Summary: A new study has decoded the structural white matter communication networks of the brain in individuals with bipolar disorder. The meta-analysis harmonized diffusion MRI data from 449 people with bipolar disorder and 510 healthy controls across 16 international research sites via the ENIGMA Bipolar Disorder Working Group.
Leveraging graph theory network analysis to model the brain like a massive transit grid, the study unmasked widespread, subtle reductions in structural network connection density, longer signal routing lengths, and a rigid, hyper-dependence on a limited set of centralized brain hubs.
Key Facts
- The ENIGMA Consortium Scalability: By pooling cross-sectional imaging data from 16 global sites, the ENIGMA consortium achieved the statistical power necessary to detect subtle, system-wide white matter deviations that are completely invisible in localized, single-site clinical samples.
- Graph Theory Transport Modeling: Researchers used diffusion MRI and graph theory to map the brain’s white matter communication infrastructure. Brain regions were modeled as “nodes” and neural pathways as “routes,” allowing investigators to calculate the precise speed and efficiency of data moving across the central nervous system.
- Widespread Connection Deficits: Compared to healthy controls, individuals with bipolar disorder exhibit significantly less densely connected neural networks, diminished information exchange efficiency, and longer, more circuitous communication routes between vital brain sectors.
- The Vulnerable Functional Circuits: The structural degradation is not random; it heavily concentrates within critical functional networks responsible for processing internal thought and executing behavioral control, specifically fronto-limbic circuits (emotion regulation), basal ganglia pathways (motivation and reward), the default mode network (self-reflection), and the salience network (prioritizing information).
- Linking Architecture to Illness History: The study found distinct anatomical patterns tied directly to a patient’s unique clinical trajectory. A longer illness duration was linked to broad reductions in overall network routing efficiency and compromised wiring connecting the amygdala and hippocampus. Conversely, a later age of onset was tied to isolated structural modifications connecting the cerebellum, thalamus, and fronto-limbic pathways.
- Psychosis and Mania Signatures: Individuals with a history of psychosis demonstrated extensive systemic network disorganization overall. Meanwhile, those who experienced a higher count of manic episodes tended to have increased connectivity in certain fronto-limbic pathways, which may reflect an illness-related alteration or a structural compensatory adaptation by the brain.
- Biological Nomenclature Medication Profiling: This marks the first large-scale study to categorize psychiatric medications by their specific underlying biological mechanisms rather than generic names. The data revealed that selective serotonin reuptake inhibitors (SSRIs), anticonvulsants, and antipsychotics correlate with distinct structural connectivity variations across emotion-regulation and cognitive-control pathways.
Source: USC
New research from the Mark and Mary Stevens Neuroimaging and Informatics Institute (Stevens INI) at the Keck School of Medicine of USC has discovered subtle but widespread differences in the brain’s communication networks in people with bipolar disorder, offering new insight into how illness severity and treatment may relate to brain wiring.
Published in Biological Psychiatry, the study was led by Leila Nabulsi, PhD, a senior research associate at the Stevens INI, together with Dara M. Cannon, PhD, professor at the University of Galway, Ireland. The team analyzed brain scans from 449 people with bipolar disorder and 510 healthy controls across 16 international research sites through the ENIGMA Bipolar Disorder Working Group.
This work was made possible by ENIGMA, an international consortium founded and led in part by Paul M. Thompson, PhD, associate director of the Stevens INI. ENIGMA brings together researchers worldwide to pool their brain imaging and clinical data, allowing them to detect subtle patterns that would be difficult to identify in smaller studies.
Mapping the brain’s communication system
Using diffusion MRI, an advanced imaging technique that maps the brain’s neural pathways, the researchers examined how different regions of the brain are structurally connected. White matter acts as the brain’s communication infrastructure, allowing different regions to send signals to one another. In bipolar disorder, where patients have episodes of depression and mania or hypomania, changes in these communication pathways lead to disruptions in mood, emotion regulation, reward processing, and cognitive control.
“Bipolar disorder is defined by changes in mood and behavior, but those symptoms arise from complex brain circuits that do not operate in isolation,” said Leila Nabulsi, PhD, the study’s first author. “While previous studies identified changes in individual brain regions, we still know less about how these regions are connected as part of larger networks. By viewing the brain as an interconnected system, we can now see how differences in communication pathways relate to the circuits that regulate mood and to features of the illness.”
Previous large-scale MRI studies from the ENIGMA Consortium found that people with bipolar disorder tend to have differences in gray matter, the tissue that contains most neuronal cell bodies. Less is known about how white matter pathways are organized into large-scale brain networks, and how the efficiency of those networks relates to illness severity and treatment.
To address this, the research team used diffusion MRI and a network analysis approach known as graph theory. In simple terms, this approach models the brain like a transportation system: brain regions are treated as “nodes,” and the connections between them as “routes.” Researchers can then estimate how efficiently information may move across the network.
Subtle but widespread network differences
The study found that people with bipolar disorder showed subtle but consistent differences in how brain networks are organized. Compared with psychiatrically healthy controls, participants with bipolar disorder had less densely connected networks, lower efficiency in how information is exchanged, and longer routes for communication between brain regions.
Their brain networks relied more on highly connected “hub” regions, key points that help coordinate communication across the brain. This pattern may reflect the brain’s attempt to adapt to these network changes, with information flowing less directly and relying more on a limited set of pathways.
The most pronounced differences were seen in networks involved in emotion regulation, reward processing, attention, and self-reflection; functional systems known to be affected in bipolar disorder. These included fronto-limbic circuits, which help regulate emotion; basal ganglia pathways involved in motivation and reward; and regions within the brain’s default mode and salience networks, which are important for internal thought and prioritizing relevant information.
“In people with bipolar disorder, the brain’s communication system may be less efficiently organized, with information taking less direct routes across the network,” Nabulsi said. “We found consistent effects across a large, international sample, and they may help explain clinical differences and treatment effects in patients.”
“Psychiatric disorders are biologically complex, and no single research site can capture the full picture on its own,” said Thompson. “By harmonizing data across research groups worldwide, ENIGMA gives us the statistical power to identify brain signatures of mental health conditions and discover new ways to resist them.”
Linking brain wiring to illness history
The study also related these brain network differences to clinical features of bipolar disorder. Individuals who had been ill longer showed broader reductions in how efficiently large-scale brain networks communicate, along with altered connectivity between the amygdala and hippocampus, regions that are central to emotions and memory.
A later age of onset, by contrast, was linked to a different pattern: more pronounced changes in specific circuits connecting the cerebellum, thalamus, and fronto-limbic pathways also involved in emotion regulation. Individuals who had experienced psychosis showed more pronounced differences in brain network organization overall.
People who experienced a greater number of manic episodes tended to have higher connectivity in certain fronto-limbic pathways, which may reflect illness-related changes or the brain’s attempt to adapt to these network alterations.
Understanding treatment in context
The study also examined how different types of medication relate to brain network organization; the first large-scale effort to assess treatment effects on white matter connectivity using network-based approaches. In addition to grouping medications by traditional nomenclature, the researchers analyzed them based on their underlying biological mechanisms to understand how different mechanisms of action may be linked to changes in brain connectivity.
Antidepressant use, particularly selective serotonin reuptake inhibitors (SSRIs), was linked to less efficient communication across the brain overall, and to specific changes in limbic circuits involved in emotion regulation. People taking anticonvulsant and antipsychotic use were also linked to changes in circuits related to emotion regulation and cognitive control.
“This study should not be interpreted as guidance for changing treatment,” Nabulsi said. “Medications are prescribed for a variety of clinical reasons, and people receiving certain treatments may also differ in illness history or symptom severity.
“We found that treatment exposure is an important factor to consider when studying the biology of bipolar disorder, so we can separate the effects of the illness from those of the medications. We hope this encourages future studies to take these factors into account.”
Toward more personalized care
The researchers emphasize that these findings do not show that medications caused the observed brain differences. Because the study was cross-sectional, meaning participants were scanned at a single point in time, it cannot determine cause and effect.
Future longitudinal studies are needed that follow individuals over time to clarify how treatment, illness progression, and brain network changes are related. Ongoing work within the ENIGMA Bipolar Disorder Working Group is now addressing these questions in large-scale datasets from patients assessed repeatedly over time.
The study also shows that large-scale brain network analysis can be successfully carried out across multiple international sites, despite differences in scanners, imaging protocols, and patient populations. This type of harmonized approach brings researchers closer to identifying reliable and biologically grounded markers to inform diagnosis, prognosis, and treatment.
“Bipolar disorder affects millions of people worldwide, yet treatment response is highly variable,” said Arthur W. Toga, PhD, director of the Stevens INI. “Studies like this help us move closer to understanding the brain circuits involved, which is an essential step toward more personalized and biologically informed approaches to care.”
By identifying how large-scale brain networks are linked to illness severity and treatment exposure, the study provides a framework to understand bipolar disorder at the level of brain systems rather than isolated regions. The research team hopes this work will lay the groundwork for future longitudinal studies that follow patients over time and determine whether these network patterns predict symptom course, treatment response, and risk for future episodes.
Toga said the work also highlights the value of combining brain imaging with careful clinical evaluation.
“The more precisely we can map the brain systems involved in bipolar disorder, the better positioned we are to develop tools that may eventually help clinicians predict illness trajectories and tailor interventions,” he said. “This is the kind of foundational science we do at the Stevens INI that moves the field toward more personalized health care.”
About the study
In addition to Nabulsi, Cannon and Thompson, study authors include Melody J.Y. Kang, Neda Jahanshad, Genevieve McPhilemy, Fiona M. Martyn, Bartholomeus Haarman, Colm McDonald, Brian Hallahan, Stefani O’Donoghue, Dan J. Stein, Fleur M. Howells, Freda Scheffler, Henk S. Temmingh, David C. Glahn, Amanda Rodrigue, Edith Pomarol-Clotet, Eduard Vieta, Joaquim Radua, Raymond Salvador, Andriana Karuk, Erick J. Canales-Rodríguez, Josselin Houenou, Pauline Favre, Mircea Polosan, Arnaud Pouchon, Paolo Brambilla, Marcella Bellani, Philip B. Mitchell, Gloria Roberts, Udo Dannlowski, Tiana Borgers, Susanne Meinert, Kira Flinkenflügel, Jonathan Repple, Elisabeth J. Leehr, Dominik Grotegerd, Tim Hahn, Michèle Wessa, Mary L. Phillips, Lea Teutenberg, Tilo Kircher, Benjamin Straube, Olaf Steinstraeter, Frederike Stein, Florian Thomas-Odenthal, Nina Alexander, Paula L. Usemann, Andreas Jansen, Michael Berk, Orwa Dandash, Nadine Parker, Chao Suo, Sophia I. Thomopoulos, Ole A. Andreassen, and Christopher R.K. Ching for the ENIGMA Bipolar Disorder Working Group.
Funding: This work was supported by a 2025 NARSAD Young Investigator Grant, the Milken Institute Baszucki Brain Research Fund, the Irish Research Council, the Health Research Board, the National Institutes of Health, and other international funding sources.
Key Questions Answered:
A: Graph theory allows scientists to model the human brain exactly like a giant municipal transportation network. Individual brain regions are treated as transit stations (“nodes”) and the white matter pathways connecting them function as the train tracks (“routes”). This enables researchers to mathematically measure how efficiently or poorly information travels across the mind’s grid.
A: The study revealed that the brain’s internal communication lines are less efficiently organized. Signals cannot travel along direct paths; instead, they are forced to take longer, more roundabout routes. Furthermore, the network over-relies on a few heavily congested central “hub” stations to pass data along, which may be the brain’s way of adapting to damaged secondary routes.
A: No, it does not show cause and effect. Because this was a cross-sectional study—meaning patients were scanned at a single point in time, it is impossible to tell if the network differences were caused by the illness itself, the brain trying to adapt, or the medications. It simply highlights that a patient’s medication history is a vital factor that scientists must map to isolate the biology of the condition.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this mental health research news
Author: Laura LeBlanc
Source: USC
Contact: Laura LeBlanc – USC
Image: The image is credited to Stevens INI
Original Research: Open access.
“Structural Brain Network Alterations in Relation to Treatment and Illness Severity in Bipolar Disorder” by Leila Nabulsi, Melody J.Y. Kang, Neda Jahanshad, Genevieve McPhilemy, Fiona M. Martyn, Bartholomeus Haarman, Colm McDonald, Brian Hallahan, Stefani O’Donoghue, Dan J. Stein, Fleur M. Howells, Freda Scheffler, Henk S. Temmingh, David C. Glahn, Amanda Rodrigue, Edith Pomarol-Clotet, Eduard Vieta, Joaquim Radua, Raymond Salvador, Andriana Karuk, Erick J. Canales-Rodríguez, Josselin Houenou, Pauline Favre, Mircea Polosan, Arnaud Pouchon, Paolo Brambilla, Marcella Bellani, Philip B. Mitchell, Gloria Roberts, Udo Dannlowski, Tiana Borgers, Susanne Meinert, Kira Flinkenflügel, Jonathan Repple, Elisabeth J. Leehr, Dominik Grotegerd, Tim Hahn, Michèle Wessa, Mary L. Phillips, Lea Teutenberg, Tilo Kircher, Benjamin Straube, Olaf Steinstraeter, Frederike Stein, Florian Thomas-Odenthal, Nina Alexander, Paula L. Usemann, Andreas Jansen, Michael Berk, Orwa Dandash, Nadine Parker, Chao Suo, Sophia I. Thomopoulos, Paul M. Thompson, Ole A. Andreassen, and Christopher R.K. Ching for the ENIGMA Bipolar Disorder Working Group. Biological Psychiatry
DOI:10.1016/j.biopsych.2026.04.020
Abstract
Structural Brain Network Alterations in Relation to Treatment and Illness Severity in Bipolar Disorder
Background
Large-scale T1-weighted MRI studies have established grey-matter abnormalities in bipolar disorder (BD), with our group contributing to consensus findings. However, structural connectivity, particularly within emotion- and reward-related circuits, remains poorly understood. Diffusion-weighted MRI (dMRI) enables investigation of white-matter pathways, yet prior work is constrained by small samples, methodological heterogeneity, and unclear medication effects. We conducted the largest dMRI network analysis in BD, relating symptom burden and polypharmacy to tractography-derived connectivity and graph-theoretic metrics.
Methods
Cross-sectional structural and diffusion MRI scans from 449 individuals with BD (35.7±12.6 years) and 510 controls (33.3±12.6 years), aged 18–65, were analyzed across 16 ENIGMA-BD sites. Standardized segmentation/parcellation and constrained spherical deconvolution tractography generated individual structural connectivity matrices. Graph-theoretic metrics of global and subnetwork organization were related to symptom severity and medications.
Results
BD showed widespread network alterations (lower density and efficiency, longer path length, and higher betweenness centrality), altered microstructural organization in a limbic–basal ganglia circuit, and abnormal streamline counts in a default-mode/salience/fronto-limbic–basal ganglia network. Longer illness duration, later onset, and psychosis history were associated with greater abnormalities in network architecture, whereas more manic episodes were associated with greater fronto-limbic connectivity. Antidepressant (particularly SSRI), anticonvulsant, and antipsychotic use related to poorer global and fronto-limbic connectivity; no clear lithium effects emerged.
Conclusions
As the largest structural connectivity study in BD, we reveal widespread disruption in reward and emotion-regulation networks influenced by illness severity and medication use. Results show that multisite harmonization is feasible and highlight ENIGMA-BD as a scalable framework for identifying reproducible neurobiological markers.
