Introduction: Cognitive models of psychosis highlight the role of underlying differences in cognitive function and information processing in the development and maintenance of psychotic symptoms. As a result there is now an interest in developing a greater understanding of these cognitive changes, in order to guide the development of evidence-based therapeutic interventions. An influential cognitive model of psychosis suggests that the core underlying cognitive difference in psychosis may be one of altered contextual processing. Recent work has suggested that this may be reflected in differential perception of visual illusions. However, it is not clear if such differences are present early in the development of psychosis. Such differences have also been reported to be associated with cannabis use. Further, it has been suggested that, in addition to being risk factors for the development of psychosis, psychoactive substances such as cannabis may provide a useful model for understanding psychosis. The current thesis thus investigated, in two separate studies: (1) the consequences of cannabis use on contextual visual processing and (2), whether reported contextual processing differences in psychosis are present at illness onset. Study One: Two main hypotheses were tested. A. That THC, a key cannabis compound would reduce contextual visual suppression as measured using the Chubb illusion, and that this effect would be reduced via pre-treatment with another cannabis compound, cannabidiol (CBD). B. That THC would transiently induce symptoms of psychosis and that this increase would be reduced via CBD pre-treatment. No evidence was found to support the primary hypothesis. However, the secondary hypothesis was supported by the data. Study Two: The primary hypothesis was that contextual visual suppression, again measured with the Chubb illusion, would be reduced in patients with first episode psychosis relative to a control group. Although not significant, the data supported this hypothesis. Discussion: The results of Study One indicate that THC does not reduce visual contextual suppression as measured by the Chubb illusion. This is in contrast to evidence from other illusions, and may reflect different neural mechanisms underlying contextual visual processing. However, the study provided clear evidence that THC can induce psychotic symptoms and that this effect can be reduced by CBD pre-treatment. Study Two replicates previous findings of reduced contextual processing in psychosis and provides evidence that this may be present from the onset of illness. These findings are discussed and interpreted with regards to study limitations, clinical implications and future work.

Although cognitive deficits are recognised as a core feature of schizophrenia, their trajectory over the course of the illness is still debated. The primary objectives of this thesis were: (i) to determine the neuropsychological profile at the first episode of psychosis (Chapter 3); (ii) to examine gender differences in the profile of neuropsychological performance (Chapter 4) and (iii) to determine if cognitive abilities decline, remain static or modestly improve throughout the course of psychotic illness (Chapter 5). The first (baseline) analyses form part the AESOP (Aetiology and Ethnicity in Schizophrenia and Other Psychoses) study, a population based, case-control study of first-episode psychosis recruited over a three-year period between September 1997 to August 2000 in London, Bristol and Nottingham. The second analyses (follow-up, where I followed 108 patients and 103 healthy controls) also form part the AESOP study, conducted 6 to 10 years afterwards but only in London and Nottingham. In this thesis I found that early in the course of psychotic illness cognitive deficits are present in all psychotic disorders, but are most severe and pervasive in schizophrenia and least pervasive in bipolar. There was strong evidence for gender differences in neuropsychological performance but these differences were disorder specific. Gender related factors appear to moderate the severity of cognitive deficits in bipolar/mania and depressive psychosis patients. Future schizophrenia patients had an early static developmental impairment on measures relying on knowledge acquisition (verbal IQ). This static deficit increases in size starting in late adolescence or early adult life. There was also an increase in deficit on memory functions. Future bipolar patients had normal or slightly above normal knowledge acquisition scores (verbal IQ) which started to deteriorate starting in early adult life. Thus, both schizophrenia and bipolar/mania patients show dynamic changes in general and specific cognitive functions which start early in childhood and continue across the life span. Both depressive psychosis and other psychotic disorders show different age related changes than schizophrenia or bipolar/mania. However because of small sample size and lack of studies examining the premorbid period therefore it is difficult to provide a life course model.

Playfulness offers the potential to change perceptions of children with profound and multiple learning disabilities (PMLD) from passive children who are 'done to' to children who are inherently playful, have personalities, strengths, likes and dislikes. In a climate of concerns about the decline in free play for children in the UK and rising numbers of children with profound and complex needs, it is timely that attention be paid to these areas. The design of this multi-method, qualitative study was influenced by a range of factors. This included the nature of the children with PMLD, the theoretical stance and collaboration with a group of young disabled people. It involved an on-line survey, visits to and/or interviews at sites in the UK where there are professionals and families who were knowledgeable about people with PMLD (including arts-based professionals) and, of central importance, observations of five children with PMLD aged 3 to 7 years old in three different settings. Interviews were also conducted with the children's caregivers at home and school. This study has shown that children with PMLD are playful and that encouraging playfulness may have many significant benefits both for the children and for those around them. These include increasing stimulation, communication skills and responsiveness. Playfulness may also make the child 'feel good', possibly through providing a close connection to others, in what is termed here as 'mindful interdependency'. It was established that active playfulness stimulated and excited the children and may be linked to an increase in the number of strengths that they demonstrated. This has not previously been found in other research, as far as can be determined. A number of outputs such as a Play Passport have been produced which, it is hoped, will contribute to the encouragement of playfulness in children with PMLD.

The study aimed to determine whether staff working in residential care for the elderly could identify insidiously abusive care practice, and whether this could be predicted by staff attitudes towards older people or work locus of control. Data from 81 nurses and health care assistants was analysed using multiple regression. Results indicated that staff had more positive attitudes towards older people than negative attitudes, and that negative attitudes alone were the greatest significant predictor of staff identification of abuse, i.e. greater negative attitudes were associated with poorer recognition of abuse. 65% of participants were able to identify insidiously abusive care practice, suggesting that a significant proportion (35%) of participants were not. It was hypothesised that negative attitudes reflected a paternalistic approach to care rather than intentionally abusive practice. There was some indication that abuse-related training may moderate the relationship between negative attitude and recognition of abuse, but that training alone does not improve recognition of abuse. These hypotheses require further testing.

Chapter 1 introduced psychopathy and evaluated theories attempting to explain this disorder. Experiment 1 assessed passive avoidance learning in children with psychopathic tendencies. Results replicated previous findings with psychopathic adults indicating that this disorder is associated with poor passive avoidance learning. Experiment 2 developed a connectionist model of passive avoidance learning, the output of which was compared with the results obtained in experiment 1. The intact model successfully simulated performance of the comparison children whilst a model impaired in the formation of stimulus- punishment associations most successfully captured the performance of the children with psychopathic tendencies. Experiment 3 assessed the Blood Oxygen Level-Dependent (BOLD) responses associated with passive avoidance learning in healthy adults. Results revealed that successful passive avoidance learning was associated with activation within rostral anterior cingulate cortex, insula, caudate, hippocampus, and the amygdala. Experiment 4 assessed the performance of children with psychopathic tendencies on a novel probabilistic reversal learning paradigm. Results revealed that children with psychopathic tendencies presented with impairment only on the probabilistic contingencies. Further, it was revealed that the children with psychopathic tendencies committed more win-shift errors in the reversal phases. Experiment 5 assessed the performance of adult individuals with psychopathy using a similar task. Results revealed that adults with psychopathy were impaired in both the simple and probabilistic conditions. The adults with psychopathy also committed more win-shift responses in the reversal phases. Experiment 6 assessed the BOLD responses associated with probabilistic reversal learning in healthy adults. Results revealed that errors in both acquisition and reversal phases were associated with activations within dorsomedial and ventrolateral PFC and caudate, and deactivations within medial OFC cortex, amygdala and hippocampus. Chapter 6 re-evaluated the theories of psychopathy in light of the empirical work presented in this thesis, and discussed the implications of these results along with future research directions.

This thesis is an exploration of the symptom of delusions and the factors that are deemed important in determining their outcome. The thesis consists of three parts. Firstly, I present a systematic literature review concerning the factors that have been found to be predictive of outcome in people who experience delusions as part of psychosis. This review covers 47 studies, the majority of which employ quantitative methodology. Holding a diagnosis of schizophrenia and the presence of other positive symptoms were found to be strongly predictive factors. The influence of social factors as opposed to phenomenological and clinical factors in predicting outcome was also emphasised. Secondly, I present a qualitative study of the experiences of 10 participants who described themselves as being "recovered" from their delusions. An Interpretative Phenomenological Analysis approach is taken and 17 themes were identified as representative of the groups' shared experiences. The themes converged with previous literature in emphasising that recovery is a process which impacts upon mood, self-identity and social relationships; that isolation maintains distress and that occupation seems to promote recovery: that interactions with others are valued but also difficult; and that recovery entails loss but also reinvention, growth, enlightment and pride. It was concluded that clinicians should facilitate the appraisals of those in recovery as accomplishments and self-development and focus on individualised recovery plans. Thirdly, the final part of the thesis is a critical appraisal of the empirical study and focuses on the issues involved in judging whether qualitative research is of good quality and has good validity.

Schizophrenia (SCZ) and bipolar disorder (BPD) are common, highly heritability psychiatric disorders. Genome-wide association studies have found evidence of shared genetic susceptibility to both diseases. The most notable example is CACNA1C which encodes the 1 subunit of L-type calcium channels. Several other calcium channel genes have also been implicated in BPD and/or SCZ and together there is support for a role for these genes in both diseases. The primary function of several  subunit calcium channel genes appears to be the regulation of AMPA receptor localisation and function. Collectively these are known as Transmembrane AMPA receptor Regulatory Proteins (TARPs). This thesis aimed to identify disease relevant genetic variation in one such TARP, CACNG5, and to study the effect of these variants. CACNG5 variants in the exons and promoter region were identified in 1098 BPD, 618 SCZ, and 1087 control individuals. Four novel non-synonymous SNPs (nsSNPs) and four nsSNPs were identified. Burden analysis of nsSNPs in BPD and SCZ found evidence for association (p=0.0022). This association was strengthened by inclusion of data from European samples in the 1000 Genomes project (p=0.00057). However, combined data with the UK10K and Swedish exome sequence studies founds a weakened association signal (p=0.0082). Functional analyses using co-expression of AMPAR2 and CACNG5 constructs containing the eight nsSNPs were used to analyse changes in the expression and/or trafficking of 5 and AMPA receptors. Four of the variants were associated with decreased AMPAR2 expression as a consequence of altered trafficking to the cell surface. V146M (identified in 2 SCZ patients) overexpression increased AMPAR2 trafficking to the cell surface (p<0.005); conversely, T164L (identified in one SCZ patient) overexpression decreased the expression of AMPAR2 and its cell surface trafficking (p<0.05). Our results suggest a role for CACNG5 variants in SCZ and/or BPD and that this may be mediated via dysregulation of AMPARs.

This study describes an in-depth investigation into the pathogenic mechanisms of inherited mutations that lead to disorders of inhibitory glycinergic transmission, primarily the rare human disorder known as startle disease/hyperekplexia. I also investigated mutations causing a similar startle phenotype in cows, mice and zebrafish. Using molecular genetics techniques, I identified pathogenic mutations in the genes that encode for proteins involved in glycinergic neurotransmission, specifically the postsynaptic glycine receptor (GlyR) subunits and the presynaptic glycine transporter GlyT2. Using homology modelling and other computational biology methods, I examined the structural and functional impacts of mutations on protein function, revealing key motifs and amino acids crucial for receptor and transporter activity. Using cDNA cloning and site-directed mutagenesis, I also generated expression constructs for wild-type and mutant proteins that were used in functional tests to measure the impact of pathogenic mutations on glycine receptor and transporter function. For certain animal models of startle disease, I was also able to develop diagnostic PCR tests for pathogenic mutations, which can be used to alleviate further animal suffering by preventing ‘at risk matings’ of carrier animals. Taken together, my findings reveal several new pathogenic mechanisms of GlyR and GlyT2 mutations in startle disease in humans and animals, revealing insights into receptor and transporter function that may be applicable to other neurological disorders.

This thesis addresses the interpersonal determinants of paranoid ideation. Part one presents a systematic review of the empirical literature on the relationship between childhood adversity and paranoia. It considers whether there is an association between direct experience of interpersonal adversity and paranoia, and whether paranoia is specifically associated with a type of early interpersonal adversity. The data suggest that there is inconclusive evidence of a relationship, but support was found for a more consistent association between paranoia and emotional abuse. Possible mechanisms are considered, and results are discussed in the context of methodological limitations and the need for further investigations in this area. Part two reports on an empirical virtual reality study investigating the role of paranoia and interpersonal contingency in the experience of trust and perceived empathy in encounters with unfamiliar others, and exploring associations with other factors, including attachment history and anxiety. Sixty-one healthy male participants with varying levels of trait paranoia interviewed a virtual flatmate whose non-verbal responses were either contingent, or non-contingent in relation to them. Evidence was found for an interaction between interpersonal contingency and trait paranoia in influencing explicit trust of the avatar, but not implicit trust (as assessed by distance kept from the avatar) or perceived avatar empathy. Those high in paranoia perceived the contingent avatar as more trustworthy, whereas there was evidence of greater trust towards the non-contingent avatar in those low in paranoia. Dismissive attachment predicted greater explicit trust and perceived empathy, but reduced implicit trust. Part three is a critical appraisal of the empirical research, including personal reflections on the research process. It considers the challenges of recruitment across the paranoia continuum and the use of virtual reality technology in research investigating interpersonal encounters.

This paper examines the effectiveness of psychological interventions for patients undergoing major surgery. It reviews which of these help patients either in their psychological adaptation to or physical recovery from surgery. A total of 26 intervention studies were identified. These were divided into five categories of intervention: (1) psycho-education (2) social support (3) relaxation (4) expressive writing and (5) supportive and mixed interventions. Twenty two studies analysed measures of psychological outcomes of these, eleven reported improvements. Twenty two studies reported on measures of physical recovery of these, 10 studies reported improvements. Three further studies found better physical health outcomes in control than in intervention groups. Overall, there is mixed evidence that psychological interventions improve physical and psychological outcomes after major surgery. More good quality studies are needed to explore how psychological interventions could be incorporated into surgical ward practises.