The Old Map Was Wrong
For decades, neuroscience treated intelligence like a real estate problem. You had a fixed amount of square footage — grey matter volume in specific regions — and the question was always which rooms mattered most. Prefrontal cortex here, parietal lobe there. Find the smart room, measure its volume, predict the person.
That framing was incomplete in ways researchers are only now fully confronting.
In March 2020, a team led by Dr. Laurena Holleran, Lecturer in Clinical Neuroscience at NUI Galway, and Professor Gary Donohoe, Established Professor at NUI Galway's School of Psychology and Centre for Neuroimaging Cognition and Genomics, published findings in The American Journal of Psychiatry that pushed the field decisively in a different direction. Their conclusion, drawn from the largest meta-analysis of brain structure and cognitive function in schizophrenia to date: general cognitive ability isn't about any single brain region. It's about the wiring between them. Across the whole brain. All at once.
Diffusion Tensor Imaging and the Hidden Highway
The imaging technique at the center of this work is diffusion tensor imaging — DTI. Unlike structural MRI, which shows you the shape and volume of grey matter, DTI tracks how water molecules diffuse through the brain's white matter tracts. White matter is the cabling. Axons wrapped in myelin, bundled into highways that connect distant cortical regions. DTI gives you a proxy measure called fractional anisotropy, which describes how directionally constrained that water diffusion is. Higher fractional anisotropy means more organized, more efficient tract structure. Think of it as the difference between a well-laid fiber optic cable and a tangled extension cord thrown across a room.
Holleran's team asked a straightforward question: does variation in fractional anisotropy across the whole brain correlate with variation in measured IQ? Not in one tract. Not in ten. Across the entire white matter architecture simultaneously.
The answer was yes. Higher fractional anisotropy was associated with higher IQ. Efficient connection pathways across the entire brain provided the neural network that supported general cognitive function.
1,717 Brains and 40 Scientists
This wasn't a single-site study with fifteen undergrads in a scanner on a Tuesday afternoon. Over 40 scientists from institutions around the world contributed brain MRI scans and cognitive function measures from 1,717 participants. The sample included both healthy individuals and patients diagnosed with schizophrenia.
Getting that many brains measured consistently is genuinely difficult. Scanner manufacturers differ. Protocol sequences differ. Voxel sizes differ. A tract that looks fast in one scanner might look slow in another for purely technical reasons. So the team had to develop a new harmonization method — a way to standardize data collection and analysis across all contributing sites. That methodological contribution is arguably as important as the biological finding itself, because it opens the door to meta-analyses at scale that simply weren't possible before.
The work ran under the umbrella of ENIGMA: Enhancing Neuroimaging Genetics through Meta-Analysis, specifically the Schizophrenia Working Group. If you've been following large-scale neuroimaging collaborations, ENIGMA is the consortium that turned small-sample neuroscience into something resembling population science.
Intelligence Looks the Same in Health and Illness
Here's where the study gets genuinely interesting for clinicians. The team didn't just look at healthy people. They deliberately compared the structural-cognitive relationship in two groups: healthy controls and people living with schizophrenia.
Schizophrenia reliably involves cognitive impairment. Not psychosis per se, the hallucinations and delusions are more visible, but it's the cognitive deficits that typically predict long-term functional outcomes. Employment. Social functioning. Whether someone can manage their own affairs. If white matter efficiency underpins IQ, and schizophrenia damages white matter, you'd expect the structural-cognitive relationship to look different in patients. Weaker, maybe. Or restricted to different tracts.
It didn't.
The relationship between whole-brain white matter integrity and cognitive ability held across both groups. The same wiring principles that explain why one healthy person scores higher on an IQ test than another appear to explain cognitive performance in schizophrenia patients too. The absolute values differ, patients showed reduced fractional anisotropy on average, but the architecture of the relationship was conserved.
That's a quiet but consequential finding. It means the neural infrastructure of intelligence isn't fundamentally broken in schizophrenia. It's degraded, and that degradation carries functional cost. But the system still operates by the same rules.
Why This Matters Beyond the Scanner
Holleran framed the clinical implication plainly in her comments on the findings: understanding the neural basis of cognitive function is essential so that effective therapies can address cognitive difficulties in schizophrenia. Not just the positive symptoms that bring people into hospital, but the quieter deficits that determine whether someone can work, maintain relationships, or navigate daily life independently.
If you know that whole-brain white matter efficiency predicts real-world outcomes, and this study strengthens that link, you get a measurable target. You get a biomarker you can track longitudinally. You get a way to ask whether an intervention actually improves the thing that matters, rather than just improving the thing that's easy to measure on a rating scale.
The study doesn't claim causation. DTI shows association, not mechanism. Fractional anisotropy could be a cause of higher IQ, a consequence of more cognitive stimulation over a lifetime, or both operating in parallel. But association across 1,717 brains in a harmonized dataset is a stronger signal than any single-site study has produced before.
What the Whole-Brain Finding Displaces
The older "regionalist" framing of intelligence, that specific lobes carry specific cognitive loads, isn't wrong. The prefrontal cortex does matter for working memory. The parietal lobe does matter for spatial reasoning. But those regions don't compute in isolation. They're nodes. The intelligence lives in the edges.
Holleran and Donohoe's work makes that metaphor measurable. You can now quantify the efficiency of the edges at scale, and show that the sum total tracks general cognitive ability more robustly than any individual node's properties. That's a shift from mapping rooms to mapping the wiring closet.
And the fact that this pattern holds in both healthy brains and diseased ones suggests something reassuring: the architecture of cognition is more stable than we assumed. The system bends under the load of schizophrenia, but it doesn't break into a different shape.
For researchers building cognitive remediation programs, that's actionable. For patients wondering whether their cognitive difficulties represent some fundamental reorganization of their brain, that's hopeful. The same wiring that carries a healthy thought carries theirs, too. Just with more resistance along the way.
And resistance, unlike broken infrastructure, is something you can potentially work around. Or train through. That's the therapeutic opening this whole-brain perspective creates.
Further reading
- Generalists, Not Specialists, The Brain’s Real Wiring
- Mice Don’t See Clearly, So They Move Like Detectors
