Attention and Inference in Melancholic Depression

Download files
Access & Terms of Use
open access
Copyright: Hyett, Matthew
Altmetric
Abstract
Melancholia has long been positioned as a quintessentially biological depressive condition. Impairments in attention are prominent, particularly in shifting attention away from internal states, but a detailed neurocognitive understanding of these deficits is lacking. This thesis hence sought to clarify the cognitive and neurobiological mechanisms of attention deficits in melancholia. Analytic methods spanning cognitive and brain network modelling were employed to explore the biases and inflexibility of attention in melancholia, compared to non-melancholic depressed and healthy individuals. The first study (Chapter 2) investigates disrupted attentional inference to emotional stimuli across sub-types of depression and healthy participants. I hypothesised that both depressed groups would show impaired discriminability of emotional signals, and that melancholia would be characterised by decreased sensitivity to emotional stimuli. Signal detection data from an attentional control task were modelled using hierarchical (Bayesian) statistics. Melancholia was associated with disrupted sensitivity of emotional signals, and poorer discriminability of neutral signals, hence likely reflecting distorted attentional inference. The second study (Chapter 3) explores resting state functional brain network effective connectivity across melancholic, non-melancholic and control groups. Interactions between cortical systems corresponding to attention, executive control and interoception – derived from independent component analysis (ICA) – were modelled using dynamic causal modelling (DCM). Analyses supported the hypothesis that relationships amongst networks subserving attention and interoception would be disrupted in melancholia. This study revealed a specific ‘dysconnectivity’ between brain regions underpinning attention and interoception in melancholia. In the third study (Chapter 4), I advanced an in-scanner naturalistic film viewing paradigm to quantify brain networks underling the shifting of attention from rest to dynamic processing of exogenous emotional stimuli, employing the same groups as Chapter 3. I hypothesised that cortical systems would remain in an “at-rest” state in melancholia, reflecting impaired attentional shifting to exogenous stimuli. Surprisingly, neuronal activity in systems supporting attention and interoception were increased in melancholia compared to controls during negative film viewing. I speculate that these findings reflect ineffective neuronal adaptation during attentional resource allocation to emotional material in melancholia. Preliminary analyses (presented in Chapter 5) highlight that impaired attentional set-shifting performance is associated with disruptions to these neuronal systems, hence pointing to a disorder-specific behavioural analogue of the neurobiological findings. The studies comprising this thesis offer a unique cognitive and neurobiological explanation for attentional deficits in melancholia, and act to explain aspects of its clinical presentation in terms of impaired redirection of attention away from persistent and dysphoric internal states.
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Hyett, Matthew
Supervisor(s)
Breakspear, Michael
Parker, Gordon
Rae, Caroline
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2015
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download public version.pdf 2.87 MB Adobe Portable Document Format
Related dataset(s)