A forthcoming study from genetic testing giant 23andMe shows that a person’s genetic code could be connected to how likely they are to catch Covid-19 — and how severely they could experience the disease if they catch it. It’s an important confirmation of earlier work on the subject.

People whose blood group is O seemed to test positive for Covid-19 less often than expected when compared to people with any other blood group, according to 23andMe’s data; people who tested positive and had a specific variant of another gene also seemed to be more likely to have serious respiratory symptoms.

The study, which was released on a preprint server and which has not yet been peer-reviewed, could extend and confirm earlier work on the subject; 23andMe’s study relied on a larger dataset than earlier work and included a more diverse set of participants, the company said. Experts who aren’t affiliated with 23andMe praised the study design and the work.


“They clarify further what our data could only vaguely hint at,” said Tom Hemming Karlsen, a physician at Oslo University Hospital who published an article in the New England Journal of Medicine on genetic links with Covid-19 severity in June, and who was not associated with 23andMe’s work.

But the outside experts also cautioned that the research won’t change treatment decisions.


“It doesn’t have practical implications. There’s no treatment decisions that will be made from it — it’s just an interesting finding,” said Jennifer Lighter, a pediatrician and epidemiologist at NYU Langone who was not involved in the research.

Unlike the study Karlsen and his colleagues ran, which only included people with severe Covid-19 symptoms, 23andMe included people who had both mild and severe cases — which allowed them to draw stronger conclusions, Karlsen said.

The company’s study participants are also more diverse than Karlsen’s, which only studied people in Spain and Italy. However, the 23andMe study’s demographics still don’t fully reflect the population of the United States. A little more than 11% of the people in 23andMe’s studies said they were Latino; less than 3% said they were Black. (Latinos represent about 16% of the U.S. population, while Black people account for about 13% of the population.)

Both Karlsen and 23andMe’s team found that the genes that code for a person’s blood type seemed to be linked to whether a person would test positive for Covid-19; another section of chromosome 3 — referred to in both papers as chr3p21.31 — seemed to be linked to how severe a person’s response would be to a Covid-19 infection.

Janie Shelton, a senior scientist at 23andMe and a lead author of the paper, and her colleagues noted that genetic associations did not seem to explain all the differences between populations; public health experts have noticed that people of color seem to be particularly at risk for Covid-19 due to some of the direct and indirect health effects of inequality and discrimination.

Both studies suggested one gene found in that area on chromosome three — SLC6A20 — might be particularly related to worse outcomes; however, it’s not yet clear how a particular gene could make a meaningful difference in a person’s response to an infection.

Both 23andMe and Karlsen ran the same kind of genetic analysis — a genome-wide association study. This particular method, which tries to find similar patterns in the genetics of people with a particular condition, has significant limitations. Scientists have suggested that the method is most useful when used to analyze hundreds of thousands of genomes.

For most scientists, getting that many samples would be difficult and expensive. But 23andMe has an obvious advantage — it has already sequenced more than 12 million people, according to the company’s website; over a million people agreed to participate in the company’s Covid-19 study.

“I do think that because of the power of our large sample size, we were able to detect that association pretty strongly,” said Shelton.

Without a clear understanding of which genes matter — and why — the impact of genetic studies on Covid-19 treatment plans will be limited.

“We’d have to find out why it’s significant — is it significant because it’s affecting blood clotting?” Lighter asked. “Unless we find out why there’s a difference, we wouldn’t target therapies or [adjust] a risk category.”

Leave a Comment

Please enter your name.
Please enter a comment.

  • Article was good about the science of DNA, but you lost credibility with me when you through in Blacks discrimination and inequality. That has nothing to do with genes / DNA. Why get political?

    • That’s not political, it’s a fact – related to many years of systemic inequality and discrimination that most white people don’t understand.

    • You think this is political?
      “public health experts have noticed that people of color seem to be particularly at risk for Covid-19 due to some of the direct and indirect health effects of inequality and discrimination.”
      Really? SMH Shame on you.

  • Very interesting information! Has there been any research involving Rhesus Negative factor & covid 19?

    As an O- carrier, the rh- negative factor ( – ) means blood is copper based, instead of nickel based like all ( + ) positive blood. RH- blood is also without protein.

    Could a potential covid antagonist be created using the RH- factor or RH- based copper blood?

    • Thank you for sharing that’s very intriguing! Metal based! Perhaps that is why people often describe blood as having a metallic taste. While some specify copper or penny? Obvious, I realize.
      I will have to read more

  • Thank you for raising awareness about the the interesting article you cited, I appreciate the precautions shared that treatment implications are limited. And if course some of the details suffering during generalization, so thanks to those who commented with details about O+ and O-.

    For those commentors who do not understand the social determinants of health and epigenetics, which no one yet completely understands, basically, stress hormones that are raised due to systemic racism or explicit discriminatory events, which can help us vote with such events, also have an impact on diseases like diabetes and heart diseases which also make one more susceptible to serious COVID-19 symptoms.

    A simple analogy by Camara Jones likens it to raising flowers, some with nutrients and some without. Y’all can look it up for yourself.

    And good on @Marc. There is only one human race, homo sapiens sapiens. The habit of continuing to use a false science as a category market is problematic and also complex, but we do know that 1. Latinx represents diverse regional colonized cultures by language(s) and that 2. All the phenotypic/skin color and physical features variations of humanity exist in the latinx category. At the same time, people are suffering undue burden of risk from this disease because of their social locations which correspond to either ethnicity or “race.” Unfortunately, in public health & demography we have not yet prioritized addressing this problem of how racialized perceptions have impacted the language and habits of research.

    As to poor Jake, suspicion is the outcome of untrustworthy behavior on the part of a society and the individuals within it. The best cure for such attitudes is justice and social transformation! Until then, love, forgiveness and overlooking the faults of others while working on our own is a more useful response.

  • It seems to me that this information strengthens the supercomputer results suggesting that Covid 19 is a vascular disease involving a Bradykinin storm, not primarily a respuratory disease involving a cytokine storm.

  • Why in 2020 are you referring to Latinos as a race? “ Latinos represent about 16% of the U.S. population, while Black people account for about 13% of the population.” There are self identified black Latinos living in The United States.

Sign up to receive a free weekly opinions recap from our community of experts.
Privacy Policy