Letters: 23andMe and Your Biodata

On an episode of The Time of Monsters, Jeet Heer spoke with Professor Myles W. Jackson about the implications of DNA databases and the question of who “owns” this information, particularly as it pertains to 23andMe and its customers [“Big Pharma Wants to Own Your DNA Info,” December 21, 2022]. This is an important topic and one that deserves healthy debate. We’ve always maintained at 23andMe that our customers own their DNA, and that testing is first and foremost a choice. Those who are not comfortable with DNA testing, for any reason, should not get tested. We carry this ethos of choice—and transparency—throughout our product and the manner in which we conduct business. Choices presented to customers when they decide to test include what types of information they’d like to receive from our test, whether or not they’d like 23andMe to store their DNA sample (they can elect to have it discarded), and, crucially, how their data may or may not be used by 23andMe or third parties. We talk publicly about our research collaborations, such as those with academic institutions, nonprofit organizations, and pharmaceutical companies, so that customers are aware of how their information may be used, if they do in fact choose to opt in to our research program.

At 23andMe we conduct research to help advance science, develop our product to return new insights to customers, and develop new medicines for serious unmet medical needs. Thus far, 23andMe research has led to more than more than 200 publications, most of which are in collaboration with academic researchers. This program is conducted pro bono for the purpose of publishing genetic insights that may allow the field to advance research in a number of areas, including the genetics of cancer, Covid-19, Parkinson’s, and many more. We’ve also been able to expand our ancestry composition to include areas of the globe that historically have been underrepresented in genetics. These include regions in Africa, Asia, and the Americas. Research is conducted in-house by 23andMe scientists, studying de-identified, aggregate information from customers who have consented to participate under our Main Research Consent. The process of consenting to research is separate from the terms of service and is entirely voluntary. Notably, this is an opt-in process, meaning customers must proactively elect to participate and are not defaulted into participation. Customers have the option to change this consent at any time. Our research is overseen by a third party institutional review board, or IRB, to ensure that the work is conducted to the highest ethical standards and guidelines and that proper, informed consent is carried out. 23andMe will not share any customer data with third parties for research purposes without a customer’s explicit consent.

In addition to providing this context on our research program, we also wanted to provide clarity on several areas of the discussion that included misstatements about 23andMe:

• 23andMe does not sell, or otherwise share, any customer information to/with insurance companies or employers.
• 23andMe does not sell genetic data for advertising purposes.
• Third party collaborators do not directly access the 23andMe database. As noted above, research is conducted in-house at 23andMe, and results are then shared with collaborators. These results do not include any individual-level information unless customers have consented separately and explicitly to share that information.
• There was also mention of 23andMe trucks giving blood tests in Ukraine; 23andMe does not give blood tests and does not operate in Ukraine.

We believe in providing people with access to their DNA so that they may learn and benefit from the human genome. As part of this mission, stewardship of customer data is of paramount importance. We strive to operate transparently so that we can continue to empower our customers to make informed choices on if and how their data are used and shared. Being worthy of the trust of our customers is absolutely critical to our success as a company.

Joyce Tung, PhD
Vice President, Research
23andMe
sunnyvale, calif.

Myles Jackson Replies

I would like to thank Dr. Tung for her very thoughtful letter. I truly do appreciate the chance to enter into a dialogue with her. As for not selling or providing information to insurance companies, Dr. Tung is absolutely correct. I apologize to 23andMe for lumping them with other direct-to-consumer (DTC) genetic-testing companies. I certainly have never accused 23andMe of selling consumers’ genetic information to companies without their permission. This type of transaction would be a direct violation of the Genetic Information Nondiscrimination Act signed into law in 2008 by then-President George W. Bush.

As I mentioned on the program, the customer can opt out of the sharing of information. When I lecture on the topic, I highly recommend to those interested in genetic testing to read the terms of service. The data, of course, are always aggregated and anonymized; I never said anything to the contrary. Despite the aggregation and anonymity, it is still possible to identify individuals. A chilling article in The New York Times back in 2018 stated, “Already, 60 percent of Americans of Northern European descent—the primary group using these sites—can be identified through databases whether or not they’ve joined one themselves,” citing a study from the journal Science. Researchers say it will soon be possible to identify up to 90 percent of European Americans from genealogical databases. As has been well documented, in the US a number of individuals have been arrested for crimes committed by comparing DNA at the crime scene with a relative’s genetic sequences deposited by the individuals themselves (not the DTC genetic-testing companies) with GEDmatch and MyHeritage, for example. And there was a case where GEDMatch was hacked back in 2020. It is more difficult, however, to hack the larger databanks of 23andMe and AncestryDNA.

Indeed, I am interested in historical notions of privacy, ownership, and ownability. In November 2017, US Senator Chuck Schumer of New York called on the Federal Trade Commission to investigate DTC genetic-testing companies over concerns about privacy. In 2018 the FTC conducted an investigation, and in 2019 they offered suggestions to those companies. According to a 2020 Consumer Reports white paper, “The Federal Trade Commission can control the marketing practices of the DTC genetic testing company, but they are not currently using their enforcement powers to protect the privacy of the information.” The report went on to recommend a policy approach to protect the privacy of genetic data. In short, I agree with Dr. Tung that 23andMe has followed the requisite guidelines and best practices. The question is if the guidelines need to be rendered more stringent.

The discussion on advertising was not meant to be about 23andMe selling its data. Rather, I am interested in how DTC genetic companies advertise and market their wares aimed at specific groups. I think the commercials of AncestryDNA, which I specifically mentioned in the broadcast, speak for themselves. Back in 2008, 23andMe advertised on its website a technique to find African American ancestry called “Ancestry Painting.” At other times, it has advertised this service specifically for Native Americans, Asian Americans, and European Americans. In 1993 Congress passed the NIH Revitalization Act, which established the guidelines to include women and people of color in clinical research. The issue is about giving a medical voice to those who have been marginalized historically. For a brilliant analysis on that topic, see Steven Epstein’s Inclusion: The Politics of Difference in Medical Research (University of Chicago Press, 2007), which I briefly mentioned in the interview. In a very real sense, the “white male” should no longer serve as the exclusive model for medical care. And I certainly find that to be a good thing. That said, as numerous sociologists and anthropologists of science and medicine have argued, such a move has, in part, re-established (some argue reified) race as a biological category in highly problematic ways. As many molecular biologists tell us, there are numerous ways in which one can account for human variation. Race is not the only one. Yet despite what many molecular biologists argue is a very poor proxy for human variation, race is privileged for important historical reasons in the United States.

I thank Dr. Tung for pointing out that 23andMe’s research is in-house, and that third parties do not have access to their databanks. Those proprietary databanks are critical both scientifically and economically. I am interested in the ways in which the consumer is simultaneously the product. So, as advertised, the $250 million collaboration between GlaxoSmithKline and 23andMe, which was extended about a year ago for $50 million, is about the use of genetic sequences of individuals—again, personal information is removed and the individual’s permission must be granted—with a view to treat illnesses based on precision medicine, also known as personalized medicine. That was the point I was attempting to get across. 23andMe CEO Anne Wojcicki said early in 2022 that health, not ancestry testing, will drive growth long-term for the company. She added: “The ancestry market is something people can relate to and it’s substantial, but health is just much larger.” 23andMe is also entering into the telehealth market with its $400 million buyout of Lemonaid Health.

Finally, as for the story about the “trucks in Ukraine” gathering blood samples, as I mentioned on the program, that was an anecdote told to me circa eight-to-10 years ago by a former colleague who had been researching in Ukraine. Indeed, 23andMe uses cheek swabs to collect samples of DNA. I should have confirmed the story with another party before I mentioned it.

Again, many thanks to Dr. Tung for taking the time and participating in this dialogue.

Myles W. Jackson
princeton, n.j.