A new study published in Communications Biology suggests that people who perform better on intelligence tests may have brains that communicate more flexibly across distant regions. The research found that more diverse connections between key brain areas, along with more complex patterns of brain activity over time, were linked to higher intelligence scores.
Researchers have long sought to understand the biological basis of intelligence. Early studies often focused on identifying specific brain regions responsible for higher cognitive abilities. In particular, the frontal and parietal lobes have frequently been associated with reasoning, problem-solving, and complex decision-making. More recent theories suggest that intelligence might also rely on how flexibly brain networks can shift between different patterns of activity when solving problems.
Led by Jonas A. Thiele from the University of Würzburg, Germany, the researchers behind the study wanted to test these new ideas more directly. Specifically, they aimed to provide the first empirical test of the newly proposed Multilayer Processing Theory (MLPT) of intelligence, which suggests that human intelligence relies on processes operating across multiple spatial and temporal scales.
Instead of studying brain activity while people rested or performed simple tasks, they examined brain activity while participants completed a well-known intelligence test called Raven’s Progressive Matrices. In this test, participants look at patterns of shapes and must determine which piece correctly completes the logical sequence.
To explore these multi-scale brain processes, the scientists analyzed two different sets of data collected in separate laboratories. In the first dataset, brain activity from 67 participants (26 females, average age 23) was measured using functional magnetic resonance imaging (fMRI) while they worked through the intelligence test. This method tracks blood flow in the brain, allowing researchers to see which regions are communicating with each other across different spatial networks during a task.
In the second dataset, 131 participants (65 females, average age 24) completed the same type of reasoning test while their brain activity was recorded using electroencephalography (EEG), which measures electrical signals produced by the brain. Unlike fMRI, EEG can capture very rapid changes in brain activity, allowing researchers to study how the complexity of brain signals changes over time.
The fMRI results highlighted a crucial nuance: individuals who scored higher on the intelligence test did not simply have “stronger” overall brain connectivity. Instead, they demonstrated more diverse communication between regions in the frontal and parietal parts of the brain. These areas appeared to act as highly efficient “connector hubs,” linking entirely different brain networks together to coordinate information when participants were solving complex problems.
The EEG analysis revealed that individuals with higher intelligence scores exhibited greater signal complexity at longer (coarser) timescales, suggesting richer and more flexible large-scale brain dynamics. At the same time, there was a weaker, non-significant trend indicating lower complexity at very short (finer) timescales, which may reflect simpler, more efficient local processing within smaller brain circuits.
Together, the findings support the idea that intelligence does not come from a single brain area, but rather from how effectively different regions work together across various scales of space and time.
Thiele and colleagues concluded: “[O]ur findings provide the first empirical evidence for the key assumptions of the Multilayer Processing Theory (MLPT), which posits that higher intelligence emerges from more flexible global long-range processes operating at coarser timescales, coordinating simpler short-range processes within smaller neuronal assemblies at finer timescales.”
Despite these insights, the researchers caution that the study has several limitations. For instance, the fMRI scans and EEG recordings were taken from different groups of participants, meaning the two datasets could not be directly compared. Furthermore, the sample sizes were relatively small, which can limit statistical power, and all participants were young adults, meaning the results may not necessarily generalize to children or older adults.
The study, “Decoding the human brain during intelligence testing,” was authored by Jonas A. Thiele, Joshua Faskowitz, Olaf Sporns, Adam Chuderski, Rex Jung, and Kirsten Hilger.
