When it comes to intelligence, what factors distinguish the brains of the exceptionally smart from those of average humans?
New research by post-doctoral fellow Michael Cole and colleagues suggests as much as 10% of individual variance in human intelligence can be predicted based on the strength of neural connections between the lateral prefrontal cortex and other regions of the brain.
The authors used functional magnetic resonance imaging (fMRI) to measure brain activity. fMRI is a non-invasive neuro-imaging technique that allows neural activity to be measured indirectly by detecting changes in blood oxygenation, using the so-called blood-oxygen-level-dependent contrast.
The idea, in a very simplified form, is that the neural activity that occurs when participants are engaged in a cognitive task needs energy by means of glucose metabolism; and this, in turn, requires oxygen supply. This means that the more oxygenated blood is present in a brain region (which means a better signal), the more likely it is that some neural activity has taken place.
Their connectivity analysis additionally involved looking at the co-variation of resting-state activity in one region with activity in other regions to indirectly estimate how these regions might be connected.
In their study, the authors first asked participants to perform a task that required working memory resources to identify brain regions that are involved in higher-level “cognitive control” functions.
They then used brain scans from periods when participants were not engaged in a task (“resting state”) to determine the connectivity of these cognitive control regions with several other regions in the brain.
Finally, they correlated their connectivity measures with the average score of two intelligence tests to determine whether connectivity and fluid intelligence were related.
An intelligent guess
Of course, none of these measures can determine or measure intelligence. It is even complicated to determine how these measures are related to intelligence.
The reason for this is that intelligence is not a unified concept. There is more than a century of research and debate on what intelligence is and how it can be measured, with a great number of concepts and tests emerging.
Fluid intelligence is one concept that goes back to the late behavioural scientist Raymond Cattell and refers to a general capacity for logical thinking and problem solving, independent from a person’s knowledge.
While it is true that the authors of the new research used two well established tests for fluid intelligence, it seems quite arbitrary to simply average these two scores to undoubtedly measure it.
There is also an ongoing debate on how resting-state activity should be measured. Unfortunately, it is unclear what participants in Cole’s study did (or thought about) during the chosen rest periods and whether they were instructed to stop thinking about the task.
But this would be important to know for the interpretation of these connectivity measures.
Finally, cognitive control and intelligence are strongly related concepts. One would expect that cognitive control regions are involved in functions that are related to what we believe that intelligence is about.
Given previous research has already demonstrated the involvement of these brain regions in intelligence, the current paper shows that communication of cognitive control regions with other regions in the brain is also important.
But does this reveal a novel major predictive factor of intelligence? I believe this study surely adds to the growing body of literature investigating the neural basis of intelligence.
Unsurprisingly, though, as the authors point out themselves, it shows a broad and complex concept such as intelligence seems to be supported by a variety of neural factors of which global connectivity of core regions is just one, and whose interplay is still poorly understood.