New research published in Scientific Reports provides evidence that the path from genetic predisposition to general intelligence travels through specific, frequency-dependent networks in the brain. The findings indicate that these neural pathways are not static but appear to shift significantly between early adulthood and older age.
Intelligence is a trait with a strong biological basis. Previous scientific inquiries have established that genetic factors account for approximately 50% of the differences in intelligence between individuals. Genome-wide association studies have identified hundreds of specific variations in the genetic code that correlate with cognitive ability.
These variations are often aggregated into a metric known as a polygenic score, which estimates an individual’s genetic propensity for a certain trait. Despite this knowledge, the specific biological mechanisms that translate a genetic sequence into the ability to reason, plan, and solve problems remain unclear.
Scientists have hypothesized that the brain’s functional connectivity acts as the intermediary between genes and behavior. Functional connectivity refers to how well different regions of the brain communicate with one another. While past studies using functional magnetic resonance imaging (fMRI) have attempted to map these connections, the results have been inconsistent.
fMRI is excellent at locating where brain activity occurs but is less precise at measuring when it occurs. The authors of the new study opted to use electroencephalography (EEG). This technology records the electrical activity of the brain with high temporal resolution, allowing researchers to observe the speed and rhythm of neural communication.
“We already know that intelligence is highly heritable, which is why we are especially interested in the role of the brain as a ‘neural pathway’ linking genetic variation to cognitive ability,” said study author Rebecca Engler of the Leibniz Research Centre for Working Environment and Human Factors (IfADo).
“The lack of integrative approaches combining genetics, brain network organization, and intelligence motivated us to take a closer look at resting-state EEG markers, with a particular focus on differences between young and older adults.”
“In a recent large-scale study (Metzen et al., 2024) using resting-state fMRI, we found no robust association between functional architecture of specific brain regions and intelligence. This motivated our shift toward resting-state EEG, which captures brain dynamics at much higher temporal resolution. EEG measures brain activity as oscillations across different frequencies, allowing us to study frequency-specific brain networks that may carry distinct information relevant to cognitive ability.”
For their study, the researchers recruited a representative sample of 434 healthy adults from the Dortmund Vital Study. The participants were categorized into two distinct age groups. The young adult group consisted of 199 individuals between the ages of 20 and 40. The older adult group included 235 individuals aged 40 to 70.
To measure intelligence, the research team administered a comprehensive battery of cognitive tests. These assessments covered a wide range of mental capabilities, including verbal memory, processing speed, attention span, working memory, and logical reasoning. The scores from these tests were combined to calculate a single factor of general intelligence, often denoted as g. This factor serves as a reliable summary of an individual’s overall cognitive performance.
Genetic data were obtained through blood samples. The researchers analyzed the DNA of each participant to compute a polygenic score for intelligence. This score was calculated based on summary statistics from previous large-scale genetic studies. It represents the cumulative effect of many small genetic variations that are statistically associated with higher cognitive function.
Brain activity was recorded while participants sat quietly with their eyes closed for two minutes. This “resting-state” EEG data allowed the researchers to analyze the intrinsic functional architecture of the brain.
The team employed a method known as graph theory to quantify the organization of the brain networks. In this framework, the brain is modeled as a collection of nodes (regions) and edges (connections).
The researchers calculated metrics such as “efficiency,” which measures how easily information travels across the network, and “clustering,” which measures how interconnected specific local neighborhoods of the brain are. These metrics were analyzed across different frequency bands, including delta, theta, alpha, and beta waves.
The study employed complex statistical modeling to test for mediation effects. A mediation analysis determines whether a third variable—in this case, brain connectivity—explains the relationship between an independent variable (genetics) and a dependent variable (intelligence). The researchers looked for instances where the polygenic score predicted a specific brain network property, which in turn predicted the intelligence score.
The results showed that global measures of brain efficiency did not mediate the link between genetics and intelligence. This suggests that simply having a “more efficient” brain overall is not the primary mechanism by which genes influence cognition.
In other words, “there is no single brain region responsible for intelligence,” Engler told PsyPost. “Instead, cognitive ability relies on efficient and dynamic communication across a broad network of regions throughout the brain, and this network organization changes as we age.”
The specific neural pathways identified varied substantially by age. For young adults, the connection between genetics and intelligence was mediated by brain activity in the beta and theta frequency bands. These effects were predominantly located in the frontal and parietal regions of the brain.
The frontal and parietal lobes are areas traditionally associated with executive functions, such as decision-making, working memory, and attention. This aligns with prominent theories that attribute intelligence to the efficient integration of information between these higher-order brain regions.
But for older adults, the mediating effects were found primarily in the low alpha and theta frequency bands. Furthermore, the specific brain regions involved shifted away from the frontal cortex. The analysis identified the superior parietal lobule and the primary visual cortex as key mediators. These areas are largely responsible for sensory processing and integration.
This shift suggests that the neural architecture supporting intelligence evolves as people age. In younger adulthood, cognitive ability appears to rely heavily on the rapid, high-frequency communication of executive control networks in the front of the brain. As the brain ages, it may undergo a process of reorganization.
The reliance on posterior brain regions and slower frequency bands in older adults implies a strategy that prioritizes the integration of sensory information. This finding is consistent with the concept of neural dedifferentiation, where the aging brain recruits broader, less specialized networks to maintain performance.
The researchers also found that certain brain areas, such as the primary visual cortex, played a consistent role across both groups, though the direction of the effect varied. In both young and older adults, higher nodal efficiency in the visual cortex was associated with higher intelligence.
However, a higher genetic predisposition for intelligence was associated with lower efficiency in this region. This complex relationship highlights that the genetic influence on the brain is not always a straightforward enhancement of connectivity.
“When comparing the two age groups, we were surprised that the brain regions consistently mediating the link between genetic variation and intelligence are primarily involved in sensory processing and integration,” Engler explained. “One might expect such stable neural anchors to be associated with higher-order executive functions like reasoning or planning, typically located in frontal networks. Instead, our results suggest that sensory and associative regions play a more central role in maintaining cognitive ability than is typically emphasized in dominant models of intelligence.”
As with all research, there are some limitations to note. The study utilized a cross-sectional design, meaning it compared two different groups of people at a single point in time. It did not follow the same individuals as they aged.
Consequently, it is not possible to definitively prove that the observed differences are caused by the aging process itself rather than generational differences. Longitudinal studies that track participants over decades would be necessary to confirm the shift in neural strategies.
The study focused exclusively on resting-state EEG. While intrinsic brain activity provides a baseline of functional organization, it does not capture the brain’s dynamic response to active problem-solving.
It is possible that different network patterns would emerge if participants were recorded while performing the cognitive tests. Future research could investigate task-based connectivity to see if it offers a stronger explanatory link between genetics and performance.
“A crucial next step would be to replicate our findings in independent samples to ensure their robustness and generalizability,” Engler said. “Furthermore, it would be interesting to investigate age-related changes in functional network organization from a longitudinal rather than from a cross-sectional perspective. A further long-term goal is to investigate the triad of genetic variants, the brain’s functional connectivity, and intelligence by analyzing task-based EEG data rather than resting-state EEG data.”
The study, “Electrophysiological resting-state signatures link polygenic scores to general intelligence,” was authored by Rebecca Engler, Christina Stammen, Stefan Arnau, Javier Schneider Penate, Dorothea Metzen, Jan Digutsch, Patrick D. Gajewski, Stephan Getzmann, Christoph Fraenz, Jörg Reinders, Manuel C. Voelkle, Fabian Streit, Sebastian Ocklenburg, Daniel Schneider, Michael Burke, Jan G. Hengstler, Carsten Watzl, Michael A. Nitsche, Robert Kumsta, Edmund Wascher, and Erhan Genç.
Leave a comment
You must be logged in to post a comment.