Psychosis in Alzheimer’s Disease (AD) is defined as the presence of delusions, i.e. fixed false beliefs, or visual or auditory hallucinations (Cummings et al., 2020, Fischer et al., 2020). These symptoms are present in ≈41% of AD patients (Ropacki and Jeste, 2005) and associate with worse cognitive function, greater decline of cognition over time and increased mortality (Ismail et al., 2022, Jeste et al., 1992, Koppel et al., 2012, Koppel et al., 2014b, Ropacki and Jeste, 2005, Scarmeas et al., 2005, Vilalta-Franch et al., 2013, Wilson et al., 2006, Zahodne et al., 2015). They can be highly distressing to both patients and caregivers, resulting in increased rates of patient institutionalization (Scarmeas et al., 2005).

Clinical and research criteria have been recently developed to guide both the diagnosis of psychosis in a neurodegenerative context and to anchor research in a neurobiological framework based on currently available biomarkers hallucinations (Cummings et al., 2020, Fischer et al., 2020). Indeed, the neuropathological mechanisms underlying the emergence of these symptoms are not yet well understood and therapies for these symptoms are lacking.

One possibility is that psychotic symptoms are a consequence of impaired cognition, resulting from an inability to remember things or orient oneself in space and time, without necessarily reflecting a different pathophysiology. On the other hand, it has been shown that psychotic symptoms are more frequent in AD patients with increased typical neuropathology load, especially tau accumulations (Zubenko et al., 1991), cerebral amyloid angiopathy (Vik-Mo et al., 2019), concomitant Lewy Body (LB) pathology (Farber et al., 2000) or arteriosclerotic leukoencephalopathy (Fischer et al., 2016, Vik-Mo et al., 2019). These studies suggest that the presence of psychosis reflects either the presence of comorbidities, such as LBs, which associate with psychotic symptoms in Lewy Body dementia (LBD), or an increase in typical AD neuropathology burden, namely tau aggregates in the form of neurofibrillary tangles (NFTs) and amyloid-beta (Aβ) deposition in the form of neuritic plaques (NPs). Importantly, neuropsychiatric-neuropathology studies have found that psychosis is more common in AD+LBs versus those with pure LBD and pure AD (Chung et al., 2015, Devanand et al., 2022, Gibson et al., 2022), suggesting interactions between these two pathologies.

Differential associations between different neuropsychiatric symptoms and specific neuropathology has been supported not only in these neuropathology studies, but also by the recent finding of specific genetic risk loci for psychosis in AD in a genome-wide association study (DeMichele-Sweet et al., 2021). Moreover, neuroimaging studies of psychosis in AD have revealed increased atrophy most consistently in frontotemporal regions, substantiated by single-photon emission computed tomography and positron emission tomography studies indicating hypoperfusion of right frontal and temporal lobes (Ismail et al., 2022, Murray et al., 2014).

These findings suggest there is pathophysiological specificity to psychosis and open an avenue to explore the underlying mechanisms associated with these symptoms. However, to date, volumetric neuroimaging findings have not been assessed in a neuropathologically defined population, rendering it possible that differences could be due to imbalances in tau propagation. Here, we explored the cognitive profiles, cortical and subcortical volumes and concomitant neuropathology of a population of patients with AD neuropathological change (ADNC) with and without psychosis. We hypothesized that psychotic symptoms would be associated with a higher burden of Alzheimer’s disease and Lewy body co-pathology and therefore with a worse disease progression which might reflect on regional brain atrophy.