“Looking across the whole genome,” James Fowler from University of California, San Diego, says in a news release, “we find that, on average, we are genetically similar to our friends. We have more DNA in common with the people we pick as friends than we do with strangers in the same population.”
He and Nicholas Christakis of Yale conducted a genome-wide analysis using data from the Framingham Heart Study, which monitored heart disease in thousands of people and their children over decades. That study also included information about who is friends with whom, and one of its known limitations turned out to be an advantage for this present study: Most of the data come from people in the same population and of European descent. That means if there are genetic similarities, it’s not just because we tend to become friends with people from similar backgrounds.
Focusing on the same 1,932 individuals, the duo shuffled them around to compare pairs of unrelated friends against pairs of unrelated strangers. Friends, they found, share one percent of their genes. Buddies are as related as fourth cousins, or people who have the same great-great-great grandparents.
“Most people don’t even know who their fourth cousins are,” Christakis says. “Yet we are somehow, among a myriad of possibilities, managing to select as friends the people who resemble our kin.” Their work was published in Proceedings of the National Academy of Sciences this week.
This “functional kinship” may have offered evolutionary advantages. Not only would friends work together better, but some traits might only work if someone else also has them. “The first mutant to speak needed someone else to speak to. The ability is useless if there’s no one who shares it,” Fowler speculates. “These types of traits in people are a kind of social network effect.”
In particular, they found that genes affecting the sense of smell are the most similar in friends. They found the opposite for genes controlling immunity: Friends have different genetic protection against diseases, which could help reduce the spread of pathogens.
The team also developed a “friendship score” to predict who will be friends. They say they can predict friendship with as much confidence as doctors can predict your likelihood of obesity or schizophrenia based on genes alone. “We can do better than chance at predicting if two people are going to be friends if all we have is their genetic data,” Fowler tells Washington Post.
Other researchers have expressed concerns about various factors that could be affecting the results, BBC explains, such as relationships that are difficult to confirm and factors that drive both friendship and genetic similarities. “We don’t know if the results can be generalized to other ethnic groups,” Fowler acknowledges. “My expectation is that it will, but we don’t know.”