Project 5: Neuropsychology

Wagner et al.
Cognitive deficits and cognitive biases underlying delusional symptoms and therapeutic change

Background and Aims
Cognitive behaviour therapy (CBT) in schizophrenia needs to be informed by knowledge about cognitive limitations and specific cognitive biases of patients with schizophrenia for two reasons:

1) Well described cognitive limitations or deficits, eg. of declarative memory and attention span, arise from enduring (trait) and transient (state) neurobiological alterations in schizophrenia patients, with considerable variation being present between subjects.

Such cognitive limitations
a) underlie the development of specific symptoms (e.g. delusions, Gilleen & David, 2005),
b) are correlated with the degree of insight (Keshavan et al. 2004), and
c) are limiting the success of any therapy which, like CBT, is based on verbal learning and requires sufficient attention.

2) In addition to quantitative cognitive deficits, qualitative cognitive biases, like e.g. the false attribution of intentionality (hyperintentionality), or the salience of seemingly minor details, can be assessed with experimental psychological techniques. In general, psychological theories of delusions emphasize either altered perceptions (like illusions), altered attention (selectively attending to evidence in favour of the delusions), and particularly disturbances in making inferences (Bentall, 2003). Inferences may relate to the intentions of other minds (theory of mind) and to event causes (causal attributions). Theory-of-mind skills and attributional style together with social perception (i.e. social cue perception and facial affect recognition) are considered as the main sub-processes of social cognition (Pinkham et al. 2003). However, it has never been studied whether specific interventions, if successful, alter these biases towards normality (Blackwood et al. 2001), and whether the success of CBT critically depends on cognitive and social-cognitive skills.

While the present evidence suggests that CBT in schizophrenia may be both feasible and effective, almost no accompanying research has addressed the interaction of neurocognitive deficits and cognitive biases with therapeutic interventions. Zimmermann et al. (2005) recently emphasized that the role of neuropsychological deficits as a moderator of CBT effects on positive symptoms in schizophrenia deserves further study.

The proposed project, together with the linked projects on neuroimaging and event-related brain potentials, is aimed at studying such interactions in order
1) to predict which patients profit from a relatively costly intervention,
2) to measure the impact of a CBT intervention on cognitive mechanisms which are putatively linked to delusional and other symptoms of the disease, and thus to test hypotheses of mediating mechanisms underlying therapeutic success.

We expect, and will test statistically
1) that CBT will specifically affect cognitive biases (attributional style, cognitive impulsivity, and recognition of intentions),
2) better cognitive performance at baseline is related to better outcome in all groups, and
3) cognitive biases and the change of cognitive biases are unrelated to basic cognitive capacities (attention and verbal memory).
4) cognitive biases are related to delusional symptoms at baseline.

1. For the assessment of cognitive limitations, a short and reliable neuropsychological battery will be administered before and after therapy. This battery, similar in scope to many others, has been widely employed in the BMBF funded German Research Network on Schizophrenia (Häfner et al. 2003) and a substantial data base on the relationship of cognitive deficits with psychopathological symptoms, course of illness, and response to therapy has been gathered (n > 150 first episode patients). A similar battery has been employed to measure cognitive effects of different classes of atypical antipsychotics (Wagner et al. 2005) and cognitive deficits in putatively prodromal subject (Wagner et al. 2004).

2. The assessment of cognitive biases or cognitive styles is an emerging field of research, and several experimental studies have found evidence of specifically altered performance in delusional subjects. Therefore, the sub-processes of social cognition will be assessed in detail in the present study in order to investigate such mediating mechanisms.
The attributional style of paranoid patients, which can be assessed with the Attributional Style Questionnaire (ASQ) is altered in that they, similar to depressed patients, make global and stable explanations for negative events, but, unlike depressives, they preferentially assume external causes, and particularly other people to be responsible (Kinderman & Bentall 1997, for review see Bentall et al. 2001). Another aspect of disturbed inferential thinking relates to the evaluation of hypotheses. In tasks requiring one to make a good guess based on prior evidence, paranoid patients jump to conclusions prematurely, as if they need less evidence to be sure (this style has been termed epistemiological impulsivity). One method to assess such a reasoning style is the beads in a jar task of Garety and colleagues (1991), which will be adopted here as computer task to study patients before and after therapy. The Reading the Mind in the Eyes task (Baron-Cohen et al, 2003) will be used to assess the recognition of intentions. Standardized pictures from the Pictures of Facial Affect set (PFA, Ekman & Friesen 1978) will be used to assess facial affect recognition. These tasks will also be presented by PC.