Dementia symptoms are similar to the behavioral and cognitive signs of severe psychotic illnesses. Shared brain changes were still, and it hadn’t been determined how important they are for patients with at-risk disease stages. For a study, researchers sought to compare the structural magnetic resonance imaging (MRI) patterns of behavioral-variant frontotemporal dementia (bvFTD), Alzheimer’s disease (AD), and schizophrenia using machine learning; to estimate predictability in bvFTD and schizophrenia patients using sociodemographic, clinical, and biological data; and to look at prognostic significance, genetic underpinnings, and progression in patients with clinical high-risk (CHR) states for psychosis or recent-onset depression (ROD).

In order to identify and compare diagnostic patterns, the study included 1,870 participants from 5 cohorts, including patients with bvFTD (n = 108), established AD (n = 44), mild cognitive impairment or early-stage AD (n = 96), schizophrenia (n = 157), or major depression (n = 102). Additionally, patients with CHR (n = 160) or ROD (n = 161) were included to examine the prognostic relevance and progression of the patterns. Healthy people (n = 1,042) were used to calibrate age- and cohort-related data. Between April 2020 and April 2022, data analysis was done on data acquired between January 1996 and July 2019. Case assignments were based on diagnostic patterns, sociodemographic, clinical, and biological data, 2-year functional outcomes, genetic separability of patients with CHR and ROD who express certain patterns more or less, pattern progression from baseline to follow-up MRI scans in patients with nonrecovery vs. preserved recovery, and case assignments are based on these factors in patients with nonrecovery or preserved recovery.

The mean (SD) age of the 1,870 included patients was 38.0 (19.3) years, with 902 (48.2%) of them being female. In contrast to the temporo-limbic AD patterns, which were more prevalent in patients with schizophrenia (65 of 157; 41.2%) and severe depression (22 of 102; 21.6%), the bvFTD pattern, which includes prefrontal, insular, and limbic volume decreases, was more prominent in the individuals. High body mass index, psychomotor slowness, affective disinhibition, and paranoid ideation all contributed to the prediction of bvFTD expression (R2 = 0.11). In 92 out of 108 patients (85.5%) with bvFTD, the schizophrenia pattern was present, and it was associated with the C9orf72 variation, oligoclonal banding in the cerebrospinal fluid, cognitive impairment, and younger age (R2 = 0.29). Polygenic risk scores for frontotemporal dementia, Alzheimer’s disease, and schizophrenia predicted 2-year psychosocial deficits in patients with CHR and were predicted by bvFTD and schizophrenia pattern expressions. Findings had nothing to do with either accelerated brain aging or AD. Last but not least, the advancement of the bvFTD/schizophrenia pattern over a year separated patients with nonrecovery from those with intact recovery.

There may be neurobiological connections between bvFTD and psychosis, with changes to the prefrontal and salience systems being the main emphasis. More transdiagnostic studies are required to find similar pathophysiological mechanisms underlying the neuroanatomical interface between the two disease spectra.