New research published in Imaging Neuroscience suggests that general intelligence is supported by the brain’s ability to maintain stable, efficient, and typical connectivity patterns. The study indicates that individuals with higher cognitive abilities tend to sustain specific brain states longer and reconfigure their neural networks more efficiently than those with lower scores. These findings provide evidence that general intelligence relies on optimizing brain communication dynamics rather than simply having faster or more flexible connections overall.
The concept of general intelligence, often denoted as “g”, reflects an individual’s capacity to perform well across a variety of cognitive tasks. A person who excels in verbal reasoning often performs well in spatial or mathematical tasks as well. This observation has led scientists to look for shared biological mechanisms in the brain that drive this generalized performance.
Early theories focused on the size of specific brain regions or the strength of static connections between them. However, the brain is not a static organ. It constantly reorganizes its activity to meet changing demands. This led to the development of the Network Neuroscience Theory of Human Intelligence. This theory posits that intelligence arises from the brain’s dynamic ability to reconfigure its network topology.
Scientists use functional magnetic resonance imaging, or fMRI, to measure these changes. Most previous research looked at “static functional connectivity,” which averages brain activity over a long period. This approach misses the rapid changes that occur from moment to moment. Newer methods analyze “dynamic functional connectivity” to see how the brain transitions between different states of activity over time.
Previous dynamic studies largely focused on how frequently the brain switches between states. While informative, these metrics do not fully capture the nuance of network reconfiguration. They do not measure how consistent the connectivity is within a state or how similar a person’s brain patterns are to the general population. The authors of the current study sought to fill this gap by examining the stability, efficiency, and typicality of these dynamic patterns.
“We set out to explore a long-standing question: What is the biological basis of general intelligence? In other words, why do people who perform well on certain types of cognitive tasks like memory also tend to perform well across other cognitive tasks like attention and reasoning?” said study author Colin Hawco, an associate professor at the University of Toronto and a scientist in the Brain Health Imaging Centre at the Centre for Addiction and Mental Health.
“The Network Neuroscience Theory of Human Intelligence suggests that general intelligence relates to the brain’s capacity to flexibly change connections between different regions in response to different cognitive demands. We looked at how the brain transitions through different ‘brain states’, which represent patterns of connectivity across brain networks that change over time, but also repeat.”
“Previously, the frequency of transitions between these brain states was linked to executive functioning; the frequency of these changes might reflect the ‘flexibility’ proposed by the theory,” Hawco continued. “However, we looked beyond these more standard frequency measures, and also explored the nature of these state changes. We were especially interested in the ability to hold stable brain states, and make ‘clean’ transitions between states, as well as how typical the pattern of brain connectivity for each person was in these states.”
For their study, the researchers utilized data from the Human Connectome Project. This is a large-scale initiative designed to map the neural pathways of the human brain. The final sample included 950 young adults. The participants ranged in age from 22 to 36 years old.
The researchers analyzed fMRI data collected while the participants were in a resting state. During these scans, individuals lay still without performing any specific task. This allows scientists to observe the intrinsic functional architecture of the brain. The team also accessed scores from ten different cognitive tests taken by the participants. These tests measured abilities such as working memory, processing speed, reading decoding, and fluid intelligence.
The analytical approach involved a method called Leading Eigenvector Dynamics Analysis. This technique allows for the identification of recurring patterns of brain connectivity, referred to as “states,” at a specific point in time. The researchers identified six distinct states that the brain cycles through. State 1 represented a baseline pattern with uniform signal coherence. States 2 through 6 represented various configurations of complex networks, such as the Default Mode Network and the Frontoparietal Network.
After identifying these states, the researchers calculated several metrics for each participant. They looked at “frequency,” or how often and how long a person stayed in a specific state. They measured “transition distance,” which quantifies how much the brain’s connectivity pattern changes when moving from one state to another. Finally, they assessed “idiosyncrasy,” which measures how different an individual’s brain state is from the group average.
“We explored three main measures of brain function and related them to general intelligence: 1) frequency of changing states, measuring flexibility; 2) how well people could hold and transition brain states, measuring stability and control, and 3) how typical the connectivity was in each state, measuring how much they deviated from the ‘normal’ average,” Hawco explained. “This diverse set of measures let us characterize novel aspects of brain connectivity flexibility that relate to general intelligence.”
The researchers found that individuals with higher intelligence scores tended to maintain stable connectivity in specific states involving higher-order cognitive networks. Specifically, they spent more time in states characterized by interactions between attention and control networks. This suggests that the ability to sustain complex network configurations is a marker of higher cognitive ability.
The researchers also found a relationship between intelligence and reconfiguration efficiency. High-scoring individuals exhibited smaller connectivity changes when transitioning between similar states. Conversely, they showed larger connectivity changes when moving to distinctly different states. This implies a neural system that is precise. It conserves energy for minor adjustments but is capable of substantial reconfiguration when necessary.
Another significant finding was related to the concept of typicality. The researchers found that higher general intelligence was associated with having brain patterns that closely resembled the group average. In other words, the most “typical” brain patterns were linked to the best performance. This supports the idea that evolutionary processes may converge on an optimal functional organization.
“Finding that have a more ‘typical’ pattern of connectivity in each state was related to higher cognition was somewhat surprising,” Hawco told PsyPost. “We often think of people with higher cognition as having more unique brains, but this may not be the case at all. Our strong findings that stability when holding a brain state were also quite exciting. This aspect of brain function is generally not captured by current methods, and we think its an important window into brain functions which also has implications for our research on mental health.”
The researchers observed a different pattern when analyzing processing speed. While general intelligence was linked to stability, processing speed was associated with flexibility. Individuals with faster processing speeds tended to switch between states more frequently. They also showed higher idiosyncrasy, meaning their brain patterns were more unique compared to the group average.
This divergence suggests that different cognitive domains rely on different dynamic properties. General intelligence appears to benefit from prolonged, stable engagement of specific networks. Processing speed appears to benefit from the ability to rapidly cycle through different configurations. The contrast highlights that “better” brain function is context-dependent.
“Our findings suggest that higher general intelligence is linked to a person’s capacity to efficiently achieve and sustain typical connectivity patterns in states highlighting the interactions of ‘higher-order’ cognitive-processing networks,” Hawco explained. “Our study corroborates prior studies and theories linking these higher-order cognitive-processing networks to general intelligence, prior neural efficiency hypotheses for general intelligence, and adds to the hypothesis that group-averages represent optimal characteristics.”
“A lot of research into the brain focuses exclusively on how strong connections are on average over a longer period; this work moves into understanding patterns, and how well brain connectivity is controlled by an individual over time. It’s a really different way to think of brain function, and we think it may better capture some important aspects of cognitive function.”
“The effect sizes were moderate, indicating that a large portion of individual variability in intelligence is still not explained,” Hawco noted. “While expected of brain-behavior relationships, this indicates that our measures do not completely capture the aspects of brain network flexibility that relate to general intelligence. We still have a lot to understand in what drives generalized intelligence.”
As with all research, there are some limitations. The data relies on resting-state fMRI, which may not perfectly reflect how the brain functions during active problem-solving. The relationships observed are correlational and do not prove that specific brain patterns cause higher intelligence. Additionally, the participants were all healthy young adults, limiting the generalizability of the results to other age groups or clinical populations.
“Intelligence research has long faced skepticism, in part because of its complex history and the way results have sometimes been misused,” Hawco said. “Early intelligence testing was tied to social hierarchies, educational inequality, and even discriminatory policies — associations that understandably left a lasting stigma. In modern times, people may also view IQ research as reductionist, assuming it tries to capture the richness of human thought or potential in a single number.”
“However, contemporary intelligence research is far more nuanced. It examines the neural, genetic, and environmental factors that shape reasoning, learning, and problem-solving — not to rank individuals, but to understand how the mind works. Findings from this field inform education, cognitive training, and even clinical approaches to conditions that affect thinking.”
“In reality, studying intelligence is not about labeling people; it’s about uncovering the biological and psychological foundations of human cognition — one of the most important scientific questions there is.”
Future research will likely explore these dynamics during active tasks. Observing how the brain reconfigures when challenged with a difficult problem could provide stronger evidence for the neural efficiency hypothesis. The researchers also aim to investigate these patterns across different timescales. They hope to determine if these dynamic fingerprints can predict changes in cognitive health over time.
“Currently, we are in the process of analyzing the relationship between brain connectivity flexibility and general intelligence both across different contexts and scales,” Hawco told PsyPost. “This study was conducted on flexibility metrics calculated from individuals during resting-state (i.e., they were not asked to do anything). Having participants do cognitively challenging tasks in the MRI may provide a ‘stress test’ of the brain for cognitive performance, so brain-behavior relationships may be stronger and thus better observed.”
“Furthermore, brain modulation between task and rest may be another measure of adaptive flexibility which may be of interest. In addition, this study was conducted on high-level whole-brain measures of flexibility in connectivity patterns. Flexibility at the level of regions and connections may reveal even more interesting relationships with intelligence.”
The study, “Higher general intelligence is associated with stable, efficient, and typical dynamic functional brain connectivity patterns,” was authored by Justin Ng, Ju-Chi Yu, Jamie D. Feusner, and Colin Hawco.
