For decades, executive function has been measured behaviorally, described developmentally, and treated clinically, while its molecular basis stayed mostly dark. We have known that the prefrontal cortex does much of the work, that the skills develop across childhood and adolescence, and that they run in families. What we have not had is a map of which specific stretches of the human genome are associated with how well an adult plans, switches, and holds information in mind. In May 2026, a team reporting in Nature Communications published the first large-scale attempt to draw that map, and the result is more interesting than a simple “genes for executive function” headline would suggest.

The study, led by M. S. Rahman, A. Frkatovic-Hodzic, J. van den Ameele and colleagues, did three things at once. It identified specific regions of the genome associated with executive function in adults. It connected those regions to differences in brain structure. And it traced the genes back to when, during the long arc of brain development, they are actually switched on. That last step is what gives the paper its title and its central claim: that the genetic basis of adult executive function has what the authors call a cell-type-specific developmental origin. The genes that matter for how a 40-year-old performs on a test of cognitive control are, to a striking degree, the same genes that were busy shaping the brain before that person was born.

What the Study Actually Did

The method at the center of the paper is a genome-wide association study, or GWAS. The logic of a GWAS is straightforward even when the execution is not. Researchers take a large group of people, measure some trait, read millions of common spelling variations across each person’s genome, and then ask which of those variations show up more often in people who score higher or lower on the trait. No single variant explains much. The power comes from scale, and from looking across the whole genome at once rather than guessing in advance which genes to check.

The team ran the analysis in two separate groups of people and then combined the results. The larger group came from the UK Biobank, a long-running British research resource, and included 84,238 participants. The second came from the NIHR BioResource’s Genes and Cognition study and included 9,932 participants. Running the same analysis in two independent groups is a safeguard. A signal that appears in one cohort might be a fluke of that particular sample. A signal that appears in both, pointing the same direction, is far more likely to be real. The researchers pooled the two in a meta-analysis to sharpen the picture.

Executive function is not one thing, and the study did not pretend it was. The team examined three different measures. The one that carried the clearest genetic signal was a trail-making alphanumeric task. The trail-making test is a long-standing tool in neuropsychology. In its simpler form, a person connects numbered dots in order. In the harder, alphanumeric form, they alternate between numbers and letters, 1 to A to 2 to B and onward, which forces the brain to hold a rule in mind and switch between two sequences. That switching is a fairly pure demand on executive function, which is part of why it produced the most useful genetic signal of the three measures tested.

The Heritability Number Is Modest, and That Matters

One of the first numbers the study reports is how heritable the trail-making measure was, and it is worth slowing down on what that number means. The estimate landed in a range of roughly 7 to 26 percent. In plain terms, the common genetic variants the study could measure accounted for somewhere between about a fourteenth and a quarter of the differences between people on this task. That is a real signal. It is also a long way from “executive function is genetic.”

This is where the new study sits in productive tension with an older and well-known body of research. Twin studies, most famously work by Naomi Friedman, Akira Miyake and colleagues, had previously suggested that the shared core of executive function is highly heritable, with some estimates approaching the idea that individual differences in the common executive function factor are almost entirely genetic in origin. So why does this new genome-wide study find a far more modest figure?

The answer is that the two methods are measuring heritability in different ways, and the gap between them is one of the most familiar patterns in modern genetics. Twin studies estimate the total contribution of all genetic differences, common and rare, captured indirectly by comparing identical and fraternal twins. A GWAS measures only the contribution of the common variants it can directly read, and only for the specific trait as measured. Common-variant heritability is almost always lower than twin heritability. Researchers call the difference the missing heritability problem, and it shows up for height, for intelligence, and now for executive function. The honest reading is not that one number is right and the other wrong. It is that genes clearly matter for executive function, that common variants explain a modest and measurable slice of that, and that a great deal of the variation between people still runs through rarer variants, through development, and through environment.

Eighteen Places on the Genome, and 178 Genes

When the two cohorts were combined, the trail-making measure was associated with 18 independent regions of the genome. These are sometimes called loci, and each one is a stretch of DNA where the spelling differences between people track with differences on the task. The fact that these regions pointed in the same direction across both the UK Biobank and the smaller study is the kind of internal consistency that makes geneticists trust a result.

A genome-wide region is not the same as a gene, though. A signal sitting in a particular spot might act on a gene right next to it, or on a gene some distance away whose regulation it influences. To move from regions to candidate genes, the researchers used a set of computational follow-up methods, which the field calls in-silico analysis. That step pointed to 178 genes that could plausibly be doing the work behind the 18 signals. Two of them, a gene called NT5DC2 and a less familiar transcript labeled RP11-579E24.2, stood out because they replicated on their own, in the independent data, even before the two cohorts were combined. NT5DC2 in particular has surfaced in earlier psychiatric and neurodevelopmental genetics, which makes it a reasonable anchor for the broader finding rather than a one-off curiosity.

None of this means a single “executive function gene” has been found. The opposite is closer to the truth. What the study describes is a distributed architecture, dozens of regions and well over a hundred candidate genes, each contributing a small push, which together add up to the modest heritable signal the study detected.

The Biphasic Clue

The most thought-provoking part of the paper is not the list of genes. It is when those genes are active. The researchers looked at expression patterns, the record of which genes are switched on in which tissues and at which stages of life, and found that the genes implicated in adult executive function tended to show what they describe as a biphasic profile. They are expressed early, during the formative period of brain development, then quiet down, then become active again later in life.

That pattern carries an implication worth stating plainly. A gene that does its work in the developing fetal brain and then again decades later is not a gene that finished its job at birth. Its influence is spread across the lifespan, with one chapter written during the construction of the brain’s circuitry and another written during the long period of use, maintenance, and eventual aging. The study frames this as a cell-type-specific developmental origin, meaning the genetic signal traces back to particular kinds of cells doing particular jobs at particular developmental moments, rather than to the adult brain in some static, finished state.

This is the bridge between two literatures that have not always talked to each other. Developmental psychologists have argued for years that executive function is built through experience across the first two decades of life. Geneticists have argued that the capacity is substantially heritable. A finding that the relevant genes are developmentally timed does not resolve that tension so much as dissolve part of it. The genetic contribution and the developmental story may be describing the same process from two angles: genes that set up the developmental program, and a developmental program that then has to actually run, in a real brain, in a real environment, over real time.

Why This Connects Executive Function to Disorders

Executive dysfunction is a feature of a long list of conditions, including ADHD, schizophrenia, and several neurodegenerative diseases. One reason a study like this matters beyond the lab is that a genuine molecular handle on executive function could, in principle, help explain why these conditions so often share executive deficits despite differing in almost everything else. If the same developmentally timed genes that shape ordinary variation in cognitive control are also disrupted in these disorders, that would point toward shared biological ground beneath conditions that look very different on the surface.

That is a hypothesis the study supports rather than proves. The link the authors draw most firmly is between the trait and brain structure: the trail-making signal was associated with broad, pan-cerebral differences in how the brain is built, not just with one isolated region. That is consistent with executive function being a property of distributed networks rather than a single hub, which is what most neuroscience already expected, now with genetic and structural evidence pointing the same way.

What the Study Does Not Mean

It is worth being precise about the limits, because findings like this are easy to oversell. The study does not produce a genetic test for executive function, and nothing in it supports the idea that a person’s capacity for planning or self-control is fixed by their DNA. The heritable signal is modest. The 178 genes are candidates, not confirmed mechanisms. The participants in both cohorts were overwhelmingly of European genetic ancestry, which is a real limitation of most large genetic resources and means the findings cannot be assumed to generalize cleanly across all populations. And the trait that carried the signal was performance on one specific switching task, not the whole of what people mean when they talk about executive function in daily life.

There is also a familiar ethical caution that comes with any advance in the genetics of cognition. As more of the genome’s relationship to mental traits is mapped, the temptation toward genetic determinism grows, along with real questions about privacy and the misuse of such information in education or employment. The authors themselves are careful to stress that genetics lays groundwork rather than dictating outcomes, and that the interaction between genetic predisposition and environment is where the final picture is actually decided.

What It Means for the People Who Work on Executive Function

For the educators, clinicians, coaches, and parents who deal with executive function as a practical reality rather than a research variable, the most useful takeaway is not the gene list. It is the developmental framing. If the genetic contribution to executive function is itself developmentally timed, then the long-running case for treating executive function as something that is built, supported, and scaffolded over years gains rather than loses force. Genes that act during development still need a developing brain to act on, and that brain is shaped by everything from sleep and movement to the structure of a child’s daily demands.

This dovetails with what other recent research keeps finding from the opposite direction. Studies on the slowed development of children’s executive function since 2020 point to how sensitive these skills are to environmental conditions during the years they are forming. A genetic study that locates much of its signal in developmental windows and an environmental study that documents disruption during those same windows are not in conflict. They are describing the two halves of a single process. Nature sets a range. Development, experience, and environment decide where in that range a given person lands.

The honest summary is that executive function is partly heritable, that the heritable part is more modest at the level of common genetic variants than twin studies once implied, and that the genes involved are most active during the building and the long use of the brain rather than at a single fixed moment. That is not a tidy headline. It is a more accurate one, and it points the field toward the questions that actually matter: not whether executive function is genetic or developmental, but how the two work together across a life.