There’s no single ‘gay gene,’ according to vast DNA analysis
A new study that analyzed the DNA of nearly half a million people has found that while genetic differences play a significant role in sexual preference, there is no single gene responsible.
The findings, described Thursday in the journal Science, looked at sexual behavior and not sexual identity. Still, the results debunk the idea of a so-called singular “gay gene,” call into question such sexual orientation frameworks as the Kinsey scale — and hint at the complex factors that influence human sexuality, including society and the environment.
“The findings themselves reinforce this idea that diversity of sexual behavior across humanity is really a natural part of our overall diversity as a species,” said Benjamin Neale, a geneticist at the Broad Institute of the Massachusetts Institute of Technology and Harvard and one of the study’s senior authors. “That’s a really meaningful and important result.”
While estimates of same-sex experiences vary, a 2016 CDC study found that 6.2% of men and 17.4% of women between the ages of 18 to 44 in the U.S. reported at least one same-sex experience in their lifetimes. A smaller portion, 1.3% of women and 1.9% of men, identified as lesbian or gay, and 5.5% of women and 2% of men said they were bisexual — underscoring the difference between sexual behavior and sexual orientation identity.
Scientists have long probed the nature of same-sex behavior, finding some evidence in twin studies that genetics plays a role. But such research has been typically involved small numbers of people, and it hasn’t used modern methods of genomic analysis, scientists said.
Neale and an international team of researchers performed what’s known as a genome-wide association study. Using this method, scientists use statistical methods to search for connections between SNPs — single nucleotide polymorphisms, or individual differences in a single building block in the genetic code — and a particular trait.
Finding clear and verifiable patterns in that genetic data requires a huge sample, and the scientists knew where to find it. They pulled 408,995 individual records from the UK Biobank as well as 68,527 records from the U.S.-based personal genomics company 23andMe. This gave them an overall sample size of 477,522 people, 26,827 of whom reported same-sex sexual behavior.
The researchers found two significant spots in the genome that were linked to same-sex behavior across people of both sexes. And when they analyzed male and female genomes separately, they found three more — two specifically for men and one specifically for women — bringing the total number of significant genetic markers up to five.
Nonetheless, when taken altogether, these five locations on the genome could account for much less than 1% of same-sex sexual behavior on a population level, the researchers said.
Another analysis in the paper, which did not focus on DNA but on familial relationships between 106,979 pairs of individuals, suggested that a slightly larger share of the variation in same-sex behavior, 32.4%, could be attributed to genetics. That number may take into account other complex genetic effects beyond SNPs, though it might also be influenced by some assumptions baked into the framework, the scientists said.
Among the five significant SNPs they found, the ones specific to men were linked to smell receptor genes, sensitivity to certain scents, and regulation of the sex hormones such as testosterone.
That finding “makes a certain amount of sense,” Neale said, “but again, we don’t have much more to say beyond that sort of high-level description.”
The incomplete overlap between the genetic markers linked to male and female same-sex behaviors is a sign that slightly different processes may be at work in men and women when it comes to expressing sexual preference and behavior. It may also speak to differing influences of gendered social norms, said Mills, who wrote a commentary on the results.
If less than one-third of a person’s sexual preference and behavior is linked to genetics, where does the rest come from? Environment, culture and other factors may play a significant role, Neale said.
It’s somewhat akin to traits like height, which have a certain genetic component but can also be influenced by a complex array of other factors, such as nutrition and environment.
Exactly which environmental and cultural factors play a role is unclear, because those are varied and complex and are much harder to pin down and study than specific genetic markers, the study authors said.
The scientists also looked specifically at the “nonheterosexual” subjects in the study — those who had at least one same-sex experience — and asked them what proportion of their sexual partners were of the same sex. Responses varied across a 6-point scale, from “other sex mostly” to “same sex only.”
In this respect, the researchers found, genetics had a stronger influence on same-sex behavior in men than in women.
They also saw that the genetic factors influencing the proportion of same-sex to other-sex partners a person had were different from the ones that separated those who had any same-sex experiences from those who had only other-sex experiences.
This means that the Kinsey scale and other frameworks for sexual behavior that assume that more same-sex attraction means less opposite-sex attraction are not accurate. They must be based on a misunderstanding or an oversimplification of the processes at work, the scientists said.
This confirmation of the wide diversity of sexual behavior echoes what the researchers said they heard in discussions of the results with representatives of the LGBTQ community.
“The LGBTQ+ community has been arguing for a long time that there’s this range of sexualities, it’s not binary zero and one,” Mills said.
The scientists were quick to point out that the findings were population-based and could not be applied on an individual level. They also warned that the work should not in any case be used to try to “convert” people who engage in same-sex behaviors, and that to consider doing so would be a gross misrepresentation of the study’s findings.