Summary: A new meta-analysis has identified a distinct brain network that connects the diverse patterns of brain atrophy observed in schizophrenia. Analyzing neuroimaging findings from more than 90 published studies and over 8,000 participants, researchers generated an atrophy-connectivity map that converges on regions long implicated in schizophrenia—most notably the insula, hippocampus and fusiform cortex.
This network pattern was consistent across different illness stages and symptom clusters, and it was distinct from networks associated with aging, neurodegenerative disorders and other psychiatric conditions. The discovery supports efforts to develop targeted, network-informed interventions and will inform an upcoming clinical trial that evaluates transcranial magnetic stimulation (TMS) sites connected to this schizophrenia network.
Key Facts
- Unified Network: Schizophrenia-related atrophy across studies localizes to a single, reproducible brain network.
- Core Regions: Overlap in the atrophy connectivity map includes the bilateral insula, hippocampus and fusiform cortex.
- Clinical Implications: Results provide a rationale for trials that target network nodes with noninvasive brain stimulation to refine therapeutic strategies.
Source: Brigham and Women’s Hospital
Overview
Investigators at Mass General Brigham led a large-scale synthesis of neuroimaging findings to resolve divergent results about where and how the brain is affected in schizophrenia. By pooling spatial coordinates of reported atrophy from 90 studies, the team built an integrated map and then linked those anatomical locations to functional brain networks using a coordinate network mapping (CNM) approach. The combined dataset included thousands of individuals across diagnostic groups, enabling robust comparisons and specificity testing.
Published in Nature Mental Health, the study sought to identify whether seemingly heterogeneous sites of brain shrinkage in schizophrenia converge on a common connectivity signature. “We looked for common threads among reports on how schizophrenia affects the brain,” said Ahmed T. Makhlouf, MD, corresponding author and medical director of the Brigham and Women’s Hospital Psychosis Program. “We saw atrophy in many places, but those locations mapped back to one interconnected network.”
Senior author Shan H. Siddiqi, MD, explained the motivation for the work: previous studies often described different focal changes, which may reflect different perspectives on the same underlying network. Using CNM across many datasets allowed the team to reconstruct that broader circuit-level picture.
The aggregated dataset included 1,636 participants with recent-onset schizophrenia, 2,120 with chronic illness, more than 6,000 healthy control participants, and separate cohorts of individuals at genetic or clinical high risk for psychosis. After mapping reported atrophy sites, the authors used the human connectome as a reference wiring diagram to derive the schizophrenia atrophy network.
The derived network reliably overlapped with brain regions previously implicated in schizophrenia, particularly the bilateral insula, hippocampus and fusiform cortex. Importantly, the schizophrenia network was distinct from networks associated with normal aging, Alzheimer’s disease and several other psychiatric disorders, supporting its specificity.
Across symptom subgroups and illness stages, the atrophy connectivity pattern remained stable and showed little change associated with antipsychotic treatment, indicating a robust signature of the disorder. Participants classified as high risk shared some atrophy features with diagnosed patients, but those who progressed to clinical schizophrenia showed a distinctive connectivity profile, suggesting potential predictive value of network patterns for disease conversion.
The authors emphasize that individualized connectome mapping could further tailor insights to single patients, and they report that a clinical trial is planned to assess whether TMS targeting sites that are functionally connected to the schizophrenia network can modulate symptoms or circuit function.
“There is ongoing debate about whether schizophrenia represents a neurodegenerative process,” said Makhlouf. “Our findings identify a unique, unified network that may capture a core anatomical characteristic of the disorder.”
Authorship and disclosures
Lead authors include Ahmed T. Makhlouf and Shan H. Siddiqi. Additional Mass General Brigham contributors listed are William Drew, Jacob L. Stubbs, Joseph J. Taylor, David Silbersweig and Michael D. Fox; other coauthors include Donato Liloia and Jordan Grafman.
Disclosures noted by the authors include intellectual property and consulting relationships related to individualized resting-state mapping and TMS targeting held by some investigators, as well as occasional research funding and speaking fees from companies involved in neuromodulation. These disclosures were reported as unrelated to the present analysis and have not generated royalties tied to this work.
Funding
The work was supported by a range of institutional awards, fellowships and grants, including multiple NIH grants, foundation awards and research funds from academic institutions and philanthropic sources. These funding sources enabled the multi-cohort meta-analytic effort.
About this schizophrenia research news
Author: Cassandra Falone
Source: Brigham and Women’s Hospital
Contact: Cassandra Falone, Brigham and Women’s Hospital
Image credit: Neuroscience News
Original Research: Closed access. “Heterogenous Patterns of Brain Atrophy in Schizophrenia Localize to A Common Brain Network” by Ahmed T. Makhlouf et al., Nature Mental Health.
Abstract
Heterogenous Patterns of Brain Atrophy in Schizophrenia Localize to a Common Brain Network
Neuroanatomical findings in schizophrenia have been inconsistent across studies, complicating efforts to define the disorder’s circuit-level basis. Using a coordinate network mapping meta-analysis that leverages the human connectome, this study synthesized results from 90 published atrophy studies (total n > 8,000) and identified a reproducible connectivity pattern—the schizophrenia atrophy network—that unites diverse anatomical findings.
This network was specific to schizophrenia when compared to atrophy patterns in individuals at high risk for psychosis, normal aging populations, neurodegenerative disorders and other psychiatric conditions. The network remained stable across disease progression and among different symptom clusters. Additionally, schizophrenia atrophy patterns were negatively correlated with lesion locations previously linked to psychosis-related thought processes in an independent sample (n = 181). Overall, these results present a unique, stable and unified network that helps account for heterogeneity observed in prior atrophy research on schizophrenia.