Cognitive regulation of negative affect in schizophrenia and bipolar disorder

Psychiatry Research 208 (2013) 21–28

Contents lists available at SciVerse ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Cognitive regulation of negative affect in schizophrenia and bipolar disorder Jesseca E. Rowland a, Meelah K. Hamilton 1,a, Bianca J. Lino a, Patricia Ly a, Kelsey Denny a, Eun-Ji Hwang a, Philip B. Mitchell a,b, Vaughan J. Carr a,c, Melissa J. Green a,b,c,n a

School of Psychiatry, University of New South Wales, Sydney NSW 2031, Australia Black Dog Institute, Prince of Wales Hospital, Randwick NSW 2031, Australia c Schizophrenia Research Institute, Darlinghurst NSW 2010, Australia b

a r t i c l e i n f o

abstract

Article history: Received 17 May 2012 Received in revised form 14 February 2013 Accepted 17 February 2013

Schizophrenia (SZ) and bipolar disorder (BD) exhibit common cognitive deficits that may impede the capacity for self-regulating affect. We examined the use of particular cognitive strategies for regulating negative affect in SZ and BD, and their associations with levels of mood symptomatology. Participants were 126 SZ, 97 BD, and 81 healthy controls (HC) who completed the Cognitive Emotion Regulation Questionnaire (CERQ), the Depression Anxiety Stress Scales (DASS) and the Hypomanic Personality Scale (HPS). Patients with SZ and BD reported more frequent rumination, catastrophising and self-blame, and less use of putting into perspective, relative to HC. Additionally, SZ patients were more likely to engage in other-blame, compared to HC. The most consistent predictors of symptomatology for SZ were self-blame and catastrophising, while for BD were rumination and reduced positive reappraisal. These findings demonstrate maladaptive use of cognitive strategies to self-regulate negative affect in SZ and BD, resembling those reported previously for unipolar depression. The ineffective use of adaptive cognitive reframing strategies in both patient groups may reflect the impact of their shared cognitive deficits, and requires further investigation. Remediation of cognitive capacities contributing to ineffective self-regulation may facilitate reduced mood symptomatology in SZ and BD. Crown Copyright & 2013 Published by Elsevier Ireland Ltd. All rights reserved.

Keywords: Emotion Cognition Cognitive emotion regulation Schizophrenia Bipolar disorder

1. Introduction Schizophrenia (SZ) and bipolar disorder (BD) share some genetic vulnerability (Lichtenstein et al., 2009) as well as neuropsychological dysfunction in some cognitive domains (namely, attention, memory, and executive function) (Reichenberg et al., 2009; Bora et al., 2010). These cognitive deficits appear to be related to abnormal structure and function of prefrontal, limbic, and striatal networks known to subserve regulation of affect and motivated behaviour (Green et al., 2007). Differential disturbances in these functional brain networks (Morris et al., 2012) may underpin the disparate manifestations of dysregulated affect that distinguish these disorders: while BD is typically characterised by oscillating mood states (mania, depression)

n Corresponding author. Present address: UNSW Research Unit for Schizophrenia Epidemiology, St. Vincent’s Hospital, O’Brien Centre, Level 4, 394-404 Victoria Street, Darlinghurst NSW 2010, Australia. Tel.: þ61 2 8382 1584; fax: þ61 2 8382 1402. E-mail address: [email protected] (M.J. Green). 1 This author is deceased.

and heightened emotional reactivity (Malhi et al., 2004a, 2004b), emotion dysregulation in SZ manifests in blunted (flat) or inappropriate affect, reflecting a lack of context-appropriate emotional expressivity (Gur et al., 2006), and disjunction between reported emotional experience and expression (Ellgring and Smith, 1998; Aghevli et al., 2003). Despite these opposing clinical manifestations of dysregulated emotional expression in the context of shared cognitive disturbances, few studies have examined cognitive strategies for emotion regulation in SZ and BD. Understanding the role of cognitive biases in emotion regulation may highlight underlying cognitive skills that could be targeted for remediation to improve the capacity for effective emotion regulation in these groups. Emotion regulation refers to a range of voluntary and involuntary processes used to modulate the occurrence, intensity, and duration of internal feeling states and physiological processes that occur in response to external events and, optimally, in accord with one’s goals (Thompson, 1994; Gross, 1998; Eisenberg, 2000). Effective emotion regulation is an important factor in determining mental health and wellbeing, and may entail conscious or unconscious processes in attempts to up- or down-regulate

0165-1781/$ - see front matter Crown Copyright & 2013 Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.psychres.2013.02.021

22

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

subjective emotional feelings and behaviours (Gross and Munoz, 1995; Sloan and Kring, 2007). Common strategies for selfregulation include attempts to actively suppress emotional behaviours and physiological responses, or cognitively control the type and extent of emotional responses via techniques to re-frame the meaning of the event (such as cognitive reappraisal, refocusing attention) (Gross, 1998; Ochsner and Gross, 2005). One recent study of cognitive reappraisal in individuals with a history of psychosis revealed less frequent use of this regulatory strategy relative to non-patient controls (Livingstone et al., 2009), but did not investigate other cognitive regulatory strategies previously associated with depression and anxiety (rumination, catastrophising, self- vs. other blame). In this study we therefore focus specifically on a range of cognitive strategies that are commonly used to regulate emotional feelings and expression, with consideration of the potential for known cognitive deficits in SZ and BD to impede effective regulation of emotion and behaviour. For example, deficits in executive control may limit the capacity to engage in cognitive reappraisal (requiring the generation and maintenance of alternative explanations for events) (Green and Malhi, 2006), consistent with the association between inhibitory deficits and excessive rumination in depression (Joorman and Gotlib, 2010). We chose the Cognitive Emotion Regulation Questionnaire (CERQ) to index the extent to which a range of adaptive and maladaptive cognitive strategies are employed to regulate emotion in response to threatening or stressful life-events (Garnefski et al., 2001). Previous research using this scale has shown increased use of maladaptive strategies (in particular, rumination, self-blame and catastrophising), coupled with decreased use of adaptive cognitive reframing strategies (such as positive reappraisal), in association with both clinical and subclinical levels of depression and anxiety symptoms (Garnefski et al., 2001, 2002, 2005; Garnefski and Kraaij, 2006). Similarly, rumination has been consistently implicated as an important predictor of depression and hypomania in BD (Van Der Gucht et al., 2009; Green et al., 2011) and in ostensibly healthy adolescents (Thomas and Bentall, 2002; Knowles et al., 2005). For SZ, it has been proposed that excessive suppression (i.e., increased down-regulation of behavioural manifestations of emotion) could account for affective blunting (Kring and Werner, 2004). However, direct evidence has not supported this proposal (Henry et al., 2007, 2008), and instead suggests that SZ patients may be limited in their ability to amplify (express) emotions, representing a dysfunction in emotional ‘upregulation’ rather than excessive down-regulation (Henry et al., 2007). Yet, this study examined the propensity to regulate positive affect (i.e., happiness) only, and no study has directly addressed the broader range of cognitive strategies used in schizophrenia during attempts to voluntarily control subjective levels of negative emotion. This is despite the overtly dysregulated affect evident in emotional blunting (Henry et al., 2007) and the role of anxiety as implicated in models of paranoia (Green and Phillips, 2004), in which SZ patients with paranoid ideation and persecutory delusions in particular show a cognitive style involving a heightened vigilance for threat, followed by overt avoidance of threatening stimuli (Phillips et al., 2003; Green et al., 2004). Further, paranoia has been consistently associated with the use of an external ‘personalising’ attributional style of ‘blaming others’ for negative events, rather than acknowledging other potential situational or own contribution to the causes of negative life events (Bentall et al., 2001). 1.1. Aims of the study This study tested the following hypotheses: first, that both SZ and BD groups would report greater frequency of use of

maladaptive cognitive strategies for emotion regulation as previously associated with depression (i.e., self-blame, rumination, catastrophising), including a propensity to under-use adaptive cognitive reframing strategies (e.g. positive reappraisal) in comparison to the HC group; it was specifically predicted that the two clinical groups would show subtle differences in the use of maladaptive strategies: we expected that SZ would demonstrate increased use of other-blame on the basis of previous evidence for external-personal attributional style in SZ patients with paranoid features, while BD would show greater employment of rumination and self-blame, based on previous findings in depressed and BD samples. Our second aim was to determine the utility of these cognitive strategies in predicting levels of depression, anxiety, stress, and propensity for (hypo)mania in these groups; we hypothesised that greater use of maladaptive cognitive strategies and less use of cognitive reframing strategies would be associated with higher levels of mood disturbance in both patient groups. 2. Methods Study procedures were approved by the Human Research Ethics Committees of the University of New South Wales (HREC UNSW Protocol no. 07067) and the South Eastern Sydney Illawarra Area Health Service (SESIAHS Protocol no. 08/192). 2.1. Participants The sample comprised 126 participants with a DSM-IV diagnosis of schizophrenia (SZ), 97 participants with a DSM-IV diagnosis of bipolar I disorder (BD), and 81 healthy controls (HC) with no personal history of a DSM-IV Axis 1 disorder (except anxiety disorders), and no history of psychosis in their first-degree biological relatives. Exclusion criteria included inability to communicate sufficiently in English, current neurological disorder, a diagnosis of substance abuse or dependence in the past 6 months, and/or having been treated with electro convulsive therapy (ECT) in the previous 6 months. The SZ participants were recruited from the Australian Schizophrenia Research Bank (ASRB), with diagnoses confirmed using the OPCRIT algorithm (McGuffin and Farmer, 1991) applied to interviewer ratings on the Diagnostic Interview for Psychosis (Castle et al., 2006), and consisted of 73 males (57.9%) and 53 females (42.1%), aged 26–67 years (M ¼ 45.46, S.D. ¼10.96). The group of BD participants were recruited predominantly from the Bipolar Disorder Family Study (Mitchell et al., 2009) and the Sydney Bipolar Disorder Clinic (Mitchell et al., 2009), with Best Estimate Diagnoses (BED) of BD-I (for whom a history of mania is a requirement for diagnosis) confirmed by a psychiatrist (PBM), based on all available data from the Diagnostic Interview for Genetic Studies (DIGS) (Nurnberger et al., 1994), the Family Interview for Genetic Studies (FIGS), and medical records. The BD group comprised 36 males (34.9%) and 61 females (59.2%), aged 24–70 years (M ¼51.26, S.D. ¼12.10). The HC subjects were recruited from a number of sources, including advertisements in the local community and newspaper, and the ASRB, and consisted of 37 males (45.7%) and 44 females (54.3%), aged 23–69 years (M ¼ 44.65, S.D. ¼12.86). There were missing data on less than 10% of items for one HC participant and 31 clinical participants; missing data were replaced with the group median for each item. 2.2. Materials 2.2.1. Cognitive emotion regulation questionnaire (CERQ) The CERQ measures various types of cognitive strategies employed to regulate emotion in response to the experience of threatening or stressful life events (Garnefski et al., 2001). The CERQ is a 36-item questionnaire, consisting of 9 conceptually distinct subscales (4 items each), each pertaining to a particular type of regulatory strategy. A person’s tendency to engage in each strategy is measured on a 5-point Likert scale ranging from 1 (almost never) to 5 (almost always). Individual subscale scores are obtained by summing the scores for each strategy (ranging from 4 to 20); the higher the subscale score, the more often the cognitive strategy is used. The four maladaptive subscales of the CERQ include: self-blame (thoughts of blaming yourself for what you have experienced), otherblame (thoughts of blaming another person for what you have experienced), rumination (thinking about feelings and thoughts associated with the negative event), and catastrophising (thoughts that over-emphasize the significance and extent of the experience). The five positive subscales include: putting into perspective (thoughts that minimise the seriousness of the event relative to other life events), positive refocusing (distracting oneself from thinking about the event by focusing on positive thoughts or issues), positive reappraisal (reframing the event in a positive light), acceptance (accepting the experience and resigning

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

23

Table 1 Demographic information and DASS scores for each group.

Age (years) Gender (M/F) DASS Depression DASS Anxiety/Stress Composite HPS DIP Positive symptoms DIP Negative symptoms

SZ (n ¼126)

BD-I (n ¼97)

HC (n¼ 81)

Significant group differences

45.46 (10.96) 73/53 12.61 (10.73) 11.80 (8.78) 13.08 (8.63) 2.89 (3.63) 0.94 (1.23)

51.26 (12.10) 36/61 11.89 (11.18) 10.89 (9.04) 17.23 (11.28) na na

44.65 (12.86) 37/44 5.02 (8.81) 4.77 (5.55) 8.86 (7.62) na na

BD 4SZnn, BD 4HCnn SZ more likely to be male; BD more likely to be female SZ 4HCnnn; BD 4HCnnn SZ 4BDn 4HCnnn BD 4SZnnn 4HCnnn

Note. DASS ¼ Depression Anxiety Stress Scales; HPS ¼Hypomanic Personality Scale; DIp ¼Diagnostic Interview for Psychosis. p o 0.05. p o0.01. nnn p o0.001. n

nn

oneself to what has happened), and refocus on planning (thinking about how to handle the negative event and what steps to take). Internal consistencies of CERQ subscales range from 0.68 to 0.8312 (Garnefski et al., 2001), and evidence for discriminant and convergent validity has been reported (Garnefski et al., 2004, 2005).

Individual packages of questionnaires and invitation letters were sent, via postal mail, to 513 ASRB participants and 651 BD patients (Mitchell et al., 2009). Response rates were 15% for BD and 37% for ASRB participants. Anonymity of responses was maintained through the use of de-identified codes that allowed the collation of data with existing demographic information within each of the original sources of participants.

2.2.2. Symptomatology measures 2.2.2.1. Depression Anxiety Stress Scales (DASS). The DASS-42 is comprised of a set of three self-report scales designed to measure the negative emotional states of depression, anxiety, and stress (Lovibond and Lovibond, 1995). Each of the three DASS scales contains 14 items, divided into subscales of 2–5 items with similar content. The depression scale assesses dysphoria, hopelessness, devalued life, selfdeprecation, lack of interest/involvement, anhedonia, and inertia. The anxiety scale assesses autonomic arousal, skeletal muscle effects, situational anxiety, and subjective experience of anxious affect. The stress scale is sensitive to levels of chronic non-specific arousal and assesses difficulty relaxing, nervous arousal, and becoming easily upset/agitated, irritable/over-reactive and impatient. Participants are asked to use 4-point frequency scales to rate the extent to which they have experienced each symptom over the past week. Scores for depression, anxiety, and stress are calculated by summing the scores for the relevant items and demonstrate high internal consistency (Lovibond and Lovibond, 1995).2 2.2.2.2. Hypomanic Personality Scale (HPS). The HPS is a self-report scale comprising 48 true-false items to measure hypomanic temperament (Chapman et al., 1976). The scale demonstrates good reliability, with a reported coefficient alpha of 0.87 in an undergraduate sample (n ¼1519),2 and sound test–retest reliability (a ¼ 0.81) following an interval of 15 weeks (n¼ 89) (Meyer, 2002). Construct validity of this scale is supported by evidence that shows an association between high scores on the HPS and meeting more diagnostic criteria for mania (Meyer, 2002). 2.2.2.3. Diagnostic Interview for Psychosis (DIP). We used specific DIP items (Castle et al., 2006), completed at initial assessment within the ASRB, to determine the severity of positive and negative symptoms in schizophrenia. A positive symptom count was obtained by computing the total score of past year hallucination and delusion ratings (DIP items 49–53 and 58–64, respectively), with a possible range of scores from 0 to 25. A negative symptom count was obtained via summation of DIP items assessing restricted affect, blunted affect, and negative formal thought disorder (DIP items 90, 91 and 97), making a possible range of scores from 0 to 6. In addition, a paranoid subgroup of SZ patients was determined based on the past year scores from DIP item 60, rating the presence of ‘persecutory delusions’. 2.3. Procedure A letter of invitation and the set of self-report questionnaires were sent to all participants who had indicated interest in participating, or willingness to be contacted for subsequent research, following participation in the Australian Schizophrenia Research Bank (ASRB), an established register of participants and research data collected by scientific collaborators across five Australian states and territories (Loughland et al., 2011) or the Bipolar Disorder Family Study and Sydney Bipolar Disorder Clinic (Mitchell et al., 2009). The willing participants completed the questionnaires and returned the documents anonymously via reply-paid envelopes. Fifteen of the HC participants completed the questionnaire booklet as part of their participation in a larger study on emotion regulation.

2 For the internal consistencies (Cronbach’s alpha) of each subscale in the current sample please refer to the Supplementary material.

3. Results 3.1. CERQ and DASS subscale inter-correlations within each group To investigate the legitimacy of analysing relationships between each of the DASS and CERQ subscales separately, zero-order Pearson’s product–moment correlations were carried out to investigate subscale inter-correlations and potential multicollinearity (where any Pearson’s r value above 0.70 is considered of serious concern), and are reported in the Supplementary material. The only correlation consistently above this threshold for each group was between DASS Anxiety and DASS Stress (SZ Pearson’s r¼0.841, po0.0005; BD Pearson’s r¼ 0.838, po0.0005; HC Pearson’s r¼ 0.745, po0.0005). A composite ‘‘DASS anxiety/stress score’’ was therefore used in all subsequent analyses. 3.2. Demographic data Descriptive statistics for demographic data are presented in Table 1. A one-way analysis of variance (ANOVA) was conducted to investigate group differences in age. There was a significant main effect of group (F2,303 ¼8.89, p o0.0005); Tukey’s HSD tests revealed that the BD group was significantly older than both the HC (p ¼0.001) and SZ groups (p¼ 0.001), but the SZ and HC groups did not significantly differ in age (p¼ 0.882). Sex distribution was significantly different between the three groups (w2 ¼9.76, p ¼ 0.008), with participants in the SZ group more likely to be male, and BD patients more likely to be female, relative to HC. Age and gender were thus employed as covariates in subsequent analyses. Additionally, the paranoid and non-paranoid SZ subgroups (determined from the DIP) consisted of 39 SZ participants (age: M¼43.13, S.D. ¼10.65; gender: 21 males, 18 females), and 85 SZ participants (age: M ¼46.44, S.D. ¼10.83; gender: 52 males, 33 females), respectively (there were missing data for two participants). There was no difference in the age (p¼0.115) or gender distribution (p¼0.441) of these subgroups. 3.3. Clinical symptomatology Descriptive statistics for clinical data are also presented in Table 1. To investigate group differences in symptom scores, we conducted a 3 (group: SZ, BD, HC)  3 (symptom scale: DASS depression, DASS anxiety/stress, HPS) multivariate analysis of covariance (MANCOVA), controlling for age and gender. There was a

24

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

significant main effect of group for both DASS subscales: depression (F4,303 ¼8.45, po0.0005), and anxiety/stress (F4,303 ¼13.06, po 0.0005), as well as HPS total score (F4,303 ¼16.26, po0.0005). A series of univariate analyses of covariance (ANCOVAs) were subsequently conducted for each subscale to investigate specific group differences in clinical symptoms, controlling for age and gender. The SZ group reported significantly higher levels of symptoms than the HC group on all scales: depression (F3,206 ¼11.24, po0.0005; partial Z2 ¼ 0.142), anxiety/stress (F3,206 ¼14.40, po0.0005; partial Z2 ¼0.175), and HPS (F3,206 ¼ 8.20, po0.0005; partial Z2 ¼0.108). The BD group also reported significantly higher levels of depression (F3,177 ¼6.99, po0.0005; partial Z2 ¼0.108), anxiety/stress (F3,177 ¼14.09, po0.0005; partial Z2 ¼0.195), and hypomanic temperament (F3,177 ¼20.97, po0.0005; partial Z2 ¼0.266) than HCs. For the clinical groups, SZ participants had significantly higher scores on the DASS composite anxiety/stress subscale (F3,222 ¼3.84, p¼0.01; partial Z2 ¼0.050) than BDs, while BD participants showed greater hypomanic tendencies (F3,222 ¼9.24, po0.0005; partial Z2 ¼0.112) than SZ. Group differences between the paranoid and non-paranoid SZ participants also revealed significantly higher levels of anxiety/stress (F1,123 ¼6.26, p¼0.014; partial Z2 ¼0.049) and hypomanic tendencies (F1,123 ¼8.76, p¼0.004; partial Z2 ¼0.067), but not depression (p¼0.093), in the paranoid subgroup. Furthermore, paranoid SZ patients had significantly more positive psychotic symptoms (F1,123 ¼75.59, po0.0005; partial Z2 ¼0.385) but not negative symptoms (p¼ 0.614).

significantly more likely to employ the regulatory strategy of putting into perspective (F3,222 ¼4.88, p ¼0.003; partial Z2 ¼0.063) and showed a strong trend towards greater employment of blaming others (F3,222 ¼4.88, p ¼0.059; partial Z2 ¼0.033), while BD reported greater use of rumination (F3,222 ¼3.26, p ¼0.023; partial Z2 ¼0.043) and self-blame (F3,222 ¼3.90, p¼ 0.01; partial Z2 ¼0.051) in comparison to SZ. In comparison to the HC group, both SZ and BD participants reported greater use of rumination (SZ: F3,206 ¼ 8.05 p o0.0005, partial Z2 ¼0.106; BD: F3,177¼14.16, po0.0005, partial Z2 ¼ 0.196), catastrophising (SZ: F3,206 ¼10.17, po0.0005, partial Z2 ¼0.131; BD: F3,177¼4.77, p ¼0.003, partial Z2 ¼0.076) and self-blame (SZ: F3,206 ¼5.12, p¼ 0.002, partial Z2 ¼0.070; BD: F3,177¼9.25, p o0.0005, partial Z2 ¼0.138), but less putting into perspective (SZ: F3,206 ¼4.65, p ¼0.004, partial Z2 ¼0.064; BD: F3,177¼3.10, p ¼0.028, partial Z2 ¼0.051). Compared to HC, SZ participants were more likely to blame others (F3,206¼3.35, p ¼0.02, partial Z2 ¼ 0.047). A comparison of the paranoid and non-paranoid SZ subgroups did not reveal any significant group difference in the use of other blame (p ¼0.656). To investigate the association of psychotic symptomatology with the frequency of use of CERQ strategies, zero-order correlations were carried out using Pearson’s product–moment correlations. There were no significant associations between levels of positive or negative symptomatology and frequency of use of CERQ strategies in SZ patients. Unfortunately, the equivalent information on current symptomatology of the BD group was unavailable to undertake comparable analyses.

3.4. Cognitive emotion regulation (CERQ) in SZ and BD groups 3.5. CERQ predictors of depressed mood, anxiety and stress Means and standard deviations for the CERQ subscales for all groups are presented in Table 2. To investigate group differences in the frequency of use of CERQ strategies (represented by total scores on each subscale), we conducted a MANCOVA with group (SZ, BD, HC) as the independent variable and total scores on the nine CERQ subscales entered as dependent variables, controlling for age and gender. This revealed a significant main effect of group for rumination (F4,303 ¼10.49, po0.0005; partial Z2 ¼0.123), blaming others (F4,303 ¼3.54, p¼0.008; partial Z2 ¼0.045), catastrophising (F4,303 ¼7.89, po0.0005; partial Z2 ¼ 0.095), self-blame (F4,303 ¼7.62, po0.0005; partial Z2 ¼0.060) and putting into perspective (F4,303 ¼4.75, p¼0.001; partial Z2 ¼0.093), but not for reappraisal (F4,303 ¼1.95, p¼0.103; partial Z2 ¼0.025), acceptance (F4,303 ¼ 0.68, p¼0.603; partial Z2 ¼0.009), positive refocusing (F4,303 ¼2.12, p¼ 0.08; partial Z2 ¼ 0.027), or refocus on planning (F4,303 ¼1.23, p¼0.0.299; partial Z2 ¼0.016). For the five significant CERQ subscales, a series of ANCOVAs, controlling for age and gender, were subsequently undertaken to examine pairwise group differences in the frequency of use of these strategies. Compared to the BD group, SZ participants were

Zero-order correlations between CERQ and DASS subscales were carried out using Pearson’s product–moment correlations, and are reported in the Supplementary material. The CERQ subscales of rumination, catastrophising and self-blame were moderately and positively associated with DASS depression in all groups (Pearson’s r ranging from 0.240 to 0.486, all po0.007), while positive reappraisal and putting into perspective showed moderate negative associations with DASS depression in all groups (Pearson’s r ranging from  0.288 to 0.435, all p o0.004). For DASS anxiety/stress, rumination, catastrophising and selfblame revealed moderate positive associations in all groups (Pearson’s r ranging from 0.296 to 0.499, all po0.001). These correlations also show the existence of collinearity among CERQ subscales. Diagnostics for collinearity of CERQ predictors were therefore considered in subsequent regression analyses, but no Variance Inflation Factor (VIF) score was found to be above 5 (VIF scores above 10 are considered of serious concern). To examine the most parsimonious set of CERQ strategies for predicting mood symptomatology in the three experimental groups,

Table 2 Means and standard deviations for all groups on the CERQ. CERQ subscale

SZ (n¼126)

BD (n¼ 97)

HC (n¼ 81)

Significant group differences

Rumination Catastrophising Self-blame Other-blame Positive reappraisal Acceptance Putting into perspective Positive refocusing Refocus on planning

12.55 10.39 10.96 9.84 13.60 12.73 12.92 10.97 13.40

13.38 9.46 11.86 8.64 13.30 13.08 12.80 9.93 13.23

10.64 7.64 9.42 8.34 14.59 12.41 14.40 10.56 14.01

BD 4SZn 4 HCnnn SZ 4HCnnn; BD 4HCnn BD 4SZnn 4HCnn SZ 4BD; SZ 4HCn ns ns SZ 4BDnn; SZ oHCnn; BD oHCn ns ns

(3.61) (3.99) (3.27) (3.96) (3.77) (2.87) (3.61) (3.60) (3.42)

Note. CERQ ¼Cognitive Emotion Regulation Questionnaire. p o 0.05. p o0.01. nnn po 0.001. n

nn

(3.39) (3.64) (4.30) (2.74) (4.06) (3.14) (3.83) (4.07) (3.39)

(3.47) (3.25) (2.69) (2.82) (3.44) (3.00) (3.37) (3.73) (2.80)

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

25

Table 3 Results of (full model) backward step-wise regression analyses for SZ, BD, and HC groups (unstandardised b coefficients). CERQ subscales

DASS

HPS total score

Depression

Rumination Positive reappraisal Other-blame Acceptance Catastrophising Putting in perspective Positive refocusing Refocus on planning Self-blame Gender Total variance explained (Adjusted R2)

Anxiety/Stress composite

SZ

BD

HC

SZ

BD

HC

SZ

BD

HC

0.192  0.616 0.280 0.026 0.345  0.371 0.003  0.493 1.026nnn  3.709n 38.7%nnn

1.648nnn  0.734  0.160 0.862n  0.356  0.221 0.373  0.574  0.022 2.163 30.5%nnn

0.809n  0.522  0.304 0.239 0.330  0.314 0.288  0.598 0.577  3.819n 43.2%nnn

0.044 0.076  0.080  0.225 0.724nn  0.297 0.273  0.441 1.000nnn  2.409 33.0%nnn

1.009nn  0.965n  0.214  0.265 0.495 0.598 0.307 0.231 0.228  2.329 29.2%nnn

0.523n 0.143  0.055  0.016 0.404 0.031 0.305  0.786nn 0.282  2.449n 35.1%nnn

0.277  0.108 0.052  0.738 0.579n 0.144 0.407 0.128 0.694nn  2.963 21.2%

1.136nn  0.042 0.101  1.000nn 0.632 0.169  0.025 0.156 0.431  1.431 24.6%

0.733n 0.865n 0.349  0.251 0.178 0.339 0.220  0.783  0.057  4.108n 20.3%

Note. CERQ ¼ Cognitive Emotion Regulation Questionnaire; DASS¼ Depression Anxiety Stress Scales; HPS¼ Hypomanic Personality Scale. p o 0.05. p o0.01. nnn p o0.001. n

nn

a series of three stepwise backward multiple regression analyses were conducted using the symptom subscales as the dependent variables, controlling for gender. All CERQ subscales were entered as independent variables in the first step; those that did not contribute significantly (p40.01) were removed at each step following previously reported procedures (Garnefski et al., 2001, 2002; Garnefski and Kraaij, 2006). A summary of the results of regression analyses are presented in Table 3. 3.5.1. SZ group results DASS Depression: the combination of all CERQ subscales yielded a significant model (F10,125 ¼ 8.89, po0.0005), which explained 38.7% of the variance in depression scores (adjusted R2 ¼0.387). Step-wise removal of the least predictive variables yielded a final model (F2,125 ¼33.59, po0.0005) comprising increased self-blame (b¼1.33, po0.0005) and decreased positive reappraisal (b¼  1.27, po0.0005), explaining 34.3% of the variance in depression scores (adjusted R2 ¼ 0.343). For the paranoid subgroup, the final model (F1,38 ¼8.33, p¼0.006) consisted of increased self-blame (b¼1.34, p¼0.006) only, explaining 16.2% of the variance in depression scores (adjusted R2 ¼0.162). DASS Anxiety/Stress: the combination of all CERQ subscales yielded a significant model (F10,125 ¼7.14, po0.0005), explaining 29.3% of the variance (adjusted R2 ¼0.330) in the anxiety/stress composite score. Step-wise removal of the least predictive variables yielded a final model (F2,125 ¼22.12, p o0.0005) comprising increased self-blame (b¼ 1.97, po0.0005) and catastrophising (b¼1.35, p o0.0005), and decreased refocus on planning (b¼ 0.98, p¼0.01) explaining 33.7% of the variance in anxiety/ stress scores (adjusted R2 ¼0.337). In the paranoid subgroup, the final model (F1,38 ¼9.14, p ¼0.005) consisted only of increased self-blame (b¼1.07, p ¼0.005), explaining 17.6% of the variance (adjusted R2 ¼0.176) in anxiety/stress scores. HPS: the combination of all CERQ subscales yielded a significant model (F10,125 ¼ 4.36, po0.0005), which explained 21.2% of the variance in HPS scores (adjusted R2 ¼0.212). Subsequent stepwise removal of the least predictive variables yielded a final model (F2,125 ¼ 15.36, po0.0005) comprising increased self-blame (b¼0.67, p ¼0.005) and catastrophising (b ¼0.60, p¼0.002), explaining 18.7% of the variance in hypomania scores (adjusted R2 ¼0.187). The final model (F1,38 ¼13.26, p ¼0.001) for the paranoid subgroup consisted of increased self-blame only (b¼1.33, p ¼0.001) and explained 24.4% of the variance in hypomania scores (adjusted R2 ¼ 0.244).

3.5.2. BD group results DASS Depression: the combination of all CERQ subscales yielded a significant model (F10,96 ¼5.21, po0.0005), which explained 30.5% of the variance in depression scores (adjusted R2 ¼0.305). Subsequent step-wise removal of the least predictive variables yielded a final model (F2,96 ¼20.62, po0.0005) comprising increased rumination (b¼1.55, po0.0005) and decreased positive reappraisal (b¼  0.73, p¼ 0.003), explaining 29% of the variance in depression scores (adjusted R2 ¼0.290). DASS Anxiety/Stress: the combination of all CERQ subscales yielded a significant model (F10,96 ¼4.95, p o0.0005), which explained 29.2% of the variance in the anxiety/stress composite score (adjusted R2 ¼0.292). Subsequent step-wise removal of the least predictive variables yielded a final model (F1,96 ¼16.15, po0.0005) comprised of increased rumination (b¼1.15, po0.0005) and decreased positive reappraisal (b¼ 0.53, p¼0.009), explaining 24% of the variance in anxiety/stress scores (adjusted R2 ¼0.240). HPS: The combination of all CERQ subscales yielded a significant model (F10,96 ¼4.13, po0.0005), which explained 24.6% of the variance in HPS scores (adjusted R2 ¼0.246). Subsequent stepwise removal of the least predictive variables yielded a final model (F1,96 ¼27.37, po0.0005) comprised only increased rumination (b¼1.58, po0.0005), explaining 21.6% of the variance in hypomanic personality scores (adjusted R2 ¼0.216). 3.5.3. HC group results DASS Depression: the combination of all CERQ subscales yielded a significant model (F10,80 ¼7.08, po0.0005), which explained 43.2% of the variance in depression scores (adjusted R2 ¼0.432). Subsequent step-wise removal of the least predictive variables yielded a final model (F3,80 ¼17.97, p o0.0005) comprising increased rumination (b ¼1.20, po0.0005), decreased positive reappraisal (b ¼  1.13, p o0.0005) and gender (b¼  4.28, p¼0.008; such that males had higher scores), explaining 38.9% of the variance in depression scores (adjusted R2 ¼0.389). DASS Anxiety/Stress: the combination of all CERQ subscales yielded a significant model (F10,80 ¼5.32, po0.0005), which explained 35.1% of the variance in the anxiety/stress composite score (adjusted R2 ¼0.351). Subsequent step-wise removal of the least predictive variables yielded a final model (F10,80 ¼11.87, po0.0005) that comprised increased rumination (b¼0.84, po0.0005), decreased positive refocusing (b¼ 0.38, p¼0.01) and refocus on planning (b¼ 0.82, po0.0005), and gender (b ¼  2.93, p¼0.005; such that males had higher scores),

26

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

explaining 35.2% of the variance in anxiety/stress scores (adjusted R2 ¼0.352). HPS: the combination of all CERQ subscales yielded a significant model (F10,80 ¼ 3.03, p ¼0.003), which explained 20.3% of the variance in hypomania scores (adjusted R2 ¼0.203). Subsequent step-wise removal of the least predictive variables yielded a final model (F2,80 ¼8.04, p ¼0.001) comprising increased rumination (b¼0.63, p ¼0.006) and positive reappraisal (b¼ 0.63, p¼ 0.008), explaining 15% of the variance in HPS scores (adjusted R2 ¼ 0.150).

4. Discussion This study examined whether the propensity to use particular cognitive strategies to regulate negative affect differed among patients with SZ and BD, as well as the utility of these cognitive strategies in predicting current levels of depression, anxiety, stress, and propensity for (hypo)mania in these groups. We firstly examined the frequency of use of particular cognitive strategies that would normally be employed to regulate subjective affect in response to a negative event. As predicted, both the SZ and BD groups reported greater employment of maladaptive emotion regulation strategies (namely rumination, catastrophising and self-blame), as well as less use of the adaptive strategy of putting into perspective, relative to healthy controls. In addition, we found that the SZ group employed other-blame more frequently than both HC and BD groups. There were no group differences for any of the other adaptive reframing strategies (i.e. positive reappraisal, refocus on planning, positive refocusing, acceptance). With regard to the clinical utility of the CERQ subscales: within the SZ group, increased self-blame most parsimoniously predicted current levels of symptomatology, with the addition of decreased positive reappraisal in depression, and increased catastrophising and reduced refocus on planning in anxiety, stress and hypomania. In contrast, BD patients’ increased use of rumination was the most consistent predictor of all symptom levels, with decreased use of positive reappraisal also predicting depression, anxiety and stress. Collectively, these results confirm differential patterns of dysfunctional emotion regulation in SZ and BD (Livingstone et al., 2009), the shared features of which (rumination, catastrophising, selfblame) have been previously associated with depression (Garnefski and Kraaij, 2006). In SZ, the additional tendency to over-engage in other-blame that emerged is consistent with the self-serving (external-personalising) attribution style consistently demonstrated in psychotic or manic patients with paranoia (Bentall and Kinderman, 1999). However, an examination of SZ patients with paranoid ideation revealed no difference in the use of this strategy to those with no paranoia. Interestingly, neither clinical group reported decreased use of reframing strategies in this study (besides putting into perspective), despite previous evidence of under-utilisation of cognitive reframing techniques in both psychotic and mood disorders, relative to healthy controls (Garnefski and Kraaij, 2006; Livingstone et al., 2009; Green et al., 2011). This supports the proposal that the capacity for SZ and BD patients to effectively implement cognitive reframing strategies for self-regulation of negative affect may reflect underlying cognitive deficits in frontal executive function that have been consistently reported in both disorders (Daban et al., 2006; Green et al., 2007). Evidence in support of this proposal has been recently demonstrated in major depressive disorder, where inhibitory deficits were associated with excessive rumination (Joorman and Gotlib, 2010); it is thus conceivable that known impairments in frontal neurocognitive systems in SZ and BD may directly effect the ability to manipulate and reinterpret the significance of emotional stimuli, or consider alternate cognitive coping strategies. Direct tests of these hypotheses are yet to be carried out.

Despite this broadly similar over-use of maladaptive cognitive strategies and ineffective use of adaptive reframing strategies to regulate negative affect, distinct patterns of cognitive strategies in each group were uniquely predictive of mood symptomatology, and thus distinguished the disorders: in schizophrenia, increased self-blame and catastrophising were most predictive of current symptoms; in BD increased rumination and reduced positive reappraisal were most predictive of symptomatology. More generally, these findings are consistent with a recent study which demonstrated ‘emotion-oriented’ coping (i.e. self-blame, catastrophising, rumination) as most predictive of psychotic-like experiences in adolescents followed over a 3-year period, with the employment of more adaptive, ‘task-oriented’ coping (i.e. positive reappraisal, refocus on planning) associated with a decrease in these experiences over time (Lin et al., 2011). 4.1. Limitations Limitations of this study include the use of self-report measures: with respect to emotion regulation, individuals are not always consciously aware or sure of their own use of cognitive emotion regulation strategies in stressful situations—these may be context-dependent, or unconsciously triggered, and memory biases may also impact upon the self-report of these data. Questionnaire methods therefore introduce potential reporting biases that could be avoided by the employment of moment-tomoment experience-sampling technologies. The use of simple mail-out methodology in this study also meant that there was no information about illness stage and level of insight for both clinical groups, as well as limited information on current psychotic symptoms of the bipolar disorder group. Use of this methodology also meant that information on current psychotic symptoms of the SZ group, and determination of a paranoid subgroup (from the DIP), were obtained at initial assessment within the ASRB sample, which took place approximately 12 months prior to the completion of self-report questionnaires. In addition, given that no information was received from non-responders to the questionnaire mail-out, there was no means of determining systematic clinical or demographic differences among responders and non-responders for any of the experimental groups. Future studies would benefit from face-to-face interviews to determine any specific clinical or demographic features that might contribute to the use of particular cognitive strategies for emotion regulation. Finally, the cognitive emotion regulation measure used in the study assessed only the regulation of negative affect; strategies utilised in response to positive events might be important in the development of manic symptoms. 4.2. Clinical implications In consideration of the study’s limitations, the present findings have important clinical implications. Self-regulatory dysfunction is implicated in depression and anxiety, and these symptoms are present at significant levels in patients with psychotic disorders. The delineation of cognitive factors associated with risk for such symptomatology may therefore assist in determining appropriately targeted intervention. For BD patients, rumination appears to be the key issue contributing to higher levels of negative affect and propensity for hypomania, while for SZ patients it is over-use of self-blame. These findings suggest that differential interventions to improve clinical symptoms may be needed for those with a particular pattern of illness. Furthermore, even though both SZ and BD groups reported engaging in adaptive reframing more often than maladaptive strategies these were not effective in reducing affective symptoms, indicating the potential role of underlying general cognitive dysfunction (Kring and Werner,

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

2004). Therapeutic interventions to improve contextual awareness (i.e., to reduce maladaptive self-blame and catastrophising, and improve various cognitive reframing skills) may be useful for patients to cope with day-to-day stressors or more serious adverse experiences, in the context of other illness symptoms. 4.3. Conclusion In summary, this study highlights the predominant similarities in cognitive strategies used by SZ and BD individuals in attempts to regulate negative affect, and the extent to which specific combinations of cognitive strategies predicted negative affect and hypomanic temperament in these groups. Competency in the execution of specific emotion regulation strategies (e.g., cognitive reappraisal) may further rest on other component neuropsychological skills (e.g., executive functioning), for which remediation may also facilitate access to a wider range of emotion regulation capacities in various emotional contexts. Further investigation of the association between neurocognitive skills and emotion regulation capacities in psychotic and mood disorders is warranted.

Funding source This research was supported by the National Health and Medical Research Council of Australia (Project grant 630471 and Programme grant 510135) and the Australian Research Council (Future Fellowship FT0991511, held by MJG). The funding bodies had no role in the design of the study, collection and analysis of data, or the decision to publish.

Acknowledgements This research was supported by the Australian National Health and Medical Research Council (NHMRC) Project Grant held by Green (#630471), and used data from the Australian Schizophrenia Research Bank, funded by the NHMRC Enabling Grant (#386500 held by V. Carr, U. Schall, R. Scott, A. Jablensky, B. Mowry, P. Michie, S. Catts, F. Henskens and C.Pantelis; Chief Investigators), and the Pratt Foundation, Ramsay Health Care, the Viertel Charitable Foundation, as well the Schizophrenia Research Institute, using an infrastructure grant from the NSW Ministry of Health. MJG was supported by an Australian Research Council Future Fellowship (FT0991511). We acknowledge Carmel Loughland, Kathryn McCabe, and Jason Bridge for management and quality control of data obtained from the Australian Schizophrenia Research Bank. We would also like to thank the volunteers who participated in this study.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.psychres.2013. 02.021.

References Aghevli, M.A., Blanchard, J.J., Horan, W.P., 2003. The expression and experience of emotion in schizophrenia: a study of social interactions. Psychiatry Research 119 (3), 261–270. Bentall, R., Corcoran, R., Howard, R., Blackwood, N., Kinderman, P., 2001. Persecutory delusions: a review and theoretical integration. Clinical Psychology Review 21 (8), 1143–1192. Bentall, R.P., Kinderman, P., 1999. Self-regulation, affect and psychosis: the role of social cognition in paranoia and mania. In: Dalgleish, T., Power, M.J. (Eds.),

27

Handbook of Cognition and Emotion. John Wiley & Sons Ltd, Chichester, UK, pp. 353–381. ¨ Bora, E., Yucel, M., Pantelis, C., 2010. Cognitive impairment in schizophrenia and affective psychoses: implications for DSM-V criteria and beyond. Schizophrenia Bulletin 36 (1), 36–42. Castle, D.J., Jablensky, A., McGrath, J.J., Carr, V., Morgan, V., Waterreus, A., Valuri, G., Stain, H., McGuffin, P., Farmer, A., 2006. The Diagnostic Interview for Psychoses (DIP): development, reliability and applications. Psychological Medicine 36 (1), 69–80. Chapman, L.J., Chapman, J.P., Raulin, M.L., 1976. Scales for Physical and Social Anhedonia. Journal of Abnormal Psychology 85 (4), 374–382. Daban, C., Martinez-Aran, A., Torrentm, C., Tabare´s-Seisdedos, R., Balanza´-Martı´nez, V., Salazar-Fraile, J., Selva-Vera, G., Vieta, E., 2006. Specificity of cognitive deficits in bipolar disorder versus schizophrenia. A systematic review. Psychotherapy & Psychosomatics 75 (2), 72–84. Eisenberg, N., 2000. Emotion, regulation, and moral development. Annual Review of Psychology 51, 665–697. Ellgring, H., Smith, M., 1998. Affect Regulation During Psychosis. Oxford University Press, New York, NY. Garnefski, N., Kraaij, V., Spinhoven, P., 2006. Relationships between cognitive emotion regulation strategies and depressive symptoms: a comparative study of five specific samples. Personality and Individual Differences 40 (8), 1659–1669. Garnefski, N., Kraaij, V., Spinhoven, P., 2001. Negative life events, cognitive emotion regulation and emotional problems. Personality and Individual Differences 30 (8), 1311–1327. Garnefski, N., Kraaij, V., van Etten, M., 2005. Specificity of relations between adolescents’ cognitive emotion regulation strategies and internalizing and externalizing psychopathology. Journal of Adolescence 28 (5), 619–631. Garnefski, N., Teerds, J., Kraaij, V., Legerstee, J., van den Kommer, T., 2004. Cognitive emotion regulation strategies and depressive symptoms: differences between males and females. Personality and Individual Differences 36 (2), 267–276. Garnefski, N., Van Den Kommer, T., Kraaij, V., Teerds, J., Legerstee, J.S., Onstein, E., 2002. The relationship between cognitive emotion regulation strategies and emotional problems: comparison between a clinical and a non-clinical sample. European Journal of Personality 16 (5), 403–420. Green, M.F., Kern, R.S., Heaton, R.K., 2004. Longitudinal studies of cognition and functional outcome in schizophrenia: implications for MATRICS. Schizophrenia Research 72 (1), 41–51. Green, M.J., Cahill, C.M., Malhi, G.S., 2007. The cognitive and neurophysiological basis of emotion dysregulation in bipolar disorder. Journal of Affective Disorders 103 (1–3), 29–42. Green, M.J., Lino, B.J., Hwang, E.J., Sparks, A., James, C., Mitchell, P.B., 2011. Cognitive regulation of emotion in bipolar I disorder and unaffected biological relatives. Acta Psychiatrica Scandinavica 124 (4), 307–316. Green, M.J., Malhi, G.S., 2006. Neural mechanisms of the cognitive control of emotion. Acta Neuropsychiatrica 18, 144–153. Green, M.J., Phillips, M.L., 2004. Social threat perception and the evolution of paranoia. Neuroscience & Biobehavioral Reviews 28 (3), 333–342. Gross, J.J., 1998. The emerging field of emotion regulation: an integrative review. Review of General Psychology 2 (3), 271–299. Gross, J.J., Munoz, R.F., 1995. Emotion regulation and mental health. Clinical Psychology: Science and Practice 2 (2), 151–164. Gur, R.E., Kohler, C.G., Ragland, J.D., Siegel, S.J., Lesko, K., Bilker, W.B., Gur, R.C., 2006. Flat affect in schizophrenia: relation to emotion processing and neurocognitive measures. Schizophrenia Bulletin 32, 279–287. Henry, J.D., Green, M.J., de Lucia, A., Restuccia, C., McDonald, S., O’Donnell, M., 2007. Emotion dysregulation in schizophrenia: reduced amplification of emotional expression is associated with emotional blunting. Schizophrenia Research 95 (1–3), 197–204. Henry, J.D., Rendell, P.G., Green, M.J., McDonald, S., O’Donnell, M., 2008. Emotion regulation in schizophrenia: affective, social, and clinical correlates of suppression and reappraisal. Journal of Abnormal Psychology 117 (2), 473–478. Joorman, J., Gotlib, I.H., 2010. Emotion regulation in depression: relation to cognitive inhibition. Cognition and Emotion 24 (2), 281–291. Knowles, R., Tai, S., Christensen, I., Bentall, R., 2005. Coping with depression and vulnerability to mania: a factor analytic study of the Nolen-Hoeksema (1991) Response Styles Questionnaire. British Journal of Clinical Psychology 44 (Pt 1), 99–112. Kring, A.M., Werner, K.H., 2004. Emotion Regulation and Psychopathology. Lawrence Erlbaum Associates, Publishers, Mahwah, NJ. ¨ Lichtenstein, P., Yip, B.H., Bjork, C., Pawitan, Y., Cannon, T.D., Sullivan, P.F., Hultman, C.M., 2009. Common genetic determinants of schizophrenia and bipolar disorder in Swedish families: a population-based study. Lancet 373 (9659), 234–239. Lin, A., Wigman, J.T.W., Nelson, B., Vollebergh, W.A.M., van Os, J., Baksheev, G., Ryan, J., Raaijmakers, Q.A.W., Thompson, A., Yung, A.R., 2011. The relationship between coping and subclinical psychotic experiences in adolescents from the general population—a longitudinal study. Psychological Medicine 41 (12), 2535–2546. Livingstone, K., Harper, S., Gillanders, D., 2009. An exploration of emotion regulation in psychosis. Clinical Psychology and Psychotherapy 16, 418–430. Loughland, C., Draganic, D., McCabe, K., Richards, J., Nasir, A., Allen, J., Catts, S., Jablensky, A., Henskens, F., Michie, P., Mowry, B., Pantelis, C., Schall, U., Scott, R., Tooney, P., Carr, V., 2011. Australian Schizophrenia Research Bank: a database of comprehensive clinical, endophenotypic, and genetic data for

28

J.E. Rowland et al. / Psychiatry Research 208 (2013) 21–28

aetiological studies of schizophrenia. Australian and New Zealand Journal of Psychiatry 44 (11), 1029–1035. Lovibond, S.H., Lovibond, P.F., 1995. Manual for the Depression Anxiety Stress Scales, 2nd ed. Psychology Foundation, Sydney. Malhi, G.S., Lagopoulos, J., Sachdev, P., Mitchell, P.B., Ivanovski, B., Parker, G.B., 2004a. Cognitive generation of affect in hypomania: an fMRI study. Bipolar Disorders 6 (4), 271–285. Malhi, G.S., Lagopoulos, J., Ward, P.B., Kumari, V., Mitchell, P.B., Parker, G.B., Ivanovski, B., Sachdev, P., 2004b. Cognitive generation of affect in bipolar depression: an fMRI study. European Journal of Neuroscience 19 (3), 741–754. McGuffin, P., Farmer, A., 1991. A polydiagnostic application of operational criteria in studies of psychotic illness: development and validation of the OPCRIT system. Archives of General Psychiatry 48, 764–770. Meyer, T.D., 2002. The Hypomanic Personality Scale, the Big Five, and their relationship to depresssion and mania. Personality and Individual Differences 32 (4), 649–660. Mitchell, P.B., Johnston, A.K., Corry, J., Ball, J.R., Malhi, G.S., 2009. Characteristics of bipolar disorder in an Australian specialist outpatient clinic: comparison across large datasets. Australian and New Zealand Journal of Psychiatry 43 (2), 1009–1017. Morris, R.W., Sparks, A., Mitchell, P.B., Weickert, C.S., Green, M.J., 2012. Lack of corticolimbic coupling in bipolar disorder and schizophrenia during emotion regulation. Translational Psychiatry 2, e90.

Nurnberger, J.I., Jr, Blehar, M.C., Kaufmann, C.A., York-Cooler, C., Simpson, S.G., Harkavy-Friedman, J., Severe, J.B., Malaspina, D., Reich, T., 1994. Diagnostic interview for genetic studies. Rationale, unique features, and training. NIMH Genetics Initiative. Archives of General Psychiatry 51 (11), 849–859. (discussion 863-844). Ochsner, K.N., Gross, J.J., 2005. The cognitive control of emotion. Trends in Cognitive Sciences 9 (5), 242–249. Phillips, M.L., Drevets, W.C., Rauch, S.L., Lane, R., 2003. Neurobiology of emotion perception II: implications for major psychiatric disorders. Biological Psychiatry 54, 515–528. Reichenberg, A., Harvey, P.D., Bowie, C.R., Mojtabai, R., Rabinowitz, J., Heaton, R.K., Bromet, E., 2009. Neuropsychological function and dysfunction in schizophrenia and psychotic affective disorders. Schizophrenia Bulletin 35 (5), 1022–1029. Sloan, D.M., Kring, A.M., 2007. Measuring changes in emotion during psychotherapy: conceptual and methodological issues. Clinical Psychology: Science and Practice 14 (4), 307–322. Thomas, J., Bentall, R.P., 2002. Hypomanic traits and response styles to depression. British Journal of Clinical Psychology 41 (Pt 3), 309–313. Thompson, R.A., 1994. Emotion regulation: a theme in search of definition. Monographs of the Society for Research in Child Development 59 (2–3), 25–52. Van Der Gucht, E., Morriss, R., Lancaster, G., Kinderman, P., Bentall, R.P., 2009. Psychological processes in bipolar affective disorder: negative cognitive style and reward processing. British Journal of Psychiatry 194, 146–151.