Journal of Affective Disorders 179 (2015) 142–147
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
Does a history of substance abuse and illness chronicity predict increased impulsivity in bipolar disorder? Isabelle E. Bauer 1, Thomas D. Meyer n,1, Marsal Sanches, Giovana Zunta-Soares, Jair C. Soares University of Texas Health Science Center at Houston, Department of Psychiatry and Behavioral Sciences, 1941 East Road, Houston, TX 77054, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 25 November 2014 Received in revised form 3 March 2015 Accepted 5 March 2015 Available online 14 March 2015
Background: Impulsivity is a common feature shared by bipolar disorder (BD) and substance use disorder (SUD). SUD and recurrent mood episodes are considered to be risk factors for poor outcome in BD. However, the association between impulsivity, illness chronicity and SUD in BD remains unexplored. Methods: 103 BD patients with and without a lifetime history of SUD (36.82 711.34 years, 40 males) were recruited. Participants completed the SCID interview and were administered measures of impulsivity including the Barratt Impulsivity Scale (BIS) and selected tests of the Cambridge Neuropsychological Test Automated Battery (CANTAB). Hierarchical regression analyses explored the relationship between illness chronicity, SUD, and impulsivity. Results: Variance in the BIS, number of false alarms on the Rapid Visual Processing task and other impulsivity indicators of the Cambridge Gambling Task (CGT) was not explained by the chosen variables. Only an increased number of commission errors in the negative condition of the Affective Go/No Go task was significantly associated with illness chronicity. Furthermore there was a trend suggesting a relationship between a lifetime history of SUD and increased propensity to risk-taking during the CGT. Limitations: Potential limitations include medication and patients' remission status from SUD. Conclusions: Contrary to our expectations impulsivity was generally not predicted by indicators of illness chronicity or SUD. While impulsivity could still be a marker of BD that is present before the onset of the disorder, the link between the number of mood episodes and specific indicators of impulsivity may be related to mechanisms of neuroprogression. & 2015 Elsevier B.V. All rights reserved.
Keywords: Impulsivity Bipolar disorder Substance abuse Illness chronicity
1. Introduction Bipolar disorder (BD) is a serious mental illness clinically characterized by mood dysregulation, brain abnormalities, and cognitive deficits such as poor response inhibition that in some cases persist during the euthymic and acute phases (Bora et al., 2009; MacQueen et al., 2001; Quraishi and Frangou, 2002). The majority of BD patients also present with additional mental health conditions such as anxiety and substance use disorder (SUD) (Asaad et al., 2014; Merikangas et al., 2007) that are considered to be predictors of poor response to treatment (Swann, 2010) and low remission rates (Ostacher, 2011) in BD and SUD patients.
n Corresponding author at: University of Texas Health Science Center at Houston, Department of Psychiatry and Behavioral Science, 1941 East Road, Houston, TX 77054, United States. Tel.: þ1 713 486 2638; fax: þ1 713 486 2553. E-mail addresses: [email protected] (I.E. Bauer), [email protected] (T.D. Meyer). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jad.2015.03.010 0165-0327/& 2015 Elsevier B.V. All rights reserved.
Impulsivity is a common feature shared by BD and SUD (Powers et al., 2013; Swann et al., 2007) and has been associated with poor cognitive control and disregard of the long-term implications of one's behavior, and may therefore result in risky and disorganized behaviors (Evenden, 1999; Moeller et al., 2001). Impulsivity can be considered as being a multidimensional concept and assessed via self-reports and behavioral paradigms. Previous studies showed that the Barratt Impulsivity Scale (BIS) total score (Patton et al., 1995) – a widely used self-rating measure of impulsivity in psychiatry – discriminates well between healthy subjects and euthymic BD patients (Ekinci et al., 2011; Etain et al., 2013). Additionally, BIS scores have been found to be significantly elevated in BD patients with a history of SUD (Etain et al., 2013). The comparability and equivalence of self-rated and behavioral indicators of impulsivity is still equivocal (for review (Newman and Meyer, 2014)). A study found that euthymic BD patients encountered difficulties on behavioral measures of impulsivity such as the Hayling Sentence Completion (HSCT) and the Iowa Gambling Task (Christodoulou et al., 2006). Furthermore, the BIS score was found to correlate positively with the number of
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
commission errors on the HSCT and the Iowa Gambling task. Two studies using the BIS alongside the Immediate Memory Task– Delayed Memory Task (IMT–DMT) – a modified version of the Continuous Performance Task (CPT) found that while the BIS scores were elevated in BD patients compared to healthy controls (the two studies included euthymic BD and BD patients with SUD respectively), the number of commission errors of the IMT–DMT was comparable between the two groups (Swann et al., 2004,, 2003). Thus, not all facets of impulsivity discriminate between BD and healthy controls. Furthermore, the BIS and behavioral tasks may not capture the same aspects of impulsivity and might be differentially related to outcome measures such as the course or duration of the bipolar disorder. The literature on the relationship between impulsivity, BD, and SUD is sparse, and it is unclear whether impulsivity is a marker of the bipolar disorder or rather the result of neural damage associated with repeated mood episodes and/or drug use. Theories of neuroprogression in mood disorders suggest that ongoing mood episodes lead to a state of chronic inflammation and eventually result in neurocognitive impairment (Berk et al., 2010, 2011; Kapczinski et al., 2008). Abusing drugs has negative effects on the brain reward mechanisms and on the dopaminergic and serotonergic neurotransmitter systems (Koob and Le Moal, 1997) and may induce hazardous reward seeking behaviors and poor decision making (Kirby et al., 2011). Therefore, a positive association between measures of impulsivity and illness chronicity in BD would support the notion that the levels of impulsivity increase as the bipolar disorder progresses. Similarly, a link between increased impulsivity levels and SUD would be related to the deleterious effects of drugs on the brain. By contrast if there was no relation among these measures, prior research would suggest that impulsivity is more likely to be a marker of BD, being present before the onset of the disorder. Based on the evidence reviewed above the aim of this crosssectional study is to investigate the predictive power of indicators of illness chronicity such as illness duration and the number of prior mood episodes on impulsivity in BD, and determine whether a history of SUD additionally explains elevations in impulsivity above and beyond indicators of illness chronicity. Further, we aim to determine whether self-ratings (i.e. BIS) and behavioral measures from the CANTAB battery measure heterogeneous or similar facets of impulsivity. Based on the literature we predict that 1. indicators of illness chronicity will explain variance in both selfrated and behavioral impulsivity measures and 2. that a history of SUD will significantly add to this effect.
2. Methods 2.1. Sample The sample included 103 adult BD patients (M7SD: 36.82711.34 years; 40 males, 63 females; 74 BD-I, 20 BD-II, and 9 BD-NOS) (see Table 1). Participants were recruited from inpatient and outpatient clinics of the University of North Carolina at Chapel Hill (UNC) and the University of Texas Health Science Center at Houston (UT). All patients met the DSM-IV-R criteria for BD. The diagnosis of BD among patients were ascertained by the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders Axis I (SCID I) (First et al., 2012), which was administered to all participants by an independent psychiatrist or trained research assistant. The interview also included the Hamilton Depression Rating Scale (HAM-D) (Hamilton, 1960) and the Young Mania Rating Scale (YMRS) (Young et al., 1978). Illness chronicity was defined by the number of mood episodes (including manic, depressive and mixed episodes) and illness duration. 44 Participants had a lifetime history of SUD and had
143
Table 1 Demographic and clinical characteristics of the sample. Group
Mean (S.D.)
Age (years) Sex
36.82 7 11.34 40 Males 63 Females 13.9 73.3 110.747 13.71 18.45 7 8.16 76.497 5.5 5.08 7 4.85 12.677 8.46
Education (years) Full scale IQ (WASI) Age of onset BIS YMRS HAMD a Number of mood episodes o 13 4 13 Illness duration (years) a Lifetime history of substance and/or alcohol use disorder
n¼ 24 n¼ 79 18.357 11.71 Yes – n¼ 44 No – n¼59
a Notes: Absolute numbers might vary due to missing data; YMRS ¼Young Mania Rating Scale; HAMD: Hamilton Depression Rating Scale; BIS: Barratt Impulsivity Scale.
previously met criteria for substance abuse or dependence. Subjects were excluded if they fulfilled criteria for SUD in 6 months prior to the beginning of the study or had any current serious medical problems including cardiovascular and neurological disorders. 43 BD participants took psychotropic medication at the time of enrollment. The study protocol was approved by the local Institutional Review board and informed consent was obtained from all the participants. 2.2. Impulsivity The Barratt Impulsive Scale (BIS) (Patton et al., 1995) is a 30-item self-report measure of impulsivity that includes three scales: cognitive impulsivity – e.g. difficulties with concentrating, motor impulsiveness, e.g. motor restlessness and non-planning impulsiveness – e.g. selfcontrol. The BIS has been often used in research in BD and was therefore chosen as a self-report. Behavioral aspects of impulsivity were measured by the Affective Go/No-Go task (AGN), the Cambridge Gambling Task (CGT), and the Rapid Visual Processing Task (RVP) of the computerized Cambridge Neurocognitive Test Automated Battery (CANTAB – http://www.cantab.com). This battery was chosen based on its established sensitivity to cognitive impairment in psychiatric disorders (Sweeney et al., 2000). The tasks of the CANTAB are briefly described below and a detailed description is given elsewhere (Robbins et al., 1994). AGN evaluates inhibition control. Participants are instructed to respond to either happy (e.g. joyful, warmth, courage) or negative words (e.g. mistake, hopeless, burden), and inhibit their response to stimuli of opposite valence. Stimuli were presented in both a random (negative, positive, positive, negative) and sequential way (positive, positive, negative, negative). The outcome measure included in this report to reflect impulsivity is the number of commission errors in response to distracting stimuli. CGT evaluates impulse control and risk-taking behavior. Primary outcome measures include deliberation time prior to making a bet, risk taking (mean proportion of the number of points risked on trials with a more likely outcome), quality of decision making (proportion of bets on gamble trials with a more likely outcome), and delay aversion (ratio score representing the ability to bet large amounts depending on the odds of winning/losing). RVP evaluates sustained visual attention. The main outcome measure of this task in relation to the concept of impulsivity is the number of false alarms. Because none of the transformation positively affected its distribution towards greater normality (see below) this variable was coded into a dichotomous variable (1¼presence of false alarms, 2¼ absence of false alarms).
144
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
impulsivity the first block of independent variables of these regression analyses included illness chronicity – measured by the number of mood episodes coded into a dichotomous variable (1: ≤13 episodes, 2: 413 episodes) and illness duration. The second block of variables included SUD – this variable was coded as a dichotomous variable (1 ¼SUD, 2 ¼ no SUD). Based on Tabachnick and Fidell's (2007) formula a sample of 74 subjects was found to be sufficient to provide adequate statistical power (80%) to detect an association between predictors. Statistical significance was defined as p o.05 and p-values ranging from .05 to .10 were considered to be statistical trends.
2.3. Statistical procedure Statistical analyses were performed using IBM SPSS statistics (Version 21.0). Normality assumptions for continuous variables were examined. Where appropriate, outliers were winsorised and log, square root or reciprocal transformations applied to achieve normality if appropriate. If normality could not been achieved, either untransformed or dichotomized scores were used. Correlational analysis was used to investigate the relationship between potential predictors of impulsivity, demographics and clinical variables such as illness duration and number of mood episodes. First, we conducted a series of regression analyses to examine whether the variables sex, age, current symptom levels (HAM-D, YMRS) affected the dependent variable, i.e. self-rated and behavioral indicators of impulsivity and to determine whether these variables needed to be included in the final hierarchical regression models. Then, we conducted a series of hierarchical regression analyses to examine our main hypothesis with them either BIS or CANTAB measures as the dependent variable. The dichotomous dependent variable “false alarms” (RVP task) was analyzed using a logistic regression. If none of the before mentioned covariates (e.g. sex, age, symptom levels) proved to be significantly related to
3. Results First we explored the correlations between self-rating and CANTAB measures of impulsivity (Table 2). In line with prior research overall correlations were low and ranged from r ¼ .13 to .19 between BIS and behavioral measures of impulsivity and from r ¼ .14 to .30 between indices from different behavioral tasks.
Table 2 Coefficients of correlation among self-rating (BIS) and CANTAB measures of impulsivity. BIS
AGN total CE AGN negative CE AGN neutral CE AGN positive CE CGT risk taking CGT delay aversion CGT deliberation time CGT QDM RVP FAa
.03 .11 .03 .11 .18 .06 .13 .10 .18
AGN total CE
AGN negative CE
AGN neutral CE
AGN positive CE
CGT risk taking
CGT delay aversion
CGT deliberation time
CGT QDM
.92nn .60nn .29nn .11 .18 .08 .14 .06
.56nn .31nn .09 .19 .08 .09 .08
.09 .02 .08 .06 .26nn .07
.01 .08 .08 .10 .29nn
.34nn .07 .08 .22n
.20n .34nn .03
.08 .05
.07
Acronyms: AGN¼ Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE ¼commission errors; CGT ¼Cambridge Gambling Task; FA ¼False Alarms; QDM ¼Quality Decision Making; RVP¼Rapid Visual Processing. n
p o.05. p o.01. For the dichotomous variable “False alarms” Spearman's coefficients of correlation are reported.
nn
a
Table 3 Results from the hierarchical regression analyses for Model 1 (illness duration, number of mood episodes) and 2 (Model 1 plus SUD). Variable
BIS AGN total commissions AGN total commissions – positive AGN total commissions – negative AGN total commissions – neutral CGT – delay aversion CGT – risk taking CGT – deliberation time (ms) CGT – quality of decision making
RVP – false alarmsa Mood episodes Illness duration SUD
Mean (SD)
76.37 (5.44) 13.45 (9.03) 4.16 (3.34) 3.25 (2.67) 4.13 (4.24) 0.44 (0.24) 0.54 (0.14) 2324.24 (953.01) 0.89 (0.14)
With errors¼75% No errors¼ 35%
Model 1a
Model 2b
ΔR2
p-Value
R2
p-Value
R2
p-Value
.014 .030 .028 .070 .003 .030 .054 .005 .000
.59 .21 .29 o .05 .89 .26 o .10 .93 1.00
.014 .042 .035 .096 .018 .044 .090 .002 .000
.78 .24 .37 o .05 .67 .27 o .05 .05 .99
.000 .011 .007 .026 .015 .014 .035 .000 .000
.85 .28 .42 .12 .25 .26 o .10 .92 .89
Model 1 B (SE)
R2
Model 2: B (SE)
R2
χ2 model 1
χ2 model 2
3.7
14.15 (p ¼ .08)
3.69 (p ¼ .88)
.002 .025 (.57) .008 (.02)
.05 (.58) .007 (.02) .802 (.46)
Acronyms: AGN ¼ Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE ¼commission errors; CGT¼ Cambridge Gambling Task; FA ¼False Alarms; QDM¼ Quality Decision Making; RVP¼Rapid Visual Processing; SUD ¼ Substance Use Disorder. a
Note: The dichotomous variable “False alarms” was analyzed using a logistic regression; R2 refers to Cox & Snell's estimate.
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
Next we ran separate regression analyses for all indicators of impulsivity to determine whether the potential confounders age, sex, HAMD and YMRS scores were associated with impulsivity. Since none of them were associated with the dependent variables (all Fo1.2, all R2 o5%, available on request from the authors), they were not included in the final regression model. The results of the hierarchical regression analysis showing the association between illness chronicity, substance abuse disorders and impulsivity measures are shown in Tables 3 and 4. Self-rated impulsivity using the BIS was not significantly predicted by indicators of illness chronicity and a history of SUD. The overall number of commission errors across the positive, negative or neutral conditions of the AGN task was not predicted by these variables. However, when considering only the total number of commission errors during the negative condition of the AGN, Model 1, which included illness duration and the number of mood episodes as predictors, was found to be statistically significant (R¼.265, R2 ¼ .070, F (1,88)¼3.323, p¼.041). Model 2, which additionally included SUD, still explained a significant amount of variance [F (3,87)¼3.076, p¼.032], but the addition of SUD did not significantly increase the amount of explained variance (ΔR2: F(3,87)¼.026, p¼ .12). Only the number of prior mood episodes emerged as a predictor, albeit not significant (po.10), of AGN commission errors in response to negative words (see Table 3). For the variable “risk taking” of the CGT task, Model 1 was not significant (p¼ .086), while model 2 including SUD proved to be significant therefore explaining a significant amount of the variance in risk taking (F (3, 87) ¼2.852, p ¼.042). However, adding SUD as a predictor in model 2 only lead to a non-significant change in R2 (p¼ .07). When looking at the beta coefficients the number of mood episodes was found to be the only significant predictor of risk taking (β ¼ .246, p ¼.02). There was only a trend suggesting a positive association between SUD and risk taking (β ¼ .188, p¼ .07). Other impulsivity measures of the CGT task such as deliberation time, quality of decision making and delay aversion, were not predicted by the selected variables. Similarly, the number of false alarms in the RVP task was not associated with either SUD or illness chronicity.
4. Discussion The current study aimed to investigate the association between illness chronicity, SUD, and indicators of impulsivity in BD. Based on the neuroprogression model (Berk et al., 2011) we expected to observe a strong relationship between the course of the bipolar disorder and levels of self-rated and CANTAB measures of impulsivity. However, two out of the 10 indicators of impulsivity showed some associations: the number of commission errors to distracting stimuli during the negative condition of the AGN task and risk taking in the CGT. Both were related to illness chronicity, more specifically to the number of prior mood episodes, but not to illness duration. Further, history of SUD did not generally predict elevated impulsivity in our sample and contributed only minimally to risk-taking. The current findings are important for three reasons. First, the lack of a strong relationship between impulsivity and illness chronicity, and SUD, suggests that neither of these variables contribute to the increased levels of impulsivity observed in BD (Najt et al., 2007; Peluso et al., 2007). Based on prior research this finding may suggest that impulsivity is less a consequence of BD but rather a core trait of the disorder and is possibly associated with the neuroanatomical and neurochemical abnormalities previously observed in BD (Hajek et al., 2005) such as the dysregulation of the fronto-limbic system and/or abnormalities in the dopaminergic projections to the frontal, limbic
145
and striatal regions as they are involved in self-monitoring and reward seeking behaviors (Mason et al., 2014). Second, both the regression and correlation analyses indicate that self-rating and behavioral measures of impulsivity tap into different facets of impulsivity. This result is not surprising if one considers that, in the literature, the BIS score is viewed as being a trait measure of self-control (Swann et al., 2004), while neuropsychological measures of impulsivity appear to be state-dependent and more prone to fluctuate across phases of illness (Swann et al., 2003). Along the same line, a recent systematic review of studies in impulsivity showed that while self-rating measures of impulsivity differentiate between euthymic BD and healthy volunteers, neuropsychological measures of impulsivity do not
Table 4 Beta coefficients for the predictors of impulsivity in Model 2: illness duration, number of mood episodes and SUD. β BIS Mood episodes Illness duration SUD
.090 .059 .022
AGN total commissions Mood episodes Illness duration SUD
.155 .037 .106
AGN total commissions – positive Mood episodes Illness duration SUD
.139 .122 .086
AGN total commissions – negative Mood episodes Illness duration SUD AGN total commissions – neutral Mood episodes Illness duration SUD CGT – delay aversion Mood episodes Illness duration SUD
.175** .154 .160 .050 .005 .123 .075 .137 .119
CGT – risk taking Mood episodes Illness duration SUD
.246n .088 .188**
CGT – deliberation time Mood episodes Illness duration SUD
.025 .024 .010
CGT – quality of decision making Mood episodes Illness duration SUD
.002 .003 .015
RVP – false alarmsa Mood episodes Illness duration SUD
.048 .007 .802
Acronyms: AGN ¼Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE¼ commission errors; CGT ¼ Cambridge Gambling Task; FA¼ False Alarms; QDM¼ Quality Decision Making; RVP¼Rapid Visual Processing; SUD¼ Substance Use Disorder. a Note: The dichotomous variable “False alarms” was analyzed using a logistic regression. n po.05. nn po.10.
146
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
(Newman and Meyer, 2014). Furthermore, another study did not find significant correlations between the BIS and the BART – a CANTAB risk taking measure (Reddy et al., 2014). The authors argued that, unlike behavioral measures of impulsivity, the BIS score is influenced by the individual's past experience and selfperception including phases of euthymia, depression and mania. This means that while self-rated impulsivity reflects how individuals evaluate their own past and present behavior in everyday life, behavioral measures of impulsivity assess certain elements of performance in a pre-designed task in a specific situation on that day and time. Third, our results suggest that while illness chronicity and SUD may not be generally linked to impulsivity, specific facets such as ‘commission errors’ in response to negative stimuli and ‘risk taking behaviors’ may be. Considering first the relationship between illness chronicity measured by the number of mood episodes, and commission errors in response to negative stimuli in the AGN task, one explanation is that patients with BD experience difficulties in suppressing the prepotent response triggered by the presentation of negative stimuli. Although we only found a non-significant trend for the predictor ‘number of mood episodes’ this finding is consistent with other research providing evidence of an attentional bias towards negative information among individuals with BD (García-Blanco et al., 2014; Gopin et al., 2011). Due to this bias participants with BD may struggle to process neutral or positive stimuli after being exposed to negative triggers. Such findings have been corroborated by fMRI studies that revealed altered patterns of neural activity in the frontal, cingulate and limbic regions during the Emotional Stroop and Go/Nogo tasks (Houenou et al., 2011; Malhi et al., 2005; Wessa et al., 2007). Additional research is, however, needed to determine the functionality and neurobiological underpinning of “commission errors” and “risktaking” in BD. Looking at risk taking next, our findings suggest that the number of prior mood episodes is a strong predictor of risk taking. Based on the discussed neuroprogressive nature of the bipolar disorder, it could be hypothesized that repeated mood episodes are associated with increased neurotoxicity along with brain changes and behavioral abnormalities. It is noteworthy that there was also a trend suggesting an association between a lifetime history of SUD and increased risk taking. Because the variable “risk taking” refers to the proportion of points that the participants chose to gamble on a highly likely outcome, this can be interpreted as an index of reward seeking. Given the deleterious effects of drugs, such as alcohol, opioids and psychostimulants on the brain reward system (Di Chiara and Bassareo, 2007), it is possible that a lifetime history of SUD increases the sensitivity to potential rewards (Alloy et al., 2015; Johnson, 2005) and, as a result, more likely to take risks to win. This is consistent with previous findings that associated increased risk taking tendencies with elevated activity in the ventrostriatal circuits (Mason et al., 2014). Abnormalities in these brain regions have been associated with increased preference for immediate rather than delayed reward (McClure et al., 2004). The current result is, however, less robust than that highlighted in a previous study in which SUD was associated with an increased number of commission errors in a sample of BD patients performing a sustained attention task (Swann et al., 2004). The trend observed in the current study may be due to the fact that our sample had been in remission from SUD and did not fulfill criteria for SUD at the time of testing. It is notable that, the number of mood episodes had a more prominent role than illness duration in predicting the variability of both commission errors and risk taking. This finding provides further support to the cumulative negative effects of mood episodes on the brain. The small sample size did not enable us to conduct separate analyses on the number of depressive and manic episodes. It is therefore still unclear whether the main
polarity of prior mood episodes is essential in predicting impulsivity. There is, however, evidence that a higher number of manic episodes is a predictor of cognitive impairment (Martínez‐Arán et al., 2004). Before drawing final conclusions some limitations need first to be mentioned. Some we already hinted at such as that our participants were in full remission from a SUD at the time of enrollment or that we observed a substantial level of depressive symptoms which might have affected responsivity to negative emotional stimuli (Joormann and Gotlib, 2007; Rinck and Becker, 2005). Additionally, while we have information about who took medication, the information was not sufficient to control for any potential effects of medication on motor and cognitive processing speed. It is however notable that there was no significant difference in impulsivity levels (BIS and CANTAB measures) between medicated and non-medicated individuals. Furthermore, this study did not collect information on other BD-related comorbidities such as anxiety disorders which may lead to increases in impulsive behavior. Comparisons with other studies such as Ibanez et al. (2012) and Kaladjian et al. (2009) suggest that our patients performed similar to other samples, however variables such commission errors often show reduced variability which makes it more difficult to find significant associations. From a statistical viewpoint, a sample of 74 subjects was considered to provide sufficient power to detect an association between predictors (Tabachnick and Fidell, 2007). Therefore, it is unlikely that the current sample size had a major impact on our findings. Furthermore, we also used a range of measures of impulsivity including the widely used BIS and selected tasks of the CANTAB to assess the multiple facets of impulsivity. Last but not least, the proportion of variance explained in the regression analyses included in this paper is small (o 10%), i.e. other predictors might be more relevant than the ones we had selected. Despite the limitations, to the authors' best knowledge this is the first study that investigates the impact of both SUD and illness chronicity on self-rating (BIS) and behavioral aspects of impulsivity (CANTAB measures) in a sample of BD patients. We found no strong link between the chronicity of the bipolar disorder, SUD and measures of impulsivity. This suggests that general levels of impulsivity may not be a direct consequence of neuroprogression but it does not rule out that elevated impulsivity could be a trait marker of the bipolar disorder. Because our work suggests highly specific associations between disorder-related factors and facets of impulsivity, future research should focus less on trait impulsivity or general levels of impulsivity but rather examine theoretically derived hypothesis under which conditions actual impulsive behaviors are expected to be triggered and which consequences they might have. This would also be highly informative for clinical interventions.
Role of funding source This work was partly supported by the Stanley Medical Research Institute, the John S. Dunn Foundation, NIH grant MH 085667 (JCS), and by the Pat Rutherford Jr. Chair in Psychiatry (UTHealth).
Conflict of interest Drs. Bauer and Zunta-Soares have no conflicts of interest. Dr. Sanches has received research grants from Janssen. Professor Thomas D. Meyer has been a speaker for Pfizer and Lundbeck. Professor J. Soares has received grants/research support from Forrest, BMS, Merck, Stanley Medical Research Institute, NIH and has been a speaker for Pfizer and Abbott.
Acknowledgment N/A.
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
References Alloy, L.B., Nusslock, R., Boland, E.M., 2015. The Development and Course of Bipolar Spectrum Disorders: An Integrated Reward and Circadian Rhythm Dysregulation Model. Annu. Rev. Clin. Psychol. 11 (1). Asaad, T., Okasha, T., Ramy, H., Fekry, M., Zaki, N., Azzam, H., Rabie, M.A., Elghoneimy, S., Sultan, M., Hamed, H., Refaat, O., Shorab, I., Elhabiby, M., Elgweily, T., ElShinnawy, H., Nasr, M., Fathy, H., Meguid, M.A., Nader, D., Elserafi, D., Enaba, D., Ibrahim, D., Elmissiry, M., Mohsen, N., Ahmed, S., 2014. Correlates of psychiatric co-morbidity in a sample of Egyptian patients with bipolar disorder. J. Affect. Disord. 166, 347–352. Berk, M., Conus, P., Kapczinski, F., Andreazza, A.C., Yücel, M., Wood, S.J., Pantelis, C., Malhi, G.S., Dodd, S., Bechdolf, A., 2010. From neuroprogression to neuroprotection: implications for clinical care. Med. J. Aust. 193, S36–S40. Berk, M., Kapczinski, F., Andreazza, A., Dean, O., Giorlando, F., Maes, M., Yücel, M., Gama, C., Dodd, S., Dean, B., 2011. Pathways underlying neuroprogression in bipolar disorder: focus on inflammation, oxidative stress and neurotrophic factors. Neurosci. Biobehav. Rev. 35, 804–817. Bora, E., Yucel, M., Pantelis, C., 2009. Cognitive endophenotypes of bipolar disorder: a meta-analysis of neuropsychological deficits in euthymic patients and their first-degree relatives. J. Affect. Disord. 113, 1–20. Christodoulou, T., Lewis, M., Ploubidis, G., Frangou, S., 2006. The relationship of impulsivity to response inhibition and decision-making in remitted patients with bipolar disorder. Eur. Psychiatry 21, 270–273. Di Chiara, G., Bassareo, V., 2007. Reward system and addiction: what dopamine does and doesn't do. Curr. Opin. Pharmacol. 7, 69–76. Ekinci, O., Albayrak, Y., Ekinci, A.E., Caykoylu, A., 2011. Relationship of trait impulsivity with clinical presentation in euthymic bipolar disorder patients. Psychiatry Res. 190, 259–264. Etain, B., Mathieu, F., Liquet, S., Raust, A., Cochet, B., Richard, J., Gard, S., Zanouy, L., Kahn, J.-P., Cohen, R.F., 2013. Clinical features associated with traitimpulsiveness in euthymic bipolar disorder patients. J. Affect. Disord. 144, 240–247. Evenden, J., 1999. Impulsivity: a discussion of clinical and experimental findings. J. Psychopharmacol. 13, 180–192. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B., 2012. Structured Clinical Interview for DSM-IVs Axis I Disorders (SCID-I), Clinician Version, Administration Booklet. Am. Psychiatr. Pub. García-Blanco, A., Salmerón, L., Perea, M., Livianos, L., 2014. Attentional biases toward emotional images in the different episodes of bipolar disorder: an eyetracking study. Psychiatry Res. 215, 628–633. Gopin, C.B., Burdick, K.E., DeRosse, P., Goldberg, T.E., Malhotra, A.K., 2011. Emotional modulation of response inhibition in stable patients with bipolar I disorder: a comparison with healthy and schizophrenia subjects. Bipolar Disord. 13, 164–172. Hajek, T., Carrey, N., Alda, M., 2005. Neuroanatomical abnormalities as risk factors for bipolar disorder. Bipolar Disord. 7, 393–403. Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56. Houenou, J., Frommberger, J., Carde, S., Glasbrenner, M., Diener, C., Leboyer, M., Wessa, M., 2011. Neuroimaging-based markers of bipolar disorder: evidence from two meta-analyses. J. Affect. Disord. 132, 344–355. Ibanez, A., Cetkovich, M., Petroni, A., Urquina, H., Baez, S., Gonzalez-Gadea, M.L., Kamienkowski, J.E., Torralva, T., Torrente, F., Strejilevich, S., 2012. The neural basis of decision-making and reward processing in adults with euthymic bipolar disorder or attention-deficit/hyperactivity disorder (ADHD). PLoS One 7, e37306. Johnson, S.L., 2005. Mania and dysregulation in goal pursuit: a review. Clin. Psychol. Rev. 25, 241–262. Joormann, J., Gotlib, I.H., 2007. Selective attention to emotional faces following recovery from depression. J. Abnorm. Psychol. 116, 80. Kaladjian, A., Jeanningros, R., Azorin, J.-M., Nazarian, B., Roth, M., MazzolaPomietto, P., 2009. Reduced brain activation in euthymic bipolar patients during response inhibition: an event-related fMRI study. Psychiatry Res.: Neuroimag. 173, 45–51. Kapczinski, F., Vieta, E., Andreazza, A.C., Frey, B.N., Gomes, F.A., Tramontina, J., Kauer-Sant'Anna, M., Grassi-Oliveira, R., Post, R.M., 2008. Allostatic load in bipolar disorder: implications for pathophysiology and treatment. Neurosci. Biobehav. Rev. 32, 675–692.
147
Kirby, L., Zeeb, F., Winstanley, C., 2011. Contributions of serotonin in addiction vulnerability. Neuropharmacology 61, 421–432. Koob, G.F., Le Moal, M., 1997. Drug abuse: hedonic homeostatic dysregulation. Science 278, 52–58. MacQueen, G.M., Young, L.T., Joffe, R.T., 2001. A review of psychosocial outcome in patients with bipolar disorder. Acta Psychiatr. Scand. 103, 163–170. Malhi, G.S., Lagopoulos, J., Sachdev, P.S., Ivanovski, B., Shnier, R., 2005. An emotional Stroop functional MRI study of euthymic bipolar disorder. Bipolar Disord. 7, 58–69. Martínez‐Arán, A., Vieta, E., Colom, F., Torrent, C., Sánchez‐Moreno, J., Reinares, M., Benabarre, A., Goikolea, J., Brugue, E., Daban, C., 2004. Cognitive impairment in euthymic bipolar patients: implications for clinical and functional outcome. Bipolar Disord. 6, 224–232. Mason, L., O'Sullivan, N., Montaldi, D., Bentall, R.P., El-Deredy, W., 2014. Decisionmaking and trait impulsivity in bipolar disorder are associated with reduced prefrontal regulation of striatal reward valuation. Brain. McClure, S.M., Laibson, D.I., Loewenstein, G., Cohen, J.D., 2004. Separate neural systems value immediate and delayed monetary rewards. Science 306, 503–507. Merikangas, K.R., Akiskal, H.S., Angst, J., Greenberg, P.E., Hirschfeld, R.M., Petukhova, M., Kessler, R.C., 2007. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Arch. Gen. Psychiatry 64, 543–552. Moeller, F.G., Barratt, E.S., Dougherty, D.M., Schmitz, J.M., Swann, A.C., 2001. Psychiatric aspects of impulsivity. Am. J. Psychiatry 158, 1783–1793. Najt, P., Perez, J., Sanches, M., Peluso, M., Glahn, D., Soares, J.C., 2007. Impulsivity and bipolar disorder. Eur. Neuropsychopharmacol. 17, 313–320. Newman, A.L., Meyer, T.D., 2014. Impulsivity: present during euthymia in bipolar disorder? – a systematic review. Int. J. Bipolar Disord. 2, 2. Ostacher, M.J., 2011. Bipolar and substance use disorder comorbidity: diagnostic and treatment considerations. FOCUS: J. Lifelong Learn. Psychiatry 9, 428–434. Patton, J.H., Stanford, M.S., Barratt, E.S., 1995. Factor structure of the Barratt impulsiveness scale. J. Clin. Psychol. 51, 768–774. Peluso, M., Hatch, J., Glahn, D., Monkul, E., Sanches, M., Najt, P., Bowden, C., Barratt, E., Soares, J., 2007. Trait impulsivity in patients with mood disorders. J. Affect. Disord. 100, 227–231. Powers, R.L., Russo, M., Mahon, K., Brand, J., Braga, R.J., Malhotra, A.K., Burdick, K.E., 2013. Impulsivity in bipolar disorder: relationships with neurocognitive dysfunction and substance use history. Bipolar Disord. 15, 876–884. Quraishi, S., Frangou, S., 2002. Neuropsychology of bipolar disorder: a review. J. Affect. Disord. 72, 209. Reddy, L.F., Lee, J., Davis, M.C., Altshuler, L., Glahn, D.C., Miklowitz, D.J., Green, M.F., 2014. Impulsivity and risk taking in bipolar disorder and schizophrenia. Neuropsychopharmacology 39 (2), 456–463. Rinck, M., Becker, E.S., 2005. A comparison of attentional biases and memory biases in women with social phobia and major depression. J. Abnorm. Psychol. 114, 62. Robbins, T., James, M., Owen, A., Sahakian, B., McInnes, L., Rabbitt, P., 1994. Cambridge Neuropsychological Test Automated Battery (CANTAB): a factor analytic study of a large sample of normal elderly volunteers. Dement. Geriatric Cogn. Disord. 5, 266–281. Swann, A.C., 2010. The strong relationship between bipolar and substance‐use disorder. Ann. N. Y. Acad. Sci. 1187, 276–293. Swann, A.C., Dougherty, D.M., Pazzaglia, P.J., Pham, M., Moeller, F.G., 2004. Impulsivity: a link between bipolar disorder and substance abuse. Bipolar Disord. 6, 204–212. Swann, A.C., Gerard Moeller, F., Steinberg, J.L., Schneider, L., Barratt, E.S., Dougherty, D.M., 2007. Manic symptoms and impulsivity during bipolar depressive episodes. Bipolar Disord. 9, 206–212. Swann, A.C., Pazzaglia, P., Nicholls, A., Dougherty, D.M., Moeller, F.G., 2003. Impulsivity and phase of illness in bipolar disorder. J. Affect. Disord. 73, 105–111. Sweeney, J.A., Kmiec, J.A., Kupfer, D.J., 2000. Neuropsychologic impairments in bipolar and unipolar mood disorders on the CANTAB neurocognitive battery. Biol. Psychiatry 48, 674–684. Tabachnick, B., Fidell, L., 2007. Multivariate analysis of variance and covariance. Using Multivar. Stat. 3, 402–407. Wessa, M., Houenou, J., Paillère-Martinot, M.-L., Berthoz, S., Artiges, E., Leboyer, M., Martinot, J.-L., 2007. Fronto-striatal overactivation in euthymic bipolar patients during an emotional go/nogo task. Am. J. Psychiatry 164, 638–646. Young, R., Biggs, J., Ziegler, V., Meyer, D., 1978. A rating scale for mania: reliability, validity and sensitivity. Br. J. Psychiatry 133, 429–435.
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
Does a history of substance abuse and illness chronicity predict increased impulsivity in bipolar disorder? Isabelle E. Bauer 1, Thomas D. Meyer n,1, Marsal Sanches, Giovana Zunta-Soares, Jair C. Soares University of Texas Health Science Center at Houston, Department of Psychiatry and Behavioral Sciences, 1941 East Road, Houston, TX 77054, United States
art ic l e i nf o
a b s t r a c t
Article history: Received 25 November 2014 Received in revised form 3 March 2015 Accepted 5 March 2015 Available online 14 March 2015
Background: Impulsivity is a common feature shared by bipolar disorder (BD) and substance use disorder (SUD). SUD and recurrent mood episodes are considered to be risk factors for poor outcome in BD. However, the association between impulsivity, illness chronicity and SUD in BD remains unexplored. Methods: 103 BD patients with and without a lifetime history of SUD (36.82 711.34 years, 40 males) were recruited. Participants completed the SCID interview and were administered measures of impulsivity including the Barratt Impulsivity Scale (BIS) and selected tests of the Cambridge Neuropsychological Test Automated Battery (CANTAB). Hierarchical regression analyses explored the relationship between illness chronicity, SUD, and impulsivity. Results: Variance in the BIS, number of false alarms on the Rapid Visual Processing task and other impulsivity indicators of the Cambridge Gambling Task (CGT) was not explained by the chosen variables. Only an increased number of commission errors in the negative condition of the Affective Go/No Go task was significantly associated with illness chronicity. Furthermore there was a trend suggesting a relationship between a lifetime history of SUD and increased propensity to risk-taking during the CGT. Limitations: Potential limitations include medication and patients' remission status from SUD. Conclusions: Contrary to our expectations impulsivity was generally not predicted by indicators of illness chronicity or SUD. While impulsivity could still be a marker of BD that is present before the onset of the disorder, the link between the number of mood episodes and specific indicators of impulsivity may be related to mechanisms of neuroprogression. & 2015 Elsevier B.V. All rights reserved.
Keywords: Impulsivity Bipolar disorder Substance abuse Illness chronicity
1. Introduction Bipolar disorder (BD) is a serious mental illness clinically characterized by mood dysregulation, brain abnormalities, and cognitive deficits such as poor response inhibition that in some cases persist during the euthymic and acute phases (Bora et al., 2009; MacQueen et al., 2001; Quraishi and Frangou, 2002). The majority of BD patients also present with additional mental health conditions such as anxiety and substance use disorder (SUD) (Asaad et al., 2014; Merikangas et al., 2007) that are considered to be predictors of poor response to treatment (Swann, 2010) and low remission rates (Ostacher, 2011) in BD and SUD patients.
n Corresponding author at: University of Texas Health Science Center at Houston, Department of Psychiatry and Behavioral Science, 1941 East Road, Houston, TX 77054, United States. Tel.: þ1 713 486 2638; fax: þ1 713 486 2553. E-mail addresses: [email protected] (I.E. Bauer), [email protected] (T.D. Meyer). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jad.2015.03.010 0165-0327/& 2015 Elsevier B.V. All rights reserved.
Impulsivity is a common feature shared by BD and SUD (Powers et al., 2013; Swann et al., 2007) and has been associated with poor cognitive control and disregard of the long-term implications of one's behavior, and may therefore result in risky and disorganized behaviors (Evenden, 1999; Moeller et al., 2001). Impulsivity can be considered as being a multidimensional concept and assessed via self-reports and behavioral paradigms. Previous studies showed that the Barratt Impulsivity Scale (BIS) total score (Patton et al., 1995) – a widely used self-rating measure of impulsivity in psychiatry – discriminates well between healthy subjects and euthymic BD patients (Ekinci et al., 2011; Etain et al., 2013). Additionally, BIS scores have been found to be significantly elevated in BD patients with a history of SUD (Etain et al., 2013). The comparability and equivalence of self-rated and behavioral indicators of impulsivity is still equivocal (for review (Newman and Meyer, 2014)). A study found that euthymic BD patients encountered difficulties on behavioral measures of impulsivity such as the Hayling Sentence Completion (HSCT) and the Iowa Gambling Task (Christodoulou et al., 2006). Furthermore, the BIS score was found to correlate positively with the number of
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
commission errors on the HSCT and the Iowa Gambling task. Two studies using the BIS alongside the Immediate Memory Task– Delayed Memory Task (IMT–DMT) – a modified version of the Continuous Performance Task (CPT) found that while the BIS scores were elevated in BD patients compared to healthy controls (the two studies included euthymic BD and BD patients with SUD respectively), the number of commission errors of the IMT–DMT was comparable between the two groups (Swann et al., 2004,, 2003). Thus, not all facets of impulsivity discriminate between BD and healthy controls. Furthermore, the BIS and behavioral tasks may not capture the same aspects of impulsivity and might be differentially related to outcome measures such as the course or duration of the bipolar disorder. The literature on the relationship between impulsivity, BD, and SUD is sparse, and it is unclear whether impulsivity is a marker of the bipolar disorder or rather the result of neural damage associated with repeated mood episodes and/or drug use. Theories of neuroprogression in mood disorders suggest that ongoing mood episodes lead to a state of chronic inflammation and eventually result in neurocognitive impairment (Berk et al., 2010, 2011; Kapczinski et al., 2008). Abusing drugs has negative effects on the brain reward mechanisms and on the dopaminergic and serotonergic neurotransmitter systems (Koob and Le Moal, 1997) and may induce hazardous reward seeking behaviors and poor decision making (Kirby et al., 2011). Therefore, a positive association between measures of impulsivity and illness chronicity in BD would support the notion that the levels of impulsivity increase as the bipolar disorder progresses. Similarly, a link between increased impulsivity levels and SUD would be related to the deleterious effects of drugs on the brain. By contrast if there was no relation among these measures, prior research would suggest that impulsivity is more likely to be a marker of BD, being present before the onset of the disorder. Based on the evidence reviewed above the aim of this crosssectional study is to investigate the predictive power of indicators of illness chronicity such as illness duration and the number of prior mood episodes on impulsivity in BD, and determine whether a history of SUD additionally explains elevations in impulsivity above and beyond indicators of illness chronicity. Further, we aim to determine whether self-ratings (i.e. BIS) and behavioral measures from the CANTAB battery measure heterogeneous or similar facets of impulsivity. Based on the literature we predict that 1. indicators of illness chronicity will explain variance in both selfrated and behavioral impulsivity measures and 2. that a history of SUD will significantly add to this effect.
2. Methods 2.1. Sample The sample included 103 adult BD patients (M7SD: 36.82711.34 years; 40 males, 63 females; 74 BD-I, 20 BD-II, and 9 BD-NOS) (see Table 1). Participants were recruited from inpatient and outpatient clinics of the University of North Carolina at Chapel Hill (UNC) and the University of Texas Health Science Center at Houston (UT). All patients met the DSM-IV-R criteria for BD. The diagnosis of BD among patients were ascertained by the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders Axis I (SCID I) (First et al., 2012), which was administered to all participants by an independent psychiatrist or trained research assistant. The interview also included the Hamilton Depression Rating Scale (HAM-D) (Hamilton, 1960) and the Young Mania Rating Scale (YMRS) (Young et al., 1978). Illness chronicity was defined by the number of mood episodes (including manic, depressive and mixed episodes) and illness duration. 44 Participants had a lifetime history of SUD and had
143
Table 1 Demographic and clinical characteristics of the sample. Group
Mean (S.D.)
Age (years) Sex
36.82 7 11.34 40 Males 63 Females 13.9 73.3 110.747 13.71 18.45 7 8.16 76.497 5.5 5.08 7 4.85 12.677 8.46
Education (years) Full scale IQ (WASI) Age of onset BIS YMRS HAMD a Number of mood episodes o 13 4 13 Illness duration (years) a Lifetime history of substance and/or alcohol use disorder
n¼ 24 n¼ 79 18.357 11.71 Yes – n¼ 44 No – n¼59
a Notes: Absolute numbers might vary due to missing data; YMRS ¼Young Mania Rating Scale; HAMD: Hamilton Depression Rating Scale; BIS: Barratt Impulsivity Scale.
previously met criteria for substance abuse or dependence. Subjects were excluded if they fulfilled criteria for SUD in 6 months prior to the beginning of the study or had any current serious medical problems including cardiovascular and neurological disorders. 43 BD participants took psychotropic medication at the time of enrollment. The study protocol was approved by the local Institutional Review board and informed consent was obtained from all the participants. 2.2. Impulsivity The Barratt Impulsive Scale (BIS) (Patton et al., 1995) is a 30-item self-report measure of impulsivity that includes three scales: cognitive impulsivity – e.g. difficulties with concentrating, motor impulsiveness, e.g. motor restlessness and non-planning impulsiveness – e.g. selfcontrol. The BIS has been often used in research in BD and was therefore chosen as a self-report. Behavioral aspects of impulsivity were measured by the Affective Go/No-Go task (AGN), the Cambridge Gambling Task (CGT), and the Rapid Visual Processing Task (RVP) of the computerized Cambridge Neurocognitive Test Automated Battery (CANTAB – http://www.cantab.com). This battery was chosen based on its established sensitivity to cognitive impairment in psychiatric disorders (Sweeney et al., 2000). The tasks of the CANTAB are briefly described below and a detailed description is given elsewhere (Robbins et al., 1994). AGN evaluates inhibition control. Participants are instructed to respond to either happy (e.g. joyful, warmth, courage) or negative words (e.g. mistake, hopeless, burden), and inhibit their response to stimuli of opposite valence. Stimuli were presented in both a random (negative, positive, positive, negative) and sequential way (positive, positive, negative, negative). The outcome measure included in this report to reflect impulsivity is the number of commission errors in response to distracting stimuli. CGT evaluates impulse control and risk-taking behavior. Primary outcome measures include deliberation time prior to making a bet, risk taking (mean proportion of the number of points risked on trials with a more likely outcome), quality of decision making (proportion of bets on gamble trials with a more likely outcome), and delay aversion (ratio score representing the ability to bet large amounts depending on the odds of winning/losing). RVP evaluates sustained visual attention. The main outcome measure of this task in relation to the concept of impulsivity is the number of false alarms. Because none of the transformation positively affected its distribution towards greater normality (see below) this variable was coded into a dichotomous variable (1¼presence of false alarms, 2¼ absence of false alarms).
144
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
impulsivity the first block of independent variables of these regression analyses included illness chronicity – measured by the number of mood episodes coded into a dichotomous variable (1: ≤13 episodes, 2: 413 episodes) and illness duration. The second block of variables included SUD – this variable was coded as a dichotomous variable (1 ¼SUD, 2 ¼ no SUD). Based on Tabachnick and Fidell's (2007) formula a sample of 74 subjects was found to be sufficient to provide adequate statistical power (80%) to detect an association between predictors. Statistical significance was defined as p o.05 and p-values ranging from .05 to .10 were considered to be statistical trends.
2.3. Statistical procedure Statistical analyses were performed using IBM SPSS statistics (Version 21.0). Normality assumptions for continuous variables were examined. Where appropriate, outliers were winsorised and log, square root or reciprocal transformations applied to achieve normality if appropriate. If normality could not been achieved, either untransformed or dichotomized scores were used. Correlational analysis was used to investigate the relationship between potential predictors of impulsivity, demographics and clinical variables such as illness duration and number of mood episodes. First, we conducted a series of regression analyses to examine whether the variables sex, age, current symptom levels (HAM-D, YMRS) affected the dependent variable, i.e. self-rated and behavioral indicators of impulsivity and to determine whether these variables needed to be included in the final hierarchical regression models. Then, we conducted a series of hierarchical regression analyses to examine our main hypothesis with them either BIS or CANTAB measures as the dependent variable. The dichotomous dependent variable “false alarms” (RVP task) was analyzed using a logistic regression. If none of the before mentioned covariates (e.g. sex, age, symptom levels) proved to be significantly related to
3. Results First we explored the correlations between self-rating and CANTAB measures of impulsivity (Table 2). In line with prior research overall correlations were low and ranged from r ¼ .13 to .19 between BIS and behavioral measures of impulsivity and from r ¼ .14 to .30 between indices from different behavioral tasks.
Table 2 Coefficients of correlation among self-rating (BIS) and CANTAB measures of impulsivity. BIS
AGN total CE AGN negative CE AGN neutral CE AGN positive CE CGT risk taking CGT delay aversion CGT deliberation time CGT QDM RVP FAa
.03 .11 .03 .11 .18 .06 .13 .10 .18
AGN total CE
AGN negative CE
AGN neutral CE
AGN positive CE
CGT risk taking
CGT delay aversion
CGT deliberation time
CGT QDM
.92nn .60nn .29nn .11 .18 .08 .14 .06
.56nn .31nn .09 .19 .08 .09 .08
.09 .02 .08 .06 .26nn .07
.01 .08 .08 .10 .29nn
.34nn .07 .08 .22n
.20n .34nn .03
.08 .05
.07
Acronyms: AGN¼ Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE ¼commission errors; CGT ¼Cambridge Gambling Task; FA ¼False Alarms; QDM ¼Quality Decision Making; RVP¼Rapid Visual Processing. n
p o.05. p o.01. For the dichotomous variable “False alarms” Spearman's coefficients of correlation are reported.
nn
a
Table 3 Results from the hierarchical regression analyses for Model 1 (illness duration, number of mood episodes) and 2 (Model 1 plus SUD). Variable
BIS AGN total commissions AGN total commissions – positive AGN total commissions – negative AGN total commissions – neutral CGT – delay aversion CGT – risk taking CGT – deliberation time (ms) CGT – quality of decision making
RVP – false alarmsa Mood episodes Illness duration SUD
Mean (SD)
76.37 (5.44) 13.45 (9.03) 4.16 (3.34) 3.25 (2.67) 4.13 (4.24) 0.44 (0.24) 0.54 (0.14) 2324.24 (953.01) 0.89 (0.14)
With errors¼75% No errors¼ 35%
Model 1a
Model 2b
ΔR2
p-Value
R2
p-Value
R2
p-Value
.014 .030 .028 .070 .003 .030 .054 .005 .000
.59 .21 .29 o .05 .89 .26 o .10 .93 1.00
.014 .042 .035 .096 .018 .044 .090 .002 .000
.78 .24 .37 o .05 .67 .27 o .05 .05 .99
.000 .011 .007 .026 .015 .014 .035 .000 .000
.85 .28 .42 .12 .25 .26 o .10 .92 .89
Model 1 B (SE)
R2
Model 2: B (SE)
R2
χ2 model 1
χ2 model 2
3.7
14.15 (p ¼ .08)
3.69 (p ¼ .88)
.002 .025 (.57) .008 (.02)
.05 (.58) .007 (.02) .802 (.46)
Acronyms: AGN ¼ Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE ¼commission errors; CGT¼ Cambridge Gambling Task; FA ¼False Alarms; QDM¼ Quality Decision Making; RVP¼Rapid Visual Processing; SUD ¼ Substance Use Disorder. a
Note: The dichotomous variable “False alarms” was analyzed using a logistic regression; R2 refers to Cox & Snell's estimate.
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
Next we ran separate regression analyses for all indicators of impulsivity to determine whether the potential confounders age, sex, HAMD and YMRS scores were associated with impulsivity. Since none of them were associated with the dependent variables (all Fo1.2, all R2 o5%, available on request from the authors), they were not included in the final regression model. The results of the hierarchical regression analysis showing the association between illness chronicity, substance abuse disorders and impulsivity measures are shown in Tables 3 and 4. Self-rated impulsivity using the BIS was not significantly predicted by indicators of illness chronicity and a history of SUD. The overall number of commission errors across the positive, negative or neutral conditions of the AGN task was not predicted by these variables. However, when considering only the total number of commission errors during the negative condition of the AGN, Model 1, which included illness duration and the number of mood episodes as predictors, was found to be statistically significant (R¼.265, R2 ¼ .070, F (1,88)¼3.323, p¼.041). Model 2, which additionally included SUD, still explained a significant amount of variance [F (3,87)¼3.076, p¼.032], but the addition of SUD did not significantly increase the amount of explained variance (ΔR2: F(3,87)¼.026, p¼ .12). Only the number of prior mood episodes emerged as a predictor, albeit not significant (po.10), of AGN commission errors in response to negative words (see Table 3). For the variable “risk taking” of the CGT task, Model 1 was not significant (p¼ .086), while model 2 including SUD proved to be significant therefore explaining a significant amount of the variance in risk taking (F (3, 87) ¼2.852, p ¼.042). However, adding SUD as a predictor in model 2 only lead to a non-significant change in R2 (p¼ .07). When looking at the beta coefficients the number of mood episodes was found to be the only significant predictor of risk taking (β ¼ .246, p ¼.02). There was only a trend suggesting a positive association between SUD and risk taking (β ¼ .188, p¼ .07). Other impulsivity measures of the CGT task such as deliberation time, quality of decision making and delay aversion, were not predicted by the selected variables. Similarly, the number of false alarms in the RVP task was not associated with either SUD or illness chronicity.
4. Discussion The current study aimed to investigate the association between illness chronicity, SUD, and indicators of impulsivity in BD. Based on the neuroprogression model (Berk et al., 2011) we expected to observe a strong relationship between the course of the bipolar disorder and levels of self-rated and CANTAB measures of impulsivity. However, two out of the 10 indicators of impulsivity showed some associations: the number of commission errors to distracting stimuli during the negative condition of the AGN task and risk taking in the CGT. Both were related to illness chronicity, more specifically to the number of prior mood episodes, but not to illness duration. Further, history of SUD did not generally predict elevated impulsivity in our sample and contributed only minimally to risk-taking. The current findings are important for three reasons. First, the lack of a strong relationship between impulsivity and illness chronicity, and SUD, suggests that neither of these variables contribute to the increased levels of impulsivity observed in BD (Najt et al., 2007; Peluso et al., 2007). Based on prior research this finding may suggest that impulsivity is less a consequence of BD but rather a core trait of the disorder and is possibly associated with the neuroanatomical and neurochemical abnormalities previously observed in BD (Hajek et al., 2005) such as the dysregulation of the fronto-limbic system and/or abnormalities in the dopaminergic projections to the frontal, limbic
145
and striatal regions as they are involved in self-monitoring and reward seeking behaviors (Mason et al., 2014). Second, both the regression and correlation analyses indicate that self-rating and behavioral measures of impulsivity tap into different facets of impulsivity. This result is not surprising if one considers that, in the literature, the BIS score is viewed as being a trait measure of self-control (Swann et al., 2004), while neuropsychological measures of impulsivity appear to be state-dependent and more prone to fluctuate across phases of illness (Swann et al., 2003). Along the same line, a recent systematic review of studies in impulsivity showed that while self-rating measures of impulsivity differentiate between euthymic BD and healthy volunteers, neuropsychological measures of impulsivity do not
Table 4 Beta coefficients for the predictors of impulsivity in Model 2: illness duration, number of mood episodes and SUD. β BIS Mood episodes Illness duration SUD
.090 .059 .022
AGN total commissions Mood episodes Illness duration SUD
.155 .037 .106
AGN total commissions – positive Mood episodes Illness duration SUD
.139 .122 .086
AGN total commissions – negative Mood episodes Illness duration SUD AGN total commissions – neutral Mood episodes Illness duration SUD CGT – delay aversion Mood episodes Illness duration SUD
.175** .154 .160 .050 .005 .123 .075 .137 .119
CGT – risk taking Mood episodes Illness duration SUD
.246n .088 .188**
CGT – deliberation time Mood episodes Illness duration SUD
.025 .024 .010
CGT – quality of decision making Mood episodes Illness duration SUD
.002 .003 .015
RVP – false alarmsa Mood episodes Illness duration SUD
.048 .007 .802
Acronyms: AGN ¼Affective Go/No-Go; BIS ¼ Barratt Impulsivity Scale; CE¼ commission errors; CGT ¼ Cambridge Gambling Task; FA¼ False Alarms; QDM¼ Quality Decision Making; RVP¼Rapid Visual Processing; SUD¼ Substance Use Disorder. a Note: The dichotomous variable “False alarms” was analyzed using a logistic regression. n po.05. nn po.10.
146
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
(Newman and Meyer, 2014). Furthermore, another study did not find significant correlations between the BIS and the BART – a CANTAB risk taking measure (Reddy et al., 2014). The authors argued that, unlike behavioral measures of impulsivity, the BIS score is influenced by the individual's past experience and selfperception including phases of euthymia, depression and mania. This means that while self-rated impulsivity reflects how individuals evaluate their own past and present behavior in everyday life, behavioral measures of impulsivity assess certain elements of performance in a pre-designed task in a specific situation on that day and time. Third, our results suggest that while illness chronicity and SUD may not be generally linked to impulsivity, specific facets such as ‘commission errors’ in response to negative stimuli and ‘risk taking behaviors’ may be. Considering first the relationship between illness chronicity measured by the number of mood episodes, and commission errors in response to negative stimuli in the AGN task, one explanation is that patients with BD experience difficulties in suppressing the prepotent response triggered by the presentation of negative stimuli. Although we only found a non-significant trend for the predictor ‘number of mood episodes’ this finding is consistent with other research providing evidence of an attentional bias towards negative information among individuals with BD (García-Blanco et al., 2014; Gopin et al., 2011). Due to this bias participants with BD may struggle to process neutral or positive stimuli after being exposed to negative triggers. Such findings have been corroborated by fMRI studies that revealed altered patterns of neural activity in the frontal, cingulate and limbic regions during the Emotional Stroop and Go/Nogo tasks (Houenou et al., 2011; Malhi et al., 2005; Wessa et al., 2007). Additional research is, however, needed to determine the functionality and neurobiological underpinning of “commission errors” and “risktaking” in BD. Looking at risk taking next, our findings suggest that the number of prior mood episodes is a strong predictor of risk taking. Based on the discussed neuroprogressive nature of the bipolar disorder, it could be hypothesized that repeated mood episodes are associated with increased neurotoxicity along with brain changes and behavioral abnormalities. It is noteworthy that there was also a trend suggesting an association between a lifetime history of SUD and increased risk taking. Because the variable “risk taking” refers to the proportion of points that the participants chose to gamble on a highly likely outcome, this can be interpreted as an index of reward seeking. Given the deleterious effects of drugs, such as alcohol, opioids and psychostimulants on the brain reward system (Di Chiara and Bassareo, 2007), it is possible that a lifetime history of SUD increases the sensitivity to potential rewards (Alloy et al., 2015; Johnson, 2005) and, as a result, more likely to take risks to win. This is consistent with previous findings that associated increased risk taking tendencies with elevated activity in the ventrostriatal circuits (Mason et al., 2014). Abnormalities in these brain regions have been associated with increased preference for immediate rather than delayed reward (McClure et al., 2004). The current result is, however, less robust than that highlighted in a previous study in which SUD was associated with an increased number of commission errors in a sample of BD patients performing a sustained attention task (Swann et al., 2004). The trend observed in the current study may be due to the fact that our sample had been in remission from SUD and did not fulfill criteria for SUD at the time of testing. It is notable that, the number of mood episodes had a more prominent role than illness duration in predicting the variability of both commission errors and risk taking. This finding provides further support to the cumulative negative effects of mood episodes on the brain. The small sample size did not enable us to conduct separate analyses on the number of depressive and manic episodes. It is therefore still unclear whether the main
polarity of prior mood episodes is essential in predicting impulsivity. There is, however, evidence that a higher number of manic episodes is a predictor of cognitive impairment (Martínez‐Arán et al., 2004). Before drawing final conclusions some limitations need first to be mentioned. Some we already hinted at such as that our participants were in full remission from a SUD at the time of enrollment or that we observed a substantial level of depressive symptoms which might have affected responsivity to negative emotional stimuli (Joormann and Gotlib, 2007; Rinck and Becker, 2005). Additionally, while we have information about who took medication, the information was not sufficient to control for any potential effects of medication on motor and cognitive processing speed. It is however notable that there was no significant difference in impulsivity levels (BIS and CANTAB measures) between medicated and non-medicated individuals. Furthermore, this study did not collect information on other BD-related comorbidities such as anxiety disorders which may lead to increases in impulsive behavior. Comparisons with other studies such as Ibanez et al. (2012) and Kaladjian et al. (2009) suggest that our patients performed similar to other samples, however variables such commission errors often show reduced variability which makes it more difficult to find significant associations. From a statistical viewpoint, a sample of 74 subjects was considered to provide sufficient power to detect an association between predictors (Tabachnick and Fidell, 2007). Therefore, it is unlikely that the current sample size had a major impact on our findings. Furthermore, we also used a range of measures of impulsivity including the widely used BIS and selected tasks of the CANTAB to assess the multiple facets of impulsivity. Last but not least, the proportion of variance explained in the regression analyses included in this paper is small (o 10%), i.e. other predictors might be more relevant than the ones we had selected. Despite the limitations, to the authors' best knowledge this is the first study that investigates the impact of both SUD and illness chronicity on self-rating (BIS) and behavioral aspects of impulsivity (CANTAB measures) in a sample of BD patients. We found no strong link between the chronicity of the bipolar disorder, SUD and measures of impulsivity. This suggests that general levels of impulsivity may not be a direct consequence of neuroprogression but it does not rule out that elevated impulsivity could be a trait marker of the bipolar disorder. Because our work suggests highly specific associations between disorder-related factors and facets of impulsivity, future research should focus less on trait impulsivity or general levels of impulsivity but rather examine theoretically derived hypothesis under which conditions actual impulsive behaviors are expected to be triggered and which consequences they might have. This would also be highly informative for clinical interventions.
Role of funding source This work was partly supported by the Stanley Medical Research Institute, the John S. Dunn Foundation, NIH grant MH 085667 (JCS), and by the Pat Rutherford Jr. Chair in Psychiatry (UTHealth).
Conflict of interest Drs. Bauer and Zunta-Soares have no conflicts of interest. Dr. Sanches has received research grants from Janssen. Professor Thomas D. Meyer has been a speaker for Pfizer and Lundbeck. Professor J. Soares has received grants/research support from Forrest, BMS, Merck, Stanley Medical Research Institute, NIH and has been a speaker for Pfizer and Abbott.
Acknowledgment N/A.
I.E. Bauer et al. / Journal of Affective Disorders 179 (2015) 142–147
References Alloy, L.B., Nusslock, R., Boland, E.M., 2015. The Development and Course of Bipolar Spectrum Disorders: An Integrated Reward and Circadian Rhythm Dysregulation Model. Annu. Rev. Clin. Psychol. 11 (1). Asaad, T., Okasha, T., Ramy, H., Fekry, M., Zaki, N., Azzam, H., Rabie, M.A., Elghoneimy, S., Sultan, M., Hamed, H., Refaat, O., Shorab, I., Elhabiby, M., Elgweily, T., ElShinnawy, H., Nasr, M., Fathy, H., Meguid, M.A., Nader, D., Elserafi, D., Enaba, D., Ibrahim, D., Elmissiry, M., Mohsen, N., Ahmed, S., 2014. Correlates of psychiatric co-morbidity in a sample of Egyptian patients with bipolar disorder. J. Affect. Disord. 166, 347–352. Berk, M., Conus, P., Kapczinski, F., Andreazza, A.C., Yücel, M., Wood, S.J., Pantelis, C., Malhi, G.S., Dodd, S., Bechdolf, A., 2010. From neuroprogression to neuroprotection: implications for clinical care. Med. J. Aust. 193, S36–S40. Berk, M., Kapczinski, F., Andreazza, A., Dean, O., Giorlando, F., Maes, M., Yücel, M., Gama, C., Dodd, S., Dean, B., 2011. Pathways underlying neuroprogression in bipolar disorder: focus on inflammation, oxidative stress and neurotrophic factors. Neurosci. Biobehav. Rev. 35, 804–817. Bora, E., Yucel, M., Pantelis, C., 2009. Cognitive endophenotypes of bipolar disorder: a meta-analysis of neuropsychological deficits in euthymic patients and their first-degree relatives. J. Affect. Disord. 113, 1–20. Christodoulou, T., Lewis, M., Ploubidis, G., Frangou, S., 2006. The relationship of impulsivity to response inhibition and decision-making in remitted patients with bipolar disorder. Eur. Psychiatry 21, 270–273. Di Chiara, G., Bassareo, V., 2007. Reward system and addiction: what dopamine does and doesn't do. Curr. Opin. Pharmacol. 7, 69–76. Ekinci, O., Albayrak, Y., Ekinci, A.E., Caykoylu, A., 2011. Relationship of trait impulsivity with clinical presentation in euthymic bipolar disorder patients. Psychiatry Res. 190, 259–264. Etain, B., Mathieu, F., Liquet, S., Raust, A., Cochet, B., Richard, J., Gard, S., Zanouy, L., Kahn, J.-P., Cohen, R.F., 2013. Clinical features associated with traitimpulsiveness in euthymic bipolar disorder patients. J. Affect. Disord. 144, 240–247. Evenden, J., 1999. Impulsivity: a discussion of clinical and experimental findings. J. Psychopharmacol. 13, 180–192. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B., 2012. Structured Clinical Interview for DSM-IVs Axis I Disorders (SCID-I), Clinician Version, Administration Booklet. Am. Psychiatr. Pub. García-Blanco, A., Salmerón, L., Perea, M., Livianos, L., 2014. Attentional biases toward emotional images in the different episodes of bipolar disorder: an eyetracking study. Psychiatry Res. 215, 628–633. Gopin, C.B., Burdick, K.E., DeRosse, P., Goldberg, T.E., Malhotra, A.K., 2011. Emotional modulation of response inhibition in stable patients with bipolar I disorder: a comparison with healthy and schizophrenia subjects. Bipolar Disord. 13, 164–172. Hajek, T., Carrey, N., Alda, M., 2005. Neuroanatomical abnormalities as risk factors for bipolar disorder. Bipolar Disord. 7, 393–403. Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56. Houenou, J., Frommberger, J., Carde, S., Glasbrenner, M., Diener, C., Leboyer, M., Wessa, M., 2011. Neuroimaging-based markers of bipolar disorder: evidence from two meta-analyses. J. Affect. Disord. 132, 344–355. Ibanez, A., Cetkovich, M., Petroni, A., Urquina, H., Baez, S., Gonzalez-Gadea, M.L., Kamienkowski, J.E., Torralva, T., Torrente, F., Strejilevich, S., 2012. The neural basis of decision-making and reward processing in adults with euthymic bipolar disorder or attention-deficit/hyperactivity disorder (ADHD). PLoS One 7, e37306. Johnson, S.L., 2005. Mania and dysregulation in goal pursuit: a review. Clin. Psychol. Rev. 25, 241–262. Joormann, J., Gotlib, I.H., 2007. Selective attention to emotional faces following recovery from depression. J. Abnorm. Psychol. 116, 80. Kaladjian, A., Jeanningros, R., Azorin, J.-M., Nazarian, B., Roth, M., MazzolaPomietto, P., 2009. Reduced brain activation in euthymic bipolar patients during response inhibition: an event-related fMRI study. Psychiatry Res.: Neuroimag. 173, 45–51. Kapczinski, F., Vieta, E., Andreazza, A.C., Frey, B.N., Gomes, F.A., Tramontina, J., Kauer-Sant'Anna, M., Grassi-Oliveira, R., Post, R.M., 2008. Allostatic load in bipolar disorder: implications for pathophysiology and treatment. Neurosci. Biobehav. Rev. 32, 675–692.
147
Kirby, L., Zeeb, F., Winstanley, C., 2011. Contributions of serotonin in addiction vulnerability. Neuropharmacology 61, 421–432. Koob, G.F., Le Moal, M., 1997. Drug abuse: hedonic homeostatic dysregulation. Science 278, 52–58. MacQueen, G.M., Young, L.T., Joffe, R.T., 2001. A review of psychosocial outcome in patients with bipolar disorder. Acta Psychiatr. Scand. 103, 163–170. Malhi, G.S., Lagopoulos, J., Sachdev, P.S., Ivanovski, B., Shnier, R., 2005. An emotional Stroop functional MRI study of euthymic bipolar disorder. Bipolar Disord. 7, 58–69. Martínez‐Arán, A., Vieta, E., Colom, F., Torrent, C., Sánchez‐Moreno, J., Reinares, M., Benabarre, A., Goikolea, J., Brugue, E., Daban, C., 2004. Cognitive impairment in euthymic bipolar patients: implications for clinical and functional outcome. Bipolar Disord. 6, 224–232. Mason, L., O'Sullivan, N., Montaldi, D., Bentall, R.P., El-Deredy, W., 2014. Decisionmaking and trait impulsivity in bipolar disorder are associated with reduced prefrontal regulation of striatal reward valuation. Brain. McClure, S.M., Laibson, D.I., Loewenstein, G., Cohen, J.D., 2004. Separate neural systems value immediate and delayed monetary rewards. Science 306, 503–507. Merikangas, K.R., Akiskal, H.S., Angst, J., Greenberg, P.E., Hirschfeld, R.M., Petukhova, M., Kessler, R.C., 2007. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Arch. Gen. Psychiatry 64, 543–552. Moeller, F.G., Barratt, E.S., Dougherty, D.M., Schmitz, J.M., Swann, A.C., 2001. Psychiatric aspects of impulsivity. Am. J. Psychiatry 158, 1783–1793. Najt, P., Perez, J., Sanches, M., Peluso, M., Glahn, D., Soares, J.C., 2007. Impulsivity and bipolar disorder. Eur. Neuropsychopharmacol. 17, 313–320. Newman, A.L., Meyer, T.D., 2014. Impulsivity: present during euthymia in bipolar disorder? – a systematic review. Int. J. Bipolar Disord. 2, 2. Ostacher, M.J., 2011. Bipolar and substance use disorder comorbidity: diagnostic and treatment considerations. FOCUS: J. Lifelong Learn. Psychiatry 9, 428–434. Patton, J.H., Stanford, M.S., Barratt, E.S., 1995. Factor structure of the Barratt impulsiveness scale. J. Clin. Psychol. 51, 768–774. Peluso, M., Hatch, J., Glahn, D., Monkul, E., Sanches, M., Najt, P., Bowden, C., Barratt, E., Soares, J., 2007. Trait impulsivity in patients with mood disorders. J. Affect. Disord. 100, 227–231. Powers, R.L., Russo, M., Mahon, K., Brand, J., Braga, R.J., Malhotra, A.K., Burdick, K.E., 2013. Impulsivity in bipolar disorder: relationships with neurocognitive dysfunction and substance use history. Bipolar Disord. 15, 876–884. Quraishi, S., Frangou, S., 2002. Neuropsychology of bipolar disorder: a review. J. Affect. Disord. 72, 209. Reddy, L.F., Lee, J., Davis, M.C., Altshuler, L., Glahn, D.C., Miklowitz, D.J., Green, M.F., 2014. Impulsivity and risk taking in bipolar disorder and schizophrenia. Neuropsychopharmacology 39 (2), 456–463. Rinck, M., Becker, E.S., 2005. A comparison of attentional biases and memory biases in women with social phobia and major depression. J. Abnorm. Psychol. 114, 62. Robbins, T., James, M., Owen, A., Sahakian, B., McInnes, L., Rabbitt, P., 1994. Cambridge Neuropsychological Test Automated Battery (CANTAB): a factor analytic study of a large sample of normal elderly volunteers. Dement. Geriatric Cogn. Disord. 5, 266–281. Swann, A.C., 2010. The strong relationship between bipolar and substance‐use disorder. Ann. N. Y. Acad. Sci. 1187, 276–293. Swann, A.C., Dougherty, D.M., Pazzaglia, P.J., Pham, M., Moeller, F.G., 2004. Impulsivity: a link between bipolar disorder and substance abuse. Bipolar Disord. 6, 204–212. Swann, A.C., Gerard Moeller, F., Steinberg, J.L., Schneider, L., Barratt, E.S., Dougherty, D.M., 2007. Manic symptoms and impulsivity during bipolar depressive episodes. Bipolar Disord. 9, 206–212. Swann, A.C., Pazzaglia, P., Nicholls, A., Dougherty, D.M., Moeller, F.G., 2003. Impulsivity and phase of illness in bipolar disorder. J. Affect. Disord. 73, 105–111. Sweeney, J.A., Kmiec, J.A., Kupfer, D.J., 2000. Neuropsychologic impairments in bipolar and unipolar mood disorders on the CANTAB neurocognitive battery. Biol. Psychiatry 48, 674–684. Tabachnick, B., Fidell, L., 2007. Multivariate analysis of variance and covariance. Using Multivar. Stat. 3, 402–407. Wessa, M., Houenou, J., Paillère-Martinot, M.-L., Berthoz, S., Artiges, E., Leboyer, M., Martinot, J.-L., 2007. Fronto-striatal overactivation in euthymic bipolar patients during an emotional go/nogo task. Am. J. Psychiatry 164, 638–646. Young, R., Biggs, J., Ziegler, V., Meyer, D., 1978. A rating scale for mania: reliability, validity and sensitivity. Br. J. Psychiatry 133, 429–435.
