Journal of Affective Disorders 176 (2015) 118–124

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research report

Further evidence of a dissociation between decision-making under ambiguity and decision-making under risk in obsessive–compulsive disorder Hae Won Kim a, Jee In Kang a, Kee Namkoong a, Kyungun Jhung b, Ra Yeon Ha c, Se Joo Kim a,n a b c

Department of Psychiatry and Institute of Behavioral Science in Medicine, Yonsei University College of Medicine, Seoul, South Korea Department of Adolescent Psychiatry, National Center for Child and Adolescent Psychiatry, Seoul National Hospital, Seoul, South Korea Department of Psychiatry, Seoul Bukbu Hospital, Seoul, South Korea

art ic l e i nf o

a b s t r a c t

Article history: Received 16 September 2014 Received in revised form 29 January 2015 Accepted 29 January 2015 Available online 7 February 2015

Background: Deficits in decision-making have been suggested as a key concept in understanding the symptoms of obsessive–compulsive disorder (OCD). However, evidence in the extant literature remains inconclusive on whether patients with OCD show inferior performance on laboratory decision-making tasks. The aims of the present study were therefore to (1) assess decision-making under ambiguity and under risk in patients with OCD and (2) study the influence of neuropsychological and clinical variables on decision-making in OCD. Methods: The sample consisted of 65 patients with OCD and 58 controls. The Iowa gambling task (IGT) and the game of dice task (GDT) were used to examine decision-making under ambiguity and decisionmaking under risk, respectively. In addition, reversal learning and executive function were assessed in terms of their relationship with decision-making tasks. Results: Patients with OCD showed impairment in the IGT, but not in the GDT. Reversal learning was neither impaired nor correlated with IGT performance. Among the clinical variables, illness severity and depression were associated with IGT scores. Executive function was impaired, but no significant relationship was found between executive function and GDT performance in OCD patients. Limitations: Almost all OCD patients were on medication when they performed decision-making tasks. Conclusions: Patients with OCD are impaired in decision-making under ambiguity, but not under risk. These findings demonstrate that decision-making processes are dissociated in OCD. & 2015 Elsevier B.V. All rights reserved.

Keywords: Decision-making Iowa gambling task Game of dice task Obsessive–compulsive disorder

1. Introduction Obsessive–compulsive disorder (OCD) is a chronic psychiatric disorder characterized by anxiety-provoking intrusive thoughts (obsession) and/or repetitive ritualistic behaviors (compulsion) aimed at reducing anxiety (American Psychiatric Association, 2000). Approaches to reveal the neural substrates of these symptoms have focused on frontal–subcortical circuits, with particular attention to increased orbitofrontal–striatal metabolism in resting state and during symptom provocation (Saxena et al., 1998; Saxena and Rauch, 2000; Whiteside et al., 2004). As the orbitofrontal cortex and ventromedial prefrontal regions are involved in reward perception and adaptation to shift in reward contingencies (Rolls,

n

Corresponding author. Tel.: þ 82 2 2228 1620; fax: þ82 2 313 0891. E-mail address: [email protected] (S.J. Kim).

http://dx.doi.org/10.1016/j.jad.2015.01.060 0165-0327/& 2015 Elsevier B.V. All rights reserved.

2000), which are critical in decision-making (Krawczyk, 2002; Rolls and Grabenhorst, 2008), disrupted decision-making in OCD has been of interest, together with orbitofrontal dysfunction. In some articles, obsessive–compulsive symptoms have even been proposed as a manifestation of an underlying decision-making deficit; seeking for an immediate relief of anxiety by compulsion (short-term reward) leads to functional impairment and compromises quality of life in the long term (negative long-term consequence) (Cavedini et al., 2006; Sachdev and Malhi, 2005). From a psychological aspect, decision-making can be classified into at least two different categories: decision-making under ambiguity and decision-making under risk (Brand et al., 2006). While ambiguous conditions provide equivocal rules for reward and punishment, risky conditions offer expected outcomes for competing alternatives. In neuropsychological research, decision-making under ambiguity and decision-making under risk are frequently examined with the Iowa gambling task (IGT) (Bechara et al., 1994) and the game

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

of dice task (GDT) (Brand et al., 2005a), respectively. The IGT requires the development of implicit preference for advantageous choice under ambiguous reinforcement contingencies (decision-making under ambiguity), which is mediated by emotional learning (Bechara et al., 2000a) and relevant structures, such as the orbitofrontal/ventromedial prefrontal cortex and amygdala (Bechara, 2004; Bechara et al., 1999, 2000b; Gupta et al., 2011). In a previous study of individuals with ventromedial prefrontal lesions, poor IGT performance was also influenced by reversal learning (Fellows and Farah, 2005). On the other hand, gains and losses in the GDT are based on explicit rules and apparent winning probabilities (decision-making under risk), which are associated with executive function (Brand et al., 2007a, 2005a, 2005b, 2007b, 2008b) and activation of several brain regions, such as the dorsolateral prefrontal cortex, posterior parietal lobe, anterior cingulate cortex, and right lingual gyrus (Labudda et al., 2008). Performance on decision-making tasks has been widely investigated in OCD, but the results remain inconclusive. In accordance with the functional alteration of orbitofrontal–striatal circuits, previous studies have frequently demonstrated reduced IGT score in patients with OCD (Cavedini et al., 2002, 2012, 2010; da Rocha et al., 2011a; Kashyap et al., 2013; Starcke et al., 2009, 2010). On the other hand, two studies reported impaired performance in only subgroups of patients with OCD (Lawrence et al., 2006; Nielen et al., 2002), and one study revealed comparable performance in medication-naïve patients with OCD (Krishna et al., 2011). As IGT taps several cognitive processes, it is unclear on which component patients with OCD principally rely to solve the IGT. Concerning the factors contributing to IGT performance, there is no agreement on whether demographic and clinical variables are involved in disadvantageous choices. The evidence for a correlation between OCD symptom severity and preference for unfavorable choices is contradictory. While three studies reported positive results (da Rocha et al., 2011b, 2008; Nielen et al., 2002), four studies did not (Cavedini et al., 2002; da Rocha et al., 2011a; Krishna et al., 2011; Lawrence et al., 2006). Furthermore, to the best of our knowledge, only two studies were conducted with the GDT (Starcke et al., 2009, 2010); patients with OCD in these studies performed similarly compared to controls. However, the small sample sizes would limit the generalizability of these results. In this study, we aimed to obtain clearer evidence on decisionmaking ability in a larger sample of patients with OCD. The aims of this study were to (1) examine the properties of decision-making under ambiguity and under risk in patients with OCD compared to controls and (2) study the influence of fundamental cognitive processes and clinical characteristics on decision-making performance. Particularly, we focused on reversal learning with regard to performance on the IGT; reversal learning depends on the orbitofrontal cortex (Rolls, 2000; Rolls et al., 1994), a region of interest in OCD. We also included executive function to explore its relationship with GDT performance. In accordance with recent studies that indicate the involvement of the dorsolateral prefrontal cortex in OCD (Menzies et al., 2008; Piras et al., 2015), we hypothesized that (1) patients with OCD would perform inferiorly in both decision-making tasks and that (2) poor performance on the IGT and GDT would correlate with impairment in reversal learning and executive function, respectively.

2. Methods 2.1. Participants Sixty-five patients with OCD were recruited from the psychiatry department of Severance Hospital, Yonsei University Health System. All participants were 19 years or older at enrollment. Diagnoses were confirmed by a trained psychiatrist with the Structured Clinical

119

Interview for DSM-IV (SCID) (First et al., 1996). Participants with comorbid psychiatric disorder were included only when their obsessive–compulsive symptoms were the primary focus of clinical attention. Seventeen of them had comorbid psychiatric disorders, including major depressive disorder (n¼8), panic disorder (n¼3), bipolar II disorder (n¼ 2), tic disorder (n¼2), social phobia (n¼1), and body dysmorphic disorder (n¼1). Any patient with a substancerelated disorder, history of psychotic symptoms, or neurological or medical disease that might influence performance was excluded. The Yale–Brown obsessive-compulsive scale (Y–BOCS) (Goodman et al., 1989) and Montgomery–Åsberg depression rating scale (MADRS) (Montgomery and Asberg, 1979) were used to assess the severity of obsessive–compulsive symptoms and depressive symptoms, respectively. Sixty-three patients were receiving medication, including selective serotonin reuptake inhibitors (SSRIs) (n¼ 63), benzodiazepines (n¼ 43), and atypical antipsychotics (n¼8). Benzodiazepines on an as-needed basis were not allowed prior to and during the task. Fifty-eight age-matched controls were recruited from the local community by advertisement. Exclusion criteria for controls were current or any lifetime history of DSM-IV axis I disorder, neurological disease, and use of medications that may affect cognitive function. The study was approved by the Institutional Review Board of Severance Hospital. All participants provided written informed consent at the beginning of the study. 2.2. Neuropsychological tasks 2.2.1. Iowa gambling task The computerized version of the IGT was conducted to examine decision-making under ambiguity (Bechara et al., 2000b). Participants were told to earn as much virtual money as possible without prior notice of when the task would end. On each trial of card selection from four decks (A, B, C, and D), a variable amount of reward was given, which was intermittently followed by unexpected loss. Decks A and B were related to high gains, but even greater losses resulted in an overall disadvantageous outcome. In contrast, decks C and D were related to small gains, but even smaller losses produced a favorable result in the long run. The outcome measure, calculated for the entire game (100 trials) and for every 20 trials (five blocks in total), was the total difference between advantageous and disadvantageous choices ([CþD] [AþB]). We also drew a distinction between the first two blocks (trials 1–40) and the last three blocks (trials 41–100) according to the previous study, which reported different aspects of decisionmaking across trials (Brand et al., 2007b). 2.2.2. Game of dice task The GDT (Brand et al., 2005a) was used to measure decisionmaking in risky situations. The goal of this game was to maximize fictitious money across 18 throws of a single die. Before each trial, participants chose between single numbers or a combination of two, three, or four numbers. Each choice was associated with specific gains or losses as well as different winning probabilities: 1/6 probability of gaining/losing $1000 (single number), 2/6 probability of gaining/losing $500 (two numbers), 3/6 probability of gaining/losing $200 (three numbers), and 4/6 probability of gaining/losing $100 (four numbers). The choices with less than 50% chance of winning (one or two numbers) were considered risky decisions while those with at least 50% chance of winning (three or four numbers) were considered non-risky decisions. The net score was calculated by subtracting the number of risky choices from that of non-risky choices. 2.2.3. Simple reversal learning task The simple reversal learning task (SRLT) was administered to measure stimulus-reinforcement learning and reversal learning

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

2.2.4. Wisconsin card sorting test The computerized version of the Wisconsin card sorting test (WSCT) (Berg, 1948) was used to examine executive functions, such as abstraction ability, set-shifting, and cognitive flexibility. The test consisted of one response card and four stimulus cards that differed in color, form, and number. Participants were told to match the response card without knowing the matching rule, and only received feedback on whether the trial was correct or not. After ten consecutive correct trials, the matching criterion was changed without notice. The game continued until the six categories were completed or all 128 cards were sorted. The number of categories completed and number of perseverative errors were the main outcomes. 2.3. Statistical analyses All statistical analyses were conducted with Statistical Package for the Social Sciences (SPSS) version 18.0 for windows (Chicago, SPSS Inc.). Although the distributions of some variables scores were not perfectly normal, we used parametric tests according to the central limit theorem, as the sample size in this study was large enough (Field, 2013). T-tests were used to assess differences between groups on continuous demographic and neuropsychological variables. Chisquare tests were conducted on non-continuous variables. The Pearson correlation coefficient was used to examine the relationship between variables in each group. Analyses of group differences in the selection of different choices were performed with a repeated measures analysis of variance (ANOVA). The level of statistical significance was set at p¼ 0.05. When the repeated measures ANOVA was statistically significant, we performed a post-hoc analysis with a Bonferroni correction, where a p-value of 0.05 divided by the number of comparisons (in the case of IGT, p¼0.05/5) was considered to indicate a significant difference.

3. Results 3.1. Demographic and clinical variables The demographic data and clinical characteristics of participants are shown in Table 1. Both groups were matched for age (p ¼0.964) and education years (p¼ 0.860), although the OCD Table 1 Demographic and clinical characteristics of patients with OCD and controls.

Age Gender (male/female) Education (years) Onset age Duration of illness Y–BOCS MADRSa

OCD (n¼ 65)

Control (n¼58)

t or χ2

df

p Value

26.62 7 9.12 51/14 13.98 7 2.04 17.25 7 7.84 9.577 7.46 22.077 7.42 16.87 7 9.04

26.56 7 6.28 36/22 14.58 71.76 – – – –

0.45 3.98  1.73

121 1 121

0.964 0.046 0.860

OCD: obsessive–compulsive disorder; Y–BOCS: Yale–Brown obsessive–compulsive scale, MADRS: Montgomery–Åsberg depression rating scale. a

Scores were available for only 52 patients.

Iowa Gambling Task 8.00 6.00

Mean Netscore

(Fellows and Farah, 2003). The task comprised two card decks, one related to a $50 gain, and the other to a $50 loss. After meeting the learning criterion of eight consecutive card selections from the advantageous deck, the contingencies were changed. This reversal phase continued until the criterion was met again. The outcome measure was the number of errors made before meeting the criterion during the reversal phase (reversal error).

Group effect p = 0.012 Block effect p < 0.001 Group Block effect p = 0.230

4.00 2.00 0.00 -2.00

Block 1

Block 2

-4.00

Block 3

Controls

Block 4

Block 5

OCD

Fig. 1. Iowa gambling task results over five blocks. Mean ( 7 SEM) is given.

Game of Dice Task 8.00 OCD

Mean Frequency

120

Controls

6.00

4.00

2.00

0.00 One Number Two Numbers Three Numbers Four Numbers Netscore

Fig. 2. Game of dice task performances on different alternatives. Mean (7 SEM) is given.

group contained more male participants (p ¼0.046). The total Y–BOCS scores ranged from 6 to 36 in OCD patients. 3.2. Decision-making tasks and neuropsychological tests The total net score of the patients with OCD on the IGT was significantly lower than that of controls (p ¼0.011). To analyze group differences in card selections over time, we conducted a 2 (group)  5 (blocks) repeated measures ANOVA. There were significant main effects for blocks (F(4,118) ¼6.381, p o0.001) and group (F(1,121) ¼6.468, p ¼0.012). The group by block interaction did not reach significance (F(4,118) ¼1.424, p ¼0.230). Fig. 1 shows the learning curves for this analysis. Separate investigations for each block revealed a significant difference between groups in only block 4 (p¼ 0.007) after correcting for multiple comparisons (adjusted p-value of 0.01). No significant differences were found in blocks 3 (p ¼0.096) and 5 (p ¼ 0.017), but the net score of the last three blocks altogether (trials 41–100) differed significantly between groups (p ¼0.007). On the GDT, patients with OCD performed similarly to controls (p ¼0.439). A 2 (group)  4 (choice) repeated measures ANOVA revealed a significant main effect of choice (F(3,119) ¼10.037, po 0.001), but not for group (F(1,121) ¼0.892, p ¼0.347). The group by choice interaction failed to reach significance (F(3,119) ¼1.232, p¼ 0.301). None of the single alternatives differed significantly between groups (one number: t¼ 1.37, df¼121, p¼ 0.173; two numbers: t ¼0.66, df¼121, p¼ 0.511; three numbers: t ¼  1.74, df¼121, p ¼0.985; four numbers: t¼0.55, df ¼121, p ¼0.585). The results are shown in Fig. 2. Group differences in SRLT performance were analyzed with a 2 (group)  2 (phase) repeated measures ANOVA. Significant main effects were found for group (F(1,121) ¼21.086, p o0.001) and phase (F(1,121) ¼7.867, p ¼0.006). The group by phase interaction did not

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

reach significance (F(1,121) ¼2.116, p¼ 0.148). No significant difference was found in a single comparison of reversal error scores between groups (p ¼0.086). Regarding the WCST measures, patients with OCD committed more total errors and perseverative errors than did controls (p ¼0.009 and p o0.001, respectively). The results of the neuropsychological tests are presented in Table 2. Since there were significantly more male participants in our patients, we also made comparisons between male participants in the OCD group and male controls in neuropsychological tests; significant differences were found in the IGT and WCST, but not in the GDT and SRLT (data not shown). 3.3. Correlation analyses of decision-making Neither IGT nor GDT scores were correlated with the scores of SRLT or WCST for the patients. Among the clinical variables, higher MADRS scores were associated with lower IGT scores. As MADRS data for 13 patients were missed at random due to loss of records, MADRS scores were available for only 52 patients with OCD. Y–BOCS score was also correlated with total net score on the IGT. In the control group, the correlations between IGT total score, scores on the

121

latter part of the IGT, and scores on the GDT reached significance. The results are presented in Table 3.

4. Discussion The aims of this study were to (1) assess decision-making properties under ambiguity and under risk in patients with OCD and (2) obtain further evidence of relevant factors in decisionmaking in OCD. The main findings of our study can be summarized as follows: (1) decision-making under ambiguity is impaired in OCD, whereas decision-making under risk is not, as measured with the IGT and the GDT, respectively; (2) reversal learning assessed with SRLT was not impaired in OCD, and the number of reversal errors did not correlate with IGT scores; (3) total and perseverative errors on the WCST were more committed in OCD, but there were no correlations between WCST measures and GDT scores. For the IGT, patients with OCD exhibited a significantly lower total net score than controls. This finding is consistent with those of most previous studies (Cavedini et al., 2002, 2012, 2010; da Rocha et al., 2011a; Kashyap et al., 2013; Starcke et al., 2009, 2010),

Table 2 Performance on neuropsychological tests.

Iowa gabling task Total net score ([C þ D]  [Aþ B]) Blocks 1–2 ([C þ D]  [Aþ B]) Block 3–5 ([C þD]  [A þB]) Game of dice task Net score Simple reversal learning task Reversal error Wisconsin card sorting test Categories completed Total error Perseverative error

OCD

Control

t

df

p Value

 4.727 25.02  3.85 7 10.13  0.747 20.58

8.107 30.01  2.247 11.07 10.34 7 24.12

 2.58  0.84  2.75

121 121 121

0.011 0.403 0.007

4.54 7 10.33

6.02 710.77

 0.78

121

0.439

8.777 6.25

6.717 6.96

1.73

121

0.086

5.28 7 1.41 25.45 7 13.87 12.83 7 7.68

5.667 1.03 18.72 7 13.97 6.95 73.13

 1.71 2.67 5.44

116.86 121 121

0.090 0.009 o 0.001

OCD: obsessive–compulsive disorder.

Table 3 Correlations between neuropsychological tests and clinical features. IGT-T

IGT 3–5

GDT-T

GDT-one

SRLT-RE

WCST-C

WCST-PE

OCD IGT 3–5 GDT-T GDT-one SRLT-RE WCST-C WCST-PE Onset age Illness duration Y–BOCS MADRS

0.918nn 0.058  0.099  0.132 0.058  0.150 0.119 0.033  0.264n  0.414nn

 0.013  0.073  0.082 0.069  0.121 0.067 0.127  0.253  0.389nn

 0.716nn  0.038 0.085  0.108 0.203  0.02  0.013  0.055

 0.056  0.140 0.209  0.164  0.068 0.073 0.074

0.009 0.247n  0.22 0.09 0.134 0.128

 0.261n  0.152 0.083  0.228 0.092

 0.171 0.068 0.171 0.332n

Controls IGT 3–5 GDT-T GDT-one SRLT-RE WCST-C WCST-PE

0.939nn 0.415nn  0.314n  0.082 0.147 0.055

0.442nn  0.346nn  0.033 0.140 0.077

 0.721nn  0.093 0.097 0.145

0.048  0.160  0.097

0.005 0.104

0.330n

OCD: obsessive–compulsive disorder, IGT-T: Iowa gambling task-total net score ([Cþ D]  [Aþ B]), IGT 3–5: Iowa gambling task block 3–5 ([Cþ D]  [Aþ B]), GDT-T: game of dice total net score, GDT-one: game of dice one number choice, SRLT-RE: simple reversal learning task-reversal error, WCST-C: Wisconsin card sorting task-categories completed, WCST-PE: Wisconsin card sorting task-perseverative error, Y–BOCS: Yale–Brown obsessive-compulsive scale, MADRS: Montgomery–Åsberg depression rating scale n

po 0.05. p o0.01.

nn

122

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

which suggest impaired decision-making under ambiguity in OCD. However, the results of the IGT should be interpreted with caution, as there have been recent suggestions that the IGT refers to both types (ambiguous and risky) of decision-making across trials (Brand et al., 2007a, 2006). Moreover, selections during the later trials may represent decision-making under risk, for the contingencies of reward and punishment may become known after many trials. In fact, when we analyzed the first and last parts of the task separately, only the scores of the last three blocks were significantly lower in patients. This finding was similar to that of da Rocha et al. (2011a), which revealed impaired performance in only the second half of the task (trials 51–100) in OCD patients. Therefore, our results of poorer IGT performance, especially in the last trials, may reflect at least partially an impairment of decision-making under risk in OCD. Given that the rules of the IGT become increasingly evident as the task progresses, our results suggest that OCD patients may not develop an explicit strategy for making advantageous choices based on their implicit learning, which is consistent with previous research (Starcke et al., 2010). Furthermore, the last blocks of the IGT were positively and negatively correlated with net score and one number choice on the GDT in controls, respectively, yet not in the patients with OCD. These results might also lend support to the aforementioned interpretation that controls improve their IGT performance by shifting to an explicit condition, a strategy that did not appear to be available to patients with OCD. Alternatively, it also could be the case that OCD patients do develop such a strategy (as do controls), with OCD patients weighing risk versus reward differently. In order to reach a clear conclusion, further research is required that evaluates whether or not the subjects are explicitly aware of the contingencies and assesses the point at which ambiguity becomes explicit awareness. Contrary to our hypothesis, our patients did not differ from controls in terms of net score and numbers of alternative choices on the GDT. As the GDT explores decision-making under risk in which potential outcomes of alternative choices and their winning probabilities depend on explicit information, dorsolateral prefrontal-related cognitive functions would be involved. Our prediction of impaired decisionmaking under risk was based on growing evidence that the dorsolateral prefrontal cortex is involved in the pathophysiology of OCD (Melloni et al., 2012; Menzies et al., 2008; Piras et al., 2015), beyond the well-documented orbitofrontal–striatal dysfunction. In OCD, to the best of our knowledge, only two studies have been conducted with GDT that investigated decision-making under risk with performance comparable to that of controls (Starcke et al., 2009, 2010). However, we hypothesized reduced GDT performance despite these negative results, for the small sample sizes (n¼18, Starcke et al., 2009; n¼23, Starcke et al., 2010) of the previous studies could have led to weak statistical power. Further, the patients in their study were not impaired on the executive function tasks. Considering that there is substantial evidence of reduced executive function in OCD (Kashyap et al., 2013; Olley et al., 2007; Shin et al., 2014; Tukel et al., 2012), a caveat must be mentioned in generalizing the previous results. However, replication of intact GDT performance in a larger sample of OCD patients in the present study may verify the ability to make beneficial choices under explicit conditions in OCD. The second objective of this study was to identify the correlates of decision-making in OCD. In this regard, we mainly focused on reversal learning and executive function. Second, we included several factors that might influence IGT and GDT performance, such as demographic variables (age and education years) and clinical characteristics (age of onset, duration of illness, severity of OCD symptoms, and severity of depression). Given that the IGT is a measure of a complex construct (Busemeyer and Stout, 2002), several component processes that might affect IGT performance have received attention, including

reversal learning. Reversal learning is an orbitofrontal/ventromedial prefrontal-dependent function to reverse and/or suppress responses when a previously rewarded stimulus is no longer rewarding (Clark et al., 2004; Fellows and Farah, 2003; Rolls, 2000; Rolls et al., 1994). In order to achieve a favorable outcome in the IGT, initially-acquired preference toward the disadvantageous decks should be suppressed, which may rely on reversal learning (Fellows and Farah, 2005; Maia and McClelland, 2004). In OCD, previous studies have suggested impaired reversal learning in line with orbitofrontal dysfunction (Abbruzzese et al., 1997; Aycicegi et al., 2003; Remijnse et al., 2006), but contradictory findings also exist (Chamberlain et al., 2007; Kuelz et al., 2004). Given the critical role of the orbitofrontal cortex in the pathophysiology of OCD, we hypothesized that impaired reversal learning would affect IGT performance in patients with OCD. However, while reversal error scores on the SRLT were reduced in these patients, they did not correlate with IGT scores, which suggests that the reduced IGT performance was not attributed to deficits of reversal learning in our patients. Considering that reversal learning would rely on successful processing of negative feedback to avoid a nonrewarding alternative, this non-significant correlation may imply that focusing on the reward, rather than ineffectively learning from the punishment, impeded solving the IGT for these patients. Other mechanisms, such as somatic markers, might also affect IGT performance in OCD (Cavedini et al., 2012; Starcke et al., 2009). To address further aspects of the relationship between reversal learning and IGT performance directly, the shuffled version of the IGT (Fellows and Farah, 2005) would be helpful in future research. One of the interesting finding of this study was the lack of significant association between GDT scores and the WCST measures in OCD patients. In fact, most previous studies that used the GDT have shown that decision-making under the explicit condition is correlated with executive functions, such as categorizing options, set-shifting, and cognitive flexibility, which are often measured with the WCST (Brand et al., 2007a, 2005a, 2005b, 2004, 2008b; Brand and Schiebener, 2013). From the results of our study, we have discerned that decision-making under risk in OCD would involve different cognitive processes other than the abovementioned executive functions. Previous studies with healthy controls have demonstrated that decision-making under risk taps several fundamental processes, such as feedback processing (Brand, 2008; Brand et al., 2009), applying strategies (Brand et al., 2008a), logical reasoning (Brand et al., 2008a; Schiebener et al., 2011), and handling probabilities (Schiebener et al., 2011), beyond the crucial role of executive function. It was noteworthy that these component processes, including executive function, interact with each other in performing the GDT. According to Schiebener et al. (2011), even a reduced level of executive function was compensated by higher scores on the handling probabilities in avoiding risky choices. Accordingly, comparable performance on the GDT in our patients might have relied on other competent subcomponents, such as the processing of probabilities. Future research is thus needed to explore the contributions of several component processes, other than executive function, to GDT performance in patients with OCD. Taken together, the observed deficient performance on the IGT yet not on the GDT may indicate that patients with OCD are less competent in making advantageous decisions when expected outcomes and their probabilities are uncertain. Seemingly, this observation appears as a failure to avert punishment under ambiguous conditions. However, the lack of relationship between ambiguous decision-making and either reversal learning or set-shifting, which both depend on intact learning from negative feedback to unlearn the previously learned rules, may imply that unfavorable decisions are derived from seeking immediate gratification, rather than from failure to avoid punishment. In accordance with Cavedini et al.'s (2006) suggestion, these observations may be the correlates of a compulsive

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

behavior that seeks immediate alleviation of anxiety, despite subsequent functional impairment and compromised quality of life. The temporary relief of anxiety would reinforce the compulsive behaviors and thus lead to repetition of ritualistic behaviors. Among the clinical variables, we found a significant correlation between IGT scores and MADRS scores, implying that depression may influence decision-making capacity under the ambiguous condition (although as the IGT proceeds, it tends to shift gradually from decision-making under ambiguity to decision-making under risk). Indeed, some previous studies have shown impaired IGT performance in patients with major depressive disorder (Must et al., 2006), and in nonclinical subjects with higher negative mood (Suhr and Tsanadis, 2007). In addition, we found a weak negative correlation between IGT scores and Y–BOCS scores, which was consistent with the finding of some previous studies (da Rocha et al., 2011b, 2008; Nielen et al., 2002). In contrast, other studies found no association between IGT performance and illness severity (Cavedini et al., 2002; da Rocha et al., 2011a; Krishna et al., 2011; Lawrence et al., 2006). This lack of agreement on the relationship between task performance and disease severity may limit the applicability of IGT in clinical use. In order to obtain evidence for the ecological validity of the IGT, an assessment of functional capacity and its relationship with decision-making is needed. No other demographic or clinical variables were associated with decisionmaking performance; these results are in accordance with those of most previous studies (Cavedini et al., 2002, 2012; da Rocha et al., 2011a, 2008; Kashyap et al., 2013; Lawrence et al., 2006; Nielen et al., 2002). There are some limitations in our study, however. First, we had more male participants among the patients than among the controls, but gender did not affect performance on the decisionmaking tasks. Second, general intellectual capacities were not assessed; intelligence may have altered decision-making capacity. Third, scores of depressive symptoms were not available for the controls. Complete information on the MADRS score would have enabled a comparison of decision-making between the groups after adjusting for the influence of depressive symptoms. Another limitation is the use of WCST as the only measure of executive function. Diverse aspects of executive function and their relationship with decision-making under risk should be noted in future research. Lastly, almost all OCD patients in this study were taking medications such as SSRIs, benzodiazepines, and atypical antipsychotics when they were performing the decision-making tasks. Therefore, we could not exclude the possible influences of these medications on patients' performance. In summary, our patients with OCD showed deficits in decisionmaking under ambiguity, but not in decision-making under risk. These results lend support to previous findings of dissociated decision-making processes in OCD. In addition, performance on the IGT and the GDT did not depend on reversal learning and executive function, respectively.

Role of funding source The funding source did not give any influences on the study design, data collection, analysis, interpretation of data, the writing of the report, and the decision to submit the paper for publication.

Conflict of interest No conflict declared.

Acknowledgment This research was supported by grants (to SJ Kim) from the National Research Foundation (NRF) of Korea funded by the Ministry of Education, Science and Technology (2010-0022363).

123

References Abbruzzese, M., Ferri, S., Scarone, S., 1997. The selective breakdown of frontal functions in patients with obsessive–compulsive disorder and in patients with schizophrenia: a double dissociation experimental finding. Neuropsychologia 35, 907–912. American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR), 4th ed. Amer Psychiatric Association, Washington, DC. Aycicegi, A., Dinn, W.M., Harris, C.L., Erkmen, H., 2003. Neuropsychological function in obsessive–compulsive disorder: effects of comorbid conditions on task performance. Eur. Psychiatry 18, 241–248. Bechara, A., 2004. The role of emotion in decision-making: evidence from neurological patients with orbitofrontal damage. Brain Cogn. 55, 30–40. Bechara, A., Damasio, A.R., Damasio, H., Anderson, S.W., 1994. Insensitivity to future consequences following damage to human prefrontal cortex. Cognition 50, 7–15. Bechara, A., Damasio, H., Damasio, A.R., 2000a. Emotion, decision making and the orbitofrontal cortex. Cereb. Cortex 10, 295–307. Bechara, A., Damasio, H., Damasio, A.R., Lee, G.P., 1999. Different contributions of the human amygdala and ventromedial prefrontal cortex to decision-making. J. Neurosci. 19, 5473–5481. Bechara, A., Tranel, D., Damasio, H., 2000b. Characterization of the decision-making deficit of patients with ventromedial prefrontal cortex lesions. Brain 123 (Pt 11), 2189–2202. Berg, E.A., 1948. A simple objective technique for measuring flexibility in thinking. J. Gen. Psychol. 39, 15–22. Brand, M., 2008. Does the feedback from previous trials influence current decisions? A study on the role of feedback processing in making decisions under explicit risk conditions. J. Neuropsychol. 2, 431–443. Brand, M., Franke-Sievert, C., Jacoby, G.E., Markowitsch, H.J., Tuschen-Caffier, B., 2007a. Neuropsychological correlates of decision making in patients with bulimia nervosa. Neuropsychology 21, 742–750. Brand, M., Fujiwara, E., Borsutzky, S., Kalbe, E., Kessler, J., Markowitsch, H.J., 2005a. Decision-making deficits of korsakoff patients in a new gambling task with explicit rules: associations with executive functions. Neuropsychology 19, 267–277. Brand, M., Heinze, K., Labudda, K., Markowitsch, H.J., 2008a. The role of strategies in deciding advantageously in ambiguous and risky situations. Cogn. Process. 9, 159–173. Brand, M., Kalbe, E., Labudda, K., Fujiwara, E., Kessler, J., Markowitsch, H.J., 2005b. Decision-making impairments in patients with pathological gambling. Psychiatry Res. 133, 91–99. Brand, M., Labudda, K., Kalbe, E., Hilker, R., Emmans, D., Fuchs, G., Kessler, J., Markowitsch, H.J., 2004. Decision-making impairments in patients with Parkinson's disease. Behav. Neurol. 15, 77–85. Brand, M., Labudda, K., Markowitsch, H.J., 2006. Neuropsychological correlates of decision-making in ambiguous and risky situations. Neural Netw. 19, 1266–1276. Brand, M., Laier, C., Pawlikowski, M., Markowitsch, H.J., 2009. Decision making with and without feedback: the role of intelligence, strategies, executive functions, and cognitive styles. J. Clin. Exp. Neuropsychol. 31, 984–998. Brand, M., Recknor, E.C., Grabenhorst, F., Bechara, A., 2007b. Decisions under ambiguity and decisions under risk: correlations with executive functions and comparisons of two different gambling tasks with implicit and explicit rules. J. Clin. Exp. Neuropsychol. 29, 86–99. Brand, M., Roth-Bauer, M., Driessen, M., Markowitsch, H.J., 2008b. Executive functions and risky decision-making in patients with opiate dependence. Drug Alcohol Depend. 97, 64–72. Brand, M., Schiebener, J., 2013. Interactions of age and cognitive functions in predicting decision making under risky conditions over the life span. J. Clin. Exp. Neuropsychol. 35, 9–23. Busemeyer, J.R., Stout, J.C., 2002. A contribution of cognitive decision models to clinical assessment: decomposing performance on the Bechara gambling task. Psychol. Assess. 14, 253–262. Cavedini, P., Gorini, A., Bellodi, L., 2006. Understanding obsessive–compulsive disorder: focus on decision making. Neuropsychol. Rev. 16, 3–15. Cavedini, P., Riboldi, G., D’Annucci, A., Belotti, P., Cisima, M., Bellodi, L., 2002. Decision-making heterogeneity in obsessive–compulsive disorder: ventromedial prefrontal cortex function predicts different treatment outcomes. Neuropsychologia 40, 205–211. Cavedini, P., Zorzi, C., Baraldi, C., Patrini, S., Salomoni, G., Bellodi, L., Freire, R.C., Perna, G., 2012. The somatic marker affecting decisional processes in obsessive– compulsive disorder. Cogn. Neuropsychiatry 17, 177–190. Cavedini, P., Zorzi, C., Piccinni, M., Cavallini, M.C., Bellodi, L., 2010. Executive dysfunctions in obsessive–compulsive patients and unaffected relatives: searching for a new intermediate phenotype. Biol. Psychiatry 67, 1178–1184. Chamberlain, S.R., Fineberg, N.A., Blackwell, A.D., Clark, L., Robbins, T.W., Sahakian, B.J., 2007. A neuropsychological comparison of obsessive–compulsive disorder and trichotillomania. Neuropsychologia 45, 654–662. Clark, L., Cools, R., Robbins, T.W., 2004. The neuropsychology of ventral prefrontal cortex: decision-making and reversal learning. Brain Cogn. 55, 41–53. da Rocha, F.F., Alvarenga, N.B., Malloy-Diniz, L., Correa, H., 2011a. Decision-making impairment in obsessive–compulsive disorder as measured by the Iowa Gambling Task. Arq. Neuropsiquiatr. 69, 642–647.

124

H.W. Kim et al. / Journal of Affective Disorders 176 (2015) 118–124

da Rocha, F.F., Malloy-Diniz, L., Lage, N.V., Correa, H., 2011b. The relationship between the Met allele of the BDNF Val66Met polymorphism and impairments in decision making under ambiguity in patients with obsessive–compulsive disorder. Genes Brain Behav. 10, 523–529. da Rocha, F.F., Malloy-Diniz, L., Lage, N.V., Romano-Silva, M.A., de Marco, L.A., Correa, H., 2008. Decision-making impairment is related to serotonin transporter promoter polymorphism in a sample of patients with obsessive– compulsive disorder. Behav. Brain Res. 195, 159–163. Fellows, L.K., Farah, M.J., 2003. Ventromedial frontal cortex mediates affective shifting in humans: evidence from a reversal learning paradigm. Brain 126, 1830–1837. Fellows, L.K., Farah, M.J., 2005. Different underlying impairments in decisionmaking following ventromedial and dorsolateral frontal lobe damage in humans. Cereb. Cortex 15, 58–63. Field, A.P., 2013. Discovering Statistics Using IBM SPSS Statistics, 4th ed. SAGE, London. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J., 1996. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Clinician Version (SCID-CV). American Psychiatric Press, Inc., Washington, D.C. Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., Fleischmann, R.L., Hill, C.L., Heninger, G.R., Charney, D.S., 1989. The Yale–Brown Obsessive Compulsive Scale. I. Development, use, and reliability. Arch. Gen. Psychiatry 46, 1006–1011. Gupta, R., Koscik, T.R., Bechara, A., Tranel, D., 2011. The amygdala and decisionmaking. Neuropsychologia 49, 760–766. Kashyap, H., Kumar, J.K., Kandavel, T., Reddy, Y.C., 2013. Neuropsychological functioning in obsessive–compulsive disorder: are executive functions the key deficit? Compr. Psychiatry 54, 533–540. Krawczyk, D.C., 2002. Contributions of the prefrontal cortex to the neural basis of human decision making. Neurosci. Biobehav. Rev. 26, 631–664. Krishna, R., Udupa, S., George, C.M., Kumar, K.J., Viswanath, B., Kandavel, T., Venkatasubramanian, G., Reddy, Y.C., 2011. Neuropsychological performance in OCD: a study in medication-naive patients. Prog. Neuropsychopharmacol. Biol. Psychiatry 35, 1969–1976. Kuelz, A.K., Riemann, D., Zahn, R., Voderholzer, U., 2004. Object alternation test—is it sensitive enough to detect cognitive dysfunction in obsessive–compulsive disorder? Eur. Psychiatry 19, 441–443. Labudda, K., Woermann, F.G., Mertens, M., Pohlmann-Eden, B., Markowitsch, H.J., Brand, M., 2008. Neural correlates of decision making with explicit information about probabilities and incentives in elderly healthy subjects. Exp. Brain Res. 187, 641–650. Lawrence, N.S., Wooderson, S., Mataix-Cols, D., David, R., Speckens, A., Phillips, M.L., 2006. Decision making and set shifting impairments are associated with distinct symptom dimensions in obsessive–compulsive disorder. Neuropsychology 20, 409–419. Maia, T.V., McClelland, J.L., 2004. A reexamination of the evidence for the somatic marker hypothesis: what participants really know in the Iowa gambling task. Proc. Natl. Acad. Sci. USA 101, 16075–16080. Melloni, M., Urbistondo, C., Sedeno, L., Gelormini, C., Kichic, R., Ibanez, A., 2012. The extended fronto–striatal model of obsessive compulsive disorder: convergence from event-related potentials, neuropsychology and neuroimaging. Front. Hum. Neurosci. 6, 259. Menzies, L., Chamberlain, S.R., Laird, A.R., Thelen, S.M., Sahakian, B.J., Bullmore, E.T., 2008. Integrating evidence from neuroimaging and neuropsychological studies

of obsessive–compulsive disorder: the orbitofronto–striatal model revisited. Neurosci. Biobehav. Rev. 32, 525–549. Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to be sensitive to change. Br. J. Psychiatry 134, 382–389. Must, A., Szabo, Z., Bodi, N., Szasz, A., Janka, Z., Keri, S., 2006. Sensitivity to reward and punishment and the prefrontal cortex in major depression. J. Affect. Disord. 90, 209–215. Nielen, M.M., Veltman, D.J., de Jong, R., Mulder, G., den Boer, J.A., 2002. Decision making performance in obsessive compulsive disorder. J. Affect. Disord. 69, 257–260. Olley, A., Malhi, G., Sachdev, P., 2007. Memory and executive functioning in obsessive–compulsive disorder: a selective review. J. Affect. Disord. 104, 15–23. Piras, F., Piras, F., Chiapponi, C., Girardi, P., Caltagirone, C., Spalletta, G., 2015. Widespread structural brain changes in OCD: a systematic review of voxelbased morphometry studies. Cortex 62, 89–108. Remijnse, P.L., Nielen, M.M., van Balkom, A.J., Cath, D.C., van Oppen, P., Uylings, H.B., Veltman, D.J., 2006. Reduced orbitofrontal–striatal activity on a reversal learning task in obsessive–compulsive disorder. Arch. Gen. Psychiatry 63, 1225–1236. Rolls, E.T., 2000. The orbitofrontal cortex and reward. Cereb. Cortex 10, 284–294. Rolls, E.T., Grabenhorst, F., 2008. The orbitofrontal cortex and beyond: from affect to decision-making. Prog. Neurobiol. 86, 216–244. Rolls, E.T., Hornak, J., Wade, D., McGrath, J., 1994. Emotion-related learning in patients with social and emotional changes associated with frontal lobe damage. J. Neurol. Neurosurg. Psychiatry 57, 1518–1524. Sachdev, P.S., Malhi, G.S., 2005. Obsessive–compulsive behaviour: a disorder of decision-making. Aust. N. Z. J. Psychiatry 39, 757–763. Saxena, S., Brody, A.L., Schwartz, J.M., Baxter, L.R., 1998. Neuroimaging and frontal– subcortical circuitry in obsessive–compulsive disorder. Br. J. Psychiatry Suppl. 35, 26–37. Saxena, S., Rauch, S.L., 2000. Functional neuroimaging and the neuroanatomy of obsessive–compulsive disorder. Psychiatr. Clin. N. Am. 23, 563–586. Schiebener, J., Zamarian, L., Delazer, M., Brand, M., 2011. Executive functions, categorization of probabilities, and learning from feedback: what does really matter for decision making under explicit risk conditions? J. Clin. Exp. Neuropsychol. 33, 1025–1039. Shin, N.Y., Lee, T.Y., Kim, E., Kwon, J.S., 2014. Cognitive functioning in obsessive– compulsive disorder: a meta-analysis. Psychol. Med. 44, 1121–1130. Starcke, K., Tuschen-Caffier, B., Markowitsch, H.J., Brand, M., 2009. Skin conductance responses during decisions in ambiguous and risky situations in obsessive–compulsive disorder. Cogn. Neuropsychiatry 14, 199–216. Starcke, K., Tuschen-Caffier, B., Markowitsch, H.J., Brand, M., 2010. Dissociation of decisions in ambiguous and risky situations in obsessive–compulsive disorder. Psychiatry Res. 175, 114–120. Suhr, J.A., Tsanadis, J., 2007. Affect and personality correlates of the Iowa Gambling Task. Person. Individ. Differ. 43, 27–36. Tukel, R., Gurvit, H., Ertekin, B.A., Oflaz, S., Ertekin, E., Baran, B., Kalem, S.A., Kandemir, P.E., Ozdemiroglu, F.A., Atalay, F., 2012. Neuropsychological function in obsessive–compulsive disorder. Compr. Psychiatry 53, 167–175. Whiteside, S.P., Port, J.D., Abramowitz, J.S., 2004. A meta-analysis of functional neuroimaging in obsessive–compulsive disorder. Psychiatry Res. 132, 69–79.