A new study published in the journal Cerebral Cortex provides evidence that differences in general intelligence may be partly explained by the density of neurites within certain brain connections, and that this brain structure is, in turn, influenced by a person’s genetic makeup. The researchers found that individuals with higher genetic scores for intelligence tended to have a higher density of neurites in specific white matter tracts, and that this structural feature mediated the relationship between genetic variation and cognitive ability.
The research addresses a longstanding challenge in neuroscience: understanding how inherited genetic differences relate to brain structure and cognitive performance. While earlier studies have shown that general intelligence is highly heritable, the new study focuses on the role of white matter.
White matter acts as the brain’s communication network, composed of vast bundles of nerve fibers that transmit signals between different regions. While it has been known that white matter is involved in cognitive functioning, it has remained unclear which specific features are most relevant and how they connect genetics to intelligence.
Past studies have often relied on a measurement called fractional anisotropy, which captures the directional movement of water molecules in the brain and is commonly used to assess white matter integrity. Higher values have been associated with better cognitive performance.
However, fractional anisotropy does not clarify whether the observed effects are due to denser neurite connections, more uniform fiber orientation, or greater levels of myelination. The new study was designed to go beyond that general measure by using more targeted neuroimaging techniques.
“In general, our motivation for this study was to examine the structural network connectivity of the brain, i.e., anatomical connections between brain regions, in relation to intelligence more thoroughly, as there is now a consensus that different regions of our brain form a network and exchange information,” said study author Christina Stammen of the Leibniz Research Centre for Working Environment and Human Fa ctors(IfADo).
“The motivation for this study in particular arose from a previous study we conducted. In that study, we found significant positive associations between fractional anisotropy values, a commonly used measure of white matter microstructure, and general intelligence. Since it remained unclear whether this association was due to axon density, parallel, homogenous fiber orientation distributions, or myelination, we conducted our second study.”
To examine the brain’s microstructural architecture in greater detail, the researchers used two advanced imaging methods. One, known as neurite orientation dispersion and density imaging (NODDI), provides estimates of both the density and alignment of neurites—the projections from nerve cells, which in white matter are primarily signal-transmitting fibers called axons.
The other, myelin water fraction imaging, estimates the amount of myelin, a fatty substance that wraps around nerve fibers and speeds up signal transmission. Together, these approaches allowed the team to separately examine neurite density, orientation dispersion, and myelination in white matter.
The study involved more than 500 healthy young adults, most of whom were university students. Each participant completed a series of well-validated intelligence tests that measured a range of cognitive abilities, including verbal, numerical, and spatial reasoning, general knowledge, and processing speed. The researchers then used a statistical approach called factor analysis to compute each participant’s general intelligence score, often referred to as the g factor.
To assess genetic influence, the researchers generated a polygenic score for intelligence for each participant. These scores summarize the small effects of thousands of genetic variants previously linked to intelligence in large-scale genome-wide studies.
DNA was obtained from cells brushed from the participants’ oral mucosa and analyzed for common single nucleotide polymorphisms. After a series of quality control steps, the polygenic scores were calculated based on how many intelligence-associated variants each person carried and the strength of those associations.
Participants also underwent MRI scans, from which the researchers extracted detailed measurements of white matter properties in 64 distinct brain tracts. For each tract, they calculated the neurite density index, orientation dispersion index, and myelin water fraction. The researchers then conducted mediation analyses to examine whether these brain features statistically explained the link between genetic variation and general intelligence.
The findings showed that neurite density, but not orientation dispersion or myelination, was significantly related to general intelligence. The analysis revealed two key connections: First, people with higher polygenic scores for intelligence tended to have higher neurite density in 28 of the 64 white matter tracts. Second, higher neurite density in 18 tracts was directly associated with higher general intelligence scores.
Among these, six tracts were found to be significant in both relationships, statistically mediating the connection between genetics and cognitive performance. These included areas such as the uncinate fasciculus, superior longitudinal fasciculus, cingulum, and middle longitudinal fasciculus, which are known to support functions like memory, language, and cognitive control.
In these regions, higher genetic scores were associated with greater neurite density, which in turn was linked to higher intelligence scores. This suggests that genetic variation may influence how densely packed the brain’s communication pathways are, and that this difference in structure may help explain why some people perform better on cognitive tasks than others.
Notably, the orientation dispersion index, which reflects how uniformly or diffusely aligned the neurites are within a tract, did not show any association with intelligence. This suggests that the arrangement or directionality of white matter fibers may not be as important as their overall density when it comes to cognitive performance.
The results for myelin content were similar. Although some genetic associations were found with myelin water fraction in a few tracts, there was no evidence that this feature was related to intelligence in the sample. In other words, the degree of myelination in white matter did not mediate the link between genetics and general intelligence.
“The fact that there were no significant associations between myelin water fraction and intelligence in predominantly young adults surprised us, as it is a long-standing hypothesis that differences in myelination could underlie differences in intelligence,” Stammen told PsyPost. “Myelin plays a crucial role in increasing conduction velocity and ensuring precise spike timing, both of which are critical for the synchronization and coupling of neuronal ensembles across distributed brain networks.”
The researchers interpret these findings to mean that the number of neurites in certain brain pathways may be more relevant for intelligence than the speed or efficiency of signal transmission along those pathways. This aligns with the idea that individuals with more densely connected brain networks may be better equipped for complex problem-solving because they have more parallel routes for information processing.
Some of the white matter regions found to be important in this study overlap with areas highlighted in previous research using more general imaging measures. For example, past work has shown associations between intelligence and fractional anisotropy in the superior longitudinal fasciculus and cingulum. The current study adds specificity to those findings by identifying neurite density as the relevant microstructural property.
The findings indicate that “our genetic makeup influences the neurite density of certain connections, i.e. fiber tracts, in our brain, which in turn has an impact on our intelligence,” Stammen explained. “The relation between our genetic variants and our intelligence is therefore partially mediated by the neurite density of certain fiber tracts in the brain, which has never been investigated before.”
“In contrast, neurite orientation dispersion and estimated myelination of the same fiber tracts showed no mediating effects between genetic variants and intelligence,” Stammen explained. “This means that the number or density of neurites is more important for intelligence than their alignment or estimated myelination. The predominantly positive mediation effects suggest that parallel information processing takes place between important areas of intelligence, as there are more neurites and thus more possible ways to think through problems.”
“Overall, the effects found are of course very small. However, given that all three interrelated variables – genetic makeup, the brain, and intelligence – are very complex constructs, even small contributions from individual factors are meaningful.”
As with all research, there are some limitations. The sample consisted mostly of young, highly educated individuals with above-average intelligence scores. This limits how broadly the findings can be applied to the general population. Additionally, while the neuroimaging methods used here offer more specificity than earlier approaches, they still rely on indirect estimates and cannot capture the full complexity of brain microstructure.
The authors also caution that their mediation analyses are exploratory and based on regularization methods designed to detect patterns in complex data. These methods can help identify promising relationships, but the results should be replicated in other samples and confirmed with different statistical techniques.
Another limitation is that the polygenic scores used in this study, while based on large-scale genetic data, still explain only a small portion of the total variation in intelligence. As larger and more diverse genetic studies become available, future research may be able to account for a greater share of cognitive differences.
Overall, this study provides new insights into the biological mechanisms linking genetics to intelligence. By focusing on specific microstructural features of white matter, it offers a more detailed view of how genetic variation may shape the brain’s structural networks and contribute to cognitive ability. The findings suggest that the density of neurites, rather than their alignment or myelination, may play a more central role in this process.
“The next steps would be to replicate our results in independent samples,” Stammen explained. “Furthermore, it would be interesting to use other new imaging methods to investigate the relation between genetic variants, the brain, and human intelligence. We would like to encourage others to conduct more mediation research in order to directly analyze the triad between genes, the brain, and behavior in one.”
The study, “Neurite density but not myelination of specific fiber tracts links polygenic scores to general intelligence,” was authored by Christina Stammen, Javier Schneider Penate, Dorothea Metzen, Maurice J. Hönscher, Christoph Fraenz, Caroline Schlüter, Onur Güntürkün, Robert Kumsta, and Erhan Genç.
