Melbourne School of Psychological Sciences - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 73
  • Item
    Thumbnail Image
    Volumetric Analysis of the Entorhinal and Transentorhinal Cortices in the Early Detection of Alzheimer’s Disease
    Quek, Yi-En ( 2024-02)
    Diagnosing Alzheimer’s disease (AD) early in the disease process can be challenging. Incorporating biomarkers into the diagnostic evaluation for AD has been recommended to increase diagnostic accuracy. Structural magnetic resonance imaging (MRI) is a particularly appealing biomarker because it is a widely available and noninvasive procedure. In view of the early pathological involvement of the entorhinal and transentorhinal cortices in the disease process, MRI-based volume measurements of these regions are promising candidate markers for early AD. Whilst significant advances have been achieved in automated segmentation of brain MRI in recent years, the entorhinal and transentorhinal cortices have been shown to be particularly challenging regions for automated segmentation. The current dissertation sought to examine the role of entorhinal and transentorhinal cortical MRI volumetric analysis to enhance the early detection of AD. Chapter 2 presented a systematic review and meta-analysis of studies comparing automated and manual regional brain volume measurements across healthy controls (HCs), individuals with amnestic mild cognitive impairment (aMCI), and individuals with AD dementia. There was no significant difference in automatic and manual volume measurements for the hippocampus, lateral ventricles, and parahippocampal gyrus. There was, however, significant heterogeneity in the effect estimates across the studies. Chapter 3 comprised an investigation into the effect of MRI scan orientation on the reliability of entorhinal and transentorhinal cortical volume measurement in order to establish a standard protocol for segmentation for subsequent Chapters. Aligning MRI scans to a common orientation improved the reliability of entorhinal and transentorhinal cortical volume measurement. Chapter 4 described an independent validation of a recently developed automated segmentation method, Automatic Segmentation of Hippocampal Subfields-T1 (ASHS-T1), for the measurement of entorhinal and transentorhinal cortical volumes in HCs, individuals with aMCI, and individuals with AD dementia. Despite differences between the volumes obtained from ASHS-T1 and those obtained from manual segmentation, ASHS-T1 was sensitive to AD-related entorhinal and transentorhinal cortical atrophy. Chapter 5 contained an examination of the contribution of selected mesial temporal lobe volumes obtained from ASHS-T1 alongside test scores derived from comprehensive neuropsychological assessment to aid identification of individuals with aMCI and individuals with AD dementia in a memory clinic cohort. Volumes of the entorhinal cortex, transentorhinal cortex, and hippocampus performed no better than chance and offered no utility over and above valid and reliable neuropsychological measures for identifying individuals with AD. Chapter 6 introduced a novel, deep learning-based automated segmentation method, TransEntorhinal Segmentation by Deep Learning (TES-DL), to segment the entorhinal and transentorhinal cortices. Despite suboptimal segmentation accuracy, TES-DL showed mostly medium to large effect sizes for the differences in entorhinal and transentorhinal cortical volumes between HC, aMCI, and AD groups in two AD cohorts. Taken together, the findings from the current dissertation highlight the difficulty of accurate automated segmentation of the entorhinal and transentorhinal cortices and challenge the clinical utility of the volumes of these regions in the early detection of AD. Importantly, whilst advances in automated neuroimaging methods promise exciting new possibilities, it is essential that these methods are thoroughly validated prior to widespread clinical application.
  • Item
    Thumbnail Image
    Understanding the Mental Health Impacts and Support Needs Following Coercive Control
    Lohmann, Susanne ( 2024-01)
    Coercive control is a complex, insidious and under-researched dimension of psychological intimate partner violence (IPV) that has been associated with trauma and mental health impacts. This thesis had two overarching aims: (1) to improve the understanding of the trauma and mental health impacts of coercive control, and (2) to explore the mental health support needs and wants of women following coercive control exposure to aid their long-term recovery. To achieve these aims this thesis employed a mixed methods approach and included three separate yet interlinked studies: a systematic review and meta-analysis, and two qualitative reports. The review aimed to examine quantitative associations involving coercive control with posttraumatic stress disorder (PTSD), depression and complex PTSD (CPTSD) and included 68 observational studies. Data from 45 studies were included in the meta-analyses. Meta-analyses showed that coercive control exposure was moderately associated with both PTSD and depression. The associations were comparable to those between broader dimensions of psychological IPV and PTSD and depression that were also investigated in the review, and with associations involving physical IPV with PTSD and with depression examined in previous meta-analyses. Only one study reported on the relationship between coercive control and CPTSD and meta-analyses could not be performed. Studies 2 and 3 addressed gaps that were identified in the systematic review and meta-analysis. To enhance the understanding of the trauma and mental health impacts of coercive control and the support needs of victim-survivors, a qualitative approach was employed for studies 2 and 3, and 16 women who had experienced coercive control were interviewed. Study 2 reported on these women’s lived experience of the mental health impacts of coercive control. Study 3 explored what types of psychological supports women need and want to aid their long-term recovery. Building on the findings of study 1, studies 2 and 3 made six key contributions to the coercive control literature: (1) Clarification of the coercive control construct by identifying entrapment (including social entrapment) and insidiousness (characterised by subtlety, intangibleness, gradual worsening), as the underlying dimensions of coercive control that likely differentiate coercive control from other dimensions of IPV, and may be associated with unique mental health reactions; (2) evidence that coercive control exposure should be recognised as a traumatic event; (3) qualitative evidence for the association between coercive control and CPTSD symptoms; (4) proposition that the ongoing threat and terror, insidiousness and entrapment of coercive control contribute to the unique and complex trauma and mental health reactions that may be reflected in CPTSD symptoms; (5) the need for a holistic recovery focus that integrates clinical and personal recovery; and (6) the need for integrated trauma-and-coercive control-informed psychosocial interventions. Taken together, this thesis has made several significant contributions to the coercive control literature that have important implications for policy, clinical practice and research.
  • Item
    Thumbnail Image
    Towards a functional action-based account of Moralization
    Rhee, Joshua Ju-suk ( 2023-12)
    Morality has traditionally been considered to reflect a set of universal and objective values. However, more recent work investigating people’s subjective moral beliefs have revealed that substantial and meaningful variance can exist within and between individuals in their tendency to see attitudes or actions as morally relevant. Moralization describes the process underlying within-person variance in attribution of moral significance to actions or attitudes. In this thesis we build on recent work investigating the mechanisms of subjective moralization by proposing the action-based moralization account – a framework that conceptualises moralization as functional response to action-based uncertainty in the context of threat-appraisal. To evaluate the utility of this account as a functional theory of moralization, we apply a multi-method approach to investigate the relationship between threat-appraisal and moralization. Specifically, we find evidence of a relationship between individual differences in threat-sensitivity and tendency for moralization. Through an ecological momentary assessment study, we also find that within-person variation in the experience threat-appraisal related affect predicts moralization. Finally, in two representative sample studies in the U.S. and U.K we find that individual differences in subjective threat-appraisal of COVID-19 is positively associated with moralization of COVID-19 preventative behaviours. Finally, we investigate the longitudinal relationship between moralization and uncertainty reduction in the context of a real-world political event. We find that moralization is associated with prediction confidence at time 1, and predicts subjective belief confirmation at time 2. Collectively, these findings provide initial support for the action-based moralization account, and highlight the value of functional theories as a framework for understanding the precursors and outcomes of moralization.
  • Item
    Thumbnail Image
    An investigation of the holistic processing of faces
    Cheng, Xue Jun ( 2023-11)
    Whether it is due to the specific configuration of faces or our particular expertise with them, faces are seen to be a special class of object and, hence, processed differently compared to most other stimuli in our environment. One influential concept - holistic processing - has been used to describe how we process faces. In general, the concept of holistic processing is based on the idea that the entire face object is perceived and processed as a single unit instead of as individual parts or an aggregate of those parts. However, there is no single clear operationalisation of holistic processing in the literature, and the various experimental paradigms used seem to assess different aspects of processing. In this thesis, I explore the different ways holistic processing has been investigated and provide an in-depth examination of the underlying processing of composite faces. According to theories of information processing, holistic processing can be operationalised as a coactive architecture where all face information is combined into a single channel, which then drives decision-making in specific tasks. Coactive processing can be distinguished from other processing architectures such as serial and parallel processing using systems factorial technology and model-fitting. Overall, our composite faces were best fit by a mixture of serial and parallel processing, indicating little support for the argument that faces are processed holistically. We further investigate this mixture model and discuss how these results can be situated on a continuum which ranges from analytic, independent processing to complete, holistic processing.
  • Item
    Thumbnail Image
    The role of disgust in moral judgement
    Donner, Michael Ryan ( 2023-10)
    The emotion disgust underlies some of our moral judgments. But it’s not clear whether this relationship is because of disgust’s role in motivating the avoidance of pathogen threats (i.e., pathogen disgust) or reproductive threats (i.e., sexual disgust) (or both). The present thesis systematically examines the respective role of pathogen and sexual disgust in moral judgment across three empirical chapters. Chapter 2 (published in the journal Emotion Review; N = 72,443), was a systematic review and meta-analysis on associations between trait disgust sensitivity and moral judgments (as per Moral Foundations Theory; Graham et al., 2013). The key finding from this investigation was that measures of trait sexual disgust were more strongly related to moral judgments (especially of the sanctity domain) than were measures of trait pathogen disgust. Building on Chapter 2’s findings, Chapter 3 (published in the journal Emotion) aimed to clarify the zero-order effects of the meta-analysis by investigating the unique (semi-partial) contribution of each disgust type on moral judgment. In two studies (N = 2,718), comprised of 10 samples (taken from Australia, the United States, and Brazil), it was found that the unique associations between trait sexual disgust and sanctity/binding moral judgments were strong and significant, whereas the unique associations between trait pathogen disgust and sanctity/binding moral judgments were weak and non-significant. Moreover, the sexual disgust-sanctity/binding association seemed to be explained by motivations to avoid promiscuous sex (a specific kind of reproductive threat that triggers sexual disgust). Together, the findings from Chapters 2 and 3 argue against popular pathogen avoidance explanations of the connection between disgust and sanctity/binding concerns, and instead argue in favor sexual avoidance explanations (e.g., promiscuity avoidance). Chapter 4 moves away from an individual differences approach and adopts an experimental approach to investigate whether inducing state sexual disgust would have a causal impact on moral judgments across two studies (N = 1,290). Although the sexual disgust induction was successful at inducing disgust in both experiments, the induction had no direct causal impact on moral judgments (compared to a control condition) in either experiment. Nevertheless, consistent with the previous two chapters, trait sexual disgust was uniquely and reliably associated with sanctity/binding moral judgments in both experiments. Inconsistencies between trait and state sexual disgust findings might be due to limitations of the experiments or because trait and state approaches test a different underlying process. In summary, the present thesis shows that sexual disgust is an important factor in moral judgments.
  • Item
    Thumbnail Image
    Addressing the Complexity of Restricted and Repetitive Behaviours: From Structure to Correlates
    Spackman, Emily Kate ( 2024-01)
    Circumscribed Interests (CI) and Insistence on sameness (IS) are two key subdomains of Restricted and Repetitive Behaviours (RRBs) – a core diagnostic domain of autism spectrum disorder. Both CI and IS are broad and complex, with CI encompassing a range of intense interests and related behaviours that may or may not be focused on highly idiosyncratic topics or objects (e.g., traffic lights or calendar dates) and IS encompassing reliance on routines, needing permanence in one’s environment, and/or for objects to be arranged in particular ways. Characterising individual differences in CI and IS presentation, including their pattern of association with key demographic, cognitive, and clinical correlates, is essential for improving screening and diagnosis, understanding the psychological processes underpinning these behaviours, and informing the development of more personalised healthcare and support options. However, previous research focused on the correlates of CI and IS has yielded inconsistent findings. Given the breadth of behaviours encompassed by CI and IS, it has been suggested that, rather than being unidimensional, both of these domains may consist of more specific subdomains that might have different trajectories and at least partially distinct underlying mechanisms. Thus, this conflation of distinct constructs under the broader CI and IS domains may have contributed to noted inconsistencies. Therefore, this thesis first sought to clarify the structure of the CI and IS domains and, secondly, to characterise the correlates of identified CI and IS subdomains. To address these questions, this program of work first summarised existing literature exploring developmental and clinical correlates of specific RRB subdomains, including CI and IS, through a systematic review and meta-analysis. The structure and correlates of CI and IS, specifically, were then further examined through four empirical papers drawing on two samples and distinct methodological approaches. Studies 1-3 utilised a range of variable- and person-centred approaches in a larger sample consisting of 1,892 autistic children recruited through the Simons Foundation for Powering Autism Research (SPARK) initiative. Study 4 combined clinical interview and qualitative approaches in a smaller sample recruited through the Stanford Autism Clinic’s research registry and social media pages. Findings from studies 1 and 2 provide preliminary evidence for a distinction between restricted interests (RI), referring to interests that, although atypical in terms of their intensity, were otherwise centred around topics commonly observed in the general population (e.g., music, videogames, etc) and unusual interests (UI) encompassing interests not salient outside of autistic populations (e.g., an interest in traffic lights or dates). Importantly, RI and UI showed distinct patterns of association with age, sex, IQ, and social motivation (study 1) and may be used to characterise individual differences among autistic individuals (study 2). Study 3 identified three subdomains of IS: IS-Ritualistic/sameness, referring to the need for element of one’s environment to be or feel “just right”, IS-Routines, and IS-Others, referring to the need for sameness and predictability during one’s interactions with others. Similar to CI subdomains, identified IS subdomains also showed unique patterns of association with key correlates, including age, sex, IQ, social abilities, sensory sensitivity, and self-regulation. Finally, study 4 provided a more in-depth, qualitative characterisation of aspects of CI and IS that are not well captured by current quantitative instruments, including their intensity, frequency, trajectory, adaptive benefits, and associated challenges. Together, findings demonstrate the complexity of CI and IS domains and highlight how precise and empirically-supported construct measurement is essential for valid and reliable characterisation. Limitations and future directions of this work are discussed.
  • Item
    Thumbnail Image
    Improving the Strength and Quality of Research Evidence in Clinical Neuropsychology
    Furey, Rachel Taneya ( 2023-10)
    Evidence replicability directly impacts the quality of care received by neuropsychology patients. While the importance of evidence-based practice has been increasingly recognized in neuropsychology over the past decades, some common practices that lack strong empirical support persist. Many of these practices can be observed in using and interpreting cognitive test scores. Such practices may have originated during a time when questioning expert opinion was uncommon and clinical expertise was relied upon for knowledge. As a result, the need to empirically investigate the assumptions behind some long-standing clinical practices has been overlooked, particularly those practices with high face validity. This thesis aims to evaluate the need and ways to improve the quality of research evidence used to inform clinical neuropsychological practice. Four empirical studies were conducted as part of this PhD project to address this objective. Each study incorporated metascientific practices to demonstrate how these practices can be employed in neuropsychology research to generate strong, high-quality evidence. Study 1 addressed the hypothesis that in the verbal paired-associates (VPA) subtests of the Weschler Memory Scales (WMS), hard items are more sensitive to left hippocampal damage than easy items, based on the assumption that hard items are “semantics-free”. The findings of Study 1 supported a latent semantic structure for the WMS-III VPA, suggesting that performance depends on the semantic features of the target words in each item. Study 2 investigated the latent structure of executive function in the Delis-Kaplan Executive Function System (D-KEFS), addressing the hypothesis that executive function constructs can be integrated with other cognitive ability constructs as defined by the Cattell- Horn-Carroll (CHC) theory. The findings of Study 2 demonstrated the previously published factor models for the D-KEFS, which lacked strong theoretical foundations, were not replicable. In contrast, a factor model consistent with CHC theory was well-fitting for the D-KEFS tests. Study 3 examined the clinical utility of separate recall and recognition testing in identifying memory disorders using a systematic review and meta-analyses. Despite popularity in clinical practice, the results of Study 3 found no evidence to support the hypothesis that recall and recognition tests provide different information. Instead, the findings suggest that recognition tests offer no incremental utility for identifying memory disorders compared to recall tests. Study 4 was a meta-research study that aimed to examine the impact of implementing reporting standards in a single neuropsychology journal, namely, Archives of Clinical Neuropsychology, which introduced mandatory reporting guidelines for submitting authors in 2017. The findings indicate that the quality of reporting in neuropsychology is currently suboptimal, potentially undermining the replicability of research evidence. Furthermore, the findings suggest that requiring authors to adhere to reporting guidelines when submitting their research may not guarantee high-quality methodological and statistical reporting. Overall, the findings of this PhD project demonstrate there is a need to improve evidence replicability in neuropsychology. The current peer review process is not sufficient to guarantee that published research evidence is strong or even valid. Broader adoption of metascientific practices is recommended across all stages of research, including before, during, and after publication. Although improving the peer review process would be a step towards improving evidence quality in neuropsychology, there are currently few incentives for journals to address this issue as a priority. Therefore, if substantial progress is to be made in improving the evidence quality in neuropsychology or other scientific disciplines, addressing the incentive system in academia may be a necessary first step.
  • Item
    Thumbnail Image
    The Associative Basis of Semantic Processing during Comprehension
    Shabahang, Kiavash ( 2023-08)
    The current thesis presents converging experimental and computational evidence for the centrality of retrieval operations during the comprehension of linguistic material, and develops an associative theory of semantic processing. Chapter 1 outlines the broad theoretical antecedents with a focus on the tension between symbolic and associative representations of meaning, arguing for a structural distinction between associations. Chapter 2 presents an associative network simulation of semantic memory and shows how assuming a spreading activation through a network of direct associations can match the variance in human free word association and word similarity data, traditionally explained by latent-representation semantic memory models. Instead of assuming generalization during encoding as in the latent-representation accounts, the Dynamic-Eigen-Net algorithm is presented as a way to defer generalization until retrieval through spreading activation in the network. Chapter 3 further extends the Dynamic-Eigen-Net to serial-order associations and shows how the Dynamic-Eigen-Net is better able to discriminate consistent and inconsistent novel bi-grams compared to standard spreading activation algorithms. Chapter 4 presents experiments that introduce challenges to the direct-encoding account of paradigmatic formation, where paradigmatic relations are assumed to correspond to directly encoded associations. Over a series of experiments, reliable evidence for the presence of syntagmatic associations is presented, but evidence for paradigmatic associations was inconclusive. Chapter 5 undertakes a comparison of the direct encoding account of paradigmatic relations and alternative theories. The results provide clear evidence in support of the retrieval account of paradigmatic relations, where paradigmatic relations are dynamically formed during retrieval. Chapter 6 revisits the Dynamic-Eigen-Net, presenting one way order-independent and -dependent associations can be integrated in an associative network model of the Syntagmatic-Paradigmatic framework. Chapter 7 concludes with a focus on challenges to overcome to pave the way towards an associative theory of semantic composition.
  • Item
    Thumbnail Image
    Associations between estradiol, brain structure and function, and extinction recall in adolescents
    Zwaan, Isabel Sterre ( 2023-08)
    Background: Adolescence is a life stage where anxiety and depressive (internalising) disorders emerge. Internalising disorders are twice as prevalent in females as males, a pattern apparent in adolescence. This sex disparity may be explained by pubertal and neurodevelopmental factors, as dramatic hormonal changes and dramatic brain changes occur during adolescence. Individual differences in pubertal hormones may interact with neurodevelopment to impact mental health. The sex steroid hormone estradiol is key in puberty, however few studies have examined associations between estradiol, neurodevelopment, and internalising symptoms, particularly in adolescents. In fact, estradiol may influence neural networks important for emotional processing and regulation, where alterations in these neural networks may underlie emotion dysregulation underlying internalising disorders. Aims: This thesis aims to investigate associations between estradiol variability (secondarily, average estradiol) and brain structure in female adolescents, as well as whether brain structure mediates longitudinal associations between estradiol variability and internalising symptoms. Further, we aim to investigate associations between estradiol variability (secondarily, average estradiol and estradiol level) and brain activation during emotion processing, as well as whether brain function mediates longitudinal associations between estradiol variability and internalising symptoms. Finally, the current thesis aims to explore associations between brain structure, fear extinction recall and trait and state anxiety symptoms, as well as moderating effects of age and sex, in a pilot study of adults and adolescents. Method: Study 1 and Study 2 of this thesis utilised the same sample of 44 female participants measured at age ~12 years (baseline) and age 13 years (follow-up). At baseline adolescents completed self-report questionnaires measuring anxiety (Spence Children’s Anxiety Scale) and depressive (Children’s Depression Inventory 2) symptoms, four saliva samples across one month to measure estradiol, and an MRI scan. In the MRI scan we collected T1-weighted images measuring brain structure for Study 1, and BOLD activation during a passive emotional face viewing task (including angry, fearful, calm faces) for Study 2. At follow-up, participants completed questionnaires measuring internalising symptoms. We utilised average estradiol and estradiol variability for both studies and estradiol on day of scan (the fourth saliva sample) for Study 2, using these to examine associations to brain structure and function during the emotional faces task and to internalising symptoms. In Study 3 we used pilot data from a sample of 18 adults and 20 adolescents who completed a fear conditioning, extinction learning, extinction recall, re-conditioning, and re-extinction fMRI task (a variant of the ‘screaming lady’ paradigm) and examined associations between skin conductance response during extinction recall and brain regions of interest in both groups combined and adolescents only, examining moderating effects of age group and sex. Additionally, we examined correlations between extinction recall, region of interest, and self-reported state and strait anxiety (State-Trait Anxiety Inventory). Results: We found that increased estradiol variability was associated with decreased right medial orbitofrontal cortex (mOFC) thickness, however there was no association between average estradiol and brain structure. Additionally, we found that estradiol on day of scan (but not average estradiol or estradiol variability) was associated with decreased precuneus and precentral gyrus and increased angular gyrus and frontal orbital cortex during processing of negatively valenced vs calm faces. We did not find associations between estradiol and internalising symptoms, or brain structure or function and internalising symptoms. In our pilot study, preliminary findings indicated that lateral OFC thickness was associated with extinction recall response in male, but not female, adolescents, while increased skin conductance response to the threat signal during extinction recall was associated with increased right mOFC thickness in adults. Further, we found a positive correlation between trait anxiety and right mOFC thickness and negative correlation between state anxiety and extinction recall response in adults and adolescents, and a positive correlation between trait anxiety and left medial OFC thickness in adolescents. Significance: To the best of our knowledge, these were the first studies examining associations between estradiol variability and brain structure and function in adolescents. We extended existing literature by demonstrating that increased estradiol variability, and not average estradiol, was associated with decreased vmPFC thickness (a region important for emotion regulation and fear extinction learning and extinction recall) in female adolescents at age 12. Further, we demonstrated that estradiol may module vmPFC activation during emotion processing, potentially facilitating emotion regulation. In our pilot study we (preliminarily) found that sex moderated the association between lateral OFC thickness and extinction recall response in adolescents, thereby extending the limited research on brain structure and extinction recall by indicating that sex differences may be present in these associations. Together, our findings support the notion that female sex steroids have organizational and activational effects on the brain, and that adolescence is a sensitive period of development.
  • Item
    Thumbnail Image
    Sleep disturbance as a feature of dementia with Lewy bodies
    Salthouse, Olivia Rose ( 2023-07)
    Dementia with Lewy bodies (DLB) is considered the second most common form of dementia after dementia of the Alzheimer’s type. While DLB has been considered its own diagnostic entity for approximately 30-years, several aspects of the clinical presentation remain enigmatic and require further research. Sleep is one such feature. Rapid eye movement (REM) sleep behaviour disorder (RBD) is a prominent feature of DLB and research over the past two decades has shown that it often precedes the onset of other clinical disease features by many years. The striking relationship between RBD and DLB has placed the REM stage of sleep at the forefront of sleep research in this population. Less consideration has been given to other aspects of sleep, yet clinical anecdote and case study evidence strongly suggest that sleep disruption is a pervasive clinical phenomenon, which plays itself out across all stages of sleep, affecting sleep quality, sleep structure, sleep electrophysiology and sleep behaviour. The aims of the current thesis were three-fold: 1) to comprehensively characterise the sleep profile of those in the early stages of probable DLB using actigraphy and PSG recording methods, 2) to characterise the most sensitive areas of cognition in DLB (attentional function and visual processing), and 3) to examine the relationship between sleep disruption and these sensitive areas of cognition in DLB. Participants were nine individuals with early stage DLB and nine age- and education-matched healthy older adults. Actigraphy and PSG sleep recording methods were used to objectively measure sleep quality and architecture. Participants underwent overnight in-home polysomnography (PSG) monitoring on 2 nights and completed 14 days of continuous actigraphy recording. Subjective sleep was measured using self-report questionnaires. Standardised neuropsychological measures assessed visuoperceptual and visuospatial function. Two computerised reaction time paradigms were completed by participants and provided a comprehensive measure of attentional function. This research documents significant disruptions to sleep quality and architecture in individuals with DLB. Actigraphy data revealed reduced sleep quality (reduced sleep efficiency and greater wake after sleep onset) and greater intra-individual variability in sleep quality from night-to-night. Greater variability in daytime sleep/inactivity was also documented. Overnight PSG revealed changes to sleep macrostructure and microstructure. Increased stage 1 sleep (N1), reduced stage 3 sleep (N3), and fewer microstructural features (sleep spindles and K-complexes) were documented. Sleep stage distribution throughout the night was also altered from the typical trajectory; as sleep progressed across the night, the relative distributions of sleep stages remained comparable, particularly for REM sleep. Examination of relationships between sleep and cognition revealed strong and consistent associations in those with DLB. Reduced sleep quality and increased variability in sleep quality were strongly associated with attentional impairment and in particular, attentional variability. Greater variability in daytime sleep/inactivity was also associated with more severe attentional impairment in DLB. Visual processing was largely independent of sleep. Pervasive sleep disruption is a clinical feature of DLB that emerges early in the course of disease. Disruption is represented across all aspects of sleep architecture, and impacts night time sleep and daytime functioning. These disturbances are characterised by prominent intra-individual variability and represent another fluctuating feature of this active and turbulent disease. These findings point to a loss of sleep macro- and micro-structure and suggest loss of structural definition in sleep, and more broadly a breakdown in the boundaries that typically separate the stages of sleep and the states of sleep and wakefulness. Empirical support for associations between sleep disruption and the quintessential cognitive features suggests a common underlying mechanism may underpin both sleep and cognition. This has key implications for understanding the origins of these dynamic features. Broad sleep disturbance is an early and prominent clinical symptom and offers a promising road for better understanding DLB and the neurological mechanisms which underly the enigmatic clinical symptoms of the disease.