By Jim Shimabukuro (assisted by ChatGPT)
Editor
The 15 April 2026 Nature paper “Ancient DNA reveals pervasive directional selection across West Eurasia,” led by Ali Akbari and senior author David Reich, is being widely viewed as one of the most consequential studies in ancient genomics since the first large-scale recovery of ancient human DNA in the 2010s. Its central thesis is straightforward but profound: human evolution in the last 10,000 years has not slowed down or effectively stopped, as many earlier researchers suspected. Instead, natural selection has been widespread, continuous, and measurable across historical populations of West Eurasia, especially after the transition from hunting and gathering to agriculture. The authors argue that earlier studies underestimated recent human evolution because they lacked both sufficiently large ancient DNA datasets and statistical methods capable of distinguishing genuine natural selection from confounding factors such as migration, population mixing, and random genetic drift [1].
The study’s scale is unprecedented. The researchers analyzed DNA from 15,836 ancient West Eurasians, including more than 10,000 newly sequenced individuals. These samples spanned approximately 18,000 years and covered broad regions of Europe and western Asia. The sheer size and temporal depth of the dataset allowed the team to observe changes in gene frequencies over time almost directly, rather than inferring them indirectly from modern populations alone [1]. Earlier evolutionary genetics studies often relied on what evolutionary biologists call “hard sweeps,” cases in which a single advantageous mutation rapidly spreads to fixation throughout a population. Those events appear to have been relatively rare in recent human history. What Akbari and colleagues instead found was pervasive “directional selection” operating simultaneously across hundreds of variants, often with individually small but cumulative effects [1].
A major innovation was methodological. The researchers developed a statistical framework called AGES that tracks allele frequency changes through time while controlling for population structure and migration. This matters because ancient Eurasia experienced repeated migrations and admixture events—including the spread of early farmers from Anatolia and later Bronze Age steppe expansions—which can mimic the appearance of selection if not properly controlled for. The AGES framework attempts to isolate consistent directional changes that are more plausibly attributable to natural selection itself [1]. The authors estimate selection coefficients for approximately 9.7 million variants, creating what amounts to a large evolutionary map of how human genomes changed across millennia [1].
The paper identified 479 variants showing strong evidence of selection. Many involve traits already suspected to be adaptive in post-Ice Age Eurasia. Alleles associated with lighter skin pigmentation increased substantially in frequency, plausibly because lighter skin aids vitamin D synthesis in northern latitudes with lower ultraviolet exposure, especially after farming populations adopted grain-heavy diets deficient in vitamin D [1,4]. Variants linked to red hair also increased, although the evolutionary explanation remains uncertain [4]. Other selected variants involved immune function and disease resistance. The study found evidence for changing frequencies in genes related to leprosy resistance, HIV-related immunity pathways, autoimmune conditions, and inflammatory diseases [1,3]. Some disease-associated alleles appear paradoxical because they increase risk for modern illnesses while apparently conferring earlier survival advantages against pathogens prevalent in ancient agricultural societies.
One of the paper’s most discussed findings concerns polygenic selection—selection acting not on a single gene but on combinations of many genes associated with complex traits. The researchers report measurable shifts in polygenic scores related to body fat, schizophrenia risk, and cognitive performance [1]. They observed long-term decreases in genetic variants associated with higher body fat and schizophrenia susceptibility, alongside increases in variants correlated in modern genome-wide association studies with educational attainment and cognitive performance [1]. This aspect of the study has drawn both scientific interest and caution. The authors repeatedly stress that these polygenic scores were derived from modern industrialized populations and therefore cannot straightforwardly be interpreted as direct measures of ancient intelligence, behavior, or mental health. Traits associated with “cognitive performance” today may have reflected entirely different adaptive pressures in ancient environments, including social coordination, planning ability, memory, or other behavioral capacities relevant to agricultural and urban societies [1].
The strongest historical pattern emerging from the paper is that selection accelerated after the rise of agriculture. Farming dramatically altered human environments: diets became narrower and carbohydrate-rich; settlements became denser; exposure to domesticated animals and pathogens increased; labor became more specialized; and fertility rates rose. These transformations created powerful new evolutionary pressures. The study therefore supports the increasingly influential view that the Neolithic agricultural revolution was not merely a cultural transition but also a major evolutionary turning point in human biology [1,4].
Razib Khan’s May 2026 essay “Humanity Is Still Evolving—and Fast” frames the study within a broader intellectual shift occurring in evolutionary genetics [2]. For decades, many scientists and popular writers assumed that technological civilization had largely insulated humans from natural selection. Medicine, social institutions, and cultural adaptation appeared to weaken Darwinian pressures. Khan argues that the new ancient-DNA evidence overturns this assumption. Human beings remain biological organisms embedded within changing environments, and cultural innovations themselves generate new selection pressures. In Khan’s interpretation, agriculture, urbanization, pandemics, migration, and social complexity repeatedly reshaped the adaptive landscape for human populations [2]. Rather than replacing evolution, civilization redirected it.
One important implication of the study is methodological. Ancient DNA has now matured from a field focused mainly on reconstructing migration histories into a tool capable of reconstructing evolutionary processes themselves. Researchers can increasingly observe selection “in motion” across time instead of inferring it retrospectively from present-day genomes. This transforms evolutionary genetics from a largely inferential science into something closer to a historical observational science [1,2].
Another major implication concerns medicine. Many modern diseases may reflect evolutionary tradeoffs that emerged under radically different environmental conditions. Genes that once improved resistance to infection or famine may predispose people today to autoimmune disorders, obesity, or metabolic disease. Understanding the evolutionary history of these variants may improve biomedical research by clarifying why harmful alleles persist at high frequencies [1,4]. The study therefore strengthens the broader evolutionary medicine framework, which views modern disease partly through the lens of ancestral adaptation and mismatch.
The paper also has important implications for discussions of human diversity and the genetics of behavior. Because the study touches sensitive subjects such as cognition-related polygenic scores, it will almost certainly become controversial in public discourse. The authors themselves are careful to emphasize several limitations. The findings apply specifically to West Eurasian populations and cannot automatically be generalized globally. The polygenic scores used are probabilistic statistical constructs derived from modern datasets, not direct measurements of ancient traits. Environmental context remains enormously important, and the interaction between genes and culture is complex [1]. Nevertheless, the study demonstrates that behavioral and cognitive-related traits were not exempt from evolutionary pressures in recent human history. That conclusion is likely to influence future research in anthropology, psychology, behavioral genetics, and history.
Perhaps the deepest implication is conceptual. The study undermines the lingering idea that humans somehow stepped outside evolution once civilization emerged. Instead, it suggests that civilization itself became one of evolution’s most powerful engines. Agriculture, cities, literacy, disease environments, state formation, and social stratification did not halt biological evolution; they accelerated and redirected it. Human history, in this view, is simultaneously cultural and biological. The ancient DNA revolution is increasingly revealing how tightly intertwined those two dimensions have always been [1,2,4].
References
[1] “Ancient DNA reveals pervasive directional selection across West Eurasia” (Nature, 15 April 2026) — https://www.nature.com/articles/s41586-026-10358-1
[2] “Humanity Is Still Evolving—and Fast” (The Free Press, 5 May 2026) — https://www.thefp.com/p/humanity-is-still-evolvingand-fast
[3] “Human evolution didn’t slow down; we were just missing the signal” (Live Science, 15 April 2026) — https://www.livescience.com/archaeology/human-evolution/human-evolution-didnt-slow-down-we-were-just-missing-the-signal-large-dna-study-reveals-natural-selection-led-to-more-redheads-and-less-male-pattern-baldness
[4] “Red hair gene favoured by natural selection over last 10,000 years, study finds” (The Guardian, 16 April 2026) — https://www.theguardian.com/science/2026/apr/16/red-hair-gene-favoured-natural-selection-study
Filed under: Uncategorized |
Educational Technology and Change Journal 
Leave a comment