Abstract
The profile of brain structural abnormalities in schizophrenia is still not fully understood, despite decades of research using brain scans. To validate a prospective meta-analysis approach to analyzing multicenter neuroimaging data, we analyzed brain MRI scans from 2028 schizophrenia patients and 2540 healthy controls, assessed with standardized methods at 15 centers worldwide. We identified subcortical brain volumes that differentiated patients from controls, and ranked them according to their effect sizes. Compared with healthy controls, patients with schizophrenia had smaller hippocampus (Cohen's d=-0.46), amygdala (d=-0.31), thalamus (d=-0.31), accumbens (d=-0.25) and intracranial volumes (d=-0.12), as well as larger pallidum (d=0.21) and lateral ventricle volumes (d=0.37). Putamen and pallidum volume augmentations were positively associated with duration of illness and hippocampal deficits scaled with the proportion of unmedicated patients. Worldwide cooperative analyses of brain imaging data support a profile of subcortical abnormalities in schizophrenia, which is consistent with that based on traditional meta-analytic approaches. This first ENIGMA Schizophrenia Working Group study validates that collaborative data analyses can readily be used across brain phenotypes and disorders and encourages analysis and data sharing efforts to further our understanding of severe mental illness.
Introduction
Despite decades of neuroimaging, the structural "fingerprint" of schizophrenia has remained elusive. Most previous studies suffered from small sample sizes, leading to inconsistent or even contradictory results. Furthermore, traditional "retrospective" meta-analyses—which pool results from already-published papers—are often hampered by publication bias and the "apples-to-oranges" problem of different sites using different analysis software and statistical models.
The ENIGMA consortium sought to overcome these hurdles by creating a coordinated, global framework to analyze raw data using identical methods. This approach provides the statistical power necessary to rank brain abnormalities by their effect sizes and investigate how factors like medication and illness duration influence brain structure.
Methods
The researchers employed a prospective meta-analysis approach. Instead of just reading old papers, they coordinated with 15 international research centers to run a standardized analysis pipeline on their local datasets.
- Image Processing: All sites used FreeSurfer, an automated tool for segmenting subcortical structures (like the hippocampus, amygdala, and thalamus) from T1-weighted MRI scans.
- Statistical Modeling: They used univariate linear regressions to compare patients and controls, adjusting for age, sex, and total intracranial volume (ICV).
- Effect Size: Results were standardized using Cohen’s d, allowing for a direct comparison of the "magnitude" of abnormality across different brain regions.
- Moderator Analysis: The team used meta-regression to see if variables like the age of onset, duration of illness, or the type of antipsychotic medication explained the differences in volume observed between different study sites.
Key Results
The study established a clear ranking of subcortical changes. The hippocampus showed the most robust deficit, while the lateral ventricles showed the most significant enlargement.
The study reported several significant differences between schizophrenia patients and healthy controls (all values are Cohen’s d effect sizes):
- Significant Volume Reductions:
- Hippocampus ($d = -0.46$)
- Amygdala ($d = -0.31$)
- Thalamus ($d = -0.31$)
- Nucleus Accumbens ($d = -0.25$)
- Intracranial Volume ($d = -0.12$)
- Significant Volume Increases:
- Lateral Ventricles ($d = 0.37$)
- Pallidum ($d = 0.21$)
No detectable differences were found for the caudate or putamen when simply comparing groups, though the putamen was linked to illness duration in later analyses.
Key Insights: Medication and Illness Duration
The findings suggest a complex interplay between the disease and its treatment. While most structures shrink in schizophrenia, the pallidum and putamen actually showed volume increases.
These increases were positively correlated with the duration of illness and antipsychotic dosage (CPZ equivalents), suggesting they may be a side effect of medication rather than a primary feature of the disorder. Furthermore, hippocampal volume loss was most severe in cohorts with a higher proportion of unmedicated patients. This suggests that untreated psychosis may be particularly damaging to the hippocampus, or conversely, that medication might have a partially "protective" effect on this region.
Conclusion
This paper serves as a "proof of concept" for the ENIGMA consortium, proving that large-scale, decentralized collaboration can produce much more reliable neurobiological markers than isolated research efforts. While the study provides a definitive "snapshot," its cross-sectional design cannot definitively say how these brain volumes change within a single individual over time. Future work by the ENIGMA group plans to expand this work to include cortical thickness, surface area, and longitudinal mapping to further our understanding of severe mental illness.
Summary updated