«ATTENTIONAL BIAS IN PATIENTS WITH IMPLANTABLE CARDIOVERTER DEFIBRILLATORS: EXAMINING MECHANISMS OF HYPERVIGILENCE AND ANXIETY By NEHA K. DIXIT A ...»
Functional neuroimaging has provided additional insight into the relationship between selective attention and emotion. Neural structures involved in the early processing of emotional stimuli include the amygdala, anterior cingulate, and frontal cortex (Dolan, 2000). There is a large degree of overlap between these structures and those involved in processing selective attention. Two main neural mechanisms exist by which emotion may guide information processing. Regions of the brain that rely on sensory information such as the visual cortex and the extrastriate cortex regulate bottom-up influences on attention. During presentation of an emotional stimulus these regions show increased activity or amplification resulting in favored attentional selection to that stimulus (Mangun, Jha, Hopfinger, & Handy, 2000). Amplification is thought to occur via bottom up input from the amygdala that reacts to emotional representation put forth. For example, Lane and colleagues (1997, 1998) found that exposing subjects to emotionally- arousing pictures increased activation in the visual cortex compared to neutral pictures. Others such as Pessosa and Ungerleider (2004) have found increased activation of areas such as the fusiform gyrus (also involved in visual processing) when showing subjects fearful verses neutral faces.
While the amygdala sends amplifying signals to sensory cortices, other regions of the brain are implicated in top-down processing of emotional information to help modulate its selection. The two major brain regions involved in selection and suppression of information from the amygdala are the dorsolateral and ventromedial corticies (Mangun et al, 2000). The dorsolateral region is involved in selecting and maintaining stimulus attributes in working memory (Cohen et al, 1999, 2000), while the ventromedical region is involved in registering the emotional significance of stimuli and is also involved in motivational and goal directed processing (Bush, Luu, Posner, 2000). The anterior cingulate which is a part of the ventromedial frontal cortex, is a structure involved in conflict detection and may play a role in what emotional information to let into decision making processing and which to leave out (Hariri et al, 2004;
Whalen & Bush, 1998). Bishop and colleagues (2004) showed using an fMRI study that, individuals who were highly anxious had higher anterior cingulate cortex (ACC) activation compared to less anxious comparison subject when viewing threatening stimuli. According to the somatic marker hypothesis of Damasio (1994) feedback from autonomatic (emotional) responses provides critical input via the amygdala to decision-making processes mediated by the frontal lobes. Given the neural mechanisms involved in emotional processing, theories of normal emotion have their parallel in theories of disordered emotions such as anxiety and depression where thinking, cognitive processing, and decision making has been shown to be distorted (Beck, 1976).
Experimental Paradigms Examining Attentional Bias Two common experimental methodologies used to examine selective attention in adults are the emotional stroop task and the emotional dot probe paradigm. In both tasks, the basic premise is to orient a subject’s attention to particular stimuli while utilizing interfering emotional stimuli to distract the subject. Together, these two tasks have been manipulated such that researchers have been able to determine the nature of both cognitive and neuroanatomical aspects of anxiety disordered individuals and their ability/inability to attend meaningfully to specific stimuli.
Since its inception the Stroop has commonly been used as the gold standard for selective attention tasks. In this particular task, a subject is asked to read a set of emotional and neutral words and then asked to name the color of the word disregarding the word’s content. It has been demonstrated by numerous researchers that subjects’ response latencies to emotional words are longer for subjects for whom the words have relevance (e.g. socially relevant words for social phobias) compared to control subjects. The results indicate that the automaticity of word naming is overridden by the emotional content of the word. In fact word naming appears to take longer when examining emotional words for people with affective disorders. It has been posited that the Stroop is a task of conflict detection and monitoring. Although attentional networks have been implicated in this detection process, the results of the emotional version of this task are often misguided and interpreted inaccurately (Algom, Chajut & Lev, 2004). In the Stroop it is often the level of emotional content within the word that drives the attentional bias and is not considered a hallmark stroop color naming effect. The emotional content of the word itself is the interfering stimulus and can vary on its level of biasing attention. The dot probe task, a derivative of Posner’s original exogenous cuing paradigm, is more commonly used to examine attentional bias in healthy and mood disordered individuals (Posner, 2000).
The dot probe paradigm orients a person’s attention to a particular spatial location by the presentation of a cue prior to a target probe. In this task individuals are asked to attend to letters or dots presented in different spatial locations on a computer screen (cue). They are then shown a target probe in the same or different location of the previously presented cue and asked to respond to the probe. The basic premise of the dot probe paradigm is that a person’s visual selective attention can be oriented differentially to spatial locations. A subject’s reaction time measured by response latencies between cue and target (i.e., probe) detection is the main measure of their attentional capture. Further analysis of the dot probe task involves examination of response latencies subjects have to valid trials and invalid trials, sometimes called the validity effect. Longer response latencies are observed for trials where probes occur in a different location from the cue (i.e., invalid or incongruent) suggesting that individuals are primed by the cue to orient their attention in one direction and have difficulty disengaging from that location in response to the probe. The dot probe paradigm has utilized the principles of Posner’s “shift” and “disengage” components of attention (1980) to describe instances of disturbed selective attention during the task. The task has been manipulated in numerous ways to examine selective attention in anxious individuals mainly with the addition of emotional cue related stimuli; usually an emotional word, face, or picture and by priming locations in a valid/invalid manner to create an attentional response bias.
Evidence of Attentional Bias in Anxiety Disorders Individuals with anxiety disorders are of particular interest when examining attentional biases because of the nature of the disease state. Mood congruent attentional biases are well established in the anxiety literature (Williams, Watts, MacLeod, &Mathews, 1997; Armony &LeDoux, 2000). It has been posited that human anxiety reflects a heightened response of the fear system (Lang & Ohman, 2000; Armony & LeDoux, 2000; Fox, Russo, Bowles, & Dutton 2001). Thus, it is adaptive for people who perceive a threat to get anxious and thereby engage neural systems to aid in the resolution of the threat. If resolution cannot be reached, higher order brain systems (frontal cortex, etc.) must come online and create alternative response options.
As such, both individuals with anxiety disorders and those with subclinical levels of anxiety may differentially strategize execution of action during the presence of threat (Derryberry & Reed, 2004).
Major findings in the dot probe literature demonstrate that anxious individuals show a bias towards threatening faces, words, and negative pictures. A study by MacLeod, Mathews and Tata (1986) demonstrated, using an emotional dot probe task, that anxious patients were faster to respond to the probe (dot) when it appeared in the location where a threat-related word has just appeared (valid cue) compared a non-threat related word. This effect was disproportionately seen in anxious individuals compared to non-anxious control participants and was specific to threat-relevant information. This result has been replicated throughout the literature (see Mogg & Bradley, 2000 for review) and suggests that threatening information captures visual attention particularly in those individuals who are especially sensitive to fearrelevant stimuli in the environment (e.g. anxious individuals). Hypervigligence to external stimuli and processes of relevance such may divert attentional resources away from non-threat related information and bias attention towards threat related information. Similar findings have also been shown in non-clinical populations (Fox et al, 2001). Individuals with subclinical levels of anxiety (high trait anxiety) have also demonstrated an attentional bias towards threatening information (Fox et. al, 2001, Wilson & McLeod, 2003) when compared to low trait- anxious individuals.
An important issue raised in the dot probe literature is one of individual differences in anxiety-disordered patients. While some patients may exhibit heightened attentional capture to threatening information, it has been demonstrated that some have difficulty disengaging from threatening stimuli. A series of studies have shown individuals with high levels of state anxiety to have difficulty disengaging from negative or threatening information (Amir, Elias, Klumpp, & Przeworski, 2003; Fox et al, 2001; Yiend & Mathews 2001) that is particularly relevant to them.
Personal or individual threat therefore is an important consideration when interpreting dot probe findings.
Other models discussing biased attentional direction (e.g. Williams, MacLeod) posit that high trait-anxious individuals orient their attention towards threatening information while low trait anxious individuals will orient away from the threatening information. The shifted attentional model account proposed by Mogg and Bradley (2000), posits that regardless of level of anxiety, all individuals will direct attention away from mild threat intensity stimuli and orient towards stimuli with a high threat intensity. The observable difference between the two groups is the intermediate levels of threat intensity. McLeod and Wilson (2003) designed a unique study in which they varied the intensity level of a variety of angry faces on a continuum of threat. Findings showed that all subjects showed greater vigilance (longer response times) to the most extreme and intense faces. They did not show any effects at very low levels of intensity.
The critical difference in this study was to intermediate levels of angry faces. High trait anxious individuals displayed a greater vigilance to threat compared to low anxious individuals. These findings appear to suggest another mechanistic view of anxious individuals. They appear to predict that high trait anxious participants reach a threshold of subjective threat at lower levels of perceived threat than low anxious individuals.
The ability to disengage from personally relevant threat has been demonstrated in medical populations. Researchers have demonstrated that patients with chronic pain show attentional bias towards pain related information (e.g. words) when compared to medical counterparts who did not have chronic pain (Dehgani, Sharpe, & Nicholas, 2003; Beck et al, 2001). In fact, pain patients show differential bias towards words that are related to their particular type of pain. For example, Van Damme, Lorenz, Eccleston, Koster, DeClercq, & Crombez, 2004 showed that patients with increased negative cognitions about their pain were more likely to have increased response latencies to affective pain words compared to sensory pain words. Patients rating their subjective pain experience as more intense (e.g., burning, stinging) showed increased response latencies to sensory words compared to affective words.
Similar findings, demonstrating disproportionately increased response latencies to clinicallyrelevant words compared to other words and compared to controls exist in literature examining social phobia, specific phobia, and generalized anxiety disorder (Compton, 2003).
In sum, the dot probe literature highlights what has been interpreted as the highly anxious individual’s inability to disengage from relevant threat information as evidenced by longer response latencies during negative verses pleasant conditions. This generally results in disproportionate slowing during clinically relevant or threatening conditions during invalid verses valid trials.
The present study furthered an understanding of the relationship between anxiety and cardiac arrhythmias, specifically the role of information processing in ICD recipients. The number and proportion of individuals being implanted with ICDs is growing in this country.
With new advances in technology, ICDs will increase in their favorability as treatment of choice in both arrhythmias and congestive heart failure. Research suggests the presence of anxiety, specifically shock-related anxiety results in an increase in hypervigilence to bodily symptoms and health-related stress (Pauli et. al, 1999; Sears et al, 2001; Godemann, 2004). In addition, shock-related anxiety is associated with depression and decreased quality of life (Sears et al., 1999, 2001, 2003). ICD recipients are, by nature of the mechanism of their device and disease state biased towards potentially threatening sensations from their bodies. Much like studies of attentional bias in pain patients, ICD recipients offer a unique perspective to examine the relationship between emotion and attentional bias. In addition, the present study has clinical significance in that findings may identify mechanisms by which ICD recipients may process information, particularly cardiac related information, and to what degree they may over engage this information is critical to treating them. This line of research may aid in the development of individually tailored psychosocial interventions and the types of patient information that is offered in a clinical setting.