Music as a probe of default mode network function in young-onset Alzheimer's disease

Background:Large-scale brain network disintegration is a unifying theme in neurodegenerative disease with converging lines of evidence implicating the default mode network (DMN) in Alzheimer’s disease (AD). Most previous functional MRI (fMRI) work in AD has assessed the DMN role in memory but emerging evidence points to a much broader pathophysiological phenotype. Sound is an effective and clinically-relevant stimulus to address this: key DMN components are involved in sound decoding and patients with AD commonly experience difficulty perceiving complex auditory environments. In this activation fMRI study we used music as a model paradigm to assess changes in DMN function in young-onset AD. Methods: 32 patients fulfilling criteria for young-onset AD and 18 healthy age-matched controls underwent sparse-acquisition fMRI whilst passively listening to 8-second musical melodies in which key informational properties of temporal regularity (isochronous versus anisochronous) and familiarity (familiar versus novel) were manipulated in a factorial design. fMRI data were analysed in SPM8 and patient and healthy control groups were compared using 2nd-level t-tests. Results:Relative to healthy controls, AD patients showed enhanced activation of temporoparietal cortex for processing anisochronous versus isochronous melodies, but reduced activation of retrosplenial cortex, temporoparietal and medial prefrontal cortices for processing unfamiliar versus familiar melodies. Conclusions:Our findings delineate a novel pathophysiological phenotype of DMN dysfunction in this young-onset AD cohort. Relative to healthy individuals, these AD patients showed a complex profile of altered DMN activity during decoding of information in music, with both abnormally enhanced and reduced engagement of DMN components when processing particular musical properties. This work illustrates the potential of music as a novel probe of DMN dysfunction and the potential utility of fMRI in uncovering bidirectional network activity shifts in AD. Future work should assess the value of this paradigm as a sensitive disease marker, including during the presymptomatic phase of genetic AD.