Plate Nº 95 · recorded October 10, 2026
Neuroscience & MindReported finding
Largest cognition genetics study raises rare-variant gene list to 26
UK Biobank researchers analyzed 455,000 people and tied 26 genes to lower cognitive test scores through rare damaging variants — up from five — with 14 already linked to neurodevelopmental conditions.
By Marcus Bennett3 min read693 words
In brief
- Analysis covered more than 455,000 UK Biobank participants, up from about 270,000 in earlier work, by imputing test scores for an additional 170,000 people.
- Common DNA changes linked to test scores rose from 390 to 550, a 40% increase.
- Genes in which rare damaging variants lower test scores jumped from 5 to 26; 14 are already linked to neurodevelopmental conditions.
- About 40% of UK Biobank participants never completed the verbal-numerical reasoning test, biasing prior samples.
- Study published October 5, 2026 in Nature Genetics (DOI: 10.1038/s41588-026-02787-5).
The largest genetic study of cognition to date analyzed DNA from more than 455,000 people of European ancestry and identified 26 genes in which rare damaging variants lower scores on a problem-solving test. The work, published October 5, 2026 in Nature Genetics, more than five-folded the prior count of five such genes.
Researchers at the Wellcome Sanger Institute and Amsterdam University Medical Center expanded an earlier UK Biobank analysis of about 270,000 participants to more than 455,000 by statistically estimating verbal-numerical reasoning scores for people who never sat the test. The approach also raised common-DNA-change signals linked to cognition from 390 to 550 — a 40% increase.
How did the team add so many participants?
UK Biobank holds health and lifestyle records for roughly half a million Britons. About 40% never completed the cognitive test, and those who did were likelier to hold higher-education credentials. That imbalance, known as ascertainment bias, can pull genetic findings toward people who are healthier or better educated.
To counter it, the team predicted missing test scores from other recorded variables, lifting the analysis from roughly 270,000 to 455,000 participants and adding information from more than 170,000 additional people.
"We were able to include information from over 170,000 additional participants to give a more representative picture of how genetics influences cognition scores and to uncover rare DNA differences linked to cognition that we previously did not have enough statistical power to find," said Dr. Daniel Malawsky, co-first author at the Sanger Institute.
What did the expanded analysis find?
The team reported three categories of signal:
- 550 common variants, each with a very small individual effect on test scores
- 26 genes in which rare damaging variants were linked to lower scores, up from five
- 14 of those 26 genes already known to cause neurodevelopmental conditions
Four more had been previously tied to measured cognitive scores or education in UK Biobank data. The remaining eight lacked strong prior links to cognition, although patients with neurodevelopmental conditions carried more damaging variants in those genes than chance would predict — a hint that they contribute to risk.
Why do the rare-variant findings matter?
All 26 genes lowered scores when damaged; none raised them. That pattern suggests the rare variants act by disrupting brain development rather than by enhancing it. Fourteen of the genes are already implicated in conditions such as intellectual disability.
Co-senior author Dr. Hilary Martin of the Sanger Institute pointed to a striking population-level pattern:
"By mapping genetic variants across nearly half a million people, we've shown that damaging DNA differences in the same genes that cause severe neurodevelopmental conditions can also affect cognitive traits across the wider population."
Most UK Biobank carriers of these rare variants have no clinical diagnosis, the authors noted, meaning the genetic effects can shape cognition without crossing a diagnostic threshold.
What are the limits?
The participants were all of European ancestry, so the findings may not generalize to other groups. Estimated scores inherit uncertainty from the imputation step, and the authors argue that better direct cognitive measurement remains the highest priority.
Co-senior author Dr. Abdel Abdellaoui of Amsterdam University Medical Center said the team's biggest analytical challenge was confirming the imputed scores still tracked cognitive ability rather than simply echoing the health, behavior and socioeconomic variables used to derive them. "We tested that extensively," Abdellaoui added.
The team stressed that the results speak to groups, not individuals. Genetic scores derived from the study cannot reliably predict any single person's cognitive ability.
What changes for researchers?
By doubling the catalog of common variants tied to cognition and quintupling the rare-variant gene list, the study gives researchers new entry points into how DNA shapes brain development and why damaging changes in the same gene can cause severe conditions in some carriers but only modest cognitive effects in others.
The paper is by David M. van den Berg and colleagues, published as "Imputation of fluid intelligence scores reduces ascertainment bias and increases power for analyses of common and rare variants" (DOI: 10.1038/s41588-026-02787-5).
via Medical Xpress (Source)
More from Marcus Bennett
Nearby plates
- ADHD Linked to More Than Fourfold Higher Dementia Risk in Large Study
- New Immune Cell Atlas Links Genetic Variants to Disease Mechanisms
- New RNA Testing Guidance Could Deliver Clearer Genetic Diagnoses
- Everyday Puzzles and Reading May Help Reverse Cognitive Decline
- Mediterranean-Style Diets Show Brain Benefits—But Only in Snapshot