Neuronal and glial dysfunction, white matter hyperintensities and cognition in ageing and Alzheimer's disease.

Abstract

This cross-sectional study examined associations between multiple fluid biomarkers of neuronal and glial dysfunction (plasma neurofilament light chain, CSF growth-associated protein 43 and CSF soluble triggering receptor expressed on myeloid cells 2), total white matter hyperintensity volume and episodic memory and executive function performance in the context of Alzheimer's disease biomarker status. A total of 563 participants (mean age = 71.9 years, standard deviation = 7.2) from the Alzheimer's Disease Neuroimaging Initiative were classified by the amyloid-β/tau/neurodegeneration framework into no Alzheimer's disease pathology (n = 176), suspected non-Alzheimer's disease pathophysiology (n = 87) or Alzheimer's disease continuum (n = 300) groups. Participants completed baseline neuropsychological assessment, plasma/CSF biomarker collection and MRI. Analyses explored the relative contributions of biomarkers to episodic memory and executive function performance and whether relationships varied by amyloid-β/tau/neurodegeneration group status. Across all participants, neurofilament light chain ( $\hat{\beta }$ = -0.14, P < 0.001) and growth-associated protein 43 ( $\hat{\beta }$ = -0.13, P < 0.001) were the strongest biomarkers associated with episodic memory performance, such that greater levels were associated with worse episodic memory. There was a group by growth-associated protein 43 interaction with episodic memory: greater growth-associated protein 43 was associated with lower episodic memory performance in participants classified as Alzheimer's disease continuum relative to the no Alzheimer's disease pathology group ( $\hat{\beta }$ = -0.26, P < 0.001). No robust associations between biomarkers and executive function performance or between soluble triggering receptor expressed on myeloid cells 2, white matter hyperintensity volume and cognition were observed. Biomarkers of neuro-axonal injury and synaptic dysfunction may independently contribute to episodic memory performance across participants with differing amyloid-β/tau/neurodegeneration profiles. Growth-associated protein 43 may predict worse episodic memory performance in participants with greater Alzheimer's disease pathology. These biomarkers of neuronal dysfunction may serve as domain-specific cognitive correlates in the context of Alzheimer's disease biomarker status.