Functional Imaging Studies of Speech and Verbal Memory in Healthy Adults and Patients with Alzheimer’s Disease

Alzheimer’s disease (AD) results in a diffuse, but characteristic impairment of cognitive function, with early involvement of verbal episodic memory. A prodromal phase of amnestic mild cognitive impairment (MCI) consists of patients with a mild, isolated impairment of episodic memory. In this thesis, I have described experiments performed on these patients and healthy volunteers using functional magnetic resonance imaging (fMRI). I aimed to investigate changes in neural activity associated with the breakdown in verbal episodic memory. Initially, I established the feasibility of using fMRI to investigate spoken responses in a study of speech production in healthy volunteers. This was important for investigating spoken retrieval of episodic memory. I also demonstrated integration of perceptual feedback and motor feedforward responses during propositional speech production within the medial planum temporale, associated with suppression of activity in secondary somatosensory cortex within the parietal operculum. In the verbal memory study, I demonstrated that successful encoding of heard sentences was associated with greater activity in cortical regions associated with semantic processing, but lower activity within early auditory cortex; implying a “top-down” effect on early perceptual cortex, related to sustained auditory attention. Patients with AD did not show this top-down effect. In addition, less activity was observed during encoding in AD patients, compared to MCI patients or controls, in regions associated with motivation. In the medial temporal lobes, there was less activity in AD compared to controls, but higher activity in MCI, consistent with previous reports. During retrieval, there was less activity in frontal executive control systems in AD compared to controls. This was seen in both performance-matched comparisons and in the neural response to a reduction in retrieval performance. MCI patients showed early changes in parietal lobe retrieval performance-related activity. Overall, the reduced verbal encoding performance in AD was related to impairments in the function of both MTL memory-related systems and sustained auditory attention, and was associated with reduced motivation. During free recall, lower performance in AD was associated with impairment of frontal cognitive control. Therefore, I have shown that verbal episodic memory impairment in AD is the consequence of altered activity in multiple cognitive networks, in addition to the well-recognised impairments in the MTL-memory network. These results have implications for future therapeutic interventions to improve memory function in this patient group, highlighting the potential use of drugs that enhance attention, motivation and frontal executive function