTSD or any other mental health diagnosis, recruited from patients and staff within the medical treatment center.

Approval to conduct this research study was granted by the Institutional Review Board at William Beaumont Army Medical Center, Fort Bliss, Texas.

Behavioral Measures

Both versions of the IGT used the same fixed schedule of monetary reinforcement for each of the decks as in previous studies (Bechara et al., 2000). There were four decks of cards each containing 100 cards.

Standard IGT

For the standard IGT, Decks A and B were disadvantageous in that they yielded large immediate earnings but even larger delayed losses. For example, the first 10 cards selected from Decks A and B resulted in total earnings of US$1,000 but losses totaling US$1,250 with a net loss of US$250. Deck A contained low, frequent losses, and Deck B contained higher, but less frequent losses (e.g., one loss of US$1,250 in the first block of 10 trials). In contrast, Decks C and D were advantageous in that they yielded low immediate earnings but even smaller delayed losses. For example, the first 10 cards selected from Decks C and D resulted in a total earning of US$500 (average loss per card $50) and also resulted in losses totaling US$250. Deck C contained relatively low, frequent losses, and Deck D contained higher, but less frequent, losses (e.g., one win of US$250 in the first block of 10 trials).

Variant IGT

The vIGT reversed the earnings and losses so that Decks C and D were disadvantageous in that they yielded small immediate losses but even smaller delayed wins. For example, the first 10 cards selected from Decks C and D resulted in total losses of US$500 and total wins of US$250. In contrast, Decks A and B were advantageous in that they yielded high immediate losses but even larger delayed winnings. For example, the first 10 cards selected from Decks A and B resulted in a total loss of US$1,000 (average loss per card US$100) but wins totaling US$1,250.

Questionnaires

Test of Memory Malingering (TOMM; Tombaugh, 2003)

The TOMM was used to assess effort in all participants. The TOMM consists of two learning trials and a retention trial that uses 50 pictures of common, everyday objects (e.g., chair, pencil, etc.). A cutoff score (< 45 correct) for the first two learning trials was used to determine eligibility.

Reward and Punishment Sensitivity subscales (Behavioral Activation and Behavioral Inhibition Scales: BAS/BIS; Carver & White, 1994)

The BAS/BIS assesses two general motivational systems that underlie behavior and affect: a behavioral inhibition system (BIS) and a behavioral activation system (BAS). The BAS/BIS scales consist of 20 items that participants rate on a 4-point scale. The measure yields four scale scores, one BIS score (Punishment Sensitivity), and three BAS scores (Reward Responsiveness, Drive, and Fun Seeking). The scales have alpha reliabilities that range from .66 to .76. For the current study, only scores of Punishment Sensitivity (BIS_pun) and Reward Responsiveness (BAS_rew) were used in our analysis because of relevance to performance on the behavioral tasks. Examples of items include, “When good things happen to me, it affects me strongly,” and “I feel worried when I think I have done poorly at something.”

PTSD Checklist-Military Version (PCL-M; Orsillo, 2001)

The PCL-M is a 17-item self-report measure of the DSM-IV symptoms of PTSD. It has a test–retest reliability of .96, internal consistency (alpha coefficient) of .93 for B symptoms (i.e., reexperiencing), .92 for C symptoms (i.e., effortful avoidance), .92 for D symptoms (i.e., hyperarousal), and .97 for all 17 symptoms. An example of an item includes “Feeling very upset when something reminded you of a stressful military experience?” A total score of 50 is considered to be PTSD positive in the military.

Beck Depression Inventory—2nd edition (BDI-II; Beck & Steer, 1984)

The BDI-II is a questionnaire designed to assess the severity of depression in adolescents and adults. The BDI-II consists of 21 symptom items which participants rate items on a 4-point scale ranging from 0 to 3 in terms of severity. The BDI-II has a high coefficient alpha (.92) and test–retest reliability (.93). An example of an item includes, “I am sad all the time, and I can’t snap out of it.” A total score of 0 to 13 is considered minimal range, 14 to 19 is mild, 20 to 28 is moderate, and 29 to 63 is severe.

Beck Anxiety Inventory (BAI; Beck, Epstein, Brown, & Steer, 1988)

The BAI is a questionnaire designed to assess the severity of anxiety symptoms in adolescents and adults. The total score is the sum of the ratings for 21 symptoms (e.g., unable to relax, fear of losing control, nervous, etc.) occurring during the past month. Each symptom is rated on a 4-point scale ranging from 0 to 3. It has a high coefficient alpha (.93) and test–retest reliability (.75). A total score of 0 to 21 is considered very low anxiety, 22 to 35 is moderate, and 36 or higher is severe.

Procedure

All participants were recruited from the garrison medical clinic and surrounding area via posters, clinician referral, and word of mouth. Participants arrived at the designated data collection location in the medical clinic. After being consented, all participants were administered the TOMM and then given instructions for completing the demographic section on the computer, followed by the behavioral tasks, and finally the remaining questionnaires.

The presentation order of the IGT and vIGT were counterbalanced between participants in both groups. Participants started the task with a facsimile gift of US$2,000. Participants were allowed to select freely from any of the decks throughout the game. After each deck selection, participants received three types of trial feedback: (a) Two messages indicating the amount of money lost and won on that trial (e.g., “You win US$100” “You lose US$1250!”), (b) a smiley face presented next to the amount won and a sad face presented next to the amount lost, and (c) green and red bars presented at the top of the screen that indicated the total amount of money won (green) or lost (red) after each selection. The size of the green and red bars was proportional to the amount won or lost. These three types of trial feedback appeared simultaneously for 3 s before an intertrial interval lasting 1.5 s during which the words “Please Wait” were displayed. In addition to the trial feedback, the colored bars remained present throughout the task. Participants selected up to 100 cards whereupon the task ended. At this point the participants were offered a short break before beginning the remaining version of the IGT. The participants were not informed when either task was going to finish.

After completion of the data collection, participants were briefed on the purpose of their participation in the study] and were given time to ask questions if needed.

Results

Data Reduction

Performance on both versions of the IGT was quantified by dividing 100 trials in to 5 blocks of 20 trials each. A proportion score for each block was calculated as the number of cards chosen from the advantageous decks in relation to the disadvantageous decks. Scores less than .50 indicate more disadvantageous selections, whereas scores greater than .50 indicate more advantageous selections. In addition, the proportion for the 100 trials for both versions of the IGT was calculated.

Due to the high multicollinearity between depression and anxiety, only depression scores (BDI) were used as the covariate in the analysis. Depression scores were adjusted with grand mean centering to increase precision of estimates of variance.

Basic Group Differences

Forty-three percent (10 of 23) of the PTSD group was on an antidepressant compared to 9% (2 of 23) in the healthy controls. These included selective serotonin reuptake inhibitors (SSRI), serotonin–norepinephrine reuptake inhibitors (SNRI), and other atypical antidepressants (e.g., norepinephrine–dopamine reuptake inhibitors; NDRI). The groups had an unequal gender distribution, χ²(1, n = 46) = 4.60, p < .05, with eight women in the control group (35%) and 2 women in the PTSD group (1%). A series of t tests were run to compare group differences in education, age, symptom severity, and motivational traits. A Bonferroni adjustment was made to control for familywise error rate (p ≤ .007). There were no significant group differences in education (years), t(44) = 0.77, p > .007, and age, t(44) = −1.85, p > .007. Furthermore, there were no significant group differences in motivational traits on Reward, t(42) = 2.75, p > .007, and Punishment, t(42) = −0.92, p > .007, subscales of the BAS/BIS. In contrast, the PTSD group scored significantly higher on PTSD symptoms, t(42) = −15.03, p < .007; symptoms of anxiety, t(42) = −10.79, p < .007; and depression, t(42) = −9.70, p < .007, relative to the Control group as assessed with the PCL-M, BAI, and BDI (Table 1).

Table 1.

Group Descriptives.

	Group	M	SD
Age(years)	PTSD	38.0	8.4
Age(years)	Controls	33.5	8.2
Education (years)	PTSD	14.5	2.0
Education (years)	Control	14.9	1.9
Males (91%)	PTSD	0.9	0.3
Males (91%)	Control	0.7	0.5
BAS_rew	PTSD	13.8	2.4
BAS_rew	Control	15.9	2.2
BIS_pun	PTSD	18.6	2.6
BIS_pun	Control	18.2	3.1
BAI	PTSD	24.4	9.5
BAI	Control	2.2	2.6
BDI	PTSD	23.0	7.6
BDI	Control	4.7	4.0
PCL-M	PTSD	58.6	8.9
PCL-M	Control	24.2	6.1

Note: Group descriptives in means and standard deviations. Education in years; PTSD = posttraumatic stress disorder; Male = proportion of men; BIS_pun = Behavioral Activation and Behavioral Inhibition System domains—Punishment; BAS_rew = Behavioral Activation and Behavioral Inhibition System domains—Reward; BAI = Beck Anxiety Inventory; BDI = Beck Depression Inventory; PTSD Checklist—Military Version = PCL-M.

Decision-Making Performance

Overall performance on both versions of the IGT was assessed using separate analysis of variance (ANOVA) with Block (5) as the within-subject factor.

To compare group differences, data from both versions of the IGT were analyzed using separate mixed-factorial analyses of covariance (ANCOVA) with Block (5) as the within-subject factor, Group (PTSD vs. Control) as the between-subject factor, and Gender (Males vs. Females) and Depression scores (BDI) as the covariates. The assumption of regression homogeneity was met for all covariates for each ANCOVA.

Standard IGT

Overall, using more conservative Greenhouse– Geisser corrections for all analyses, there was a significant main effect of Block, F(4, 180) = 2.86, p = .036, $η_{p}^{2}$ = .06, suggesting that learning occurred with our sample of participants.

The mixed ANCOVA revealed that the equality of covariance matrices was not violated, p > .05. The results indicated that the main effect of Block, F(4, 168) = 0.92, p = .437, $η_{p}^{2}$ = .02; main effect of Group, F(1, 42) = 0.42, p = .521, $η_{p}^{2}$ = .01; the Block × Gender interaction, F(4, 168) = 0.43, p = .745, $η_{p}^{2}$ = .01; and Group × Block interaction, F(4, 168) = 0.47, p = .712, $η_{p}^{2}$ = .01 were not significant, suggesting that performance between the groups varied little over the course of the task when controlling for variance associated with gender and depression.

Variant IGT

Overall, there was a significant main effect of Block, F(4, 180) = 8.58, p < .001, $η_{p}^{2}$ = .16, suggesting that, overall, learning occurred within our sample of participants.

The mixed ANCOVA revealed that the equality of covariance matrices was not violated, p > .05. The results revealed a significant main effect of Block, F(4, 168) = 3.28, p = .026, $η_{p}^{2}$ = .07, indicating that there were differences across the five blocks. The main effect of Group was also significant, F(1, 42) = 4.9, p = .032, $η_{p}^{2}$ = .11, suggesting that, overall, individuals with PTSD performed worse than controls. The Block × Gender interaction, F(4, 168) = 1.38, p = .255, $η_{p}^{2}$ = .03, and Block × Group interaction, F(4, 168) = 0.77, p = .505, $η_{p}^{2}$ = .02, were not significant. Overall, the results suggest there was a significant difference in performance between groups over the course of the task when controlling for variance associated with gender and depression (Figure 1).

Figure 1.

Group performance on the variant IGT (Iowa Gambling Task) quantified as the estimated marginal means of advantageous card selections over 5 blocks of 20 trials after adjusting for depression and gender as covariates.

Antidepressant use

By conducting a post hoc analysis using simple comparisons, the mean proportion of total advantageous selections were compared between all participants reporting antidepressant use (n = 15) versus nonmedicated individuals (n = 31). The medicated group consisted of 2 controls and 10 with PTSD, whereas the nonmedicated group consisted of 21 controls and 13 with PTSD. The results revealed that the medicated group chose significantly more cards overall from the advantageous decks (M = 2.90, SD = 0.62) on the IGT, t(44) = −2.42, p = .019 (two-tailed), compared to the nonmedicated group (M = 2.49, SD = 0.50). On the vIGT, there were no significant differences, t(44) = 0.29, p = .772, between the medicated group (M = 3.00, SD = 0.77) and the nonmedicated group (M = 3.08, SD = 0.92).

Additional comparisons were run to explore the differences in performance in the PTSD group that might be attributed to antidepressant treatment. The results indicated that medicated individuals with PTSD (n = 13) performed significantly better on the IGT, t(21) = −2.74, p = .012 (M = 2.91, SD = 0.64), compared to nonmedicated individuals (n = 10) with PTSD (M = 2.25, SD = 0.48). There were no significant differences on the vIGT, t(21) = −0.55, p = .589, between the medicated (M = 3.02, SD = 0.81) compared to nonmedicated individuals with PTSD (M = 2.82, SD = 0.93). Overall, these results suggest possible efficacy of antidepressant treatment on emotion-based decision making (Figure 2).

Figure 2.

Performance of medicated versus nonmedicated participants with PTSD (posttraumatic stress disorder) on both versions of the IGT (Iowa Gambling Task) quantified as on the total proportion of advantageous cards selected.

Discussion

The results from the current study indicate that active-duty U.S. soldiers with PTSD have emotion-based decision-making impairments that appear to be due to an inability to balance punishment and delayed reward. This was evidenced by normal performance on the standard IGT but significantly poor performance on the variant IGT compared to healthy soldiers. In addition, participants who reported taking prescribed antidepressants showed significantly better performance on the standard IGT compared to nonmedicated participants, regardless of group designation, suggesting possible efficacy of pharmacologic interventions in enhancing emotion-based decision making.

Not unlike other studies (Mueller et al., 2010), the current study used both the standard and variant versions of the IGT, both of which are posited to rely on emotion-based learning (Bechara et al., 2000). Poor performance on both versions of the IGT has been suggested to reflect a form of cognitive disinhibition to future consequences, both good and bad. In contrast, normal performance on the standard IGT and impaired performance on the vIGT, together, suggest a pattern of decision making that reflects hypersensitivity to immediate punishment and/or hyposensitivity to delayed rewards. The punishment and rewards in both versions of the IGT are in the form of facsimile monetary losses and earnings. In other words, individuals with this form of decision making show behavioral pattern that suggests they are guided by the magnitude of immediate losses presented in the vIGT, and therefore, fail to integrate information and subsequently adjust their behavior to attain delayed earnings. This is further supported in that the PTSD group appeared to be able to adjust behavior based on delayed punishments as evidenced by their normal performance on the standard IGT. The results of the current study suggest that soldiers with PTSD have a decision-making impairment that might be due to difficulty processing the delayed earnings associated with decks of cards that present large immediate losses.

One difficulty with exploring the underlying psychopathology of PTSD is that often the target population is being pharmacologically treated. Without a pretreatment baseline assessment it is difficult to ascertain whether any group effects are due to the positive pharmacodynamic properties of the medication(s) or variability in PTSD subtypes. For example, low levels of serotonin generated by a serotonin agonist (Cyproheptadine, 8 mg) have been shown to negatively influence IGT performance in the later trials (Bechara, Damasio, & Damasio, 2001). Yet more recent evidence has shown that in subjects with rapid depletion of tryptophan, a technique for transiently reducing brain serotonin, performance on other decision-making tasks improves (Talbot, Watson, Barrett, & Cooper, 2005). In contrast to the serotonergic system, O’Carroll and Papps (2005) reported that the noradrenergic system is not part of the neurobiological process underlying human decision making in that a norepheneprhine reuptake inhibitor (NRI; Reboxetine, 3 mg and 8 mg) failed to influence performance in the IGT. In the current study, 43% of the PTSD group was on an antidepressant (SSRI, SNRI, and NDRIs) compared to 9% in the healthy control group. Overall and within the PTSD group, individuals being treated with an antidepressant outperformed those that were non-antidepressant users. Therefore, it is quite possible that psychotropic factors associated with antidepressant treatment efficaciously mediated emotion-based decision-making performance on the standard IGT in our PTSD group. Future studies should investigate the effects of individual antidepressants on decision-making performance.

One limitation of the present study is that the time since their diagnosis was not collected for this study. Given that there are temporal neurobiological changes associated PTSD (Cardenas et al., 2011), this study would have a greater contribution to the literature if these data had been collected. Furthermore, maladaptive behaviors involving substance use/misuse and/or undetected mood conditions (e.g., dysthymia) may exacerbate impairments in neuropsychological functioning. Future studies should expand on such limitations by controlling for temporal changes associated with PTSD.

Sample size is always a limitation with this type of study. Although the number of participants for which data were obtained is large enough to ensure reliable and interpretable analyses, it may be limited in observing factors and interactions with small effect sizes. The sample size was, however, determined, a priori, to be large enough to detect large effect sizes associated with the effects of PTSD on IGT performance. Furthermore, similar sample sizes have been used in prior PTSD studies (Neylan et al., 2004) and studies using the IGT (Dretsch & Tipples, 2008).

The findings from the current study indicate that individuals suffering from PTSD can learn experientially to modify their behaviors to choose advantageously on the standard IGT but not the vIGT. The implications of this finding suggest that soldiers with PTSD have the ability to make decisions guided by punishing reinforcers but struggle with adjusting behaviors to attain delayed rewards. Furthermore, it appears that normal performance on the standard IGT by individuals with PTSD could be attributed to antidepressant use in that medicated individuals with PTSD chose more often from the advantageous decks overall, compared to nonmedicated individuals with PTSD. For treatment purposes, both of these findings are important in that it suggests the IGT may have sensitivity for clinical use and there is potential efficacy from different antidepressants with varying neurotransmitter targets for reconstituting emotional processes associated with decision making.

Conclusion

In conclusion, our findings suggest that soldiers diagnosed with PTSD have difficulty making optimal decisions, which appears to be attributed to hypersensitivity to immediate punishment and insensitivity to delayed reward. A supplementary finding revealed that participants reporting taking antidepressants showed significantly better performance, suggesting potential efficacy of antidepressant treatment for enhancing decision making. Our findings imply that emotion-based decision making is compromised in active-duty soldiers with PTSD.

Footnotes

Authors’ Note

The opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Army and/or the U.S. Department of Defense.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received the financial support from the U.S. Army Medical Research and Materiel Command (USAMRMC).

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

Bechara

Damasio

A. R.

(2001). Manipulation of dopamine and serotonin causes different effects on covert and overt decision-making. Social Neuroscience Abstracts, 27, 126.

Bechara

Damasio

Lee

(1999). Different contributions of the human amygdala and ventromedial prefrontal cortex to decision-making. The Journal of Neuroscience, 19, 5473-5481.

Bechara

Dolan

Hindes

(2002). Decision-making and addiction (part II): Myopia for the future or hypersensitivity to reward? Neuropsychologia, 40, 1690-1705.

Bechara

Tranel

Damasio

(2000). Characterization of the decision-making deficit of patients with ventromedial prefrontal cortex lesions. Brain, 123, 2189-2202.

Beck

A. T.

Epstein

Brown

Steer

R. A.

(1988). An Inventory for Measuring Clinical Anxiety: Psychometric Properties. Journal of Consulting and Clinical Psychology, 56, 893-897.

Beck

A. T.

Steer

R. A.

(1984). Internal consistencies of the original and revised Beck Depression Inventory. Journal of Clinical Psychology, 40, 1365-1367.

Brandes

Ben-Schachar

Gilboa

Bonne

Freedman

Shalev

A. Y.

(2002). PTSD symptoms and cognitive performance in recent trauma survivors. Psychiatry Research, 110, 231-8.

Campbell

D. G.

Felker

B. L.

Liu

Yano

E. M.

Kirchner

J. E.

Chan

. . . Chaney

E. F.

(2007). Prevalence of depression-PTSD comorbidity: Implications for clinical practice guidelines and primary care-based interventions. Journal of General Internal Medicine, 22, 711-718.

10.

Cardenas

V. A.

Samuelson

Lenoci

Studholme

Neylan

T. C.

Marmar

C. R.

. (2011). Changes in brain anatomy during the course of posttraumatic stress disorder. Psychiatry Research, 193, 93-100.

11.

Carver

White

(1994). Behavioral Inhibition, Behavioral Activation, and affective response to impending reward and punishment: The BIS/BAS Scales. Journal of Personality and Social Psychology, 67, 319-333.

12.

Crone

E. A.

Bunge

S. A.

Latenstein

van der Molen

M. W.

(2005). Characterization of children’s decision making sensitivity to punishment frequency, not task complexity. Child Neuropsychol, 11, 245-263.

13.

de Rocha

F. F.

Alvarenga

N. B.

Malloy-Diniz

Corrêa

(2011). Decision-making impairment in obsessive-compulsive disorder as measured by the Iowa Gambling Task. Arg Neuropsiquiatr, 69, 642-7.

14.

de Visser

van der Knaap

L. I.

van de Loo

A. J.

van der Weerd

C. M.

Ohl

van den Bos

(2010). Trait anxiety affects decision-making differently in healthy men and women: Towards gender-specific endophenotypes of anxiety. Neuropsychologia, 48, 1598-1606.

15.

Dretsch

M. N.

Tipples

(2008). Working memory involved in predicting future outcomes based on past experiences. Brain & Cognition, 66, 83-90.

16.

Ernst

Bolla

Mouratidis

Kurian

Cadet

Kimes

A. S.

London

E. D.

(2002). Decision-making in a risk-taking task: A PET study. Neuropsychopharmacology, 26, 682-691.

17.

Ernst

Paulus

M. P.

(2005). Neurobiology of decision-making: A selective review from a neurocognitive and clinical perspective. Biological Psychiatry, 58, 597-604.

18.

Han

Klimes-Dougan

Jepsen

Ballard

Nelson

Houri

. . . Cullen

(2012). Selective neurocognitive impairments in adolescents with major depressive disorder. Journal of Adolescence, 35, 11-20.

19.

Hankin

C. S.

Spiro

Miller

D. R.

Kazis

(1999). Mental disorders and mental health treatment among US Department Veterans Affairs outpatients: The Veterans Health Study. American Journal of Psychiatry, 156, 1924-30.

20.

Koenigs

Grafman

(2009). Posttraumatic stress disorder: The role of the medial prefrontal cortex and amygdala. The Neuroscientist, 15, 540-548.

21.

Lanius

R. A.

Bluhm

Lanius

Pain

(2006). A review of neuroimaging studies in PTSD: Heterogeneity of response to symptom provocation. Journal of Psychiatric Research, 40, 709-729.

22.

Lawrence

N. S.

Jollant

O’Daly

Zelaya

Phillips

M. L.

(2008). Distinct roles of prefrontal cortex subregions in the Iowa Gambling Task. Cerebral Cortex, 19, 1134-1143.

23.

Z. L.

D’Argembeau

Bechara

(2010). The Iowa Gambling Task in fMRI images. Human Brain Mapping, 31, 410-423.

24.

Lin

C. H.

Chiu

Y. C.

Cheng

C. M.

Hsieh

J. C.

(2008). Brain maps of Iowa Gambling Task. BMC Neuroscience, 9, 72.

25.

O’Carroll

R. E.

Papps

B. P.

(2005). Decision-making in humans: The effect of manipulating the central noradrenergic system. Journal of Neurology, Neurosurgery & Psychiatry, 74, 376-378.

26.

Orsillo

S. M.

(2001). Measures for acute stress disorder and posttraumatic stress disorder. In Antony

M. M.

Orsillo

S. M.

(Eds.), Practitioner’s guide to empirically based measures of anxiety (pp. 255-307). New York, NY: KluwerAcademic/Plenum.

27.

Miu

A. C.

Heilman

R. M.

Houser

(2008). Anxiety impairs decision-making: Psychophysiological evidence from an Iowa Gambling Task. Biological Psychology, 77, 353-358.

28.

Mueller

E. M.

Nguyen

Ray

W. J.

Borkovec

T. D.

(2010). Future-oriented decision-making in generalized anxiety disorder is evident across different versions of the Iowa Gambling Task. Journal of Behavior Therapy and Experimental Psychiatry, 41, 165-171.

29.

Neylan

T. C.

Lenoci

Rothlind

Metzler

T. J.

Schuff

A. T.

. . . Marmar

C. R.

(2004). Attention, learning, and memory in posttraumatic stress disorder. Journal of Traumatic Stress, 17, 41-6.

30.

Roca

Hart

Kimbrell

Freeman

(2006). Cognitive function and dissociative disorder status among Veteran subjects with chronic posttraumatic stress disorder: A preliminary study. Journal of Neuropsychiatry and Clinical Neuroscience, 18, 226-230.

31.

Spitzer

R. L.

First

M. B.

Wakefield

J. C.

(2007). Saving PTSD from itself in DSM-V. Journal of Anxiety Disorders, 21, 233-241.

32.

Talbot

P. S.

Watson

D. R.

Barrett

S. L.

Cooper

S. J.

(2005). Rapid tryptophan depletion improves decision-making cognition in healthy humans without affecting reversal learning or set shifting. Neuropsychopharmacology, 31, 1519-25.

33.

Tanielian

Jacox

L. H.

(2008). Invisible wounds of war: Psychological and cognitive injuries, their consequences, and services to assist recovery. Santa Monica, CA: RAND.

34.

Tombaugh

T. N.

(2003). The test of memory malingering (TOMM) in forensic psychology. Journal of Forensic Neuropsychology, 2, 69-96.

35.

Vasterling

J. J.

Duke

L. M.

Brailey

Constans

J. I.

Allain

A. N.

Sutker

P. B.

(2002). Attention, learning, and memory performances and intellectual resources in Vietnam veterans: PTSD and no disorder comparisons. Neuropsychology, 16, 5-14.

36.

Villarreal

Hamilton

D. A.

Petropoulos

Driscoll

Rowland

L. M.

Griego

J. A.

Kodituwakku

P. W.

Hart

B. L.

Escalona

Brooks

W. M.

(2002). Reduced hippocampal volume and total white matter volume in posttraumatic stress disorder. Biological Psychiatry, 52, 119-125.

37.

Werner

N. S.

Duschek

Schandry

(2009). Relationships between affective states and decision-making. International Journal of Psychophysiology, 74, 259-265.

38.

Windmann

Kirsch

Mier

Stark

Walter

Güntürkün

Vaitl

(2006). On framing effects in decision-making: Linking lateral versus medial orbitofrontal cortex activation to choice outcome processing. Journal of Cognitive Neuroscience, 18, 1198-1211.

39.

Yehuda

Golier

J. A.

Tischler

Stavitsky

Harvey

P. D.

(2005). Learning and memory in aging combat veterans with PTSD. Journal of Clinical and Experimental Neuropsychology, 27, 504-15.