It’s not easy living thousands of meters above sea level. The air holds less oxygen, there’s more harmful ultraviolet (UV) radiation from the sun, and food supplies vary dramatically from season to season. But that doesn’t stop nearly 5 million people from living on the Tibetan Plateau, the world’s highest at an average of 4000 meters. Now, scientists working with the largest-ever sample of Tibetan genomes have discovered seven new ways in which Tibetan genes have been tweaked to cope with high altitude, resulting in higher body mass index (BMI) and a boost in the body’s production of the vitamin folate.

Scientists have long known how the people of the Tibetan Plateau, including Nepal’s famous mountain-climbing Sherpa, deal with oxygen levels up to 40% less than those at sea level. Unlike most mountain climbers, whose bodies acclimatize to higher elevations by temporarily boosting hemoglobin—a blood protein that carries oxygen throughout the body—Tibetans have evolved a suite of other biochemical adaptations that let their bodies use oxygen extremely efficiently. That’s good news for the Tibetans, because too much hemoglobin makes the blood harder to pump and likelier to clot, increasing the chances of stroke and heart disease.

But the details of Tibetans’ adaptations have been a mystery. Previous studies have suggested that two genes, EPAS1 (inherited from ancient hominins known as Denisovans) and ELGN1, play roles in reducing hemoglobin and boosting oxygen use. To find out whether other genes are involved, a team of scientists led by Jian Yang at the University of Queensland in Brisbane, Australia, and Zi-Bing Jin at Wenzhou Medical University in China compared the genomes of 3008 Tibetans and 7287 non-Tibetans.

The team looked for common variants among the Tibetan genomes; they then computed whether those variants likely spread throughout the population by chance or by natural selection. EPAS1 and ELGN1 predictably popped out as strong candidates for evolutionary adaptations, they report today in the Proceedings of the National Academy of Sciences. So did seven additional genes: MTHFR, RAP1A, NEK7, ADH7, FGF10, HLA-DQB1, and HCAR2.

In Tibetans, the ADH7 gene variant is associated with higher weight and BMI scores, which could help the body store energy during particularly lean times on the hardscrabble plateau. The MTHFR variant also helps with nutrient deficiency: It boosts production of the vitamin folate, important for pregnancy and fertility. Folate breaks down when exposed to high levels of UV radiation, so high folate levels would compensate for their increased UV exposure. And HLA-DQB1 belongs to a family of genes that regulates proteins critical to the immune system, particularly important given that extreme living conditions like malnutrition can make people more susceptible to disease, Yang says. What the other four gene variants do is less clear, but they could be an evolutionary response to selective pressures besides high altitude.

The team also used its analysis to pin down a likely date for the split between Tibetans and the closely related Han Chinese population: approximately 4725 years ago, or some 189 generations back. That’s about 2000 years earlier than suggested by previous studies focusing on a different, more selective set of genes known as the exome, but it’s in line with recent archaeological findings that point to distinctly Tibetan permanent settlements appearing between 3600 and 5200 years ago, Yang says.

Lynn Jorde, a geneticist at the University of Utah in Salt Lake City who also studies high-altitude genetics, says the study’s large size lends credence to the findings. Such a large data set would help researchers detect more significant variants and weed out false positives. It could also explain why previous studies, including several by Jorde’s team, haven’t noticed these genes before.

But it will take more than just genomic studies to convince him and others in the field that any particular gene really is an evolutionary adaptation, he says. “I think statistical results, while very important, will only take us so far in searching for signatures of natural selection,” Jorde says. “We need to follow up with functional studies, such as in animal models or at least in vitro systems, to pinpoint and validate the biological basis for selection.”