Brain structure and connectivity as predictors of functional decline and neuropsychiatric symptoms in Alzheimer's disease and mild cognitive impairment

  • Malgorzata Raczek

Student thesis: Doctoral Thesis


Background: Neuropsychiatric symptoms affect most people with Alzheimer’s disease (AD) and have previously been associated with structural and functional brain changes. Decline in the ability to perform activities of daily living (ADL) is a core feature of AD. This study aimed to establish whether there were structural and functional neural correlates at baseline that predicted faster decline in ADL as well as the emergence and severity of neuropsychiatric symptoms at 3 year follow-up.

Method: One hundred and eighty (n=180) patients with a diagnosis of AD, amnestic mild cognitive impairment (aMCI), or subjective cognitive decline (SCD) had a magnetic resonance imaging (MRI) scan at their initial assessment in memory clinic. Of those, 148 patients met the inclusion/exclusion criteria for this study and were invited to participate. Seventy nine of those (44 male, 35 female), consented to participate and were followed up at 3 years. The follow-up assessment of cognitive and functional status repeated clinical measures used at baseline, including Adenbrooke’s Cognitive Examination - Revised (ACE-R) and Bristol Activities of Daily Living Scale (BADLS). Data relating to neuropsychiatric symptoms were measured with Neuropsychiatric Inventory (NPI). Additionally, the clinical assessment collected information relating to educational status, place of residence and co-morbidities.

T1 and T2* MRI were obtained at baseline using the Siemens Avanto 1.5 T scanner for volumetric and resting-state functional MRI (rs-fMRI) analysis. Pre-processing of the images was performed with statistical parametric mapping 12th edition (SPM12) voxel-based morphometry (VBM) pipeline for the structural imaging and FMRIB software library (FSL) package for the rs-fMRI. The correlation analysis was performed between the change in ADL status as measured by BADLS and neuropsychiatric symptoms as measured by NPI and the regional brain volume and activity in resting state networks.

Results At baseline, 29 participants met the diagnostic criteria for AD, whereas 50 had a non-AD diagnosis of either aMCI (n=42) or SCD (n=8). At 3-year follow-up, 53 participants met the criteria for AD and 26 had a non-AD diagnosis.

VBM analysis showed negative correlation with regional grey matter (GM) volume at baseline and presence of neuropsychiatric symptoms at 3-year follow up in people with AD. Brain regions that correlated negatively with anxiety included anterior/middle cingulate cortex, which has previously been implicated in affective processing and links to default mode network (DMN). Orbitofrontal cortex and cerebellum volumes correlated negatively with high scores in total NPI score and in Psychosis sub-syndrome.

Brain connectivity analysis in participants with AD at follow-up (n=41) showed positive correlation between neuropsychiatric symptoms and connectivity in DMN as well as in the salience, as well as left and right frontoparietal networks. There was a negative correlation between connectivity in DMN and appetite.

We did not find significant correlation between ADL and regional GM volume or brain connectivity.

Conclusions: Neuropsychiatric symptoms in AD may be associated with regional atrophy and altered connectivity in brain networks that precede the onset of symptoms. Whilst this may signal a potential of using structural and functional brain imaging in improving prognosis for people with AD, the associations are complex and may follow a non-linear trajectory of change, especially in functional connectivity changes. Studies on larger population and with a wider range of functional ability and neuropsychiatric symptoms may further improve our knowledge about the potential of imaging data in predicting the course of illness for individual patients.
Date of AwardFeb 2023
Original languageEnglish
Awarding Institution
  • University of Brighton
SupervisorSube Banerjee (Supervisor), Mara Cercignani (Supervisor) & Stephanie Daley (Supervisor)

Cite this