Summary of Cognitive behavioural processes across disorders - Harvey, Watkins, Mansell, & Shafran (2004)

Chapter 1: Introduction

Today’s research on behavioural processes in psychological disorders is mainly about one particular disorder. Researchers are trying to find out the underlying processes, maintaining factors and in the end treatment strategies for their disorder of interest. By doing this, parallel work is not taken into consideration and thus important findings are undermined or left out. A. Harvey, E. Watkins, W. Mansell and R. Shafran have come up with a different approach. Their approach is oriented across disorders so that similar structures and strategies can be found, understood and treatment plans can be adjusted accordingly. The emphasis on similarities applies to psychological (cognitive and behavioural) as well as biological theories and processes. However, here only the psychological processes are analysed because the related research is based on CBT and experimental psychology which brings along scientific value. Furthermore, the research focuses on maintaining factors rather than on predisposing factors, as these are the ones with the most exciting results.

Classification

Most dominating in the area of psychological disorders are the Diagnostic and Statistical Manual (DSM) and the International Classification of Disease (ICD). These classifications provide a common language among therapists and researchers. Similar symptoms or sets of symptoms can give an insight into contributing factors or treatment plans. Furthermore, it is perceived as a relief for some patients. There are also disadvantages to diagnosing. Firstly, a disorder can be described too simplistically and important personal information can be left out. Secondly, it cannot be assumed that clinicians know all the disorders by heart. Thirdly, patients can feel stigmatised and reduced down to a disorder. Living with a label and finding a job with one can be hard nowadays.

A further problem is that those statistics describe all disorders separately and closed off from other disorders. In reality, this is mostly not the case, as many of them only differ in frequency and severity. The approach used by the DSM and the ICD is a syndromal approach. Signs and symptoms are used to define disorders which bring knowledge about the course and treatment of said disorders.

There are more approaches though. The functional approach depends on:

• Documentation of the patient’s problematic behaviour;

• Identification of causal relationships;

• Collection of additional information based on the first two facets;

• Intervention considerations;

• Observing intervention progress for adjustments.

Problems with this approach are that it cannot easily be replicated and thus it can never be empirically verified. Furthermore, it relies mainly on clinical judgment. Hayes (1996) redefined this approach and proposed that common processes of aetiology and maintenance should be researched. His research found that self-focused attention is a core principle in experiential avoidance. Based on this, further research revealed that same principle can be applied to many disorders. The first claim being that there are transdiagnostic processes.

Other alternatives suggest segmentation of the sets of symptoms in order to examine all parts of dysfunctions.

Benefits of a transdiagnostic perspective

Comorbidity

Many patients suffer from multiple disorders at the same time. Bill, a 35-year-old man, who had been diagnosed with paranoid psychosis, was admitted into a hospital, because he could not function properly in his normal life anymore. He was worried about being ridiculed as well as being attacked. He became easily upset and had outbursts every now and then. In the hospital he was re-diagnosed with social phobia, PTSD and major depression.

Findings show very high rates of comorbidity between disorders. By looking at all disorders together, it might be possible to come up with an explanation for this. The National Comorbidity Survey states that ‘pure’ cases are the exception. The majority, suffering from a lifetime disorder, were also diagnosed with at least one other disorder (80%). Altogether, 48% of a total of 65,244 people suffered from at least one lifetime disorder. Out of those, 13% reported having two disorders and 14% reported having three lifetime disorders. While the severity with those people diagnosed with one or two disorders were respectively 2.6% and 7.9%; the severity reported for people with three disorders was 89.5%.

Comorbidity with eating and anxiety disorders as well as for GAD and depression are highest with 80%. Followed by bipolar in combination with another psychological disorder (61%), insomnia with another (52%), drug-dependence with another (45%) and schizophrenia with another (32%).

It may be the case that poor discriminant validity was the reason for the great comorbidities, but this cannot be claimed for anxiety disorders, as they have been proven to be accurately distinctive. A second reason could be that one disorder may put you in risk for another one.

When it comes to comorbidities and treatment, the clinicians decide which of the disorders the primary disorder is by assessing:

• Time of occurrence

• Causality (one caused the onset of the other)

• Level of distress (the highest being the principal diagnoses)

Treating the primary disorder is assumed to make the second one vanish at the same time. In case of transdiagnostic analysis, a fourth manner of assessment can be done by testing the common processes of dysfunction.

Treatment improvement

Bearing in mind that similar (if not the same) processes are the reason for a disorder to occur, the therapies for clients should also be similarly structured. Observations or research on one treatment could be transferred to another treatment plan of a different disorder. That way, quicker plans can be made and improvements achieved. Still, it will remain unlikely that one therapy fits all disorders.

Response to treatment

Current findings show that by treating one disorder it is possible to stop the other symptoms too. This could be explained by a transdiagnostic approach. This again could be due to poor discriminant validity, but also to the same underlying process.

Drawbacks of transdiagnostic perspective

It cannot be denied that there are great differences between disorders. These can thus not be explained if we used a transdiagnostic approach. However, Klinger (1996) came up with the concept of current concerns, an unconscious processing state. It states that the different disorders only vary in the type of concerns, but what makes them similar is that their behaviour can be related to the process of being occupied by a concern or goal. Emotional responses as well as goal pursuing actions are led by it. A dichotic listening task was used to examine the current concerns theory. Two stories were told, each on a different ear. In the end it became clear that the current concerns of the participants were influential of the latter recall of the stories. Attention, memory and thought are all affected. Dream studies were also used to test the theory. Current concerns were assessed before participants went to bed. In the morning they reported having had more dreams related to the concerns of the night before.

Some examples of concerns related to disorders:

• Panic: body changes are perceived as a physical catastrophe (dying).

• Specific phobia: perceived danger of one particular object.

• Social phobia: being mortified or embarrassed by others.

• OCD: intrusive thoughts indicating responsibility for the pain of others.

• GAD: worrying about a non-specific range of imaginable hazards.

• Somatoform: bad consequences of an illness.

• PTSD: threat of danger or being hurt as a reaction to a past trauma.

• Eating disorder: excessive fear about gaining weight and concerns about body shape.

• Sleep disorder: Inadequate amount of sleep or restless sleep.

• Psychotic disorder: being persecuted by others.

• Unipolar depression: loss of self-worth and acceptance.

How to evaluate transdiagnostic perspectives?

Processes which are an aspect of cognition or behaviour and support the maintenance of a psychological disorder are considered. Disorder focused research can be easily assessed via pre and post treatment measurements. This results in theories and models of the disorders. One of them is the cognitive model of social phobia by Clark and Wells (1995).

This example illustrates the analysed processes by the authors. Processes start:

1. Prior to ingoing a situation.

2. During, and

3. While leaving a situation.

1. When people suffering from social phobia anticipate a social situation, they selectively retrieve bad information about the perception of them by others. So they think they are seen as stupid or ridiculous. They then ruminate about these thoughts, which leads to maintenance of the anxiety and thus to avoiding situations.

2. In a social situation, patients focus only on themselves. Paying close attention to internal cues, such as sweating, blushing, etc., gives them reason to believe that others see these things and think they are unable to perform well. Even though it is a type of self-focus, social phobic people see themselves through the lens of others. These recurrent images lead also to the proliferation of the disorder. Little attention is paid to the environmental cues and thus people cannot see opposing evidence. Avoidance (either overt i.e. leaving a situation, or subtle i.e. avoiding eye contact) follows. This is called safety behaviour, as it prevents patients from undergoing the feared situation. On the other hand, it is actually not a safety behaviour, as it might reinforce the anxiety.

3. Again, negative retrieval is a factor that hinders people from feeling good about a social contact. Situations are interpreted negatively and rumination about their failure goes on. This is called post-mortem.

To sum-up:

• Attentional processes: self-focus, internal cue awareness.

• Memory processes: selective retrieval, recurrent images.

• Reasoning processes: interpretative bias.

• Thought processes: rumination.

• Behavioural processes: avoidance, safety behaviours.

The psychological disorders

The psychological disorders that are presented here are:

• Anxiety.

• Somatoform disorders.

• Dissociative disorders.

• Sexual and gender identity disorders.

• Eating disorders.

• Sleep disorders.

• Impulse control disorders.

• Mood disorders.

• Schizophrenia and other psychotic disorders.

• Substance related disorders from the DSM.

To state the quality of the evidence, different terms are used:

Good quality evidence is achieved when two or more conducted studies are based on:

• Psychometrically valid instruments.

• At least two different testing methods (interviews, self-report, experiments).

• Good methods.

• Reported by at least two researchers.

Moderate quality evidence criteria are met when one or more conducted studies are based on:

• No use of psychometrically valid instrument.

• Only one of the testing methods was applied.

• Only one research group conducted the study.

• At least one study is methodologically valid and strong.

Tentative quality evidence is given when the conducted studies are based on:

• Non patient samples.

• Methodologically problems.

• Single case studies.

To describe the evidence of the studies the terms ‘positive’ (+), ‘negative’ (-) and ‘mixed’ (+/-) are used.

Evidence referring to the transdiagnostic approach is either labelled as:

• Definite; this is the case when most research shows that the process is found in all disorders.

• Possible; if at least two disorders show the process in question. More research is then needed.

Chapter 2: Attention

A man became anxious when thinking about sex, because he thought others could read his mind. In situations with strangers he would notice sounds or movements, such as coughing or turning heads and take them as signals that those people knew what he was thinking. He tried to avoid looking at people and as a result lived more or less in isolation, focussing only on his thoughts.

This example of a person diagnosed with schizophrenia demonstrates selective attention. He was detecting all those ‘signs’ that others could read his mind, but what he did no longer hear or see is the context in which it occurred or conversations around him. Instead he was listening to his own body signals and stimuli. This is self-focused attention. These processes are systematic rather than random, even though they can become automatic, so that it is hard for people to change their focus again. Sometimes threats are the main focus of attention; some patients try to avoid any kind of threat. In order to catch all the signals, his/her attention switches all the time. That is why patients are highly distractible. Selective attention is the process by which stimuli from the external, as well as the internal world, are detected and thus further processing is possible.

On the other hand, attentional bias is the tendency to systematically attend to, or avoid, some stimuli at the expense of others.

Selective attention can be controlled or automatic. Some theorist’s say these are distinct processes, others say they lie on a continuum. What is clear is that they are both happening.

Self-focused attention

Ingrim (1990) defined this process as “an awareness of self-referent, internally generated information that is in contrast with an awareness of external generated information which derives from sensory input”. Attention is paid to physical states, memories, thoughts, feelings and emotions.

Measuring self-focused attention

External validity may not be consistent when it comes to self-focussed attention measurements as they are either based on lab experiments or self-reports. Internal validity on the other hand is strong.

Self-report measures are questionnaires about attention. They can either assess trait measures (so how much attention is paid to a stimulus on a day-to-day basis) or state measures (when a specific time period is the subject of report). To name a few measures:

• Private Self-consciousness Scale: a ten-item trait measure of Inner State Awareness and Self-Reflectiveness.

• Focus of Attention questionnaire: ten-item scale for social phobia, state measure, and one scale measuring self - the other focused attention.

• The Pain Vigilance and Awareness Questionnaire: 16-item scale, measuring pain attention in the last two weeks.

• The Somatosensory Amplification Scale: ten-item scale, assessing attention and fear to bodily changes.

• The Autonomic Perception Questionnaire: 21-item scale, assessing dispositions to perceived bodily sensations.

Benefits of this measurement are that they are easy to conduct and to apply in clinical settings. Furthermore, they may be the only way to assess these sensations. However, disadvantages are that they do not only measure attention, but also other processes. They are easily biased, because self reports are based on retrospective memories. Another drawback is that they cannot assess automatic processes.

The emotional Stroop Task is the method most commonly used in assessing attention. Participants have to name either the colour of the ink in which a word is presented or read the word, which is always the name of a colour. Sometimes the word and the ink match the same colour, sometimes they are contradicting. So the word it written in yellow, but it spells ‘red’. Contradictions take longer to process, as selective attention to reading interferes with naming the colour of the ink. To measure emotional interferences, the word colours were replaced by words describing emotions and neutral words for verification. Experiments showed that patients with GAD needed more time to be able to name emotional words.

Explanations for that are:

• Threats may trigger emotional responses, which lead to inhibition of another reaction.

• It could also indicate limitations of cognitive capacities, because people are trying to avoid the threat.

• Another explanation is that the task may reflect the pre-occupation with that type of thought, which then distracts from the ink colour.

To measure automatic processes, mask tasks were established. The threat word was only briefly presented, so no conscious processing could have happened, and then scrambled letters of that word were presented. It still took GAD patients longer to colour name the word.

• Detection tasks are another way to measure selective attention. Detection of visually presented stimuli related to the current concerns, should be easier and thus quicker to be detected than neutral stimuli.

• Dichotic listening tasks are also used to measure selective attention. Two different stories are told in each ear. The story similar to your concern can be better recalled afterwards.

• Dot-probe tasks are also used in measuring processes. Participants focus on a screen on which a dot will be reflected. They have to press a button as soon as they detect the dot. After this, they see two words; one neutral, one emotional. Another trial of dot spotting follows. The response time taken to detect the dot in the spatial location of either the neutral or the emotional word is calculated as an indication of selective attention to one of them. The dot-probe task is the most direct study in relation to selective attention.

• Eye tracker as a measure of selective attention, is a form of visual scan path studies. It measures the time a patient focuses on a picture. A disadvantage is that it can only measure those things that are in direct vision, but we can also detect things from the corner of our eye. The eye tracker cannot measure covert attention processes.

Are there stimuli that stand out when it comes to detection?

Stimuli that are novel attract attention independent of modality. Furthermore, those stimuli that have striking features, such as bright colours, sudden movements or those that are painful are especially attractive of our attention. In the perspective of evolution this makes sense as those objects could mean either danger or reward. For the general population it follows that mild threats are usually attempted to be avoided while stronger threats are paid more attention to and angry faces are easier to detect than happy ones. People react to words related to their current concerns. Detection leads to the urge to react either in an avoiding or approaching manner. For patients with anxiety, it is suggested that they detect stimuli faster than others, but then choose to avoid it, which in turn prevents normal processing of the stimuli and thus further anxious reactions.

Selective attention

Three processes belong to selective attention:

1. Attention to concern-relevant external stimuli.

2. Attention to concern-relevant internal stimuli.

3. Attention to avoidance and attention towards sources of safety.

Anxiety disorders

Panic disorder with and without agoraphobia

Patients suffering from these disorders are slower in naming body threat words, such as collapse or coronary. Stroop tasks showed evidence for general anxiety and depression. The primary concern of a panic disorder patient is the fear of internal physical sensations, such as a slower or faster heartbeat. Self-report measures can support this assumption. Furthermore, researchers found a selective attention bias to bodily changes. It is a core feature for panic disorder but also found throughout the range of anxiety disorders.

Specific phobia

It is not yet clear whether selective attention in specific phobia is due to automatic responses to concern related objects or not. Findings are mixed. This might only be the case for certain stimuli. Visual search tasks displayed that for concern relevant (feared) stimuli (e.g. spiders, if the phobia is about spiders) the detection time was faster than for non-fearful stimuli. Still people try to avoid threat stimuli and so viewing times for patients were reduced. In real life scenarios, it is therefore likely that patients detect fearful stimuli but at the same time they will pay attention to safety locations as they are motivated to avoid threat. That is exactly what was found by the study of Salkoviskis (1998). People suffering from arachnophobia were fastest to detect a presented light in a room when it appeared at the place where threat and safety were placed together (spider next to the open door).

Social phobia

People suffering from social phobia showed reduced response times when presented with social threat words (embarrassment, blushing), but not to physical threats. There were some studies that failed to replicate this finding. It is assumed that comorbid depression may be the reason for this. Studies that controlled for comorbidity showed reduced response times in patients with social phobia without depression. A visual search task revealed that angry facial expressions were faster to be detected and emotional faces led to greater distraction.

Contrasting findings state that people with social phobia try to avoid faces and other social cues. A visual scan path study confirmed this as it was found that patients would avoid faces and especially looking someone in the eye.

It is therefore assumed that only if participants of the different studies had extended display times that would give them time to process, they were able to look away from the threat. Though patients would always detect social threats, they may be able to cope with the fear by avoiding looking at it. On the other hand, this brings about the problem that they will no longer perceive the environmental cues, but focus on their internal ones instead. People with social phobia reported the highest rates of public and private self-consciousness. Own thoughts and behaviours are closely observed and analysed. In highly threatening situations, patients also managed to avoid looking at the threats. Furthermore, as with all anxiety disorders, social phobic patients have a higher awareness of bodily changes. On the contrary, the accuracy of an increased heartbeat or other sensations does not seem to be given.

Obsessive compulsive disorder

The findings on the attention bias towards concern related objects are mixed. One well-controlled research found positive evidence for selective attention to contamination words (dirty) in patients with contamination anxiety. It also found positive evidence for the attentional bias for social threat words in those OCD patients high in anxiety.

Post-traumatic stress disorder

It is one criterion of the disorder that patients have a high vigilance to signals of threat. However the findings of the Stroop emotional task are mixed. It is therefore not clear if it is an automatic process or not. It is assumed that there is an attentional bias for threat words but here again it cannot be said with definite evidence. Real life studies are very important in this field but hard to conduct.

Generalized anxiety disorder

People with this diagnosis worry about all kinds of threat and thus it is assumed that they will have an attentional bias. This is in fact so; all of them show selective attention to threatening information. Those who also worry about physical threat also show selective bias in this area. Additionally, it is said that GAD is associated to automatic as well as controlled processing of stimuli. At first they attend to threat stimuli, but in the second half of most studies, if enough time is given, they search for safety signs and attend to those. Bodily sensations are carefully examined, but the accuracy is rated as higher in comparison to phobic patients.

Somatoform disorder

Pain disorder

Five study findings of heightened attention to pain-related words showed significant effects, but one dot probe task failed to support the assumption. Fear and anxiety of pain however are associated to selective attention to those words. Intervention that distracts a patient from paying attention to their pains showed significant reduction in pain reported.

Hypochondriasis

Patients with hypochondria exhibit an attentional bias towards somatic symptoms. They further reported more discomfort and disability relating to their symptoms. Instructions such as: “observe your symptoms closely” led to an increase of symptoms reported. Patients with depression and anxiety also have the attentional bias for somatic symptoms.

Body dimorphic disorder

Observational studies show longer mirror viewing and higher frequencies of body checking. Not their actual reflection, but rather their own interpretation of themselves was watched and parts they disliked were in focus. Emotional words interfered with naming ink colours and thus the attention bias is supported. The bias was higher for positive words.

Sexual disorders

Those patients suffering from sexual disorders report inaccurate estimations of their functioning. When asked to pay attention to their genitals while getting an erection, they got less aroused in contrast to controls. Also presentations of a highly attractive girl led to a decrease of erection in men. This might be due to avoidance of internal as well as external stimuli related to sex as those suppose a threat. No research has been conducted, so nothing can be said for sure.

Eating disorders

There are several eating disorders, but the authors have decided to group them together as it is better to detect common results. All eating disorder patients showed heightened interfering processes in the Stroop task. Words related to food and body shape made it harder to concentrate on the ink. More attention was paid to threatening self-relevant information not only related to body weight and shape. Bulimic patients in particular showed distractibility when confronted with words invoking separation or ridicule. In relation to body shape they attended more to words referring to big shapes and avoided those referring to a thin appearance. They looked longer at those body parts with which they were dissatisfied. For all participants (normal and eating disorders) there was an attentional bias towards high-calorie foods when they were fasting. Thus binge eating might be the consequence of a longer period of non-eating. However, no clear evidence could be found to support attention towards either internal or external stimuli.

Sleep disorders

When it comes to sleeping disorders, evidence is vague. One study supports attention towards internal as well as external sources of threat, while the other does not. Internal threats would be checking ones muscles for tension or the heartbeat. External stimuli on the other hand are not checked.

Mood disorders

Unipolar depression

Stroop task on depression shows strong evidence for attentional bias towards depression related words (negative valence). However other studies show mixed findings. This might be correlated to the comorbid anxiety disorder. It might be that depression leads to a more controlled attentional bias than anxiety. A lack of protection bias was confirmed, meaning that depressed people would not avoid bad information or negative words. Furthermore, a visual scan path showed avoidance patterns of facial features. Little research has been conducted on internal attention, such as thought, pain or lethargy.

Bipolar disorder

These patients also show the attentional bias for negatively-valenced words. Disengaging from it is harder. In the phases of mania, neuropsychological tests found that it was easier (faster response times) for patients to respond to positive words compared to healthy controls. Research is only getting started in this area.

Psychotic disorders

Increased inactivity on the Stroop task was found, when patients were faced with threatening words, paranoia-related, low and high-self-esteem-related words or words related to their delusions. Even though measures are limited, attention to internal stimuli seems to be heightened, as they report high self-consciousness. Avoidance is found via the scan path in relation to faces and here especially to eyes. This would go hand in hand with their social withdrawal.

Substance-related disorder

The attentional bias towards drug related stimuli is quite obvious in this disorder. The stronger the craving for a substance, the greater the interference with colour naming or thoughts.

Sum-up

It can be said that there is satisfactory support to assume three definite transdiagnostic processes among psychological disorders:

1. Selective attention towards concern-related external stimuli.

2. Selective attention to concern-related internal stimuli.

3. Attentional avoidance or attention towards safety sources

There is additional evidence that a process across all disorders is involved, which leads to high distractibility by neutral stimuli. Sudden movements or noise might seem neutral to us, but could be perceived as a trigger for people with a disorder. Another explanation for this is that the learning history of a patient possibly shows a connection between the seemingly neutral stimulus and the actual concern, so that it is no longer neutral. However, it could also mirror a vulnerability to distraction.

Definite Transdiagnostic processes

Selective attention towards concern-related external stimuli

Those stimuli that are somehow related to our current goals, thoughts or beliefs, catch our attention to a much higher degree than neutral stimuli. In psychological disorders this is most clearly seen in anxiety disorders, but also in pain and BDD, eating disorders, insomnia, unipolar depression, the psychotic disorders and substance-related disorders. There is no convincing proof of the absence of this process in any study. Also, it is believed that this process can underlie an automatic attention bias towards threat, especially in GAD and PTSD patients.

Selective attention towards concern-related internal stimuli

Internal stimuli’s, being self-focused, is also a definite process across disorders despite some difficulties in methodology. More research is needed on the particular disorders, but today it is known that specifically threatening internal states are in focus of anxiety patients.

Attentional avoidance and attention towards sources of safety

This process also meets the criteria for a definite process calling. Reduced attention to faces in depression, schizophrenia and social phobia are found. The firmest evidence is found in social and specific phobia. The only disorder in which it was not established is GAD. Several explanations (attending to threat as distracting themselves from worse threats) are suggested, but nothing could be scientifically agreed upon.

May attention processes have a causal weight?

Most crucial perhaps is the question of causality; are those attentional processes a maintaining factor of the disorder? Increased emotional reaction was revealed in studies of attention biases in anxiety patients. People reacted stronger to stressful events than control groups. A further study showed that distress levels can be heightened by making people attended to negative stimuli. The reverse was used in a therapy session with specific phobia. Facing a fearful object and manipulating attention (giving descriptions about a spider’s appearance) led to greater fear reduction than for those who did not answer any questions by looking at it. When it came to social phobia, patients were trained to focus their attention on external stimuli, on the environment, instead of internal body states and thoughts. Thus current research tends to support the causal role of attention in psychological disorders.

Theoretical matters

There are specific concerns related to a disorder but also concerns that seem to occur across disorder such as social threats. Emotional words also seem to interfere with normal processing in many patients throughout the range of disorders. It could be overlapping concerns or comorbid disorders. There are three dominating models which try to explain this phenomenon. M. Williams et al. (1997) say that high levels of anxiety (state and trait) prime patients automatically to threats. It goes hand in hand with the claim that selective attention is tied to automatic processing. Depression is seen to follow a controlled process and so it makes sense that selective attention is not part of the disorder. However this model is not flawless as it cannot explain the tendency of patients to attend to safety signals.

Mogg and Bradley (1998) suggest an automatic process but they distinguish between high levels of anxiety and lower level anxiety patients. Those high in anxiety (can also be temporarily induced by a situation) attend to threat while those less fearful will avoid a threat. You pay more attention to an obscure object in the darkness, when you are more scared than during the day. In patients with depression, the motivational goals are mostly diminished and thus they always attend to threats. It also gives an explanation for why this motive to look away can be reinforcing the anxiety. Looking away will prevent patients from determining real threat and imagined. Thus all objects are still seen as danger. However this model lacks an explanation for internal attention.

Wells and Mathew (1994) stated a model which contains an executive system S-REF (controls automatic process). It is tied and influenced by self-beliefs and procedural plans. In patients with psychological disorders, this monitoring might be too excessive and thus new information that could give reason to erase negative thoughts cannot be processed. Ruminating over bad thoughts and internal states occurs. The environment is only scanned for concern-related stimuli. This model can explain internal and external focus of attention and it can also explain why distractions from concern-relevant objects help overcoming fear. The criticism that the automatic process evidence is overlooked is answered with the assumption that all monitoring can be habituated and thus become automatic.

Altogether it must be said that a model containing parts of all three models would be most efficient. The motivational factors play a role as well as automatic threat appraisal but also threat avoidance and safety search. All could be incorporated into the first model (S-REF) but this has not been done yet.

Clinical implications

In disorders such as GAD, social phobia, and specific phobia treatments focusing on attentional bias worked. However it is not clear whether it was a direct process or just a secondary effect that occurred while targeting another cognitive or behavioural process. Five way in which selective attention may contribute to the maintenance of a disorder are suggested:

1. People only focus on consistent information with their current concern so that no disconfirming information is found.

2. Learning new skills or gaining knowledge might be restricted through limited attention to other sources.

3. Attention is more focused on internal stimuli which leads to attributing any event to internal reasons, such as blaming oneself for an earthquake.

4. Selective attention also affects memory and its encoding.

5. Through habituation an automaticity is established which may lead to the feeling of being out of control or going mad.

Being aware of these five processes can help to treat the disorders. New information has to be made salient and therapists need to be patient because not all information might be perceived as they intend it. Conditions can be created where an internal focus is prevented (reducing eye contact). Attentional training can help to gain control back over ones thoughts. Situations could be analysed and feelings as well as thoughts changed. When this is applied to real life, it is called a behavioural experiment. This can be a great relief to the patients as they can see the ‘normal’ view again.

This training is important to disconfirm beliefs and disengage from unwanted thoughts. It is not a coping strategy but rather a resetting process. Wells (2000) came up with such a training device. After self-focused attention has been explained with its disadvantages, patients are asked to fixate on a visual stimulus while hearing different sounds. Each sound has to be taken in and attention has to shift between the sounds. Later they are asked to hear as many as possible simultaneously. External attention is trained.

There are some more therapies that may help. One of them is mindfulness training. People are trained to pay attention to a single facet of the environment so that no bad thoughts can intervene. Cognitive rehabilitation is another form of therapy. This is especially effective for patients with schizophrenia. Another form is computer based, where patients have to form new associations with non-threatening stimuli. Positive data log is a similar strategy but here patients have to actively look for positive stimuli. This may be helpful to patients with BDD or eating disorders. Automatic thought records are another strategy to identify bad thoughts and analyse their truth content. Then alternative evaluations are found.

Future projects

Most research on attentional bias is based on anxiety disorders. More studies have to be conducted on other disorders. Further on there are no studies on comorbid disorders. One factor that makes research difficult however is ethical concerns about exposing patients to their internal stimuli and bad memories (e.g. in PTSD). Personal relevance of the stimuli, is shown to have a great impact on attentional bias in disorders. This could be the reason for the differences in attentional biases among disorders. Difficulties in disengaging from threatening stimuli lead to performance deficits. This is even stronger in acute psychosis and mania. Studying this relationship could help finding a proper treatment for that patient and help them to better function again.

Chapter 3: Memory

It is not unlikely that there is a memory gap after an accident or any other traumatic event. Many patients cannot recollect the entire situation or only after a longer time period. However what most patient experience are short flashbacks of the incident or nightmares. PTSD patients suffer from these symptoms. They wish to forget but somehow cannot.

The construct of memory

The human memory has three stages of processing. First is encoding, we attend to information through selective attention so that we can process the environment for its most important facets. Schemas can help us process the world around us easier but it can also lead to biases as it is a sort of selective attention. The second stage is storage, which is limited, as all of us have already experienced. Information is first stored in short-term memory and will eventually be processed into long-term memory. This is then divided into declarative and non-declarative memory. The latter is observable in behaviour that is automatic. We do not need to engage in active thinking while doing these tasks. Declarative memory on the other hand is different. We go through our memory to find the right answers for facts. This is further divided into semantic and episodic memory. Semantic memory is responsible for factual knowledge, knowledge about names, places etc. Episodic memory refers to personal knowledge. We engage in that kind of memory when we recollect latest personal events (type of yesterdays’ lunch). The third process of memory is retrieval; we try to remember which means we access our long-term memory. Recall tests or recognition tasks can check for this memory. These three processes are closely connected and thus it is hard to separate them from each other in order to define where the disorder occurs.

Explicit and implicit memory

Explicit memory tests such as free call involve patients’ consciousness as they have to name all the previous presented stimuli. They can identify conscious memory processes. Implicit memory tasks on the other hand do not require that. It is a more subliminal test which tests the influence of presented stimuli on responses. Automatic memory processes can be caught that way. Word stem completion might be one of the tasks or tachistoscopic identification (name a series of words after seeing a pattern).

Perceptually-driven and conceptually-driven tasks

Perceptually driven tasks are those which entail a description of visual stimuli. Conceptually driven tasks on the other hand are those that focus on the meaning of stimuli. It is proposed that the same tasks are needed for retrieval and encoding to get the best results of memory processes. When it comes to valence (testing positive or negative words), conceptual tasks are needed. Different kinds of memory processing of a traumatic event show different effects on later symptoms. If a trauma has been encoded perceptually, patients show worse symptoms of PTSD a few months later. It is suggested that not being able to intentionally access memories and thus inability to integrate them into their lives makes it harder to process.

Working memory

The working-memory is the short-term storage. Here a key feature is the central executive which serves as both a decision and a control mechanism. It controls the articulatory loop (verbal information) and the visuospatial sketchpad (visual and spatial information). The latter is especially active in PTSD patients.

Verbal and situational memory access

In the Dual Representational Theory of PTSD (Brewin 1996), two memories are proposed. Verbal and visual memories (VAM) can be accessed intentionally due to proper encoding during the trauma, situational accessible memories (SAM) are subconsciously stored. Those are formed at a lower level of consciousness and are more perceptually-driven.

Are there stimuli that we can remember better?

All in all we attend to stimuli that provide information about survival. Those are either threatening or pleasurable stimuli. Attention is also driven by existing schemas that we have learned throughout our lives. Even though we detect similar stimuli, we see the world through a lens that is made only for and by us. There are more factors than culture and raising style that influence memory. Internal and external environmental factors are also important. Mood-congruent memory for example is an internal factor that shifts our retrieval to mood consistent stimuli (happy-retrieve happy moments). Mood-dependence is a mechanism that leads to better recall of memories that were encoded in a specific mood, once you are in the same mood again (sad state-better recall of your mother’s funeral).

Some experiments tested this via mood inducing tasks. They found confirming evidence. Hypnosis can also be used to induce a mood state or music.

Mood-congruent memory is a robust effect but there is mixed evidence for the mood-dependent effect. Stimuli that led to catastrophizing or rumination are also more likely to be remembered. Personal relevance is another contributor of what stimuli is encoded and retrieved. If you are about to buy a dress, it will be easier for you to think about all the proms you went to. This comes back to the topic about current concerns.

Selective memory

There are two different approaches to how memory is biased in disorders. The first one is based upon Becks schema model and Brewers association theory that patients remember more information related to their current concerns. The second one states the opposite. Mogg came up with a vigilance-avoidance hypothesis. Most research used to identify memory biases use encoding tasks. Participants have to encode a list of words that are related to a disorder. After that they are asked to remember the words. The number of remembered items is compared to a comparison group.

Anxiety disorder

Panic disorder with or without agoraphobia

Most studies support an explicit memory bias for concern- relevant words. The findings for implicit memory are mixed.

Social phobia

Here many studies failed to confirm an explicit memory bias. However those studies that were conducted outside the laboratory were more successful in finding an explicit memory bias. Some studies found evidence for an implicit memory bias but others did not.

Obsessive-compulsive disorder

An explicit memory bias was not found for patients with OCD who fear contamination. Implicit memory bias on the other hand was positively tested in both patients and normal comparison groups.

Post-traumatic stress disorder

Patients show an explicit memory bias in PTSD, although it must be said that it was tested on an emotional Stroop task. Earlier studies were not so clear in their results. There was even better memory for emotional words. Implicit memory biases for words have not been found so far but sentences may show a different pattern.

Generalized anxiety disorder

Previous studies failed to show an explicit memory bias. Novel mythology however tested patients with GAD as biased. This novel measurement must be further tested. Moving to implicit memory bias, there is mixed evidence.

Testing outside the laboratory

There are five studies that tested memory in a real world setting. Lundh and Ost (1996) asked participants to rate face on photographs as either critical or accepting. The number of remembered positive and negative faces was counted. Anxiety patients showed a memory bias for critical faces. Lundh (1998) did the same but panic disorder patients had to rate the faces as either safe or unsafe. A memory bias for safe faces was detected. Mansell (1999) put social anxiety patients in a threat condition (they had to give a speech) and asked them then to encode a number of private self-referent, other-referent and public self-referent words. Less positive public self-referent words were recalled. Radomsky (1999) tested OCD patients by letting them remember what object they touched with either a clean or a contaminated tissue. Results showed a memory bias for contamination objects.

PTSD; a close-up

A longitudinal study was done for patients with PTSD. They were interviewed shortly after a trauma and again several years later. A positive correlation was revealed between the severity of a trauma and the alteration of remembrance of the event, the approximation of trauma exposure and the recollection of the number of symptoms directly after the event. These results confirm the selective memory bias and mood-congruent theory. An important aspect that goes with those results is that traumatic memories are not stable and ineffaceable. Consequently false memories can occur. There are different opinions about this issue: two strong opposites and one middle ground:

1. The experience of such a horrible event can be so distressing that an individual cuts out the memory of it. However after several years the memory can come back into the mind.

2. Several variables affect the construction of memory. Memories are possible to be influenced and newly formed by mood, selective attention or current symptoms.

For clinical settings it is important to be aware of the fact that memory is always reconstructed and never possible to be like a videotape in reverse. It must be treated as narrative truth. Suggestions or leading questions must therefore be avoided.

Somatoform disorders

Pain patients recalled more pain related words. Here it is important to note that comorbid studies found that depression cancelled this out. Only patients without depressed mood reported more pain words. Depressed pain patients showed memory bias for negative pain words when those were self-referential. Memory bias might thus be a secondary consequence of pain, not a trait factor. Hypochondriacs also show a heightened recall number of pain words. Severity of pain symptoms may play a role when it comes to pain disorders. Also in patients with somatoform disorders memory about the initial pain has to be treated carefully. Patients who stated an increase in pain at the end of a treatment rated their initial pain as higher than when they recorded it at the time. Pain congruent memories are also confirmed. More pain incidents were recalled in a current pain state.

Eating disorders

An explicit memory bias was found for anorectic patients in relation to anorexia-related words. No implicit bias was observed. In another study people with eating disorders had to imagine a scene involving them in connection with that cue word. On a free call test, a bias towards fat words was found. For bulimia patients, a memory bias towards weight/shape was revealed in comparison to emotional words. For all patients with an eating disorder a food-related memory bias was found.

Mood disorder

As expected, the more severely depressed, the more sad memories are recalled. Even state factors can influence memories of depressed people. On the sad vs. happy faces- photographs a clear tendency for sad faces was found. It is clear evidence for explicit memory bias. This bias is supposed to have crucial consequences and effect on maintenance of the disorder. Sad faces are better remembered and thus no objective evaluation of others is possible. Most studies also detected an implicit memory bias.

Bipolar-disorder

Manic patients used more positive adjectives to describe themselves. However when they had to recollect trait words, more negative words in comparison to a control group were reported. This suggests that the underlying process of manic and depressed patient is similar.

Substance related disorder

An explicit as well as an implicit memory bias was reported in alcoholics. State anxiety also enhances the bias for alcohol related cue words. As soon as memories about related cues (anxiety) are present, cravings increase.

Sum-up

For panic disorders, social phobia, OCD, PTSD, GAD, somatoform disorder, eating disorder, unipolar depression, bipolar disorder, and alcoholic disorder there is at least one study supporting an explicit memory bias. Depressed patients show a bias for depression-related words but not for general negative words. The explicit memory bias for concern-relevant information is a definite transdiagnostic process.

Fewer studies are found in regard to implicit memory bias. It is still regarded as possible transdiagnostic process as confirming evidence was found in most of the disorder. An implicit memory task is proposed to be of better validation if measured with a conceptually driven task, not a perceptually driven one. A fuller context could also help detecting the bias. Caution must be exercised in regards to state dependent memory such as being in pain or traumatized as those can influence the construction of memories. More real world studies have to be conducted as those seem to catch the biases in a more realistic way.

Forms of memory

Autobiographical memories are those remembrances about one’s own life. A distinction has been made between two kinds of autobiographical memories. The first are generic memories and refers to events that happened over a longer period of time. The second are specific memories which are bound to a particular place and time (about one day). People who attempted suicide were found to have more generic rather than specific memories. Disorders are thought to be maintained because of the same lack of recalling specific memories. This lack makes it harder to function in the world. Generic memories are further divided into two forms; categorical and extended memories. The first is a summary of events (birthday party, exams) and the latter refers to memories about events that lasted longer than one day (stay in a hospital). Psychopathology is related more to category memories, especially in depressed people.

Anxiety disorders

Social phobia

For patients with social phobia, decreased specific memories were not observed.

Obsessive-compulsive disorder

OCD patients showed difficulties recalling specific memories; however it is thought to be due to comorbidity with depression.

Post-traumatic stress disorder

In patients with PTSD, overgeneral memories were found for people who had experienced direct trauma. This could also be a consequence of comorbid depression. The severity of the trauma is also an important factor; the more severe a trauma, the more generic memories.

Generalized anxiety

The test on overgeneral memories revealed that GAD patients recall more nervous memories but not more generic ones.

Eating disorders

For eating disorder patients a tendency of using more generic memories was found. Here they found that there is a correlation between parental abuse and generic memory recall.

Mood disorder

Unipolar depression

The clearest evidence for a lack of specific memories is established in patients with depression or those who just attempted suicide. Most generic memories are categorical in their nature. Results for postnatal depression showed the same pattern but those for seasonal depression did not. The explanation for that is the strong biological component in seasonal depression. Child abuse was a strong predictor for generic memories in depression. The reason why specific memories are so important is that it gives the opportunity to analyse it in a different way (he might not have meant me when he was shouting as there was another person in the room too). Further one can retrieve them to change the behaviour in the future or plan things differently.

As already mentioned in the section on PTSD, autobiographical memories are not necessarily stable but they are modifiable. Experiments have been conducted to show this.

Bipolar disorder

In line with the findings for depression, patients with bipolar disorder also show a bias for generic memories and thus they came up with fewer solutions for problem solving-tasks.

Psychotic disorders

Those people suffering from delusions showed more generic memory recalls. The strange finding in that study was that a depressed group was also tested and less generic memories were reported.

Sum-up

Overgeneral memories are found in OCDs, PTSDs, depression, bipolar and eating disorders, suicidal patients and those with delusions. However in some disorders there was no such bias found which makes overgeneral memory bias a possible transdiagnostic process.

Overgeneral memory is strongly associated with trauma (war, child abuse). Altogether, generic memories bring problems with cognition, especially with reappraising of situations.

Avoidant encoding and retrieval

Overgeneral memories are seen as an avoidant coping style. Experiencing a trauma leads to the attempt to forget it in most cases. It reduces the intensity of specific events. The mechanisms are related to dissociation processes where patients cannot fully remember events which make it easier to survive. The drawback is that it prevents them from recovery as they cannot work on the trauma. The concepts of dissociation is still in debate as it is seen as too unspecific, may not be related to memory encoding or retrieval deficits and whether it is actually dissimilar to attentional avoidance. The process often described as strategy is misleading as patient cannot control it and no intentions could be found. To explore the mechanism further, an item method has been presented, words are presented and participants are asked to either remember or forget the word. Free recall of the items they were asked to remember is better for non-pathological participants. It is the direct-forgetting effect. A second method: list method is introduced where participants are asked to remember a list of words and halfway through they are instructed to forget the first half and only remember the second part. The same results are found as in the first method. The item method is sensitive to encoding while the list method measures more the retrieval process.

Anxiety disorder

Specific phobia

In line with the avoidance theory, people with spider phobia must have a poorer memory of spider words or actual spiders. Tests revealed a lack of anxiety evoking words but not for features of a spider. Poorer memory for spider words was only observed when patients were exposed to them in real life.

Social phobia

Poor results for recall of social threat words or sentences were recorded. This was interpreted as clear evidence for the avoidance hypotheses.

PTSD

Less PTSD symptoms after treatment were associated with more organized thoughts and reduced fragmentation of the traumatic incidence. The better an individual can tell a trauma, the better his or her chance to process and resolve it. The more –disorganized, -dissociative and -avoidant a person is directly after experiencing a trauma, the higher the chances of PTSD several years later. However one study that used the item method did not find any difference between the number of forgetting items related to trauma that were recalled by PTSD patients or controls. Acute stress disorder patients showed a bias towards forgetting items. One explanation is that it is possible to be diagnosed with PTSD without any dissociative issues but for the ASD diagnosis this is an inevitable criterion.

Dissociative disorder

It is to be assumed that dissociative patients show a clear bias to forgetting items but in fact the only study that has been done so far did not confirm this.

Sum-up

Poorer memory for threatening words has been found in social and specific phobia. The try to ignore uncomfortable feelings. This finding is in line with the vigilance-avoidance hypotheses. Furthermore PTSD and ASD patients show similar mechanisms that are observable in a disorganized and fragmented narrative of the trauma. Resolving it is only possible once a clear image of the event has been established. The process of dissociation or avoidant encoding and retrieval therefore meets the criteria for a possible transdiagnostic process. Future research is highly necessary as there have just been a few studies which also showed contradictory findings.

Intrusive memories

Intrusive memories are thoughts and images that are high in meaning and emotions. They are also very informative for the understanding and the maintenance of psychological disorders.

Anxiety disorders

Agoraphobia

All patient with this disorder reported intrusive memories, mostly about past events.

Social phobia

In social phobia people are believed to have intrusive observer perspective images which mean that they look at themselves from an outside point of view. These images continue when they enter a social situation which makes it impossible for them to see the real reaction of people. Perception of behaviours is distorted because of their pre-existing schema. In one study, interactions were recorded and participants asked about observer perspective memories. Social phobic patients reported far more of them. Assessing the onset of the illness, researchers found that there was a link between one negative event in which participants had a cruel image of and the onset of the disorder. This recurrent and intrusive image is proposed to be a factor of the disorder. When normal controls were tested, holding negative self-images increased the ratings of past failure recalls.

PTSD

For PTSD patients, visual intrusions are most often reported. The largest emotional impact has stimuli that were shortly before the incident, or during the worst part of the incident. The warning signal hypothesis states that these stimuli are warning signals to prevent an individual from experiencing it again. If this hypothesis holds, it is a good explanation of why these trigger have such a strong impact on patients. One can reduce the intrusions (dual representation theory) by engaging in a visual spatial task that competes with cognitive capacity for other memories. More tests are needed to conclude anything.

Mood disorders

Unipolar depression

Intrusive thoughts experienced in depressed people are quite similar to those in PTSD patients. A deviance was found on the helplessness dimension; PTSD patients scored higher.

Bipolar disorder

In comparison to control groups, bipolar patients are said to experience negative intrusions more often in everyday life.

Psychotic disorders

75% of psychotic patients report images as hallucinations. Of these, 71% are related to past memories. Becks theory about the essential value of the content of images is supported here. It is important to find the source of the image to resolve the symptoms.

Sum up

In all the disorders examined, recurrent intrusive memories have been found and thus intrusive memory is a definite transdiagnostic process. Further an interesting finding has been evoked that memories of depressed patients and PTSDs are of similar content.

Working memory

The Phonological loop is the processing of all verbal information while the visual spatial sketchpad is handling the visual and spatial input. They are both independent entities but are also controlled by the central executive of the working memory. One study tried to find out which of the two is more effective when occupied, to distract from worry. Worry is mostly verbal and so it is believed that distraction to the phonological loop will be more effective and indeed that is what has been found out. Letting participants come up with a letter once every second (involving the phonological loop itself and the control function of it) was most effective in distracting from worry.

Memory distrust

The more often patients with OCD are exposed to unsafe objects, the more they edge in their checking behaviour. This is because they distrust their memories by an increase to exposure to danger.

Sum up

Explicit selective memory and recurrent memory have been found to be definite transdiagnostic processes. Possible transdiagnostic processes could be implicit memory and overgeneral memory and avoidant strategies. There has not been enough research on the last topic about memory distrust so no inferences can be drawn.

Definite transdiagnostic processes

Explicit selective memory

Some of the results of the studies were weak. However this could be due to a lack of validity in the word tasks. Studies that have been conducted outside of the lab were far more promising. Better encoding for concern-related information is definite throughout the disorders and this is also supported by the leading theories about this topic; associative network model and the schema model.

One other proposal by Williams (1997) for the weak results in some disorders regarding memory is that encoding and retrieval processes are automatic as well as strategic. So it can happen via automatic that means unconscious process or via more effortful processing. These two can operate independently. Emotions such as anxiety might therefore lead to interference with automatic but not strategic processing while depressed people have problems with strategic processing rather than automatic. More recent evidence supports the first theories at expense of the latter. However further research is necessary to be certain.

Intrusive memories

The content of these recurrent memories are related to the current concerns of the patients. Thus they are different between the disorders. In PTSD it is assumed that the content of the images represent moments just before that trauma or its worst features. This would make evolutionary sense as it could serve as a warning signal that prevents an individual from reliving it. It is important to find out whether the presence of recurrent memories in different disorders is due to traumas throughout those disorders or if the concept of trauma is similar in different disorders but to a lower degree of severity (in OCD-being exposed to a contaminated objects).

Possible transdiagnostic processes

Implicit memory

Implicit memory reflects an automatic process and so the explanation for the weaker results regarding this type of memory given by Williams can be restated.

Overgeneral memory

Only one study has been conducted for each disorder. This gives reason to assume that it might be a definite process but more research is needed on autobiographical memory. A vital question is to what degree a trauma is sufficient to evoke overgeneral memories. A hierarchical structure has been proposed by Williams (19996) which puts specific memory at the lowest level, general knowledge at the middle and lifetime periods at the top. To retrieve specific memories one has to go through all the others.

The middle level has to be controlled and suppressed in order to retrieve the specific information. This develops by the age of 3-4 years. The aversive nature of some memories leads to the protection mechanism of suppressing the specific knowledge. However this abortion of retrieval leads to a chronic search for the last level information which results in an over-elaboration of the negative general memory. So by trying to avoid an extreme thought once, a constant bad general memory is retrieved this blocking cycle is called a mnemonic interlock. This hypothesis has been supported.

Avoidant encoding and retrieval

This strategy is thought to happen because of protection from the bad memories as ASD and PTSD patients show this bias. It might help for a short moment but in the long run it prevents them from resolving the memories. Future research needs to examine the null findings for dissociative disorders. The vigilance-avoidance hypothesis explains this process for the other anxiety disorders. Danger is quickly detected to be able to engage in safety behaviour which in that case is avoidance.

Distinct processes and inconclusive processes

No distinct processes could be found in the section about working memory. Memory distrust needs to be investigated further to find out whether it is present in more disorders than OCD otherwise it could be considered as a distinct process

Causality of memory processes

Dissociation, avoidant coping style and perceptually-driven processing reveal the most promising findings for a causal role of PTSD. Overgeneral memories are proposed to predict depression and sometimes PTSD too. However, all the methods for the studies of causality have to be improved to tell anything certain about it.

A recent manipulation experiment showed an effect on affect depending on what type of memory is retrieved. Further, holding negative self-images are found to predict higher rating of anxiety.

Theoretical issues

It is assumed that the attentional bias towards concern-related information hinders other information from the environment to be properly encoded and stored. Data have shown that patients have a better retrieval function of those memories related to the specific disorder but less access to other memories.

So memory access is poorly, because:

1. Specific memory retrieval is aborted at the general knowledge level.

2. Dissociations and avoidant coping stop encoding and proper storage.

3. Current-concerns dominate attention.

Why these memory impairments maintain a disorder is explained as followed:

• More attention and thus encoding of current-concern related issues will heighten the number of confirming evidence.

• Due to the availability heuristics, we know that the easier knowledge is available, the more we are influenced by it during the process of reasoning.

• Symptoms can modify memory retrieval.

• Poorer access to specific knowledge makes it harder to process emotionally loaded memories which are needed to resolve traumas

• Poor access to specific memory prevents patients from shifting and revaluating past situations.

• Impaired (specific) memory access is associated with poor problem solving.

Encoding and retrieval are the two most crucial processes of memory and consequently for its distortion. Current concerns influence encoding related to personal goals. Threatening stimuli also heighten our attention and encoding (weapon focus effect). Rumination and catastrophizing are further contributors to better encoding. Conceptual processing (deriving meaning) leads to better encoding too. When it comes to retrieval, current concerns and goals are also important factors. Overgeneral memory retrieval distorts retrieval so that a chronic tendency towards general memories occurs. Acute symptoms lead to a decrease in access to specific knowledge. Furthermore, disorganized memories make retrieval harder.

The use of restricted access to specific knowledge can only be explained by affect regulation. Some of the impairments of memory can be seen as a by-product of another process, e.g. memory distrust in OCD might be due to constant rumination.

Clinical implications

Avoidant encoding and retrieval

It is useful to educate patients about the coping strategy they are using and its possible consequences. Metaphors such as ‘try closing a cupboard into which all things have been thrown in randomly- does not close versus try sorting things out and order them neatly- doors can be closed again and nothing will fall out again’ can facilitate understanding the cycle. For resolving disorganized memories in trauma patients, exposure therapy seems to help. First, patients imagine the situation in as much detail (affect and sensory) as possible. Cognitive restructuring will take place afterwards or during the imagination. That way all emotions can come out which have been buried deep down (guilt, shame, hate). Later on they might even be exposed in vivo to them (likely specific phobias). Gaining factual information is another good step to the clarification of a situation. If accidents happened the police can be asked for the reports so that possible distorted memories can be restricted rationally.

Intrusive memories and images

It has been suggested that flashbacks may be a warning signal to prevent reliving a situation again. Stimuli are perceived as a trigger when they are similar to anything a person has perceived just before a trauma. This brings about all the emotions and fears that patients have experienced during a trauma. Therapists need to clear up that the past incident is over and will never occur again even if similar things happen to those before the trauma. Current warnings must be understood to be no actual signs of threat. For nightmares it is helpful to let patients write them down and change them afterwards in their imagination or also on paper. Rehearsing them also just before going to bed has led to less reoccurrence. Traumatic images can also be restricted. Patients are asked to take over a new perspective or add a helping person to the situation. Role plays can also help relieving patients from distress. It offers alternative ideas about an incident which make it less horrible. That way the occurrence of the images is decreased. Replacing negative self-images is necessary in patients with social phobia or depression. Perspective taking also seems to help in finding better images of one self. Focusing attention to something other than oneself is important in social phobia.

Overgeneral memory

Recognizing that those general recollections are part of the disorder already helps the client therapist relationship as frustrations can be avoided. Further, automatic thought diaries can help in detecting negative thoughts and reconstructing them. Another form is mindfulness-based cognitive therapy which helps parents to be non-judgmental and pay attention to only one aspect of a situation. This increases awareness for the here and now and has shown to help prevent relapses in depression. Too often patients are engaged in thinking processes about the past or future which make them feel helpless.

Memory distortion

Recollections of events have to be seen as narrative truths. Current mood, symptoms and fear can all modify memories. False memories can be elicited when therapist are pushing into one direction. Hypnosis must be applied carefully as it has been found to create false memories. Keeping pain diaries for chronic pain patients is useful as the current pain will influence estimates about the initial pain before the treatment.

Future research

Future research has to be done on a broad range of disorders so that we can be certain about the transdiagnostic approach across disorders. Appropriate methods have to be used such as experiments which can also draw conclusions about causes. Models of assessment must be conducted so that equivalent encoding can be tested and only then tested for effects. Furthermore the different findings of perceptual- and conceptual-driven tasks must be taken into consideration when testing for certain memory types (implicit vs. explicit). Increasing the number of field studies to increase external validity is also very important. Experimentally induced moods are shown to elicit different memories than real life situations. Control for current symptoms is also crucial.

One way to do that is to measure patients’ pain rate with and without current pain and to compare cognitive capacities in both conditions. Future research also needs to take into account comorbid disorders to separate out the effects only for one disorder. Relationships between all those processes discussed in this book also need to be investigated.

Chapter 4: Reasoning

Thoughts like: “I have made a fool out of myself, they were all staring at me so I must have looked weird, the reason why somebody stopped a conversation was because I am a terrible person etc.” are common in patients with social phobia. Their reasoning processes are mainly negative. Reasoning is thinking concerned with making judgments, conclusions and testing hypothesis. A bias in this process is observed when thinking is going into one direction over time and different situations. It is often the case that conclusions deviate from objective evaluations. Reasoning biases can always happen when we try to makes sense of anything. It must not be dysfunctional but can serve as protection of our self-worth (self-serving bias). There are different reasoning processes:

• Interpretative reasoning (concluding meaning from ambiguous situations)

• Attribution reasoning (drawing conclusions about a cause)

• Expectancy reasoning (making predictions about the future outcome)

• Detection of covariation.

• Hypothesis testing (evaluating the likelihood of a belief).

It is sometimes hard to separate the reasoning processes from each other as they can have great overlap.

Current concerns

Biases in reasoning reflect the current concerns of the specific disorders. It is the same with some biases that are present among the normal population, just that those serve as bolstering self-worth.

Interpreting ambiguous stimuli

How people interpret ambiguous situations, facial expressions or verbal expressions is an important factor of their overall functioning in the world. Interpreting everything in a negative manner will lead to extreme fear which is followed by impairments in day to day functioning. Cognitive distortions such as: arbitrary inferences, which is the tendency to draw conclusions from something without any evidence; selective abstraction, the affinity to look only at a small part of a situation and thus neglect the entire context; overgeneralization is observed when patients apply the conclusion of one isolated incident to most incidents; catastrophizing reflects exaggerations of an event or imaginations about a possible event; personalization involves inappropriate attributions to oneself that happened due to external factors; dichotomous thinking is the categorization of thoughts and events into only two aspects, bad or good, clean or contaminated.

Self-report examples

Some studies about reasoning included phrases that had to be evaluated by anxious patients. These vignettes were stated in an ambiguous manner so that different interpretations were possible. Patients had to rate them on a scale of threatening to not at all threatening. However these studies evoked some criticism about the demand effect.

Cognitive experimental examples

The critique about the self-report paradigms made people come up with experimental alternatives. The homophone spelling task is one way to measure interpretations of ambiguity. Words related to a disorder and neutral ones, with the same sound but different in spelling are presented via headphones and the number of disorder-congruent words counted. The number of disorder-congruent words is an index of how stimuli are interpreted. The lexical task is similar but here participants have to decide which word is proper English and which is not (Daisies or disease). The time it is taking them to decide is measured, the longer it takes, the fewer associations are assumed. Another form is the recognition memory task, an ambiguous sentence is presented and after a short while participants have to recall the content of it by choosing one of two sentences given. These are similar but vary in one word that is either related to current concerns or neutral (large hips vs. toned hips). However even those measurements could be biased. Response bias rather than interpretations could have elicited the responses. Online text comprehension (understanding of sentences either stated in a way consistent with current concerns or not) assess time differences which indicate interpretation of the world. Visual blink tasks measure blink reflexes which are larger when we think of negative scenarios. Ambiguous sentences are presented and the magnitude of the blink reflex points out whether someone interprets it as threatening or not.

Anxiety disorder

Negative interpretations are common among patients with any kind of anxiety.

Panic disorder without agoraphobia

More negative interpretations of stimuli have been found to be present in people with a panic disorder. Interpretations are concern-related which means situations that make patients uncomfortable are seen as threatening.

Social phobia

More negative disambiguation was made by patients with social phobia. These are also concern-related; in this case self-relevant information is interpreted negatively. Here a distinction is made between on-line reasoning biases (thinking badly during a social situation) and pre- or post a social situation. No positive or negative interpretations of on-line situations were made by patients while non-patients showed a clear positive bias. This might be the reason why social phobic people make more negative disambiguation of their social performance.

Post-traumatic stress disorder

There seems to be a difference between PTSD patients and those people who have experienced a trauma but do not suffer from PTSD. PTSD patients make more threatening interpretations while others do not.

Generalized anxiety disorder

Patients show more threatening interpretation bias on all of the measurement tasks.

Somatoform disorder

Ambiguous words were more often interpreted by patients as being health related compared to those of physiotherapist and normal-controls. If a pain disorder was present; words were seen as illness related or associated to pain. Hypochondriacs showed much more severe identifications of small body sensations. Disorder and illnesses stating sentences are chosen rather than benign causes when describing a situation of a sick person.

Eating-disorders

Interpretations related to being fat were more often chosen in body related scenarios. Health situations, however, were not biased. Negative self-evaluative sentence were related to body issues (I am bad- I weigh too much). Further feelings of hunger, weakness or dizziness in relation to diet were more frequently interpreted in terms of personal control.

Psychotic disorders

Ambiguous sensory input is more often misinterpreted in patients with schizophrenia. Meaningful sounds, presented to their ears were less likely to lead to hallucinations than random noise.

Substance-related disorders

The extent to which a person drinks influences the degree to which they interpret stimuli as substance-related. Ambiguous words (bar, shot) were more often interpreted as alcohol related in drinkers than in controls.

Sum-up

Interpretation bias seems to be shared between all disorders tested. Thus it meets the criteria for a definite transdiagnostic process. Interpretations are made for concern-related scenarios. Still, further research is needed as some responses could be biased due to lacking methodology.

Attributions

Life events are attributed to different causes, this is an inevitably. Explanations used by people suffering from psychological disorders differ from normal control groups. They differ in regard to:

• The mark to which the cause of an event is internal or external of the individual.

• The mark to which the cause of an event is stable or transient.

• The mark to which the cause is global or local

It has been validated that most people have a self-serving bias, interpreting negative events as caused by external, transient and local factors. Positive events however, are interpreted as internal, stable and global. In most disorders though, there is no such bias present but an opposing attributional style (pessimistic attribution style).

Anxiety disorder

It is not clear whether anxiety patients suffer from a pessimistic attribution style. Those findings that found significant results did not all control for comorbid depression and thus no conclusions can be made. Some studies show that GAD, panic disorder, and phobias lead to more pessimistic attributions; especially negative outcomes were seen as stable and global. However GAD patients were also found to be less pessimistic than depressed people. The study of attribution biases on the specific concerns directly detected more significant findings. In OCD patients leaking thoughts were regarded as more internally caused. Or in social phobia bad performances were perceived as internal, stable and global.

Somatoform

Findings for somatoform disorders show that symptoms are attributed to physical causes and illnesses rather than to emotional ones as is the case for anxiety disorders.

Sexual disorder

More internal causes are attributed to failure in sexual performances. In the case of sexually functional men, erectile difficulties were far more rated as externally caused (stress, loud noises etc.).

Eating disorder

These patients show a pessimistic attributional style. However, when depression was controlled as covariation, the bias was reduced or even extinguished.

Mood disorder

There is clear evidence for a pessimistic attributional style in patient with depression. Helplessness and reduced motivation are associated to it. Bipolar patients in a current manic episode experience a self-serving bias but afterwards they fall back into the opposing style. However, even in the manic episode, patients reported more internal causes for negative events. They might suffer from an implicit pessimistic attributional style.

Psychotic disorder

Patients who have had delusions attribute negative events extremely to external causes. For those who are paranoid attributions are made to external personal causes rather than to situations. Studies show an implicit negative attributional style for events but explicitly report the opposite. This might only serve as a defensive mechanism. However, there is some evidence for the external attribution of internal mental events which is then diagnosed as hallucination. A lack of self-monitoring and source monitoring is considered to be the first segment of this misattribution. Schizophrenics fail to distinguish the origins of either self-generated or experimenter-generated tactile sensations. The lack of source monitoring is displayed in the inability to recall the source from which knowledge is derived and whether it was produced externally or internally. A spontaneous thought can then be perceived as coming from outside you.

Impulse control disorder

An excessive self-serving bias has been found in these patients. Losses are attributed to bad luck and winning to skills. This might be one of the reasons which make stopping gambling so difficult for these people.

Substance-related disorder

Receiving treatment on addiction leads to fewer global and stable attributions about positive events compared to non-patient controls. There are no differences found for negative event attributions. The pessimistic attribution style was only found for the abstinence violation effect. Addicts who lapse from their abstinence will attribute this as failing all types of self-control. This interpretation brings back feelings of guilt and shame which often results in a full relapse.

Sum-up

In a few disorders, a bias in attribution is present; however, the nature of them can be different across the specific disorders. A pessimistic attributional style has been found for depression. Anxiety disorders and eating disorders show features of it but this might be due to comorbid depression. In schizophrenia (with hallucinations) external attributions take place for negative events. Future research is needed but so far it is concluded that attributional bias is a possible transdiagnostic process. Current concerns may be the reason for differences in attributions not the actual attributional style.

Expectancies and heuristics

We are always trying to find out what is happening next, which outcomes will follow certain cues. During these reasoning processes bias are likely to occur and this contributes to the maintenance of a disorder. To make it clear, imagine a person who has the expectations that drugs help overcoming negative emotions, he is predicted to take drugs. Expecting negative events on the other hand are likely to end in avoidance behaviour. Heuristic rules are the reason why we are mostly biased when we try to figure out uncertainties. The most common heuristics are:

1. Availability heuristic.

2. Representativeness heuristic.

3. Emotional reasoning heuristic.

It is not said that these rules always fail but they consistently bias our judgments, especially by underestimating statistical and abstract information. The reason for us to use them nevertheless, is because they are effortless and economical.

Availability heuristic

The availability heuristic influences reasoning by taking in the latest memories (those best available) to estimate the likelihood of an event to occur. A subtype of it is the simulation heuristic which is imagining a situation in different scenarios. The ease by which the simulation is possible is rated as the degree to which that outcome is likely to be true. Feelings of familiarity, vividness, recency, and salience of memories are all part of the availability heuristic and thus influence our judgment. Counterfactuals are the counterpart to simulations, alternative outcomes of an event in the past (what if...). They are mostly used for negative events and can influence future behaviour. Being able to imagine negative events easily will thus lead to reason that they are likely to happen.

Representativeness heuristic

Seeing five famous female names and eight male names on one list and six female names and eight male names on a second list that are not well known will lead to recall more female names on the first list. This is an example of the representative heuristic. The gamblers fallacy is another one; the belief that throwing a coin is not due to chance if head turned up five times in a row. Sequences are not believed to occur in chance derived occasions.

Emotional reasoning heuristic

This rule is applied when we make judgements on feelings about something. Being in a good mood will influence your answer about how satisfied you will be in 2 years, for example. When participants were asked about their current mood state however, these biases were eliminated. It is assumed that once we become aware of possible influence, we try to adapt. Ex-consequential reasoning is the term for this bias in relation to psychological disorders. Anxiety patients who consistently feel anxious think that there must be something dangerous causing it. Positive expectancies on the other hand are associated with problem solving behaviour.

Anxiety disorder

Panic disorder with and without agoraphobia

Patients with panic disorder came up with more negative future events and rated those as more likely to happen. They also show emotional reasoning, with rising feelings of anxiety indicating the presence of danger. They further overestimated the degree to which they would be terrified of a specific object.

Specific and Social phobia

They all misjudged the extent that bad events will happen to them in the future. This was specific to the current concerns (I will embarrass myself in a social context; I will be bitten by an animal). For specific phobias data show emotional reasoning biases.

Obsessive Compulsive disorder

Negative outcomes are overestimated in the case of OCD. Furthermore, evidence for emotional reasoning is supported. Studies found that checking and other rituals are repeated until something feels right.

Post-traumatic stress disorder

Acute stress disorder patients seem to estimate negative events to be likely to happen to them. Further, it has been tested that PTSD patients assume danger as soon as they have feelings of anxiety. Intrusive thoughts are connected to fear and thus they can also influence judgement.

Generalized anxiety disorder

Patients with GAD are worrying extensively and consequently they estimate the likelihood of dangerous happenings as high. The availability of negative events is heightened which might explain the increased estimation of upcoming bad events. Depression was controlled as a covariate in one study, with the result that even in the absence of depression, GAD patients rate positive outcomes as less likely to occur, and negative outcomes as more likely to occur. High trait anxiety is an even stronger predictor for the use of emotional reasoning.

Somatoform disorder

Pain disorder patient overestimated the degree to which they would experience pain during an exercise. This led to more fear and further to avoidance behaviour. It is believed that the expectations about pain toleration predict actual tolerance of pain. In patients with hypochondria, overestimation of becoming sick is a core feature. Questions about the likelihood of non-related health issues were normal.

Sexual disorder

Men suffering from erectile problems underestimated their sexual performance. In one study, men were said to be given erectile-increasing pills, decreasing-pills, or placebos, while in fact all of the pills were placebos. In those who were suffering from sexual problems, decreasing pills led to a decrease and increasing pills led to an increase of erection. This indicates that patients’ negative expectations about sexual performance lead to an actual decrease in performance.

Impulse control disorder

Gamblers, for example, believe they will win the games; their reasoning is biased by a feeling as well as by the representative heuristic. Further they think that gambling brings about good outcomes.

Eating-disorders

Patients with an eating disorder reported a higher likelihood of bad events consequential to their shape and or weight. Especially bulimic patients tend to believe that binging helps handling bad emotions and that thinness brings self-worth and other good outcomes.

Mood disorders

Depressed patients are more prone to anticipate bad events in their future. These estimates are given even in the presence of cognitive load, which means they are implicit. This leads to the assumption that images of negative events are highly available, resulting in biased expectations.

Psychotic disorders

Also in psychotic patients, overestimations of negative events have been found. Here the negative event expectations further applies to other people as well.

Substance-related disorders

The problem about expectations in addicted people is that they associate positive outcomes with the substance. The stronger those beliefs, the higher the rate of relapse present.

Sum-up

Biased judgment about future events seems to be common among all disorders tested and therefore it meets the criteria for a definite transdiagnostic process. The expected negative events are always related to the specific concern. Further on, emotional reasoning was also positively tested throughout most disorders and can be said to be a definite transdiagnostic process. Presence of an availability bias was observed in some of the disorders which make it a possible transdiagnostic process.

Identifying covariation and illusory correlation

Identifying covariation means detecting the co-occurrence of two events over time and place .It is an important mechanism in order to function in the world. Knowing that stealing is followed by punishment is helpful to prevent one or the other. However, a bias in covariation detection may lead to falsely believing that two events always occur together, when in fact they are totally independent. This false correlation is called illusory correlation. We tend to see such an illusion when we expect a correlation. This is very important when we are trying to understand PTSD patients. A trigger is seen as being tied to the horrible event and thus patients experience all the terrifying emotions again in the expectations the trauma will occur too.

Anxiety disorder

Participants were shown slides, displaying different stimuli (fearful as well as neutral) which were followed by an aversive stimulus such as loud noise. In the end they were asked to judge the eventuality that fearful stimuli and aversive response occurred together. Covariation bias is present when the proportion is extremely overestimated. This was the case in panic disorder and specific phobia. No such bias was observed in socially anxious people and fear of blood-injuries.

Impulse control disorders

Problem gamblers show a tendency to correlate environmental features and success or failures.

Psychotic disorders

Illusory correlations were found in patients with delusions. They see connections between their delusions and words relevant to those.

Sum-up

There is clear evidence for illusory correlation in panic disorders and specific phobia. Further it is observed that gamblers engage in the covariation bias and people with delusions. Thus it is a possible transdiagnostic process.

Hypothesis testing and data gathering

In the attempt to explain everything, we search for evidence. When the information is present we also review our explanations. However, these revisions are mostly biased, as we are prone to only see information consistent with prior beliefs. Expectancies have a similarly influential factor, evidence against it is downplayed and consistent evidence overvalued. Syllogistic reasoning tasks are mostly used to test the bias. Three statements are given, two of them premises and the last a conclusion. It must be decided whether the last one is a logical result from the two statements. This is based on logical reasoning as well as on previous knowledge. More mistakes are made when the statements are convincing and consistent with prior knowledge but not logical. Views can also be seen as conditional rules; “if… then”. Potential threats are stored as danger rules (If a spider approaches, then danger) or safety rules (if the spider is moving away, then safe. That way danger can be quickly detected. The other aspect of hypothesis testing is how much evidence is collected before something is attributed as dangerous or safe (confirming/rejecting a hypothesis). In some disorders it is believed that they are under cautious while in others they might gather far too much information before reaching a conclusion which makes them hyper cautious. Probabilistic tasks are useful to find out the degree to which people test a hypothesis.

Anxiety disorder

Specific and Social phobia

In arachnophobia, patients show a danger-confirming bias, where they were only looking for consistent evidence to their prior belief. This is an extension to the threat-confirming bias that is observed in all people. When tested on the syllogistic task, a general belief bias was found for women suffering from arachnophobia. This means they are generally less able to revise conclusions that have been disconfirmed.

Obsessive Compulsive disorder

OCD patients seem to gather more information before they will confirm a theory. However, this is not the case for neutral stimuli but only for concern-relevant ones. Inductive reasoning showed different results for OCD, GAD patients, and controls. OCD patients took longest to disconfirm a conclusion when it has been made prior to disconfirming new evidence. They may have a frailer confirmation bias compared to controls.

Generalized anxiety disorder

Patients suffering from GAD did not differ from normal controls on most of the tasks. They might have a weaker tolerance for uncertainty, because of an increase of gathered information before they decide to take on a theory.

Somatoform disorder

Threats related to health issues elicited increased activation of the danger-confirming bias. Every bodily sensation is perceived as potential danger.

Mood disorders

Depressed patients do not differ in hypothesis testing in relation to neutral stimuli. Further, data on information gathering also failed to show differences between controls and depressed patients. One study however, showed that depressed people gathered less information before reaching a conclusion. They also might suffer from deficits in discriminant-learning, which says that they have problems selecting one possible hypothesis among others.

Psychotic disorders

In patients with schizophrenia, there is a tendency towards information-gathering bias. People with delusions searched for less confirming evidence before they made their conclusions. Furthermore they tend to switch between hypotheses more often than controls. However, this was only the case when they were given insufficient information but not when they had all the information needed to conclude. Thus it is not a deficit in working memory. It is also seen that self-relevant and emotional hypothesis testing are poorer in these patients.

Sum-up

Hypothesis testing in general seems to be functional in most disorders, except for schizophrenia. Schizophrenics gather less information before they reach a conclusion. OCD patients on the other hand seem to need more information before they are sufficiently satisfied to conclude. Threat-confirming bias is found in specific phobia and hypochondriacs. Thus it meets the criteria for a possible transdiagnostic process. There is a tendency for general belief-bias across disorders but this need to be tested further.

Discussion

Biases in interpretation of new information seem to be present in most disorders. Attributional reasoning on the other hand is only biased in schizophrenic and depressed patients. A pessimistic attributional style is also observed in sexual disorders, eating disorders, and GAD. Judgment of expectancies seems to be biased among all patients. In regard to the availability heuristic, not enough studies have been conducted to make general statements, but for GAD and panic disorder and depression, it seems to be present. Emotional reasoning appears to be biased in most disorders. Covariation bias has been observed in panic disorders, specific phobias, and psychotic patients. Hypothesis testing needs to be separated into belief bias (found in specific phobia and hypochondriacs) and data gathering (extreme in OCD and not sufficient in schizophrenia).

Definite transdiagnostic processes

Interpretive reasoning meets the criteria for a transdiagnostic process. Ambiguous stimuli are interpreted in a threatening way. When more than one stimulus was used, the bias only appeared for the current concern stimuli. Judgments of expectancy also meet the criteria. The process is most likely related to the use of heuristics. The availability and simulation heuristics are especially affected. As an anxious patient, you have more threatening events readily accessible and thus more negative future occasions can be imagined. Mood also plays a role in availability biases, because of mood-congruency. The last bias that is said to be transdiagnostic despite the fact that it has not been studies excessively is emotional reasoning.

Possible transdiagnostic processes

Attributional reasoning is found in most disorders, except for two. Depression seems to elicit a pessimistic attributional style, while in psychotic patients a stronger explicit self-serving bias, where they are blaming external factors for negative events is observed. Anxiety and eating disorder show a heightened pessimistic style, but it may be due to comorbid depression. The availability also only meets the ‘possible’ criteria, as more examinations are needed. Further there is evidence for a threat confirmation bias in some disorders. Covariation biases and belief/confirmation biases are only observed in a few disorders, thus only enough evidence for a possible process among disorders. However, it is assumed that it is present in all disorders, but sometimes hard to see as results show similarities in non-patient controls. This is the case because covariation bias mostly occurs when an emotional arousal is present for a stimulus. Blood-injury-stimuli elicit fear in most people so everyone shows this bias but also blood-phobic patients. Furthermore, the danger-confirming bias was also found in normal people, as we have the tendency to detect danger quicker than pleasure. It is however a difference between the numbers of objects perceived as threat.

Distinct processes and inconclusive processes

In hypothesis testing, data gathering is important. There seems to be a difference between OCD patients and schizophrenics compared to patient and non-patient controls. OCD patients gather more information before reaching a conclusion and schizophrenic patients less; they are also quicker in switching conclusions. Specific phobias seem to have an increased belief bias; this is not absolute, as more research on other disorders is needed.

May reasoning bias cause disorders?

Prospective studies

It has been found that some expectancies, attributions, interpretations, and illusory correlation can predict future symptoms. Negative expectancies can thus elicit anxiety or another example, positive expectancies about substance-related outcomes can predict future addiction. High covariation bias was observed to be associated with more relapses. Further, pessimistic attribution styles were related to later depression. Attributing a traumatic event to internal and controllable factors, provoke more intrusive thoughts, anxiety, and depression.

Experimental studies

So far the studies taken into account have all been prospective. For causality, experiments are needed to avoid covariations. Manipulations of interpretation and attribution bias brought to light that emotional experiences can be altered. Consequently, these two might be the cause for prolonging mood states. However, intensive training was needed and negative meaning actively created to produce these effects. Repeated use of internal attributions in another study, led to feelings of guilt and shame, attributing the same events to external factors on the other hand, was followed by anger. The same was tested with sexually functional men who got negative feedback about their arousal state, providing either internal or external reasons. Those who blamed external reasons did better when watching another movie afterwards. These studies show that it is possible to induce the same biases in normal people that are experienced in psychopathological disorders, which indicates a causal role of reasoning biases.

Theory

The question that rises is why reasoning biases only affect disorder-relevant issues but not general issues. The most reasonable answer is that the current concerns play a crucial role. They guide attention and memory and thus only the salient features are interpreted and can be falsely attributed. Furthermore, emotions too play an important role for biases. Emotional reactions are responsible for mood states and mood influences reasoning processes. Expectations can influence interpretations and hypothesis testing as we only search for confirming evidence. Negative interpretations are observed to be strongest for ambiguous stimuli related to the disorder.

Clinical implications

By knowing about the causing or maintaining factors, clinical assessment can be improved and patients better understood. Recognition of the biases goes together with prediction about the consequences. As studies have found out the reason for some heuristics, the work with patients is promising to erase some of them. Interventions can be applied in three different ways:

1. Straight away trying to reduce biases.

2. Aim to change beliefs and assumption of the patients.

3. Question the current goals and concerns of the patients.

Cognitive therapy and reasoning biases

This form of therapy targets the first way of intervention. Training (Socratic questions) is provided to see alternatives and to challenge their current reasoning style. Patients are asked to pay attention to other stimuli than they usually do to get an idea of what else could have caused a specific event. Another way is to make them consider all the facts instead of relying on their emotional reactions. Imagery-based technique is another form of reducing biases, it can change the images most accessible. Just like encouraging them to modify an ending of a repeated nightmare over and over again to make it more available than the previous one. Furthermore, going through the course of a situation as detailed as possible, to get a good idea of what is can be like, helps shifting expectation. Behavioural experiments are also applied. First, patients give clear descriptions of what a situation might be like before they are asked to test them in real life. Here the use of video-feedback might be of great help. Instruction must be given to prepare patients. The shift of perspectives can lead to reduced anxiety. The similar strategy is used for eating disorder patients, they need to face mirrors. Here they are helped to bear looking at themselves and to see their beautiful body parts. Internal and external attributions are an important source of disorder, shifting those is another good way to approach them.

De-biasing approaches

De-biasing, as the term indicates, tries to reduce biases. Here patients are taught to consider statistics and the likelihood of their imagined situations. Critical thinking is trained such as making appropriate judgments about uncertainties. Also metacognitive processes are addressed, like re-evaluations of thinking processes. For a gambler, one of the problems is their misattribution of randomness.

There have been specific sessions, for teaching gamblers this principle and to correctly assess the chances of success. A contingency diagram for bad vs. good events, when patients engaged or not in their disorder maintaining mechanisms (e.g. being anxious or eating more), is another fruitful treatment to evaluate the effectiveness of the mechanisms and the likelihood of their thoughts. Furthermore, the ‘all or nothing’ thinking process is not healthy; patients are trained to see more areas than only these two when facing different events. Another aspect that leads to maintenance of disorders is the internal attribution; I am the one to blame. Pie charts are a good source to show all parts that could have had an influence on a specific event. For patients with traumatic experience, ‘chaining’ is helpful to assess the likelihood of a sequence (that, which led to the trauma), by tearing apart all parts of the sequence and calculating the chance of a recurrence of the exact same sequence. De-biasing is an approach to train opposite biases (pessimistic vs. self-serving) but this approach needs further research to show lasting effects.

Targeting goals and concerns

This approach targets the second way of intervention; aimed at challenging beliefs. Motivational counselling is one such treatment. Maladaptive goals are detected and via planning and discussion tried to be restructured.

Future research

The gap between emotional and neutral stimuli among non-patients and patients needs to be further assessed. Hypothesis testing is still in doubt as confirming bias and belief bias have not yet been tested. Decentring might be another impairment in patients with disorders that leads to a confirmation bias for neutral and emotional bias. A topic of interest is also whether shifting beliefs in the presence of disconfirming evidence is different among patients and non-patients. Some studies also show methodological weaknesses (generalisability questionable) when only specific symptoms of a disorder were taken into account (only arachnophobia or drinking). To be certain about associations, more experimental studies need to be conducted. So far, context has never been studied as covariation, especially in real-life settings. Illusory correlations have only been studies in two disorders, so more studies need to be done here too.

Chapter 5: Thought

Thought processes such as worrying about things that could go wrong or about possible accidents in which loved ones could be involved, can be maladaptive and crucial for the maintenance of a disorder. These negative thoughts are also called, intrusions. Metacognitive processes about the thought can lead to strengthening the symptoms. Most patients with GAD believe that their worrying is important for the prevention of bad events.

Intrusions

Intrusions are spontaneous, disliked, uncontrollable, and distinct thoughts, images or urges. Many of those include past memories. Examples of such intrusions are, “I am antisocial” or the picture of oneself, jumping out of a window.

Normal and abnormal intrusions

Everyone experiences such intrusions, also the unwanted ones every now and then. However, in clinical disorder, the intensity of them and the feelings of control over them differ. Furthermore, patients appraisal of and responses to intrusions seems to deviate from normal controls. Meaning is inferred to the content of intrusions, thus they cannot be ignored easily. Repeated dwelling on them as well as attempts to suppress them, are observed in patients and examined to have a negative effect.

Forms of intrusions

Images

Images take on forms of real-life content, which can also capture sensory information that make, them even more ‘real’. Most of the images are visual representation. They evoke emotions and guide behaviour. Their dynamic nature makes it possible to explain sequences of events.

Urges

An urge is a sensual internal desire to engage in something (drinking or washing hands). They are mostly triggered by the presence of an object or related cues. Images of engaging in this action can also lead to an urge.

Current concerns

Every disorder involves specific intrusions. Anxiety patients have images about catastrophes and PTSD patient’s recurrent thoughts about their trauma. They are all related to the specific current concerns. Individuals with substance-related disorders and eating disorders have to fight against urges of taking in large amounts of drugs or food. Stimulus or response cues relevant to the concerns can trigger the onset of intrusions. Intrusions seem to be triggered by and reflect the content of current concerns.

Recurrent negative thinking: worrying and rumination

One way to react to an intrusion is to dwell on it in the attempt to solve the issue. However, this is likely to end in worrying or rumination. Worries include everyday concerns such as “Is my child safe, am I too late” and catastrophic features, “what if I am sick”. Rumination means an uncontrollable, uselessly repetitive, and stereotypical thinking, “Why did this happen, and why me?” Both processes lead to being self-focussed but this self-focussed attention is different to the original, as there is a specific theme present on which they dwell.

Anxiety disorder

Specific and Social phobia

The thought control questionnaire has been established as the measure for rumination. Within the questions there are a few which also measure worry. Patients scoring high on this questionnaire show elevated scores on punishing themselves for their intrusions. Following social interactions, patients with social phobia show more rumination.

Obsessive Compulsive disorder

OCD is especially characterised by recurrent intrusions. Patients ruminate excessively about them which may result in ritualized responses (compulsions). There are specific intrusions for OCD patients, like worrying about contaminations, orderings. People can also suffer from morbid pre-occupation, which is worrying about negative events than could have happened.

Post-traumatic stress disorder

Tendencies to ruminate two weeks after people have experienced an earthquake, was a predictor of PTSD seven weeks later. Further, ruminations about a trauma are high in patients with PTSD. For patients with ASD, the worry scale showed extremely high scores.

Generalized anxiety disorder

A core feature of GAD is chronic worry about anything. It was possible to assess that worrying is mostly verbal in GAD patients.

Somatoform disorder

Panic disorder patients were found to be ruminating about the cause of their pain. They show further worries about the pain and its consequences. This worry is experienced as harder to ignore and more attention grabbing, intrusive, and exhausting than any other non-pain worry.

Eating disorder

Retrospectively, excessive rumination about an event has been found shortly before the onset of bulimia. Furthermore, dieters show far more ruminating about food than a normal-eater.

Sleep disorder

Insomnia patients report intrusions when they are trying to fall asleep. A ‘racing mind’ is attributed for their impaired sleep. Worry has also been shown to be elevated. The content of the ruminations is focused on the inability to sleep and general worries.

Mood disorder

Unipolar depression

It is generally common to ruminate about something, when we are in a negative mood. On the other hand, high scores on rumination predict depression and the maintenance of the disorder.

Psychotic disorder

Schizophrenics and people with delusions show the same scores on worry as GAD patients.

Discussion

Definite transdiagnostic processes

All the data collected are based on questionnaires, showing elevated scores for worry. This means that negative thinking is a definite transdiagnostic process. Whether it is possible to distinguish between rumination and worry is hard to say, as those two are very similar (high correlations). However, a distinction has been made between worries and obsessions. Obsessions are visual, less realistic, egodystonic, and less distracting. An egodystonic belief is one that is not compatible with one’s own beliefs and values. Even though obsessions and worry show differences, they are said to lay on a continuum.

Causal role of recurrent thinking

Prospective studies

In depressed as well as in PTSD, patient’s rumination is tested to be predictable by the levels of rumination one or two years before the onset of the disorders. It also predicts further symptoms.

Experimental studies

When non-patients were provoked to be worried, an increase in depression and anxiety was observed. Exposure to a stressful movie was also tested with elevated worrying and intrusions at least 3 days after watching it. Rumination was also tested to set up induced negative moods, poorer problem solving, and worse memory recall. These effects were only found in patients with depression. All together it can be said that negative thinking has a causal effect on the maintenance of disorders.

Theory

As always, the content of the negative thinking differs between the disorders, reflecting the specific concerns of each of them. Depression is associated with worrying about losses, anxiety with threats, and obsessions reflecting egodystonic thoughts. Recurrent thinking has been established to be a “self-regulatory attempt to resolve unfulfilled goals” (Greenberg 1997). It will continue until the goal is pursued or dismissed. The concept of worry is the attempt to avoid threatening events and to prepare for catastrophes. Rumination is the attempt to make sense of bad moods and ineffective to problem solving. 48% of worries show unresolved problems.

Functional recurrent thinking

• Recurrent thinking and problem solving. These forms are not always dysfunctional, the opposite is true, it can be helpful for problem solving, when its form is appropriate, meaning that it is focused on real, objective concerns. A positive attitude towards one’s ability to solve problems is associated with better outcomes. Defining a problem, creating alternatives to approach it, applying that strategy and the evaluation of its effectiveness, all encompass problem solving.

• Emotional processing. Emotional processing is the practice by which emotions are captivated and reduced to a degree, where it is possible to still function rationally. It allows us to remain capable of absorbing all the information provided by our environment. In most disorders, this processing is said to be declined and thus negative thinking can reach the intrusive extent, which it has. However, elaborating on emotional distress is necessary for successfully working emotional processing. Exposure in vivo show that extreme fear can only be overcome when faced with it. It might be the control of exposure (also in PTSD) which leads to successive reductions of intrusions but for that, professional help is needed.

What are the determinants for functional versus dysfunctional recurrent thinking?

One aspect that is crucial in the application is the extent to which people assume rumination and worry are useful.

• Insoluble problems. The problem itself might be different between normal controls and patients. So, when a problem is too hard to solve, people may be destroyed by it. Trauma patients ruminate about their experiences, which can be distressing for everyone.

• Mode of processing. The verbal rumination may lead to suppression of imagery which in line with the cognitive avoidance theory, may lead to less emotionality. This in turn prevents emotional processing and thus patients are stuck in rumination. The cognitive-style of rumination and worry is abstract. Abstract thoughts are characterized as being of a more spontaneous, indistinct, and unclear nature and harder to correct.

The used concreteness theory of worry, states that the reduced vividness, ease and speed in imagery, results from abstract thinking. All the other features of problem solving are impaired because of the lack of concreteness.

• Conditioning. Another reason for enduring rumination might be that new associations are created, which results in a higher number of triggers. Any internal or external cue can make them ruminate again. This may also account for the heightened attention to threats, found in GAD patients. Another could be that the absence of expected outcomes in patients with GAD or other, may lead to the assumption that worrying prevented the occurrence and thus it is reinforced.

Thought suppression

Another attempt to resolve rumination is by trying to suppress the intrusions. However, as already established, suppression can lead to a heightened availability of a thought (think of a white bear experiment). When participants were asked to not think about a white bear, the frequency, measured by pressing a button whenever they were thinking about one, was increased. This is called a paradoxical enhancement effect. Another effect has been measured; the delayed paradoxical enhancement effect which is a persistently high frequency of the previously suppressed thought, after the task has been stopped. The worry, punishment and distraction strategies used in disorders, are similar to suppression and its effect.

Anxiety disorders

Suppression of thoughts created an immediate increase in intrusions (related to their disorder) in patients with agoraphobia. In social phobia, the attempt to suppress thoughts led to immediate increase in intrusions for fear related and also non-fear related topics. It must be said that these studies were not between-subject designs but rather a sequential design.

Specific phobia

Patients with arachnophobia were found to attempt harder to suppress spider-related issues, however, no delayed enhancement effect was found. When patients suffering from dental phobia were tested, no paradoxical effects could be found either, thus it is assumed that there are none for specific phobia.

Obsessive-compulsive disorder

76% of patients with OCD reported attempts to suppress their intrusive thoughts repeatedly. A paradoxical effect for immediate suppression is observed but no delayed effect. It is assumed that OCD patients suffer from a deficit in inhibition control, meaning that they struggle when trying to suppress intrusions, which leads to even stronger occurrence of them.

PTSD and ASD

One core feature of PTSD and ASD is the suppression of memories about the trauma. A delayed enhancement effect could be detected in victims suffering from PTSD due to rape and accidents. ASD patients engage in more suppression independently of task instructions.

GAD

Patients with GAD were better in suppressing loose thoughts than worries. For non-patients, the opposite was the case.

Somatoform disorders

Pain

When instructed to think about pain, the number of pain-related thoughts rose. In the condition where they were asks to suppress those thoughts, no (delayed) enhancement effect was observed. However, it could have been the case that because patients were given advice how to best suppress thoughts (distraction).

Eating disorders

No differences were found between attempts of thought suppression in high risk patients and lower risk patients. However, this may be because high risk patients are more practiced in suppressing thoughts and thus no delayed enhancement could be found.

Sleep disorders

Patients suffering from insomnia, engaged in suppression and worry more often than normal sleepers. Suppressing thought prior to sleeping worsened the quality of sleep in both patients and normal controls.

Mood disorders

More severe patients try harder to suppress their negative thoughts; however, they are also less effective in doing so. They further try to distract themselves from one negative thought, by focusing on another negative one. This in turn makes it almost impossible not to think negatively. Inducing cognitive load revealed an increase in negative thinking in people who are at risk for depression.

Discussion

Possible transdiagnostic processes

Thought suppression is present in all disorders, however, it is not clear whether the consequences of it are the same across all the disorders. In PTSD, agoraphobia, and insomnia, suppression of thoughts led to paradoxical enhancement of intrusions. In social phobia, suppression created an increase in all kinds of intrusions. For other disorders such as specific phobias, no adverse effects were observable, or not enough studies had been conducted to conclude.

Is suppressing thoughts a cause for disorders?

Prospective studies

In PTSD and depression, thought suppression seems to predict the onset of the disorders as well as the severity of them.

Experimental studies

Most of the effects stated when discussing the different disorders, were found via experiments. Further manipulations also showed similar findings, that suppression influences symptoms. More negative thinking or pain related thoughts were observed, when healthy people were instructed to suppress negative thoughts or pain when pain was induced earlier.

Theory

Most studies investigated whether thought suppression of concern-relevant topics led to paradoxical effects. It is important to know, if suppression of general thoughts would also lead to an increase of these thoughts. Only then we could draw conclusions from the results of current studies.

There are a couple of factors that seem to be related to paradoxical effects:

1. When people were focusing on one distracter rather than on several, suppression was less successful.

2. Changing moods or context between the suppression task and the expression one, reduces the effect.

3. Increasing cognitive load, increased the effect, as the two processes (monitoring for unwanted thought and operating –thinking about a distracter) involved in suppression cannot work properly anymore (operation too occupied to control for the monitored thoughts).

So taken together, the environment and its cues, as well as feelings and the range of distracters, play a role in the paradoxical enhancement effect.

Disorders that are characterized by very specific concerns and intrusions only have a limited range of associated cues that may lead to trigger the intrusions. Disorders such as GAD, have a wider range of worries and thus of associations, which make it harder for them to successfully suppress the thoughts (more paradoxical effect). In anxiety disorders, fear can be seen as cognitive load and thus higher scores for paradoxical effect expected. In disorders like depression, the loss of motivation may account for less cognitive capacity and an increase of the adverse effect of suppression is also likely.

Metacognitive processes

We have discussed rumination and suppression, so which is chosen, and when? It is assumed that the type of processing depends on the evaluation of the intrusion. Evaluating a thought is what is called metacognition. Controlling, monitoring, and judging thoughts are all a part of this cognition. It is believed to play a role in sustaining a disorder. There is metacognitive knowledge and regulation. The first is the knowledge that people have about their own thinking processes (worry helps preventing danger). It also takes on the form of procedural rules or routines. Metacognitive beliefs can be positive (rumination will help solve a problem) or negative (worry will make me go crazy). However, there are more than those two beliefs, likelihood thought-action fusion beliefs (specific thought have influence on direct happenings), and morality thought –action fusion (thinking about… is like doing something bad). The Meta-Cognition Questionnaire tests all these kinds of beliefs.

Metacognitive beliefs

Anxiety disorder

OCD

Intrusions in these patients are evaluated as personally significant. This leads to feelings of responsibility over their thoughts. Studies show heightened responsibility in low-risk situations but normal appraisal in high-risk situations that are irrelevant to their intrusions. This displays an increase in though-action fusion beliefs. Further intolerance of uncertainty was observed in OCD patients.

PTSD

The degree to which patients find intrusions distressing depends on:

1. Their feelings of control about them and how unwanted they were.

2. If they appraise the intrusion as idiosyncratic negative (I am going crazy over these flashbacks).

Studies show elevated feelings of danger in PTSD patients when they experience intrusions.

GAD

Patients with GAD judge worrying as useful for solving problems and to avoid negative events. However, they are also tested to have more negative metacognitive beliefs than other patients. They worry about worry; this is termed Type 2 worry. They also show an elevated likelihood of thought-action fusion beliefs and intolerance of uncertainty. The latter might be the core reason of developing GAD.

Somatoform disorder

Hypochondriacs were found to have more negative metacognitive beliefs.

Eating disorder

In these patients no direct measure of metacognitive beliefs has been done. However, some other measures, also including metacognition, showed thought-action fusion metacognitive beliefs in relation to thoughts about food, shape and weight. They were seen as unacceptable, increasing actual weight or at least the feelings of being fat. An association was found between the beliefs to control weight and food and self-acceptance and feelings of anger.

Sleep disorders

Patients with insomnia rated worry in bed as more useful. There was a wide gap observed in what patients thought worry before sleeping could hold and what was actually achieved if they did so.

Mood disorders

Positive as well as negative metacognitive beliefs are found in depressed patients. Positive evaluation of worry was even higher than in panic disorder, after a depressed episode.

Psychotic disorders

Patients with acute psychosis show more negative beliefs about worry. It is also associated with increased feelings of danger and uncontrollability of those intrusions. Also a greater cognitive confidence was found in these patients and in schizophrenic patients. Patients with schizophrenia also showed more negative metacognitive beliefs when they had hallucinations.

Metacognitive regulation

Negative thinking and suppression of thoughts are both mechanisms related to metacognitive regulation. Decentring and metacognitive awareness are two further processes that are crucial for emotional disorders. They both regulate the recognition of thoughts versus reality. Decentring is also applied in cognitive therapy and mindfulness training. Another aspect of metacognitive regulation involves monitoring and controlling one’s own actions. OCD patients seem to be impaired in the monitoring of memory performances. The lack here may explain the lack of confidence of their memories and that in turn clarifies their checking behaviour is far more frequent. It is proposed that they even lack a reality monitoring, stating that the source of memory (imagination vs. reality) cannot be determined. This has also been observed in schizophrenic patients.

Discussion

Definite transdiagnostic process

There is enough evidence across all the disorders to regard positive and negative metacognitive beliefs as definite transdiagnostic processes. It is further likely that these processes play a role in persistent rumination and worry. Recurrent thinking has already been defined as definite and thought suppression is the next metacognitive regulation strategy that meets these criteria.

Distinct processes and inconclusive evidence

Less evidence is provided for other regulation strategies. However, metacognitive awareness was found as declined in residual depression. Memory confidence is decreased in OCD patients and reality monitoring impaired in OCD and patients with schizophrenia. The deficit in source monitoring might be a unique process in psychotic patients.

Is metacognition the cause for the maintenance and onset of disorders?

Prospective and experimental studies show that metacognition can be the reason for symptoms. The strategies taken on determine the maintenance of the disorder.

Negative interpretations of intrusions shortly after a trauma predicted PTSD one year later.

Inducing thought action fusing (letting participants write down: “I wish he was in a car accident”) led to more anxiety. When participants knew about that sort of belief, anxiety decreased. Another experiment showed that helping depressed patients monitor their actions and metacognition led to better problem solving.

Clinical implications

The findings that recurrent thinking and thought suppression have a direct influence on symptoms, makes it clear that treatment to resolve these maladaptive processes is crucial.

Recurrent thinking

There are three interventions that might be most efficient:

1. Changing cognitive avoidance into approach towards thoughts and feelings.

2. Trying to concretize processing.

3. Identifying and then reconstructing maladaptive metacognitive beliefs.

The first two interventions are thought to be best achieved via:

• Repeated exposure, combined with relaxation practices.

• By using vivid and detailed imagery to elicit emotional responses.

• It has also been proposed that fixed worry times can help regaining a feeling of control. It also breaks the endless cycle of worry and decreases the importance of them.

• Talking or writing about worry also helps as emotional responses are set free, breaking the rumination. This sort of integrative therapy (releasing emotional/interpersonal stress), serves to experience one’s own deep feelings, behaviours associated and the impact on others. They re-learn how to handle deep feelings and become aware of the effect this has on the people around them.

• Another way to help them break the cycle is by focussing on a more action-oriented way of thinking. Therapists can encourage using more process-oriented question (How) and replacing them with the entire Why questions. Socratic questions are also of good use to break down automatic negative thoughts.

• Finally, goals need to be identified and stated more clearly, realistically, and independently from other goals.

Thought suppression

So far we know that:

• Patients need to be educated about the native effects of suppression.

• Some forms of distraction are more effective than others.

• Reduction of suppression will lead to less intrusion.

The first one is easily explainable by the use of the “White bear” experiment. The second depends on the disorder and the type of intrusions they are experiencing. For those with intrusive images, a verbal distracter is more useful and vice versa. One consequence of distracting properly is that the fear of losing control is decreased. Thirdly, it has been tested that thinking of a good event is more helpful than trying not to think of a negative one. The ironic process is less active that way.

Metacognitive beliefs and strategies

Analysis of antecedents, consequences, and functions of metacognitive strategies need to be done to find more clarification for these processes. Behavioural activation approach focuses mainly on the functions and consequences of thoughts, rather than on its content. Past experiences are taken into account to appraise the likelihood of using any one of the strategies. They are seen as coping styles. Socratic questions can be used to assess the use of the strategies and they also help to make people aware of alternatives. That way, one can test explicit metacognitive beliefs and metacognitive procedural rules (when, where, why) can be established for each patient. The analysis can further predict that a change in environmental eventualities and a switch from avoidance to an approach direction are helpful. Mindfulness based therapy is another aid to overcome extreme worry and recurrent rumination. A third method is the Acceptance and Commitment therapy, which emphasise accepting one’s own thoughts and emotions.

Future research

Thought processes have only been scarcely investigated in sexual, bipolar, panic, and substance-abuse disorders. To call strategies transdiagnostic further investigation must done. Furthermore, additional experiments are need, especially in metacognition, where there have been mostly correlation studies, to assess the causality in more detail. Much of the current research is based on self-report data, so bias cannot be excluded. People might not be aware of what strategy they are using and thus, self-measures cannot assess them. Experimental manipulations must be made to be certain about the effects. A final need for future research is the fact that it has still not been agreed on the interactive effects of rumination, thought suppression, and intrusions.

Chapter 6: Behaviour

Carrying loads of stuff such as extra plastic bags, touching religious objects or avoiding seeing the doctor, are all examples of safety behaviour, patients with specific disorders engage in. Avoiding situations or things are the most prevalent. Patients still engage in these behaviours, despite the absence of symptoms. Cognition, emotions and behaviours are all thought to be closely linked with each other, according to cognitive behavioural theorists and therapists. Behaviours can influence disorders to the same extent to which cognitive dysfunctions can maintain a disorder. A threat might be detected first and misinterpreted as dangerous which then lead to flight reactions but avoidance of these places or fearful object is what follows such an experience. Avoidance can thus strengthen unhelpful beliefs. The behaviours which are most likely associated with disorders are further explored. These are: avoidance/escape, within-situation safety behaviours, and ineffective safety signals. Not all the behaviours are seen in disorder patients but instead serve crucial functions such as checking the oven plate or the street for cars. However in the absence of real threats, these behaviours are maladaptive and may worsen the patients concerns.

Escape and avoidance

There are several reasons why avoidance behaviour can be harmful. It extinguishes the possibility to disconfirm bad beliefs, it prevents patients from experiencing positive reinforcements, and negative moods might be chronic, further on, it reduces the person’s possibilities and social life and narrows the focus of attention down to one’s self (rumination). It also serves to increase the fear of stimuli and decreases feelings of control.

Clinical implications for avoidance behaviour

In clinical settings, therapists try to decline the avoidance behaviour step-by-step. The patient has to stay in a fear-eliciting situation, so that the anxiety can decrease and the avoidance behaviour will not be reinforced. Remaining in such a situation for a longer period of time, is needed for cognitive reappraisal. However there are contradictory findings for this assumption that do not show a difference between anxiety levels of those prolonged exposed to a fearful situation and those who were allowed to leave.

Distraction as a form of avoidance

Distraction is a more subtle form of avoidance. One study shows that there is no difference of reported somatic sensations between patients with panic disorder who engaged in distraction tasks and those who were focusing on the symptoms. Thus no advantages but also no disadvantages are suggested for distraction via exposure. This is in line with the cognitive hypothesis, which proposes the occurrence of a harmful impact of distraction only when the disconfirmation of bad beliefs is no longer possible.

Within-situation safety seeking behaviour

During sessions of exposure, therapists noticed behaviour that helped patients bear the fearful situation. Even though it helped them experience the full extent of their fear, it also prevented them from recovering.

The cognitive hypothesis

Safety behaviours prevent recovery from fears by extinguishing possible disconfirmation of the feared object or situation. Thus a panic attack (e.g. in panic disorder) is always perceived as ‘nearly’ occurring, just prevented by the pill a patient took or any other behaviour. That pattern can endure over ages. Safety behaviours can also be intrinsically dysfunctional because of the time spent thinking about it or engaging in a task to avoid a feared outcome (washing hands, dieting).

The safety signal hypothesis

Mowrer’s theory about avoidance is divided into two stages. The first stage includes classical conditioning, relating fear responses to conditioned (fear/pain induced) neutral stimuli. The second stage involves the appearance of avoidance behaviour. The theory involved danger-signals (all stimuli that had been related to pain) and safety-signals (those related to pleasure inducing stimuli). The theory was revised; however, these safety signals were further explored. The safety-signal hypothesis as opposed to the cognitive hypothesis, states that avoidance behaviour can persist even when no fear is present due to conditioned feelings of pleasure (absence of fear). In line with that hypothesis, safety signals are regarded as help during exposure sessions and would not lead to strengthening fear.

Current concerns

Some stimuli which seem more fearful than others, are not avoided, while seemingly harmless ones are. Why? Another question: why do compulsions and obsessions change over time in some patients?

Avoidance and within-situation seeking behaviour across disorders

All the disorders will be looked at; however, there is a greater extent to which we focus on anxiety, as most of the research has been done on that disorder.

Anxiety disorder

Panic disorder with and without agoraphobia

One criterion that leads to a diagnosis of panic disorder is avoidance. The cognitive hypothesis was supported by a study that tested the cognitive associations between moments of panic and several behaviours (safety and neutral behaviours). It revealed that when thinking of panic, they associated more of the safety seeking behaviours as being helpful to the situation. A more drastic experiment tested the use of safety seeking behaviour as they asked some patients to drop them during a 15 minute exposure and others to maintain them. Results showed that those who dropped them had a decrease in fear. Evidence for the safety hypothesis is provided by an experiment that induced CO2, which raises the heartbeat, in the presence or absence of a safe person. It was more distressing for the patients who were not allowed to have a person in the room. Still, there is evidence that over long-term periods, the presence of a safe person could maintain a disorder, as disconfirming evidence is prevented. There is a compromise of these two approaches which seems to have the best results; sole exposure to a feared situation but with the knowledge that a safe person is waiting inside (e.g. a supermarket or at a party).

Specific phobia

The most common treatment for specific phobia is exposure, 75% of patients have an improvement during the first session. One study tested:

1. Guided threat and reappraisal of the core threat (focus on threat and look for disconfirming evidence).

2. Safety behaviour utilization (option given to engage in safety behaviour).

3. Control (no instructions given) with weaker improvement results for those in the 2 and 3 condition.

A second study with claustrophobic patients supported the other hypothesis, that looking for safety signals can help patients relax more. In line with that, the claim that claustrophobic patients do not attend sufficiently to safety signs was confirmed. Another coping strategy for patients with specific phobia, here blood-injury, is to apply tension to your muscles before facing blood, which helps to prevent fainting when exposed to the ‘threat’. Whether or not a coping strategy is dysfunctional depends on the degree to which a threat is realistic (ants or snake) and to which a safety behaviour is a real value (opening a door, if the room feels too close).

Social phobia

The types of safety behaviours used in social phobia are related to avoiding direct social contact (avoid eye-contact, keeping hands busy to avoid handshaking etc). It is almost ironic to see, that these are the behaviours that lead people to suspect some weirdness about a person which is what those people fear most. In these patients it was more effective when they dropped their safety behaviours during exposure. It is a crucial and difficult part in cognitive therapy trying to bring social phobic patients to drop these behaviours.

Obsessive Compulsive disorder

Compulsions are real behaviours or mental acts which are perceived as preventing bad events in the absence of any real connections. They can be called safety behaviours as they serve to make patients feel save again. Exposure (even if it is mentally) is effective when these compulsions are stopped. Asking patients to drop these behaviours while being exposed to dirt (when contamination fear was present), led to a sudden increase of fear but a decrease over time and different sessions. Mental neutralizing, which is explaining the absurdity of their beliefs of a connection or responsibility, is also important for OCD patients.

Post-traumatic stress disorder and ASD

In order to resolve a trauma, patients need to activate and modify all parts of the event. However, to avoid the strong emotions associated, they try to dissociate. This safety behaviour maintains the disorder, it is still perceived as current threat. Emotionally laden images will unexpectedly occur to them.

Generalized anxiety disorder

It is assumed that continuous worry is also a form of avoidance. The abstract, verbal worries prevent focusing on one strong affect. Parallels to agoraphobia are present; they both balance between safety and threat signals. Furthermore, GAD patients do not use safety signals efficiently. As opposed to other disorders, they persistently focus on the threat instead of avoiding it.

Somatoform disorder

For patients with pain disorder, avoidance is displayed in daily activities as re-injury or excessive pain is expected otherwise. However, the avoidance of movements and stimulations also socially, can lead to worsening of the symptoms. Reduced activity was found to be related to decreased fear of pain but also with catastrophizing about it in comparison to usual activity. In hypochondriasis, patients engage in avoidance, reassurance seeking, and body checking behaviours. All of these are assumed to strengthen the symptoms. Paying exacerbated attention to inner sensations activates health concerns and interferes with experiencing pleasure. Finally, body dimorphic patients are also characterised by avoidance behaviours. They do not attend social situations as they fear others could detect their ‘weirdness’. Safety behaviour is for instance wearing extreme make-up, which might have the feared effect of being looked at. This in turn is not attributed to the make-up but to the perceived abnormality of them.

Sleep disorder

Patients with problematic sleeping engage in many safety seeking behaviours (daily napping, taking a day off after sleepless night and worrying), which have adverse effects.

Eating-disorders

A principal part of this disorder is avoiding food and any kind of weight gain. Excessive binging can also be seen as cognitive avoidance from bad feelings or thoughts. Safety seeking behaviours such as body checking, weighing, or dieting are also frequently observed.

Unipolar depression

People suffering from depression show a decline in pleasure and sport activities. This is believed to be a core factor for the maintenance of the disorder. Bringing social interactions and activities back in their lives has been found to be an effective treatment. Replacing avoidance with approach behaviour is crucial in these patients to overcome the disorder.

Psychotic disorders

Delusional patients were examined and avoidance behaviour tested to be the most common practice. It is further associated to higher levels of anxiety.

Substance-related disorders

One motive why people engage in taking drugs or drinking is because they are trying to forget or stop strong feelings (escape from awareness hypothesis). Manipulation and avoidance of mood is the core aim. A study that induced emotions into patients supported that hypothesis, as it showed that these elicited cravings.

Discussion

It must be said that the quality of evidence provided has to be improved for further claims about behavioural processes. Nevertheless, it was possible to define safety seeking and avoidance behaviour as transdiagnostic. Ineffective use of signals was found to be a possible transdiagnostic process.

Definite transdiagnostic processes

Avoidance behaviour is a core feature for some disorders (e.g. phobia and PTSD). In specific phobias, PTSD, pain disorder, unipolar depression, and psychotic disorder, avoidance behaviour could be observed and therefore it met the criteria. For safety behaviour, results show that in every examined disorder, this process was present which also makes it a definite disorder in anxiety disorders, BDD, eating disorders, insomnia, and psychosis. The content might be different for each disorders but the behaviour is the same.

Possible transdiagnostic processes

The safety signal hypothesis, applied to agoraphobia and GAD has been supportively proven, however more studies on different disorders are need to confirm a transdiagnostic process.

The causality of a disorder in relation to behaviour

Various studies supported the assumption that avoidance is predictive for symptoms of PTSD. Fear avoidance beliefs, e.g. in pain, may also lead to more pain. Experimental evidence is present for the causality of safety behaviours in the maintenance of a disorder. That it can prevent fear reduction is still a matter of debate.

Theory

There is a current concern hypothesis, stated by Klinger (1996) which says that people avoid those situations tied to their current concerns. This might shed some light on the differences between the disorders. Safety behaviours are mostly idiosyncratic, meaning they have a purpose for the person. This theory also explains why some safety behaviours change over time. One example of this fits the fear of contracting HIV. Before 1980, nobody knew about this illness, but as it got published in the papers, the rate of OCD patients afraid of contamination with HIV rose drastically. It must still be considered that people believe engaging in some behaviour can prevent bad situations, thus therapy focused on changing beliefs. However, changing behaviour (making patients drop safety objects) could also lead to fear reduction, according to the cognitive hypothesis. The safety hypothesis on the other hand makes use of safety behaviours in exposure settings which also showed positive results. The first focuses on elimination of safety signals, the latter on integration of them. Even though it appears as two different models, first, safety signals could be seen as coping response that are not believed to be the reason for the absence of catastrophes but rather as way to relax. Secondly, safety behaviours could also be useful when they:

• Prevent a real feared consequence (watch the street).

• Can be applied reliably (in every situation).

• No adverse consequences happen due to it.

Applied tension also meets all of those criteria.

Clinical implication

It has been established that exposure is not explicitly necessary for fear reduction but is still very effective in treatment. The reason for that is not clear yet. The two contradicting hypothesis both have their advantages and should be applied together in a treatment. During the first sessions, safety signals might be of great use to patients when facing their greatest fears. After a while however, they must go one step further and drop those behaviours too in order to fully disconfirm their unhelpful beliefs. In cases of social phobia for example, situations that have been avoided can be facilitated by imagery and/or role-playing. The therapist can help the imagining process by asking concrete questions about each stage. Moreover, when patients imagine counterproductively, one can interfere and suggest an alternative. After this exercise, role-playing can help to strengthen the new ways of approaching a situation. Behavioural activation in disorders can also be achieved by making lists of what activity they used to do and to what degree they enjoyed doing it. Further, they rate how difficult is seems to engage in that activity again. The one that is least difficult to start is practiced first. Depressed mood is evaluated before starting therapy and after each activity has been brought back into their lives.

The Acceptance and Commitment Therapy is another strategy with which to approach disorders. Here patients are demanded to focus on the effect of fear and every deep feeling. Unpleasant emotions are taught to be carried and accepted. Avoidance and cognitive fusion are said to be responsible for the disorder. Cognitive fusion is the belief that one’s thoughts (fear) reflect reality. The therapy is promising but further investigations needed.

Future research

Even though the effect of behaviours in the maintenance of disorders is well known, research is only weakly represented in this area. Additionally, many of the studies lack experimental basis. Well conducted experiments are needed to ground the two hypothesis and their predictions. Further questions such as: Does safety behaviour work in the same way across disorders; are there times that make using safety behaviours helpful etc.? The specific topic of each disorder related (safety) behaviour needs to be explored with the questions in mind why some of them are idiosyncratic and others prototypical? Also clinically important questions, if there is use of safety behaviours and if not, when is the time to drop them?

Chapter 7: Conclusion

The aim of this book was to point out the similarities of the disorders in regard to underlying processes. As most previous research on psychological disorders has been done on the adult Axis I disorders, we also took those into account for comparison. 12 definite, 9 possible, and 2 distinct processes were found for attentional, memory, thought, reasoning, and/or behavioural processes. The approach of comparing disorders is termed to be due to the functional diagnostic dimension. However, the dominant approach is still disorder-focused. Advantages are the clear evidence for an understanding of the disorder and matching treatment. Disadvantages, such as a lack of usefulness for patients with comorbid disorders and research on treatments being slow, as only one disorder is taken into account.

Reasons for differences

The assumption that there are the same cognitive and behavioural processes in the disorder leads to the question of why there are still so many different features of the disorders that make them distinct? It might be due to:

• Current concerns.

• The balance of different processes is different.

• The distinct processes are only present in these disorders.

These three reasons are all likely to contribute to the solution of the question.

Current concerns

Current concerns guide our attention and memory to certain stimuli, they further influence the way things are interpreted and the topic of thoughts moreover, and they determine the behaviour of a patient. This is the answer to the first question about the distinct features. The next; why do people have different concerns? For this it is believed that biology, personality, learning history, trauma, and culture play a role.

Balance of common processes

The assumption that goes along with the continuum belief, is that every person with a disorder can experience the five processes at any point along the continuum. Different disorders might be characterized by the similar severity of process impairment. Patients with social phobia score highest on the self-conscious process while patients with OCD only have intermediate levels.

Distinct processes

Besides the similar underlying processes, we have also found distinct processes for some disorders. It might be those processes that make a disorder specific. It must also be said that even if the criteria for definite transdiagnostic criteria were met, the processes do not have to be present in all disorders. While distinct processes could also account for a few disorders.

Sum-up

Despite the similarities in impaired processes, the current concerns, a balance of impairment, and distinct processes may account for the differences in disorders.

Comorbidity

It is a fact that most patients are not only diagnosed with one but rather with two or more disorders at the same time (average of 2.1). One reason to explain this is an overlap of maintaining processes. There are two ways this happens; first, one maladaptive process may maintain another disorder (not looking at other people –social phobia- will also prevent disconfirmation of being attacked –PTSD-), second, one disorder may increase the chance to set off another maladaptive process (focus on internal states leads to attributing arousal to external threat).

Clinical implication

It is also good to base a treatment on evidence based approaches. CBT has been well established and the content-focused approach shows promising results. However, there is also a use for process-focused approach when it comes to treatment. This way one can target the processes that maintain the disorder directly. It could be used for patients who:

• Do not show improvement with other treatment.

• For those where no exact treatment can be applied due to a lack of diagnosis.

• For complex cases where the primary disorder is not clear cut.

Process as a barrier to change

Another reason why some people do not react to therapy is that one of the processes seems to serve as a barrier. It is perceived as impossible to change behaviour or thoughts. Rumination has been detected as a barrier in depression. Depressed people who still engage in rumination during therapy show lower levels in recovery. The processes may interfere with the therapy as the communication might not be troubled as no full attention is paid or the relationship to the therapist is weak because of distrust. It might thus be useful to integrate process-focused and evidence-based approaches. This is presented here, but it must be noted that it is highly speculative.

Assessment and formulation

During the assessment, it can be useful to ask for underlying processes.

1. Is the process present?

2. How is the process relevant to the present problem?

3. Is the process likely to act as a barrier to change? If so, how?

Diary-keeping or behavioural experiments (going to the supermarket) might be useful to detect the processes. It is necessary to explain the bad outcomes of these processes to the patients to make them understand the need to change them. Therefore all processes need to be noted. This can be done via:

• Analogies (disorganized cupboard).

• Real world examples (supermarket visit).

• Questions.

• Behavioural experiments (monitor sensations with and without distractions), measures before, during and at the end of a treatment.

After these processes have been identified, an intervention can be formulated based on reducing these processes.

Treatment

The treatment must always be based on a trusting relationship and all the core elements of CBT. Moreover, a personalized formulation of the problem, a well evidenced basis, Socratic questions, guided discovery, monitoring, diary keeping, and educating about the processes. If there is more than one process that maintains the disorder, one needs to find out the primary one that causes the disorder. This is not always so clear cut so the patients also need to be asked which he/she wants to fight first. Furthermore, it can be established which process is mostly related to the current concerns and treat those first as they are clear and thus easier to solve. This in turn can encourage the patient for further work. It is very likely that by working on one process, another one or two are modified too.

Limitations

The treatment approach just explained needs to be empirically verified. There are more limitations to the work on the transdiagnostic approach. It is focused on adult Axis I disorders, a definite criteria can be achieved if only one disorder category is taken into account (anxiety disorder) and most importantly limited research has been done, and not on all disorders.

Future directions

This book focuses on maintaining processes; it is wished that the same will be done for precipitating and predisposing processes. Causal relationships need to be researched and by that more studies have to be done on current concerns. Furthermore, we came up with three possible answers to the differences between disorders. Those need to be tested.