Introduction

An important consideration in the cognitive analysis of emotions is their evolutionary origins. Regarding anxiety, there’s a system that ensures that attention is paid to threatening environmental an interoceptive stimuli that are relative to the person’s motivational state. This allows a quick and effective response to threatening stimuli. In the case of fear, a bias can then arise in selective attention to threatening information.

A feature of recent cognitive theories is the emphasis on the role of attentional processes in etiology and the maintenance of anxiety. Something can be said about all of these theories, which is why it’s subsequently explained from a cognitive-motivational point of view what the specific mechanisms are that are responsible for an attentional bias in anxiety.

Cognitive Theories

According to recent cognitive theories, biases in information processing play an important role in the etiology and maintenance of emotional disorders, such as generalized anxiety disorder (GAD) and major depressive disorder (MDD). Beck developed a schema model where relevant schemas have become dysfunctional and Bower proposed a semantic network theory of emotions where each emotion represents a node in an associative network in memory.

Both theories say that anxiety and depression are associated with an emotion-related bias in all aspects of information processing. However, this is not entirely true, because anxiety is mainly associated with a bias in selective attention and depression with a bias in selective memory.

Williams (1988) then developed a revised cognitive theory about anxiety and depression. Some characteristics of this theory are:

• Different emotional disorders are associated with different patterns of cognitive bias.
• People who have a permanent tendency to show preattentive, automatic vigilance for threat are more susceptible to the development of anxiety disorders when under stress.
• Trait anxiety influences the direction of preattentive and attentional biases to threat stimuli. People with high trait anxiety orient their attention towards the threat and people with low trait anxiety direct their attention away from it. This is the interaction hypothesis.
• Cognitive behavioural therapy is effective because these biases are corrected.

Williams then says that two mechanisms are responsible for the pre-attentive and attentional bias towards threats. This is the Affective Decision Mechanism (ADM) and the Resource Allocation Mechanism (RAM). The ADM assesses the degree of threat of stimuli. The output of this mechanism depends on the stimulus, but also on the degree of anxiety that the person currently has (state anxiety). The output serves as input for the RAM, which determines the allocation of resources of the processes.

The effect of RAM depends on the degree of trait anxiety of the person. As the output of ADM increases, the difference between people with high and low trait anxiety (interaction hypothesis) becomes clear.

Later, the Williams model was adapted using connectionist terms (1997). However, the interaction hypothesis remains an important key point in the model. Various other theories are an addition to this model. The main focus is on the assumption that preattentive processes play a role in anxiety. These processes are important in the evaluation of stimuli and in giving direction to the focus of selective attention and the response that is given as a result. These processes are said to be vulnerable to the development of clinical anxiety.

However, a number of comments can be made about these cognitive theories:

• The interaction hypothesis says that people with low trait anxiety turn away from a threatening stimulus. However, this is not smart when it comes to really threatening stimuli and would mean that the threat detection system is not functional.
• Several studies have found a bias in preattentive processes in anxiety disorders, but not in depression. This is strange, because depressed people also have a higher level of anxiety
• The theories make little use of recent neurobiological systems that play a role in anxiety.

Below are two biological formulations of anxiety:

1. Gray says that a vulnerability to anxiety is associated with individual differences in Behavioral Inhibition System (BIS) activity. This system compares current with expected stimuli and takes action if they do not match or if there are aversive stimuli. BIS ensures inhibition of current behavior, which increases the arousal and for greater attention to the stimuli. Anxious people have an overactive BIS.
2. LeDoux is known for his belief that the amygdala plays a major role in emotions and that it is activated by a threatening stimulus. Attention processes play a lesser role in his model.

Cognitive-motivational point of view

According to this position, there are two motivation-related systems that, in combination with each other, play a key role in mediating fear. These are the Valence Evaluation System and the Goal Engagement System. They can each be placed on an axis in a two-dimensional network.

A motivational analysis of these factors says that a bias in pre-attentive processes and initial orientation with regard to emotional stimuli depends on the combined functioning of valence evaluation and goal engagement. According to the cognitive-motivational point of view, the Valence Evaluation System ensures the value of a threatening stimulus. This depends on the stimulus itself, but also on the context, the degree of anxiety at that moment (state anxiety), previous experiences, and the degree of arousal.

In addition, the Valence Evaluation System is more sensitive to people with high trait anxiety. Trait anxiety is then seen as an increased output of the Valence Evaluation System, which reflects a bias in the valuation of threats. The output from the Valence Evaluation System serves as input for the Goal Engagement System, which is responsible for mediating preattentive and attention responses. Nothing is done about it with little threat. In the event of a high threat, this is immediately aimed at and other processes are stopped.

According to this view, there is no linear relationship between the valuation of threatening stimuli and attentional bias. When there is no threat, no attention is paid to the stimulus. With a little threat, the attention goes away from the stimulus. With a medium to high threat, there is an increased degree of attention given to the stimulus that ultimately gets stuck at a certain (high) level.

Some advantages of this system are:

The cognitive-motivational view says that an attentional bias or pre-attentive bias does not play a causal role in the etiology of an anxiety disorder. Such a bias can also be found in people with low trait anxiety when the stimuli have a high threat.

The presence of such a bias for mildly threatening stimuli may be a sign of a sensitivity to anxiety, but this may not be a determining factor. The primary factor that determines this sensitivity, according to the cognitive motivational point of view, is a bias in the Valence Evaluation System. The attention processes may possibly be important in the maintenance of anxiety.

Research into attentional bias in anxiety

Studies that investigate a bias in selective attention to emotional stimuli use the (emotional) Stroop test, probe detection task and color perception task, the latter being a variant of the probe detection task, where a word pair is represented of which a word is neutral and the other emotionally. After that, two colored fields appear instead of the words at the same time, but fearful people think that the colored field that replaces the emotional word appears earlier. The Stroop task bases the conclusions on interference, but these are complex and sometimes difficult to interpret because of confounds and "strategic override". The latter emerges when phobics are tested in the vicinity of the feared stimuli. At that moment the interference effect disappears. The probe detection task is less sensitive to this.

The problem with all these tasks is that it only exposes part of the attentional bias, because the tasks depend on the time the stimulus is presented. All of these tests provided evidence for attentional bias, but were performed with stimuli that the test subject could consciously focus on.

Studies that investigated whether there is also a bias for selective attention to threatening information that is offered unconscious (pre-attentive bias) have used dichotic listening tasks and visual masking paradigms. The latter are a better means to counteract consciousness and have been performed with the Stroop test and the dot probe task. To check whether the test subjects were really unaware of the stimuli, a consciousness check was performed in two ways: test subjects were asked whether they saw a word in the masked phase and whether the letters were displayed word or not. This showed that the thresholds of consciousness for different stimuli are different. Someone may be aware of a stimulus, but not of its content. With both test designs, it was ultimately demonstrated that there is also a pre-attentive bias in people with an anxiety disorder. This suggests that the bias is already at an early stage of processing information.

It has also been shown that the pre-attentive bias for threatening stimuli can be a cognitive vulnerability factor for emotional disorders. This means that people who have a permanent tendency to automatically selectively focus attention on threatening stimuli are more likely to develop an anxiety disorder when in stressful situations.

The influence of state and trait variables on preattentive and attentional bias

Evidence for the relative influence of trait and state variables on attentional bias contains three areas, namely correlations between bias and self-reported anxiety, the effect of short and long term stressors on preattentive and attentional bias, and the effect of reducing state anxiety through treatment. The last two are discussed.

The effect of stressors

The interaction hypothesis contains two components that are important for the mechanisms that underlie attentional bias. The first is that there is no difference between people with high and low trait anxiety when there is no stress. As the state anxiety increases, a difference becomes apparent and the people with high trait anxiety become sensitive to threatening stimuli. This sensitivity therefore appears to be a latent and not a manifest vulnerability factor for fear.

The second component deals with responses to the attention of people with low trait anxiety. These people tend to divert their attention from a threatening stimulus.

The first part of the hypothesis has been confirmed in various studies, although short-term stressors also elicited attentional bias in people with low trait anxiety, long-term stressors only elicited attentional bias in people with high trait anxiety. If evidence for the second component were also found, this means that the cognitive key factor underlying the vulnerability to anxiety is mediated by the mechanism that determines the direction of attention bias.

No significant evidence was found for the second part of the hypothesis, as a result of which the relative effects of state and trait anxiety on preattentive and attentional bias remain uncertain. It is true that no distinction is made between different options such as.

A threshold effect where the effects of state and trait variables are additive so that vulnerability to threatening words is primarily shown by people with high trait anxiety who are under stress.

A curvilinear relationship between the threatening value of the stimulus and the attention responses to the threat.

Different patterns of bias in people with high or low trait anxiety, which become apparent under increasing state anxiety. There is not enough evidence that with increasing threat people with a low trait anxiety turn away from the stimulus.

The effect of anti-anxiety treatment

It was argued that if the attentional bias for threat is a long-term cognitive vulnerability factor for anxiety, it is still present after treatment. However, studies show that this bias does disappear after therapy (in people with a generalized anxiety disorder). The question was whether a change had taken place in controlled strategies or in the automatic unconscious processes. The latter turned out to be the case.

Attention bias for illustrated stimuli

The studies described above have only been performed with the help of word stimuli and these have a limited scope in terms of threatening value. Illustrated stimuli such as emotional faces or scenes have a somewhat larger scope.

There are a number of tasks that can be used to examine an attentional bias for emotional faces, including the dot-probe detection task, popout tasks, and eye movement. Most of the studies revealed emotion-related effects, although there was a variability in the results with regard to the primary influence on attentional bias. Some claimed it was state anxiety, others trait anxiety or negative emotional valence in general. Differences have been found with regard to studies that have used word pairs. For example, with illustrated stimuli, there is no need for a certain stressor to provoke the bias for threatening faces. This is true with word stimuli. The hypothesis that an attentional bias for threatening stimuli is a characteristic for only people with a high trait anxiety therefore only applies if it concerns relatively weak threatening stimuli such as words.

The interaction hypothesis says that the more threatening the stimulus is, people with high trait anxiety focus their attention on the stimulus and people with low trait anxiety turn their attention away from it. The cognitive-motivational point of view says that an increasing threatening stimulus also causes a bias in people with low trait anxiety. If this did not happen, the Valence Evaluation System would not work.

Research with neutral and emotional scenes has indeed found that there is a general greater vulnerability to highly threatening scenes compared to mildly threatening ones. In addition, the low-fear group showed avoidance for mildly threatening scenes and this avoidance decreased as the stimulus became more threatening. People with high trait anxiety also exhibited greater vulnerability to threatening scenes than people with low trait anxiety. These results therefore speak against the interaction hypothesis and for the cognitive motivational view.

Time course of attentional bias

There is therefore sufficient evidence for a bias in pre-attentive attention and attention processes, but the research described so far focused primarily on initial orientation towards the threatening stimulus. The question now is what happens next. The hypothesis is that people with high trait anxiety or clinical anxiety first focus their attention on the stimulus and then try to avoid detailed processing taking place so as not to become overly anxious. This can ensure that the fear is maintained.

Preattentive and attentional bias in depression

The question is whether depression is associated with an attentional bias for negative stimulation similar to the bias in anxiety. Studies have shown that depressed people do not have a pre-attentive bias for negative information, so do not automatically focus on this. When the stimulus is offered longer and the information has come to the attention, they have a greater difficulty in separating themselves from it. Depressed people do show a bias with regard to later processes such as sustained attention. This is consistent with the predictions of the cognitive motivational point of view, which says that the manifestation of a bias for aversive stimuli in preattentive and orientation processes depends on two things: goal engagement and valence.