Selection was working on the ancient trait, not the modern one.
Over ten thousand years, the human body has been quietly rewritten — not by chance or migration, but by the relentless pressure of survival. Harvard researchers, drawing on the largest ancient genome survey ever conducted, have found that natural selection shaped blood types, disease vulnerabilities, and even traits we associate with modern life across Western Eurasia. The bones of the long dead, it turns out, carry a living argument: that humanity is not a fixed thing, but an ongoing negotiation between biology and the world it inhabits.
- A Harvard team sequenced DNA from nearly 16,000 ancient remains and found that genes shifted in frequency not by accident, but because some variants helped people survive and reproduce better than others.
- A gene variant causing coeliac disease was entirely absent 4,000 years ago yet now appears in one in five people — a rapid, unexplained rise that unsettles assumptions about agriculture and adaptation.
- The CCR5 delta-32 variant, which today confers complete resistance to HIV, surged in ancient populations thousands of years before HIV existed, suggesting a now-vanished pathogen once drove the same selection.
- Gene variants linked to intelligence test scores, income, and even smoking behavior show ancient selection signals — raising urgent and unresolved questions about what those variants were actually doing in prehistoric bodies.
- The study's geographic blind spot — the near-absence of South Asian ancient DNA — leaves a major chapter of human genetic history unwritten, and researchers are calling for urgent preservation of ancestral remains from underrepresented populations.
The bones and teeth of people dead for thousands of years still hold their genetic secrets, and scientists have learned to read them. At Harvard Medical School, researchers undertook the most ambitious ancient genome survey ever attempted, sequencing DNA from 15,836 skeletal remains spread across Western Eurasia and comparing those sequences against the genomes of over six thousand people living in the same regions today. Published in Nature, the study's central finding was unambiguous: many genes had shifted in frequency over the past ten millennia, and the engine of that change was natural selection.
The shifts were specific and sometimes startling. Over the last 6,000 years, the B blood-type variant grew more common among West Eurasians while the A variant declined — a balance likely maintained by shifting pathogens and environmental pressures. More puzzling was the HLA-DQB1 variant linked to coeliac disease: entirely absent 4,000 years ago, it now appears in one in five people, despite agriculture — and its gluten-heavy diets — having arrived 10,000 years ago. The lag between cause and consequence remains unexplained.
Around 8,000 years ago, selection favored lighter skin and pigmented hair, likely as an adaptation to low-sunlight environments where vitamin D synthesis became critical for farming populations with poor dietary sources. Meanwhile, the CCR5 delta-32 variant — which today renders carriers fully resistant to HIV — rose from 2% to 8% of Western Eurasians between 6,000 and 2,000 years ago, long before HIV existed. Some ancient and now-unknown pathogen must have driven the same selection.
Perhaps most thought-provoking were signals tied to traits we consider distinctly modern: performance on cognitive tests, years of schooling, household income, and even a tendency away from smoking — despite tobacco not reaching Eurasia until the late 15th century. What ancient behaviors or biological functions these gene variants once governed remains an open mystery.
The study also exposes a significant gap. South Asians carry ancestry from Iranian Neolithic farmers, steppe herders, and Eastern Eurasian populations, yet comparable ancient DNA research for that region barely exists. Understanding which evolutionary pressures were universal and which were local will require something not yet assembled: a preserved and accessible legacy of ancient remains from the full breadth of human diversity.
Thousands of years after death, the bones and teeth of ancient people still hold their secrets. Scientists have learned to extract DNA from skeletal remains, and in doing so, they've begun to read the genetic history written into our bodies over the past ten millennia.
Researchers at Harvard Medical School undertook the largest survey of ancient human genomes ever attempted. They sequenced DNA from 15,836 ancient remains scattered across Western Eurasia—a region spanning Europe, Russia, Central Asia, the Middle East, and Iran—and compared those sequences against the genetic makeup of 6,438 people living in those same countries today. The oldest skeletal material they examined dated back 18,000 years, though their analysis focused on the clearer genetic signal from the last 10,000 years. The work, published in Nature in April, revealed something striking: many genes had shifted in frequency over those millennia, and the culprit was not random chance or population movement, but natural selection itself.
To determine how old a skeleton is, scientists measure the radioactive carbon in bones and teeth. Carbon-14, created when cosmic rays strike nitrogen in the atmosphere, exists in living bodies at the same concentration as in the air and food around them. Once death occurs, that radioactive carbon begins to decay back into nitrogen, with no way to replenish it. The decay follows a predictable rhythm: every 5,730 years, the amount of carbon-14 halves. By measuring the ratio of radioactive to non-radioactive carbon using a mass spectrometer, scientists can calculate how long ago a person died. After 50,000 years, only a tiny fraction of the original radioactive carbon remains—but enough to tell the story.
The genetic shifts the Harvard team discovered were not random. Consider blood type. Humans carry two copies of the ABO gene, each a variant of A, B, or O—a system we share with other great apes. Over the last 6,000 years, the B variant became more common among West Eurasians while the A variant declined. The two variants have opposite effects on many traits, suggesting populations may benefit from maintaining a balance as pathogens and environmental pressures shift. Similarly, a variant of the HLA-DQB1 gene makes people susceptible to coeliac disease, where gluten triggers the immune system to attack the small intestine. Four thousand years ago, this disease-causing variant was absent; today it appears in one of every five people. Agriculture emerged 10,000 years ago, yet the variant's rise happened later and for reasons still unknown.
Around 8,000 years ago, humans began selecting for gene variants that produce lighter skin and pigmented hair. The researchers propose this was an adaptation to regions with low sunlight, where the body must synthesize more vitamin D—a particular challenge for farmers whose diets provided little of it naturally. Another striking finding involved the CCR5 gene: people with two copies of its delta-32 variant are completely resistant to HIV-1 infection. Between 6,000 and 2,000 years ago, this variant increased from 2 percent to about 8 percent among Western Eurasians. Yet HIV did not emerge until the early 20th century. The selection pressure must have come from other ancient pathogens, now unknown, that the variant also protected against.
Perhaps most intriguing were signals of selection for traits we think of as modern: performance on intelligence tests, household income, years of schooling, and markers of healthy lifestyle such as walking speed. The researchers also found that gene variants associated today with smoking were selected against in ancient times—despite the fact that tobacco did not reach Eurasia until Columbus brought it from the Americas fewer than 600 years ago. What ancient trait these variants actually governed remains a mystery.
The study opens a door to deeper understanding, but it also reveals a gap. South Asians carry genetic contributions from Iranian Neolithic farmers, western steppe herders, indigenous Eastern Eurasians, and East and Southeast Asian ancestors. A comparable ancient DNA study of South Asian ancestry could be equally illuminating, revealing which patterns of selection are universal and which are distinctive to particular regions and times. But such work requires something we have not yet fully assembled: a preserved legacy of our own ancestors' remains from thousands of years ago, waiting to be read.
Notable Quotes
It will be of interest to apply similar approaches to ancient DNA time series over longer times and to other world regions, to identify which patterns of selection are shared and which are distinctive to Holocene West Eurasia.— Harvard Medical School researchers
The Hearth Conversation Another angle on the story
Why does it matter that natural selection, not migration, drove these genetic changes?
Because it tells us what pressures were actually shaping human bodies. Migration would mean genes spread passively as populations moved. Selection means the environment—disease, sunlight, food—was actively favoring certain variants over others. It's the difference between drifting downstream and swimming toward something.
The coeliac disease variant increased from zero to 20 percent in 4,000 years, but agriculture started 10,000 years ago. Why the delay?
That's the puzzle. You'd expect the gluten exposure to immediately select against the variant, but it didn't. Maybe other factors were at play—population mixing, changes in how grain was processed, shifts in what else people ate. The researchers are honest about not knowing.
The smoking gene variants were selected against thousands of years before tobacco existed. How is that possible?
The variants we see today associated with smoking behavior may have controlled something entirely different back then. A gene doesn't have one fixed purpose. The same variant might have affected metabolism, or stress response, or something we haven't even identified. Selection was working on the ancient trait, not the modern one.
What makes the HIV resistance gene so strange?
It increased steadily for 4,000 years, then stopped. HIV didn't exist. So something else—maybe plague, maybe a virus we've never documented—was killing people who lacked that protection. Once that pathogen disappeared or changed, the selection pressure eased. We're seeing the ghost of an ancient threat.
Why focus on Western Eurasia? Why not study everywhere?
Partly because ancient DNA is easier to extract from cold, dry climates where bones preserve better. But also because the researchers are calling for exactly what you're asking—studies of South Asia, Africa, the Americas. Each region has its own story. We've only read one chapter.