Structural brain abnormalities and aggressive behaviour in schizophrenia: Mega-analysis of data from 2095 patients and 2861 healthy controls via the ENIGMA consortium

Jelle Lamsma, Adrian Raine, Seyed Mostafa Kia, Wiepke Cahn, Dominic Arold, Nerisa Banaj, Annarita Barone, Katharina Brosch, Rachel M. Brouwer, Arturo Brunetti, Vince D. Calhoun, Qian Hui Chew, Sunah Choi, Young‐Chul Chung, Mariateresa Ciccarelli, Derin Cobia, Sirio Cocozza, Udo Dannlowski, Paola Dazzan, Andrea de Bartolomeis, Marta Di Forti, Alexandre Dumais, Jesse T. Edmond, Stefan Ehrlich, Ulrika Evermann, Kira Flinkenflügel, Foivos Georgiadis, David C. Glahn, Janik Goltermann, Melissa J. Green, Dominik Grotegerd, Amalia Guerrero‐Pedraza, Minji Ha, Elliot Hong, Hilleke E. Hulshoff Pol, Felice Iasevoli, Stefan Kaiser, В. Г. Каледа, Andriana Karuk, Minah Kim, Tilo Kircher, Matthias Kirschner, Peter Kochunov, Jun Soo Kwon, И. С. Лебедева, Rebekka Lencer, Tiago Reis Marques, Susanne Meinert, Robin Murray, Igor Nenadić, Dana Nguyen, Godfrey D. Pearlson, Fabrizio Piras, Edith Pomarol‐Clotet, Giuseppe Pontillo, Stéphane Potvin, Adrian Preda, Yann Quidé, Amanda Rodrigue, Kelly Rootes-Murdy, Raymond Salvador, Antonín Škoch, Kang Sim, Gianfranco Spalletta, Filip Španiel, Frederike Stein, Florian Thomas‐Odenthal, Andràs Tikàsz, David Tomeček, A. S. Tomyshev, Mario Tranfa, Uyanga Tsogt, Jessica A. Turner, Theo G.m. van Erp, Neeltje E.m. van Haren, Jim van Os, Daniela Vecchio, Lei Wang, Adrian Wroblewski, Thomas Nickl‐JockschatmedRxiv 2024 · DOI

This mega-analysis from the ENIGMA consortium represents one of the largest efforts to date to uncover the neurobiological roots of aggressive behavior in schizophrenia. By pooling data from over 4,900 individuals across 20 international sites, the researchers moved beyond small-scale, inconsistent studies to provide a robust map of how structural brain abnormalities correlate with hostility and violence.

TL;DR

Aggressive behavior in schizophrenia is significantly linked to widespread reductions in cortical volume and white matter integrity. These structural deficits—particularly in the prefrontal cortex and internal capsule—largely drive aggression by increasing the severity of positive symptoms (like delusions) and undermining impulse control.

What Problem Does This Paper Address?

While the vast majority of people with schizophrenia are not violent, the condition is associated with a five-fold increase in the risk of aggressive behavior compared to the general population. Historically, research into the "why" has been fragmented. Some studies pointed to specific brain regions like the amygdala or prefrontal cortex, but small sample sizes and "group-level" analyses (comparing averages rather than individuals) led to inconsistent results.

The authors sought to determine if there is a consistent, individual-level neurobiological signature for aggression in schizophrenia. Specifically, they wanted to know if brain structure directly causes aggression or if it acts through "mediators"—clinical symptoms like hallucinations or poor impulse control that then lead to outward hostility.

Methods

The study utilized a mega-analysis approach, meaning they pooled raw individual data from 20 different sites rather than just combining the results of previous papers.

Key Technical Innovations:

  • Normative Modeling: Instead of simply comparing "aggressive patients" to "non-aggressive patients," the team used warped Bayesian linear regression to create a "normative" model based on the 2,861 healthy controls. They then calculated z-scores for each patient, representing exactly how much an individual’s brain volume or white matter integrity deviated from the expected "normal" for their age and sex.
  • Multimodal Imaging: The study combined T1-weighted MRI (to measure grey matter volume) with Diffusion Tensor Imaging (DTI) to measure white matter microstructural integrity (via Fractional Anisotropy and Mean Diffusivity).
  • Mediation Analysis: They used structural equation modeling to test whether brain abnormalities directly caused aggression or if the effect was "filtered" through positive symptoms, impulse control, or illness insight.

Datasets

The study drew from the ENIGMA Schizophrenia Working Group, involving:

  • 2,095 patients with schizophrenia.
  • 2,861 healthy controls.
  • Data spanned 20 sites in 13 countries, providing high geographic and demographic diversity.
  • Aggression was harmonized across different clinical scales, including the PANSS (Positive and Negative Syndrome Scale), SAPS, and BPRS.

Key Insights and Ideas

  • Global vs. Focal Deficits: Aggression isn't tied to a single "broken" part of the brain. Instead, it is associated with a global reduction in cortical volume and white matter integrity. This suggests that aggression in schizophrenia may stem from a widespread failure of neural systems rather than a localized lesion.
  • The "Symptom Bridge": Most structural deficits (like those in the DLPFC and Internal Capsule) didn't lead to aggression directly. Instead, they increased the severity of delusions and hallucinations and weakened impulse control, which in turn made aggression more likely.
  • The IPL Exception: The Inferior Parietal Lobule (IPL) stood out. Reductions here were associated with aggression independently of positive symptoms. The authors suggest this might be linked to "Theory of Mind" deficits—the inability to correctly infer others' intentions, leading to "defensive" aggression based on social misinterpretation.
  • White Matter Matters: The study highlights the Internal Capsule (IC), a major highway for signals between the cortex and subcortical regions. Reduced integrity here was a strong predictor of poor impulse control and subsequent aggression.

Key Results

The study found several statistically significant associations (using a stringent 1% significance level):

  • Total Cortical Volume (TCV): Lower volume was associated with higher aggression (OR = 0.88).
  • Global White Matter Integrity: Lower integrity was a strong predictor of aggression (OR = 0.72).
  • Regional Reductions: Specific associations were found for the Dorsolateral Prefrontal Cortex (DLPFC) (OR = 0.85), the IPL (OR = 0.76), and the Internal Capsule (OR = 0.76).
  • Mediation: Positive symptoms and poor impulse control mediated the relationship for TCV, GWM, and DLPFC, but not for the IPL.

Limitations and Future Work

  • Cross-Sectional Design: Because the scans and behavioral assessments happened at the same time, the study cannot definitively prove that brain changes precede the behavior.
  • Measurement Sensitivity: Aggression was measured using single items from broader psychiatric scales. The authors note that dedicated, multi-item aggression scales would be more sensitive.
  • The "Psychopathy" Gap: The dataset lacked information on comorbid Antisocial Personality Disorder or psychopathy, which are known to have their own distinct neurobiological signatures.
  • Future Directions: The authors recommend longitudinal studies to see if these brain markers can predict which patients will respond best to specific treatments, such as neuromodulation (TMS/DBS) or targeted cognitive behavioral therapy.

Summary updated

Abstract

This study provides evidence that the co-occurrence of positive symptoms, poor impulse control and aggressive behaviour in schizophrenia has a neurobiological basis, which may inform the development of therapeutic interventions.
BibTeX
@misc{lamsma2024structural,
  author = {Jelle Lamsma and Adrian Raine and Seyed Mostafa Kia and Wiepke Cahn and Dominic Arold and Nerisa Banaj and Annarita Barone and Katharina Brosch and Rachel M. Brouwer and Arturo Brunetti and Vince D. Calhoun and Qian Hui Chew and Sunah Choi and Young‐Chul Chung and Mariateresa Ciccarelli and Derin Cobia and Sirio Cocozza and Udo Dannlowski and Paola Dazzan and Andrea de Bartolomeis and Marta Di Forti and Alexandre Dumais and Jesse T. Edmond and Stefan Ehrlich and Ulrika Evermann and Kira Flinkenflügel and Foivos Georgiadis and David C. Glahn and Janik Goltermann and Melissa J. Green and Dominik Grotegerd and Amalia Guerrero‐Pedraza and Minji Ha and Elliot Hong and Hilleke E. Hulshoff Pol and Felice Iasevoli and Stefan Kaiser and В. Г. Каледа and Andriana Karuk and Minah Kim and Tilo Kircher and Matthias Kirschner and Peter Kochunov and Jun Soo Kwon and И. С. Лебедева and Rebekka Lencer and Tiago Reis Marques and Susanne Meinert and Robin Murray and Igor Nenadić and Dana Nguyen and Godfrey D. Pearlson and Fabrizio Piras and Edith Pomarol‐Clotet and Giuseppe Pontillo and Stéphane Potvin and Adrian Preda and Yann Quidé and Amanda Rodrigue and Kelly Rootes-Murdy and Raymond Salvador and Antonín Škoch and Kang Sim and Gianfranco Spalletta and Filip Španiel and Frederike Stein and Florian Thomas‐Odenthal and Andràs Tikàsz and David Tomeček and A. S. Tomyshev and Mario Tranfa and Uyanga Tsogt and Jessica A. Turner and Theo G.M. van Erp and Neeltje E.M. van Haren and Jim van Os and Daniela Vecchio and Lei Wang and Adrian Wroblewski and Thomas Nickl‐Jockschat},
  title = {{Structural brain abnormalities and aggressive behaviour in schizophrenia: Mega-analysis of data from 2095 patients and 2861 healthy controls via the ENIGMA consortium}},
  year = {2024},
  doi = {10.1101/2024.02.04.24302268},
  howpublished = {medRxiv},
}
Figure