Humans Haven't Stopped Evolving

During the last 50,000 years, human beings have transformed from nomadic hunter-gatherers roaming savannas to a species that traveled to the moon, invented the internet, and created an artificial form of intelligence. Yet many scientists believed that, during the same time frame, human DNA hadn’t changed much. Most research during the past two decades hinted that human evolution has largely slowed or changed so subtly that existing tools couldn’t detect it; evidence seemed to suggest that civilization put humans beyond the reach of natural selection.

“It’s like we had in our heads that a few hundred thousand years ago we had reached some kind of evolutionary optimization,” says professor of genetics and human evolutionary biology David Reich.

That changed with Reich’s latest work. Using the largest collection of ancient human DNA samples ever compiled, paired with new statistical methods, Reich and his colleagues discovered that the human genome has actually evolved rapidly during the past 10,000 years. The study, published in Nature in April, showed hundreds of genetic shifts in 16,000 individuals who lived in modern-day Europe across an 18,000-year period. The longer timespan provided a baseline that allowed the researchers to see how genes evolved during the most recent 10,000 years.

What’s more, the findings demonstrate that the pace of evolution is accelerating, with more intense natural selection occurring in the past 5,000 years than in the 5,000 years that came before. “Natural selection has not slowed down,” says Ali Akbari, a senior scientist at the Broad Institute and the paper’s first author. “We were just missing the signal.”

By collaborating with roughly 250 archeologists from around the world, Reich and his team amassed a collection of bones and teeth from more than 5,836 ancient humans. They extracted and analyzed ancient DNA from the samples and built a Genetic Relationship Matrix—a chart showing how genetically similar each person is to each of the other people studied. A common challenge in population genetics is the difficulty of filtering out “noise.” People migrating, mixing, and having children can produce large shifts in gene frequencies that can obscure small changes in their DNA that result from natural selection alone. For instance, certain traits, such as brown eyes, may rise in a population, but it’s hard to know whether that is because a wave of brown-eyed families migrated in or because having brown eyes confers a survival edge.

Using the matrix allows the researchers to cancel out this background noise so they can see the signal of evolution clearly. If a gene’s frequency trended upward consistently over diverse places and times (even just a little bit), Reich and Akbari could be confident it was due to the forces of natural selection.

Breaking with the pattern followed in many previous studies, the researchers did not focus on the genetics within individual demographic groups, such as “early European farmers” or “nomadic hunter-gatherers.” Instead, they looked for genetic variants that occurred within multiple populations at multiple time points, including in modern populations, letting the data tell the story of evolution without forcing it into categories.

The end result was the identification of 479 genes that showed evidence of natural selection. The sheer number of these genetic variants—which represent small changes in the DNA sequence of the gene—was staggering, even to the study’s authors. “It was a crazy result,” Reich says.

Genes involved in immunity were most likely to show evidence of selective pressure, and consequently, their prevalence rose and fell dramatically over time. A gene linked to an increased risk for multiple sclerosis, for example, emerged roughly 6,000 years ago in a region south of the Caucasus Mountains, and within 4,000 years it settled into 16 percent of the population. Then, 2,000 years ago, it began to recede. A gene linked with tuberculosis risk followed a similar arc, appearing 9,000 years ago and climbing to 9 percent during the next 6,000 years before reversing course. Today, it’s found in 3 percent of the population.

Other genes, such as those for male-pattern baldness and skin pigmentation, have also evolved. Interestingly, the baldness gene has diminished under selective pressures during the past 7,000 years, plummeting in frequency from 50 percent to 20 percent. There have also been shifts in blood type, with Type B rising from 0 to 10 percent during the past 10,000 years, while Type A simultaneously declined.

The study also contradicted some long-held beliefs. The genetic risk factor for cystic fibrosis, for example, was long thought to persist in European populations because carrying it confers protection against cholera. However, this study found no evidence of selection during periods when cholera was a persistent threat, suggesting that the explanation may need to be revised.

The results are fascinating, but Reich cautioned against drawing conclusions from individual genetic changes. The function of a gene thousands of years ago might be different from its role today: in a phenomenon called pleiotropy, a single gene can control seemingly unrelated traits, such as the gene that causes both white fur and deafness in cats.

“Each one of these variants needs a whole Ph.D. [dissertation] to understand what they do in their different contexts,” he says. He adds that genes take on new functions over time, and researchers can’t just assume that the functions of ancient genes are the same as those of their modern counterparts.

Akbari adds, “These results do not necessarily mean that people who are living today are healthier or smarter or sicker.” Because genes interact with the environment, it is difficult to draw conclusions about ancient populations, he explains, because “we don’t have this result in context.”

Next, Akbari and Reich want to gather data from additional populations, perhaps those that were geographically isolated, to gain a broader view of evolution worldwide; they also hope to probe the variants they identified in this study to better infer what their function may have been in the past. Regardless of how the individual traits have shifted, the takeaway from this work is that the human population is far from done evolving.

“Evolution is happening constantly, and it’s rapidly responding to the cultural, economic, and environmental changes we’ve imposed on ourselves over the last 10,000 years,” says Reich.