حجم و گاه شماری مکانیسم های عصبی تنظیم احساسات در زیرگروه کودکان تهاجمی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|38835||2011||11 صفحه PDF||سفارش دهید||9791 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Brain and Cognition, Volume 77, Issue 2, November 2011, Pages 159–169
Abstract Emotion regulation is a key social skill and children who fail to master it are at risk for clinical disorders. Specific styles of emotion regulation have been associated with particular patterns of prefrontal activation. We investigated whether anxious aggressive children would reveal a different pattern of cortical activation than non-anxious aggressive children and normally-developing children. We examined the magnitude and timing of source activation underlying the N2—an ERP associated with inhibitory control—during a go/nogo task with a negative emotion induction component (loss of earned points). We estimated cortical activation for two regions of interest—a ventral prefrontal and a dorsomedial prefrontal region—for three 100-ms windows over the range of the N2 (200–500 ms). Anxious aggressive children showed high ventral prefrontal activation in the early window; non-anxious aggressive children showed high ventral prefrontal activation in the late window, but only for the duration of the emotion induction; and normally-developing children showed low ventral prefrontal activation throughout. There were no group differences in dorsomedial prefrontal activation. These results suggest that anxious aggressive children recruit ventral prefrontal activation quickly and indiscriminately, possibly giving rise to their rigid, threat-oriented approach to conflict. The late ventral prefrontal activation seen for non-anxious aggressive children may underlie a more delayed, situation-specific, but ineffective response to frustration.
Introduction One of the most important tasks of childhood is learning to modify emotion-driven actions (e.g., hitting when angry or freezing when afraid) and the thoughts and feelings that go with them. This kind of cognitive control frequently falls under the rubric of emotion regulation. The development of emotion regulation is thought to be shaped by both biological factors and environmental transactions in infancy and childhood (see Gross & Thompson, 2007, for a review). With age, children acquire a repertoire of self-regulation strategies ranging from primitive methods, such as gaze aversion and distraction, to sophisticated methods requiring effortful regulation, such as reappraising emotional situations and inhibiting emotional impulses (for reviews see Bradley, 2000, Calkins and Hill, 2007 and Eisenberg et al., 2007; Lewis, Todd, & Xu, 2010; Thompson, 1994). With the development of increasingly sophisticated emotion regulation strategies, children are better able to inhibit their negative feelings and impulsive behaviors in everyday social contexts. However, not all children develop the same skill level when it comes to emotion regulation in general and response inhibition in particular. Inadequate regulatory capabilities can be seen all around us, for example, an angry outburst in a crowded train. Furthermore, the motivations underlying angry actions can vary greatly between people, requiring different emotion regulation styles even within the population of aggressive individuals. Examples of different motivations and their impact on aggressive behavior are discussed in the corpus of work focused on subtyping aggressive behavior problems (e.g., Barratt et al., 1997, Crick and Dodge, 1996, Dodge, 1991, Hoving et al., 1979, Moyer, 1968, Stieben et al., 2007 and Vitiello and Stoff, 1997). As is evident from this body of work, there are many different ways of splitting the pie. Going back to our example, the angry outburst on the train may come from somebody retaliating for being accidentally bumped or from someone who is rigid with anxiety about being late for work again. Because of the high comorbidity found between anxiety disorders and disorders of conduct (Kessler, Chiu, Demler, & Walters, 2005), we use the terms anxious aggressive and non-anxious aggressive to subtype our clinical group. Anxious aggressive children are often described as overcontrolled, cognitively inflexible, inhibited, and/or threat oriented, whereas purely aggressive children (without co-occurring anxiety) are often described as undercontrolled or impulsive (e.g., Eisenberg et al., 2005, Eisenberg et al., 2007 and Granic and Lamey, 2002; Lewis, Granic, & Lamm, 2006; Stieben et al., 2007). Thus, aggressive children with and without co-occurring anxiety problems may have very different emotion regulation styles. Because these regulatory differences can only be inferred on the basis of behavioral observations, we utilized a more direct measure of the cognitive resources recruited for emotion regulation. Specifically, we examined the real-time progression of frontocortical changes underlying emotion regulation in both groups of clinically-defined aggressive children as well as their age-matched normally-developing peers. Researchers have started examining the neural mechanisms underlying emotion regulation with a variety of technologies (e.g., Busatto et al., 2000, Hajcak and Nieuwenhuis, 2006 and Ochsner et al., 2004). However, electroencephalogram (EEG) recordings and the event-related potentials (ERPs) derived from them are particularly useful for studying self-regulation in children because they are relatively quick and nonintrusive (e.g., Hajcak and Dennis, 2009, Jonkman et al., 2003 and Lewis et al., 2006). One ERP component, the mediofrontal N2, is observed between 200 and 500 ms post-stimulus and is generally associated with regulatory processes. These regulatory processes include inhibitory control (Dimoska et al., 2003, Falkenstein et al., 1999, Jonkman et al., 2003, Overtoom et al., 1998, Schmajuk et al., 2006 and Smith et al., 2004) and conflict monitoring or response selection (Bartholow et al., 2005, Bekker et al., 2004, Dimoska et al., 2006, Nieuwenhuis et al., 2004 and Nieuwenhuis et al., 2003). Furthermore, a number of studies have shown greater N2 activation for nogo trials than go trials (e.g., Bekker et al., 2005, Donkers and van Boxtel, 2004, Falkenstein et al., 1999 and Nieuwenhuis et al., 2004), suggesting increased activation when prepotent responses need to be overcome, either through inhibition or conflict monitoring. The N2 may also be sensitive to the processing of emotional information. In a number of studies, emotionally salient faces and words give rise to an N2-like component of greater magnitude for negative stimuli, such as fearful or angry faces, than positive or neutral stimuli (e.g., Lewis et al., 2007, Li et al., 2008, Liddell et al., 2004 and Tucker et al., 2003). It is not clear what functional networks contribute to this greater activation but it may be that negatively valenced stimuli generate fear or shame, and that either the experience or regulation of these emotions underlies increased activation. Several studies from our laboratory have shown emotion-related changes in N2 activation within a go/nogo task (e.g., Lamm and Lewis, 2010, Lewis et al., 2006 and Stieben et al., 2007). This task consists of three blocks: the middle block is designed to induce negative emotion through the loss of valued points. According to a self-report measure administered directly after the task, participants felt increased negative emotions during the emotion induction block (which likely carried over into the final block). Results revealed greater nogo-N2 activation during and/or after the emotion induction block compared to the first block, which we interpreted as increases in emotion regulation over and above the regulatory demands of nogo trials themselves. Thus, not only does the N2 tap self-regulation, but it is also appears to reflect the regulation of emotional states or impulses. The cortical generators most often associated with the N2 include prefrontal and cingulate regions implicated in inhibitory control. However, the N2, like other scalp activation patterns, is derived from the sum of all underlying cortical generators active at a particular moment in time. Activation in some of these regions may not be related to inhibitory control. A number of studies modeling the activation underlying the N2 (in visual tasks) have shown generators suggestive of occipital, ventral temporal, and parietal areas (e.g., Lamm & Lewis, 2010—see Fig. 6 for activation superimposed over an average MRI; Lamm et al., 2006, Lewis et al., 2006 and Stieben et al., 2007), regions not directly related to inhibitory control. Given that activation from all these regions appear to contribute to the N2 scalp topography—either by enhancing its polarization (e.g., projecting negative current towards or near the mediofrontal scalp area) or diminishing its polarization (e.g., projecting positive current towards or near the mediofrontal scalp area)—much information can be gained by conducting a source-space analysis based on prespecified “regions of interest.” This enables one to test the activation of cortical regions directly related to regulatory functions. By performing a source-space analysis, one can address specific questions of spatial localization as well as the chronometric (temporal) properties of that activation at the millisecond level. Given the use of powerful computers and increased access to user-friendly programs, source-space analyses are becoming common practice. Recently, a few studies conducted with children and adults have modeled the generators underlying the N2 to areas suggestive of activation in the ventral prefrontal regions—including the orbitofrontal cortex (OFC), the rostral anterior cingulate cortex (ACC), and the ventromedial prefrontal cortex (vmPFC)—and/or the dorsal ACC (Bekker et al., 2005, Bokura et al., 2001, Lavric et al., 2004, Lamm and Lewis, 2010, Lamm et al., 2006, Lewis et al., 2006 and Nieuwenhuis et al., 2003; Stieben et al., 2007). Ventral prefrontal activation has been linked with ongoing, non-deliberate or non-executive regulatory or evaluative processes, such as “in-the-moment” inhibitory control and the evaluation of positive or negative aspects of current or anticipated stimuli (Blasi et al., 2006, Drevets et al., 1992, Durston et al., 2006, Eshel et al., 2007, Kaladjian et al., 2007 and Ochsner et al., 2004; for reviews see Phillips et al., 2008 and Rolls, 2000). Increases in ventral prefrontal activation have also been associated with increased negative emotion (for example, Baumgartner et al., 2006 and Kawasaki et al., 2001) and more specifically emotion regulation (for example, Ochsner et al., 2004; for a review see Quirk & Beer, 2006). Dorsal ACC activation, on the other hand, has been linked to various deliberate or “executive” regulatory functions, such as attention regulation (Crottaz-Herbette & Menon, 2006), performance or conflict monitoring (Blasi et al., 2006 and Santesso et al., 2005; for a review see Botvinick, Cohen, & Carter, 2004), and deliberate response inhibition (Blasi et al., 2006 and Tamm et al., 2004). Thus, the two prefrontal regions most consistently associated with the mediofrontal N2 are the ventral PFC and the dorsal ACC, thought to be linked with stimulus-bound and executive regulatory processes, respectively. Variation in both ventral PFC and dorsal ACC activation has been observed in clinical populations, specifically those with anxiety problems and aggressive behavior problems. A number of studies have shown greater ventral prefrontal activation in anxious patients than normal controls (e.g., Busatto et al., 2000, Johanson et al., 1992, Liotti et al., 2000 and Raunch et al., 1997). This activation may be related to emotional arousal or ongoing (inefficient) efforts to regulate that arousal. Moreover, we recently reported decreases in ventral prefrontal activation, following treatment, correlated with improved treatment outcomes for anxious aggressive children (Lewis et al., 2008). We interpreted these changes in terms of reduced recruitment of a ventrally-mediated threat-focused (inefficient) style of emotion regulation. Thus, anxiety related elevated ventral prefrontal activation may reflect increased inefficient regulatory efforts. Furthermore, a handful of studies have shown decreased dorsal ACC activation associated with anxiety problems (e.g., Lorberbaum et al., 2004 and Pillay et al., 2006; for a review see Drevets & Raichle, 1998). For example, Goldin, Manber, Hakimi, Canli, and Gross (2009) found reduced dorsal ACC activation for patients with social anxiety disorder, compared to normal controls, when instructed to regulate their emotions. Together, these findings suggest that anxious aggressive individuals may recruit greater ventral PFC but reduced dorsal ACC activation when attempting to regulate negative emotions. In contrast, reduced ventral PFC activation has been linked to non-anxious aggressive behavior. Patients with ventral PFC lesions have difficulties regulating anger when emotionally challenged (Anderson, Barrash, Bechara, & Tranel, 2006; for reviews see Blair, 2001, Davidson et al., 2000 and Hoptman, 2003). This is consistent with the type of deficit exemplified by the well-known case of Phineas Gage, a 19th century railway worker who was considered hard-working and responsible until an explosion damaged his ventromedial prefrontal region. After his recovery, he was reported to have become impulsive, impatient, and verbally aggressive. In other words, he showed a reduction in self-regulation similar to the undercontrolled prototype identified by Eisenberg and colleagues (2007) and others. Furthermore, a recent functional imaging study found a link between ventral PFC activation and aggressive behavior problems (Coccaro, McCloskey, Fitzgerald, & Phan, 2007). Specifically, participants with intermittent explosive disorder (IED) showed decreased OFC activation and diminished OFC–amygdala coupling, compared to normal controls, while viewing emotionally salient faces. Because amygdala–OFC interaction is critical for effective emotion regulation and control of aggression (Davidson et al., 2000; Izquierdo et al., 2005; Urry et al., 2006), these results suggest that participants with IED recruited insufficient OFC activation to down-regulate amygdala activation and thus displayed impulsive aggressive behavior. As well, two recent fMRI studies found decreased dorsal ACC activation in conduct disordered (non-anxious aggressive) children and adolescents while viewing emotionally salient images (Stadler et al., 2007 and Sterzer et al., 2005). Because dorsal ACC activation has been linked with cognitive control, as reviewed earlier, these results suggest that dorsal ACC activity may also be necessary to regulate angry emotions. In sum, it may be that purely aggressive children (without co-occurring anxiety) have insufficient ventral PFC-mediated and dorsal ACC-mediated regulatory capabilities to effectively inhibit their actions when emotionally charged. In the current study, we asked whether aggressive children with and without co-occurring anxiety problems recruited different regulatory mechanisms when emotionally challenged. Specifically, we wanted to see if they revealed patterns of ventral PFC and/or dorsal ACC activation that were different from each other and from an age-matched comparison group. Previous EEG studies (e.g., Jonkman et al., 2007, Lewis et al., 2006 and Stieben et al., 2007) have modeled such differences descriptively, but we wanted to test cortical activation differences statistically. Yet the most novel aim of the current study was to look for differences in the timing of cortical activation patterns. Very few EEG studies have examined temporal differences in cortical activities associated with emotion, despite the fine temporal resolution available with this technology. Nevertheless, a few recent studies found that the latency and duration of ERP components distinguished attentional processes induced by positive vs. negative emotions, both in adults and children (Hajcak and Olvet, 2008 and Lewis et al., 2007). Thus, an analysis of timing differences, with reference to their neural generators, seemed an important next step. To assess source activation and chronometric differences underlying emotion regulation for subtypes of aggressive children, we estimated source activation for the entire cortex using a distributed inverse model and then extracted activation values for two regions of interest (ROIs). We identified a ventral prefrontal ROI, suggestive of OFC, vmPFC, or rostral ACC activation, and a dorsomedial ROI, suggestive of dorsal ACC activation. ROI activation values were extracted for each participant and then pooled for statistical analysis. To ascertain chronometric differences between groups, we computed activation levels across three 100-ms time windows covering the typical timeframe of the nogo N2 in children. Lastly, we administered a (parent-report) questionnaire measure of effortful control so that we could assess non-task related regulatory abilities. The following predictions were made: (1) that purely aggressive children would reveal low levels of ventral PFC ROI and dorsomedial PFC ROI activation and poor effortful control scores; (2) that anxious aggressive children would reveal high levels of ventral PFC ROI activation, low levels of dorsomedial PFC ROI activation, and poor effortful control scores; and (3) that normally-developing children would reveal low levels of ventral PFC ROI activation, high levels of dorsomedial PFC ROI activation, and high effortful control scores. Since our paradigm was designed to assess the neural underpinnings of emotion regulation, we also predicted greater activation for one or both regions during the emotion induction block and/or the block following the emotion induction. Finally, since the chronometric analysis was a novel approach, we did not have specific predictions for the three groups. However, we expected anxious children to show earlier increases in ventral activation, as demonstrated by previous research (Lewis et al., 2007).
نتیجه گیری انگلیسی
3. Results 3.1. Behavioral data 3.1.1. Response time A 3 (Group) × 3 (Block) repeated-measures ANOVA was conducted, with age, medication, and stimulus duration as covariates for Go response times. (Gender was not entered as a covariate for any of the behavioral, scalp ERP, or source analyses because no significant gender differences were found.) No significant or trend-level response time effects were found (F values ranged from .23 to 1.67), indicating that any group differences in the scalp or source data could not be due to differences in task difficulty. 3.1.2. Performance accuracy As outlined in the methods section, to better capture individual differences, the task used in this study was dynamically adjusted based on each child’s performance. Thus, we did not find group or block differences in performance accuracy. 3.1.3. Effortful control In order to examine the regulatory abilities for each of the groups, a univariate ANOVA was conducted on the effortful-control cluster of the EATQ-R with age and medication as covariates. Results revealed a main effect of Group, F(2, 36) = 15.43, p < .001. Contrasts revealed greater effortful-control scores for the normal children than both the low (p < .001) and high (p < .001) anxious aggressive children. Thus, the normal children were rated by parents as being better able to regulate their actions and emotions. 3.2. N2 scalp data Visualization of the correct nogo stimulus waveform revealed clear ERP components—N1, P2, and N2—and consistent group differences: anxious aggressive children showed greatest activation for all three blocks (see Fig. 2). A 3 (Group) × 3 (Block) repeated-measures ANOVA, with age, medication, stimulus duration, and trial count as covariates was conducted on N2 scalp activation (average of Fz and FCz activation; see Fig. 3). Results revealed a main effect of Group, F(2, 32) = 4.25, p = .02. Contrasts showed greater (more negative) activation for the high anxious aggressive children than both the low anxious aggressive children (p = .01) and the normal children (p = .04) suggesting that the high anxious aggressive children recruited more cortical resources when viewing the nogo stimuli. However, there was no block effect, F(1, 32) = .72, ns. Grand-averaged ERP waveforms at site FCz. Activation is shown as negative down. Fig. 2. Grand-averaged ERP waveforms at site FCz. Activation is shown as negative down. Figure options Group and block differences in scalp activation (average of sites Fz and FCz). ... Fig. 3. Group and block differences in scalp activation (average of sites Fz and FCz). Greatest magnitude of activation is down. Figure options 3.3. N2 source-space activation 3.3.1. Ventral prefrontal ROI For each ROI independently, we conducted a 3 (Group) × 3 (Block) × 3 (Time: 200–300 ms, 300–400 ms, and 400–500 ms) repeated-measures ANOVAs, with age, medication, stimulus duration, trial count, and base-rate activation as covariates. For the three-way analyses, no significant main effects or interactions were found for either region. However, since the middle time window could be considered a transition time, we repeated the analyses omitting the middle time window. Results revealed a Block-by-Group-by-Time interaction, F(4, 62) = 2.72, p = .04. For the early time window, contrasts revealed greater (ventral prefrontal) activation for the anxious aggressive children than the typically developing children in blocks A (p = .009) and C (trend level, p = .08). Yet, as is evident from Fig. 4, there was a quadratic increase in block B compared to both blocks A and C for all three groups (p values ranging from .08 to <.001). Thus, in the early window, the anxious aggressive children showed greater activation than the other groups in the blocks that were not explicitly designed to elicit negative emotion. Group differences in ventral prefrontal ROI activation by block for the early ... Fig. 4. Group differences in ventral prefrontal ROI activation by block for the early time window. Greatest magnitude of activation is up. Figure options For the late time window, we found greater (ventral prefrontal) activation for the non-anxious aggressive children, at the trend level, than the anxious aggressive (p = .07) and normally developing children (p = .08) but only for the emotion induction block (see Fig. 5). We also found a quadratic increase in activation for block B compared to blocks A (p = .001) and C (p < .001) for the non-anxious aggressive children only. This difference represents a pattern that is essentially opposite to that of the anxious group, and it highlights differences that are specific to the emotion-induction block. Finally, we found a decrease in activation between the early and late time windows for the high-anxious group for block B (p = .009). Group differences in ventral prefrontal ROI activation by block for the late ... Fig. 5. Group differences in ventral prefrontal ROI activation by block for the late time window. Greatest magnitude of activation is up. Figure options In sum, the anxious aggressive children showed higher ventral prefrontal activation in the early window, independent of the emotion induction, and their response to the emotion induction decreased over time (across time periods); the non-anxious aggressive children showed elevated ventral prefrontal activation in the late time window, but this was specific to the emotion induction; and the normal children revealed relatively low levels of ventral prefrontal activation for all blocks and time windows. 3.3.2. Dorsomedial ROI Analysis of the dorsomedial prefrontal ROI revealed no significant main effects or interactions, even when comparing the early and late time windows. 3.4. Test of emotion induction To assess the effectiveness and timing of the emotion induction, we analyzed the self-report emotion-scale data. Given that we were not interested in individual differences, just Block effects, we dropped all the covariates (age and medication) as well as Group from the analysis. Results revealed a quadratic main effect of Block, F(2, 79) = 47.26, p < .001. Planned contrasts revealed that block B, the emotion induction block, was perceived as significantly more negative than blocks A and C (p < .001). Thus, a negative emotion induction in block B was confirmed by children’s self-report.