Background – The ability to learn new information may have a crucial impact on rehabilitation with people with aphasia (PWA). However, there has been little research on learning in PWA. Although recent studies have shed light on learning and how it might be affected by cognitive functions, the tasks involved are mostly dependent on language and their findings show much inconsistency. This gap in the existing literature inspired this thesis to examine systematically learning and cognitive functions in PWA. Aim – This thesis investigates the ability of PWA to learn new information with particular attention to whether the learning deficit(s), if any, is language-specific or general in all aspects of learning, including non-linguistic material. Also, the potential occurrence of implicit learning is examined to have a comprehensive understanding of learning among PWA. The learning outcomes are further explored in terms of how cognitive functions account for the patterns observed in the various learning tasks involved in the current study. Methods – A series of psycholinguistic experiments, with PWA and two groups (young and older) of healthy participants as controls, are included in the thesis. The experiments conducted investigated the following perspectives of learning: 1) pair-associative learning of materials of various linguistic load; 2) implicit learning in the visual modality; and 3) the effect of massed versus spaced practice on learning. In addition, cognitive profiles of PWA were built through cognitive assessments covering language, memory, attention, and executive functions. The relation between the performances on the cognitive assessments and the learning outcomes were further explored. Results and discussion – The results of the experiments have provided insight into learning in PWA in various learning tasks and how the patterns of learning differ from or resemble those in the control groups. The outcomes of learning demonstrate that, compared to the controls, PWA have reduced learning ability regardless of the type of to-be-learnt material(s). Also, the findings broadly support the evidence that learning can be enhanced through feedback and repeated practice. Further, correlations are restricted among learning tasks, indicating that learning ability in people with aphasia is independent from other cognitive functions.

Tasks designed to measure affective state in animals are valuable in researching the aetiology and treatment of affective disorders in humans. However, many traditional measures are ineffective in assessing affective valence, and judgement bias tasks were designed to overcome this problem. In this thesis I aimed to identify whether rats displayed a judgement bias during sickness, and also whether the task outcomes were translatable to an invertebrate species. A comparative study was performed with a vertebrate (rat) and an invertebrate (honeybee) model exposed to toxin-induced sickness. Behavioural indicators of sickness were assessed following toxin administration, and the animals’ expectations of reward and punishment were measured on a judgement bias paradigm. This thesis includes the first behavioural characterisation of sickness in honeybees. Quinine-induced sickness in the honeybee was accompanied by a biasing of ambiguous information consistent with a negative affective state. A judgement bias was also observed in rats treated with lithium chloride, but this finding was not repeated on replication of the experiment. Methodological problems were identified and the training protocol was revised to accelerate learning of the task and to reduce extinction of responding. In conclusion, evidence of a sickness-induced negative affect in animals was identified in this thesis. This correlates with sickness in humans, thus reinforcing the argument that negative affective states associated with sickness may have an evolutionary basis. In addition, the honeybees’ performance on the task was similar to that seen in vertebrate animals, showing the potential for the honeybee model to be used in investigations of emotion. However, alterations need to be made to the specific protocols to improve the methodology for measuring judgement bias in both honeybees and rats, and recommendations are made for future experimental designs.

Background Deficits in social functioning are a core feature of Autistic Spectrum Disorders (ASD). Naturalistic observations indicate that children and young people with ASD have difficulties in initiating and maintaining social interactions with their peers and as a result often withdraw from social situations. Difficulties in social functioning have been linked to a number of underlying cognitive and developmental factors, including deficits in theory of mind, executive functioning and difficulties experiencing interpersonal relatedness. To date there has been little attempt to draw on research based on normative models of social cognition to understand social behaviour in young people with ASD. Aim of the current study This is an exploratory study with three aims. Firstly, an overall aim to explore the utility of an information processing model proposed by Crick and Dodge (1994) to the study of social cognition in ASD. Secondly, to compare the social information processing patterns and theory of mind skills of young people with a diagnosis of Asperger syndrome/high functioning autism (AS/HFA) to a group of typically developing young people. Thirdly, to explore possible associations between social information processing patterns, theory of mind skills and social functioning in young people with AS/HFA. Methodology A matched group design was employed to compare the social information processing patterns of young people with AS/HFA to typically developing young people using a social information processing interview previously designed for this purpose. The groups were also compared on a measure of theory of mind skills. A series of correlational analyses were used to explore associations between social information processing, theory of mind and social functioning in the group of young people with AS/HFA. Results Young people with a diagnosis of AS/HFA showed significantly different patterns of information processing compared to the typically developing young people at the intent attribution, response generation and response evaluation stages of the information processing model. Young people with AS/HFA also performed less well than the typically developing young people on the task designed to tap theory of mind. Theory of mind skills were found to be significantly associated with parental ratings of peer problems in young people with AS/HFA but not with parental ratings of pro-social behaviour. There was only limited evidence of an association between social information processing and measures of theory of mind and social functioning. Discussion The results of this exploratory study are discussed in relation to the previous empirical investigations of social cognition in ASD. The study provides tentative support for the utility of a social information processing approach in understanding social functioning in ASD. Implications for clinical and support services are considered with suggestions for further research.

Studies investigating executive functioning (EF) - namely skills involving higher-order cognitive activity and/or the concurrent storage and processing of material – in individuals with Williams syndrome (WS) and Down syndrome (DS) – have not generally treated task modality (verbal vs. visuospatial) as a meaningful factor. This is surprising, given the respective ability profiles which with the two conditions have been associated. Individuals with WS display relative verbal strengths, along with visuospatial processing difficulties, while individuals with DS are often reported to display verbal difficulties more marked than their overall, or non-verbal, levels of cognitive functioning would suggest. Of the studies which have treated task modality as a meaningful factor, very few have examined skills in more than one EF domain, and none have compared individuals from either population with any other specific diagnostic group, rendering difficult considerations of whether performance patterns are syndrome-specific. The principal aim of this thesis was thus to specify the finer-grained nature of EF skills in both populations, across a wider range of EF domains than previously employed. Another aim was to better enable a consideration of the syndrome specifity of each group‟s performance, by including both populations. Four empirical research studies were undertaken. In each, children and adolescents with WS and DS were presented with an analogous verbal and visuospatial task pair, within a specific EF domain; executive-loaded working memory (ELWM), inhibition, fluency and switching respectively. The two groups were compared to typically-developing (TD) children using regression 2 techniques controlling for factors which may influence executive performance. The hypotheses were a) that individuals with WS would show relative verbal EF strengths, as indicated by comparison to the TD group, with weaker performance on the corresponding visuospatial measures, b) that individuals with DS would show relative visuospatial EF strengths, as indexed in the same way, and c) that syndrome specifity would be indicated by differences between the performance patterns of each group. Findings across the four studies were mixed. Individuals with WS showed expected relative verbal strengths with regard to both ELWM and fluency. However, verbal and visuospatial inhibitory difficulties, as well as a relative verbal switching weakness, were also observed. Individuals with DS displayed the expected relative visuospatial strength on the switching task. However, no effect of task modality was found for inhibition and fluency in this group, while difficulties were observed across both verbal and visuospatial ELWM tasks. A preliminary indication of syndrome specificity was offered by the differing patterns displayed by the two groups, across EF domains. These findings are discussed in terms of their implications for a more multi-faceted understanding of the respective EF profiles of individuals with WS and DS, which attempts to account for the effect of task modality across a number of EF domains

Speech and Language Therapists (SLTs) have a key role in the management of oropharyngeal dysphagia. Dysphagia has many adverse consequences and negatively impacts quality of life (e.g. weight loss, embarrassment, depression). Dysphagia is prevalent among the care home population, affecting between 50% and 75% of residents. However, minimal research has explored the impact of dysphagia upon the health-related quality of life (HRQOL) of care home residents. Without such data, SLTs will be illequipped to meet the needs of this client group. This investigation sought to gain insight into the HRQOL issues for care home residents with oropharyngeal dysphagia caused by an acquired neurological disorder. A qualitative methodology based on the principles of descriptive phenomenology was adopted. Interviews, mealtime observations and reflexive discussions were used to collect data from seven residents with acquired neurological oropharyngeal dysphagia and six residents without dysphagia. In terviews and discussions were transcribed verbatim and common themes were identified using a framework for phenomenological analysis. Observational data were examined for corroborations/contradictions with participants' accounts. Participants with dysphagia described the physical, social and psychological impact of dysphagia, commented on their degree of satisfaction with diet modification and provided insight into their awareness of dysphagia and its overall impact. They also discussed other aspects of their mealtime experience, referring to their enjoyment of eating and mealtimes; barriers to mealtime enjoyment; choice at mealtimes and eating desire, findings which were similar to those obtained from the control participants. The observational data mainly corroborated with participants' accounts. Based on these findings, SLTs should ensure that they consistently adopt a holistic approach to the management of care home residents with acquired neurological oropharyngeal dysphagia collaborate with and provide education and training to care home staff.

The purpose of the current study was to examine a potential mechanism in the maintenance of distress tolerance difficulties and eating pathology in individuals with eating disorders. A role for attentional bias towards emotion-related stimuli was proposed, based on the premise that these stimuli would be perceived as threatening to individuals who have difficulties tolerating distress; and use eating related behaviours to regulate their emotions. It has been suggested that difficulties tolerating distress can develop as a result of invalidating childhood experiences in this group. Therefore, the relationship between attentional bias, distress tolerance difficulties and perceived experience of an invalidating childhood environment was also examined. A visual probe detection task was used to examine attentional bias for positive and negative emotion words in individuals with and without an eating disorder. Self-report measures of distress tolerance and perceived experience of an invalidating childhood were also taken. Results did not show a significant group difference in attentional bias for positive or negative emotion words. A trend towards individuals with eating disorders being more likely to direct their attention away from negative emotion words was observed. Higher levels of distress tolerance difficulties and perceived experience of invalidating childhood environments were reported in the clinical group. Furthermore, bias away from negative emotion words was associated with a deficit in healthy distress tolerance strategies in this group. These results are consistent with the idea that cognitive avoidance of negative emotion-related information is associated with distress tolerance difficulties in individuals with eating disorders. However, further investigation of this bias would be necessary in order to determine whether or not it is a reliable finding and clarify whether it is specific to individuals with eating disorders, or to a sub group of such individuals. Implications for future research and clinical practice are discussed.

Introduction: Stroke is a leading cause of mortality and disability worldwide. There is limited evidence on the natural history, predictors, and outcomes of depression in the long term after stroke. Objectives: To describe the natural history of depression within 15 years of stroke; To identify the predictors of depression within 15 years of stroke; To identify the health outcomes of post-stroke depression. Methods: A systematic review and meta-analysis of studies of the natural history, predictors and associated outcomes of post-stroke depression was conducted. Incidence, prevalence, cumulative incidence, time of onset, duration, and recurrence of depression, within 15 years of stroke were estimated in the population based South London Stroke Register (SLSR). Predictors and outcomes of depression up to 15 years after stroke were identified. Results: The systematic review identified 49 studies. The pooled prevalence of depression was 29%. The major predictors of depression included disability and history of depression pre-stroke. The main outcomes of depression were lower quality of life and disability. The SLSR data analyses showed that the prevalence of depression was around 30% and remained stable within 15 years of stroke, with an annual incidence 7 to 21% and cumulative incidence of 55%. Depression started shortly after stroke, had a short duration and high recurrence rate. Variables associated with depression included stroke severity, disability at baseline, depression before stroke, together with disability, social isolation, and cognitive impairment at follow-up. Depression was associated with mortality, disability, cognitive impairment and lower quality of life at follow-up. Conclusion: depression is a frequent chronic and recurrent condition after stroke, with higher risk among patients with previous depression and after severe strokes, and it is associated with negative health outcomes in the long term.

Major depressive disorder (MDD) is a complex, heterogeneous disorder characterised by a pathological distortion of emotional mood. There is evidence of both genetic and environmental risk factors for MDD, and gene-environment interactions may play a particularly important role. Clinically, MDD is defined by patients meeting a number of diagnostic criteria. However, the heterogeneous nature of the disorder can make MDD difficult to diagnose, especially as it shares close similarities with other psychiatric illnesses such as bipolar disorder. Another clinical problem is that antidepressants, the first line of treatment for MDD, are ineffective in a significant proportion of patients. The projects in this thesis address four aims: (i) the identification of novel gene-environment interactions which may increase risk for MDD; (ii) the identification of diagnostic biomarkers for MDD; (iii) the identification of biomarkers for the prediction of treatment response to antidepressants; and (iv) the identification of transcriptional changes associated with antidepressant treatment and successful therapeutic response. Utilising a model of early life stress in two inbred mouse strains, we investigated the transcriptional effects of maternal separation. The top stress by strain interaction was found in the telomerase RNA component gene (Terc). We also found that a single nucleotide polymorphism in TERC (rs10936599), previously identified as a predictor of telomere length, interacted with childhood neglect to predict MDD in a human case-control cohort. A study investigating differences in the transcription of inflammatory cytokines in the blood of MDD patients, bipolar disorder patients and controls, revealed disorder-specific differences in chemokine (C-C motif) ligand 24 and CC chemokine receptor type 6 which specifically differentiated MDD patients. Furthermore, we found that transcription of tumour necrosis factor and its targets in the inflammatory cytokine pathway, and DNA methylation in interleukin-11 could be used to predict antidepressant response amongst MDD patients. Moreover, transcription of ATP-binding cassette sub-family F member 1 was found to increase on antidepressant treatment, with the magnitude of change corresponding to clinical response.

The present programme of studies investigated whether there is a positive emotional bias in the content of confabulation using the semantic-associates paradigm. This procedure comprises lists of semantically related items that are associated with a non-presented critical distracter. In three studies, 26 confabulating amnesia patients, 26 non-confabulating amnesia patients and 26 healthy controls were presented with the semantic-associates task. In study 1, this procedure was employed to induce false recall and false recognition in response to studying lists of positive, negative and neutral word lists. In study 2, a facial expressions semantic-associates procedure was constructed to examine false recognition of pictorial items. In the final study, participants were induced into positive and negative mood using a video mood induction procedure to examine the effects of mood on false recall and false recognition. Confabulating patients showed a positive emotional bias and falsely recognised a higher proportion of positive unrelated intrusions compared with non-confabulating patients and healthy controls. These findings suggested that confabulating patients’ tendency to produce pleasant false memories may represent a bias in general emotional processing. However, the positive bias was not found in the facial expressions task. This suggested that the distinctive characteristics in pictorial items may aid confabulating patients in the discrimination between studied items and non-studied intrusions. In addition, reduced false recognition of critical distracters in both confabulating and non-confabulating patients was a marker for gist memory impairment in amnesia. Finally, the video mood induction procedure demonstrated that the positive bias in confabulating patients was enhanced by, but not specific, to negative mood. However, findings from a signal detection analysis indicated that confabulating patients showed a positive bias because their memory strength for positive material was significantly weaker compared with that of non-confabulating patients and healthy controls. Future studies would need to equate for differences in memory strength between controls and amnesia patients in order to provide stronger evidence that emotional factors are playing a role in the content of confabulation.

Given the marked personal, social and economic impacts, depressive disorder creates significant demands on individuals, health service providers and society as a whole. Antidepressant drugs remain the mainstay of treatment for depression for most people in contact with healthcare services and the last 20 years have seen dramatic changes in antidepressant prescription patterns. Initially, there was an increase in the use of the selective serotonin reuptake inhibitors (SSRIs) and subsequently, other novel antidepressant agents with different pharmacological mechanisms entered the market. Given the range of choices, clinicians must decide about which is the most appropriate intervention for their patients. To this end, knowledge regarding the relative cost-effectiveness and cost-utility of individual antidepressants is important. Therefore, in this thesis, a systematic review was first conducted to assess methodological approaches in economic evaluations of pharmacological treatments using database analyses. Based on the National Health Insurance Research Database in Taiwan, a cost analysis was conducted to identify which demographic and clinical characteristics are associated with healthcare costs of patients with depression. Compared to patients prescribed SSRIs, those prescribed older antidepressants had lower total and psychiatric costs, while patients prescribed serotonin norepinephrine reuptake inhibitors (SNRIs), and other newer antidepressants had higher total and psychiatric costs. The baseline comorbidities of cardiovascular diseases (CVD) and headache were also associated with healthcare costs over the 12-month period. To further assess the longer-term economic impacts, a set of database outcome statuses (sustained treatment-free status, continuous treatment, and late re-contact) were then applied to explore factors associated with outcome status following initial treatments and to examine healthcare costs over the following three years by outcome status. Theresults showed the initial outcome status could exert an impact on total healthcarecosts in the second and third years after commencing treatments. Finally, cost-effectiveness and cost-utility were compared between different categories of antidepressant treatments and also to test whether and how the presence of CVD, the most prevalent comorbid physical illness in this cohort, affects these results. The results showed that SSRIs are more cost-effective than tricyclic antidepressants and SNRIs regardless of comorbid CVD. There are various limitations to be considered in these analyses, including the limited scope of costs, the lack of clinical information, and the adoption of utility scores from previous studies. Further efforts to elucidate the relationship between depression treatments, costs and outcomes for longer period of follow up are warranted.