Elsevier

Biological Psychology

Volume 94, Issue 3, December 2013, Pages 556-561
Biological Psychology

Attention orienting and inhibitory control across the different mood states in bipolar disorder: An emotional antisaccade task

https://doi.org/10.1016/j.biopsycho.2013.10.005Get rights and content

Highlights

  • We examined the influence of mood on inhibitory control and attentional orienting in bipolar disorder.

  • We used an emotional antisaccade task to assess the effect of mood-congruent information on the observer's attention.

  • Patients in their manic and depressive episodes showed impairments in ignoring irrelevant information, predominantly with mood-congruent facial expressions.

  • Latencies in the antisaccade task were longer in bipolar patients (regardless of the presence of affective psychopathology) than in healthy controls (i.e., a trait).

Abstract

An antisaccade experiment, using happy, sad, and neutral faces, was conducted to examine the effect of mood-congruent information on inhibitory control (antisaccade task) and attentional orienting (prosaccade task) during the different episodes of bipolar disorder (BD) – manic (n = 22), depressive (n = 25), and euthymic (n = 24). A group of 28 healthy controls was also included. Results revealed that symptomatic patients committed more antisaccade errors than healthy individuals, especially with mood-congruent faces. The manic group committed more antisaccade errors in response to happy faces, while the depressed group tended to commit more antisaccade errors in response to sad faces. Additionally, antisaccade latencies were slower in BD patients than in healthy individuals, whereas prosaccade latencies were slower in symptomatic patients. Taken together, these findings revealed the following: (a) slow inhibitory control in BD patients, regardless of their episode (i.e., a trait), and (b) impaired inhibitory control restricted to symptomatic patients (i.e., a state)

Introduction

Bipolar disorder (BD) is characterized by succeeding episodes of mania, depression, and euthymia that entail impaired mood regulation even during asymptomatic periods (Goodwin & Jamison, 2007). BD patients exhibit forms of psychological vulnerability such as an impairment interaction between cognitive and emotional networks in the brain (Strakowski, DelBello, & Adler, 2005) that yield deficits in the processing of emotional information and executive functioning. To understand the psychological vulnerability of BD patients, it is crucial to examine in detail the difficulties in cognitive functioning that can result in emotion dysregulation. In the present paper, we examined the interplay between mood symptoms and cognition in BD by assessing the inhibitory control of attention along the different episodes of the disorder (i.e., mania, depression, and euthymia). We do so by presenting emotional information (i.e., happy, sad, neutral) to which participants have to respond (e.g., see García-Blanco, Perea, & Livianos, 2013, for recent behavioral evidence with emotion words in BD patients).

Impaired attention control is an important vulnerability factor for mental disorders, supporting the hypothesis that (abnormal) emotional attention brain processes cause considerable impairment during information processing (see Berggren & Derakshan, 2013). At the theoretical level, cognitive models propose that mood disorders are characterized by impairment in overriding dominant responses and inhibiting the processing of irrelevant information that attracts attention (see Beck, 1976). Indeed, a growing body of research has associated this dysfunction for inhibiting mood-congruent stimuli with the biased processing of new information (see Joormann, Yoon, & Zetsche, 2007). Importantly, negative biases in depression or positive biases in mania may evoke extreme emotional responses that require more effortful inhibitory control, and may represent an important component of emotion dysregulation in BD (see Phillips, Ladouceur, & Drevets, 2008).

An excellent strategy for assessing inhibitory attention control is the antisaccade task (Hallett, 1978). In each trial, while the participant is fixating on a central point, a sudden-onset peripheral visual stimulus appears either to the left or right of the central point. In separate experimental blocks, participants are required to make one of two eye movements: either an eye movement toward the stimulus (prosaccade) or an eye movement away from the stimulus (antisaccade) (see Mueller et al., 2010, for review). The prosaccade task requires participants to orient their attention toward the peripheral stimulus, while the antisaccade task requires participants to inhibit the automatic prosaccade toward a target and voluntarily generate an antisaccade to the mirror position. Thus, a prosaccade involves an automatic orientation response, whereas an antisaccade involves a controlled inhibition response. Importantly, the antisaccade task provides a precise assessment of top–down cognitive processes that influence attention allocation (e.g., beliefs, mood, etc.), which is particularly relevant in individuals with psychopathology (see Hutton & Ettinger, 2006, for a review) – note that prosaccades may be instead more influenced by bottom–up processes (i.e., stimulus-driven attention; see Egeth & Yantis, 1997). Many studies have used the antisaccade task in BD using neutral stimuli (e.g., a small white light) (see Gooding and Tallent, 2001, Gooding et al., 2004, Katsanis et al., 1997). However, the majority of these experiments focused on studying inhibitory control in schizophrenia. The BD group was included merely to examine whether the deficit in the antisaccade task was an inherent feature of schizophrenia or a shared feature with other mental disorders. These studies did not report affective symptoms in BD patients. Several studies reported that BD patients, regardless of whether they were recruited from outpatient or inpatient units, committed more antisaccade errors and were slower in correct antisaccades than healthy individuals. No differences were found across groups in the prosaccade task (e.g., see Katsanis et al., 1997, for in-patients; Gooding & Tallent, 2001, for outpatients). In addition, Gooding et al. (2004) evaluated antisaccade performance at two time points (with an average interval of 33 months) in BD patients. Unlike the healthy controls, BD patients showed temporal instability in accuracy and speed in the antisaccade task, thus suggesting that this deficit in inhibitory control may be a state rather than a trait marker of BD. However, the patients’ affective symptoms were not indicated. We believe that potential confounds such as mixed-mood states or residual symptoms or heterogeneous criteria for saccade definition may have affected previous studies. In fairness to these studies, the focus was on schizophrenia, not on BD.

Of particular interest here is that the antisaccade task can be modified by the substitution of the neutral peripheral target with an emotional stimulus (e.g., a sad or a happy face) (see Derakshan, Salt, & Koster, 2009, for evidence with dysphoric individuals; see also Hardin et al., 2009, for evidence with anxious adolescents). Specifically, Derakshan et al. (2009) used facial expression (angry, happy, and neutral) in anti- and prosaccade tasks in order to examine the effects of subclinical depression (dysphoria) on attentional processing. Participants had to look toward the face (prosaccade task) or look away from the face (antisaccade task). Dysphoric individuals committed more antisaccade errors in response to emotional faces than to neutral faces (18.3% vs. 12.3%, respectively), while this effect did not occur in healthy controls (11.0% vs. 10.6%). No differences were found in the latency data for any of the groups. Derakshan et al. concluded that there is impaired attentional control in response to emotional faces in dysphoria.

We believe that it is important to examine the performance in the antisaccade task in BD patients during their distinct moods (mania, euthymia, and depression) when the peripheral stimuli are emotional images (facial expressions: neutral, happy, and sad). This manipulation allows us to examine the effects of mood on emotional information processing. In this respect, the mood-congruency hypothesis (see Bower, 1981) postulates that positive moods should facilitate orienting toward positive stimuli and hinder their inhibition, and negative moods should facilitate orienting toward negative stimuli and hinder their inhibition (see García-Blanco et al., 2013, for behavioral evidence of a mood-congruency effect with emotional words in BD patients). In addition, the present manipulation also sheds light on the question of neuropathological specificity being a state (e.g., as revealed by differences between symptomatic vs. asymptomatic BD patients) or a trait (e.g., as revealed by differences between BD patients vs. controls).

To our knowledge, the present (emotion-modified) antisaccade experiment is the first that examines the effect of mood-congruent information on inhibitory control (antisaccade task) and attention orienting (prosaccade task) among the different episodes in BD. The present experiment had two specific goals. The first goal was to assess the presence of mood-congruent biases on two different attentional processes (orientation [prosaccades] vs. inhibition [antisaccades]) in BD. If antisaccades reflect voluntary responses subject to inhibitory control that are influenced by top–down processes such as the participants’ mood state (see Hutton & Ettinger, 2006), we would expect a mood-congruent effect in symptomatic patients (i.e., antisaccade errors should be particularly pronounced in response to happy faces for manic patients or to sad faces for depressed patients). In addition, if prosaccades reflect an automatic orientation response that is mainly influenced by bottom-up processes (see Egeth & Yantis, 1997), we expect that the latencies/errors on prosaccades would be modulated by the stimulus valence (i.e., stimulus-driven attention) rather than by the participants’ mood state. The second goal was to examine whether difficulties in inhibitory control in BD patients are a trait (i.e., BD patients, regardless of their episode, should show general impaired inhibitory control reflected as slow [and error-prone] antisaccades relative to healthy individuals; see Gooding and Tallent, 2001, Katsanis et al., 1997) or a state (i.e., BD patients in depressive and manic episodes should commit more antisaccade errors and have slower antisaccades than healthy controls, while there would be differences between asymptomatic patients and healthy individuals; see Gooding et al., 2004).

Section snippets

Participants

The participants were 71 BD patients from the Psychiatry Department (42 from in-patient wards and 29 from the outpatient Bipolar Disorders Unit) at the Hospital Universitario y Politécnico La Fe (Valencia, Spain) and 28 healthy individuals recruited through advertising in the community. Patients fulfilled the DSM-IV-TR criteria for BD and were included in the manic (n = 22), depressed (n = 25), or euthymic (n = 24) group at the time of assessment. Four patients in manic episodes refused to cooperate.

Results

Each task was entered separately into a two-way repeated measures analysis of variance (ANOVA) with Group (control, euthymic, depressed, manic) as the between-subjects factor and Valence (neutral, happy, sad) as the within-subject factor. The dependent variables were the error rates and the correct saccade latencies in the prosaccade and antisaccade tasks. The averages and the standard deviations for each condition are presented in Fig. 2 and Table 2. When the effects were significant, we

Discussion

To the best of our knowledge, this is the first eye-movement experiment that examined inhibitory control when emotional faces (happy, sad, neutral) are processed by BD patients during different episodes (i.e., manic, depressive, and euthymic). The main findings can be summarized as follows. First, patients in manic episodes committed more antisaccade errors in response to happy facial expressions than to sad/neutral facial expressions, while patients in depressive episodes tended to have more

References (31)

  • A.T. Beck et al.

    Manual for the Beck depression inventory—II

    (1996)
  • G.H. Bower

    Mood and memory

    American Psychologist

    (1981)
  • M.G. Calvo et al.

    Emotional and neutral scenes in competition: Orienting, efficiency, and identification

    Quarterly Journal of Experimental Psychology

    (2007)
  • N.C. Ebner et al.

    FACES – A database of facial expressions in young, middle-aged, and older women and men: Development and validation

    Behavior Research Methods

    (2010)
  • H.E. Egeth et al.

    Visual attention: Control, representation, and time course

    Annual Review of Psychology

    (1997)
  • Cited by (40)

    • Task-irrelevant emotional faces impact BOLD responses more for prosaccades than antisaccades in a mixed saccade fMRI task

      2022, Neuropsychologia
      Citation Excerpt :

      In the current task, the emotional faces were irrelevant to saccade performance in all conditions, so participants theoretically could completely disregard the emotional content of the faces unless it captured attention automatically (Vuilleumier et al., 2001). Prior studies examining the behavioral effects of emotional stimuli on the antisaccade task have had mixed findings, partly depending on whether the face was positioned at the target location (García-Blanco et al., 2013; Geringswald et al., 2020; Llamas-Alonso et al., 2020; Reinholdt-Dunne et al., 2012; Wieser et al., 2009) or positioned at central fixation (current study; Mills et al., 2016; Malsert and Grandjean, 2016). Using the face as a target increases its relevance to the task and makes it more likely that the emotional content will impact behavior (Kanske, 2012; Pessoa et al., 2002), whereas the current study focused on participants' ability to inhibit irrelevant emotional information during a concurrent cognitive control task.

    • Variation in antisaccadic response latencies investigated with the hierarchical LATER process model

      2022, Brain and Cognition
      Citation Excerpt :

      Our study is the first to apply advanced process modeling to examine the effects of stimulus size and emotional valence on inhibitory control. Understanding how stimulus size and emotional valance affect inhibitory control is important, as problems in executive functioning, especially with emotional stimuli, are a common finding in psychological disorders, e.g. major depression (Hoffmann et al., 2019), anxiety (Chen et al., 2014), and bipolar disorder (García-Blanco et al., 2013). Our findings demonstrate differences in both accretion and caution parameters in emotional relative to neutral circle stimuli: we found faster accretion rates and lower caution thresholds in emotional face stimuli, while for scrambled faces both accretion rate and caution threshold were lowered compared to circular stimuli.

    • Attentional engagement and inhibitory control according to positive and negative symptoms in schizophrenia: An emotional antisaccade task

      2022, Schizophrenia Research
      Citation Excerpt :

      The inter-trial interval (time between the previous emotional stimulus disappearance and the onset of the initial fixation point) was 600 ms. See Fig. 2. See García-Blanco et al. (2013) for a similar experiment with bipolar disorder. An eye-tracking hardware (SMI RED250) was used to track the eye movements.

    View all citing articles on Scopus
    View full text