At the consumer genetics giant 23andMe, CEO Anne Wojcicki last week issued a remarkable statement calling her product “euro-centric” and saying her company is “part of the problem.”
Competitor Ancestry put out its own statement saying it has “a long, long way to go to make our product experience as inclusive as it can possibly be for everyone.” And the DNA testing startup Nebula Genomics published a blog post about the lack of diversity in their field — a post that was drafted months ago, but which the company held off on publishing because of concerns about how it would be received.
The reflections in the consumer genetics industry were triggered by the anti-racism protests following the killing of George Floyd. But the underlying issues are not new. The field of genetics has long been seen as too white — in its leadership, its data, and its products.
Geneticists have been grappling for years — some more urgently than others — with the underrepresentation of Black individuals and other people of color in genetic studies, databases, and the reference genome. While there’s been some halting progress, these problems have resulted in tests that only work for people of European ancestry and have undermined hopes that everyone might benefit from personalized medicine.
STAT asked five experts in genetics and in health disparities to weigh on what 23andMe and its competitors should do to address these problems. Here’s what they had to say:
Build a diverse team
In 23andMe’s statement last week, Wojcicki said she was “ashamed to say I do not have a single black employee who is at Director level or above.” She also said her company’s management team, board, and employee base need to be more diverse.
Lack of diversity among employees and executives at a company like 23andMe can send a message to potential customers — and potential hires — from underrepresented groups that they’re neither welcome nor represented, said geneticist Tshaka Cunningham.
“How can you foster trust with me as an African American if I don’t see any African Americans on your team?” said Cunningham, co-founder and chief scientific officer of TruGenomix, a startup that’s working on recruiting diverse cohorts to develop a test to gauge genetic risk for developing post-traumatic stress disorder.
Addressing inequities in hiring, promoting, and retaining employees is important on a fundamental level, Cunningham said. But a diverse team might also help a company avoid missteps such as the ad from Ancestry that was pulled last year after being widely condemned as racist; it depicted a romance between a white man and a Black woman in the antebellum South — without addressing the unequal power dynamics and exploitation involved in such relationships of that era.
Having a broader range of perspectives at the table could also help companies better navigate the shortcomings of polygenic risk score tests — analyses that gather multiple genetic variants that, together, are used to predict someone’s chance of developing a disease. So far, many commercial tests have come with warnings that they’re not very accurate or are even useless in people who are not of European ancestry. That’s due to the disproportionately low representation of other populations in the datasets used to build the tests. 23andMe’s polygenic risk score test for type 2 diabetes, introduced last year, had lower predictive power in African Americans than in any other group.
“By the time they get to posting the statement on the website that the polygenic risk score is not going to be of any use to Black people but we’re still going to take your money for it, there are many other decisions that have been made and plenty of places where that discussion should have been halted to have a deeper conversation and thought about what to do,” said Consuelo Wilkins, a physician who serves as vice president for health equity at Vanderbilt University Medical Center.
Be transparent and demonstrate the value of genetics
Before the pandemic, Cunningham traveled widely to speak at Black churches about the value of genetic research. It was part of his role as executive director and science advisory board member of the nonprofit Faith Based Genetic Research Institute.
“As I go around the country talking to people of color about genomics and encouraging them to participate and get sequenced, the questions I get are: What are they going to do with that information, and how are they going to use it to hurt me? And then why would I give my information so somebody else can make money off of it?” Cunningham said.
23andMe spells out its privacy policies online. Its customers must opt in to any of its research and it promises not to sell people’s individual-level data, though it does reserve the right to sell data that are aggregated and anonymized. But those disclosures may not be enough to build trust in communities of color that have been betrayed by medical researchers many times before.
That’s why 23andMe and other genetics researchers must go above and beyond when it comes to being transparent with minority communities about how they’re using their data, said Ky’Era Actkins, a Ph.D. student who studies human genetics at Meharry Medical College, a historically Black institution in Nashville, Tenn.
Another key step: “Try to show people the benefits of what the science community is doing in terms of helping different minority populations or diverse populations and those who may not have always been very represented in the research that has been done historically,” Actkins said.
But complicating such communication is the fact that 23andMe and other genetic testing companies aren’t making the same pitch to minority communities as, say, the National Institutes of Health’s All of Us data-gathering program, which is meant to advance precision medicine for everyone in the U.S. While these companies do have a mission to advance scientific research, they’re also trying to sell spit kits and make money. It’s not always clear that people of color will benefit from buying a 23andMe test or participating in one of the company’s research studies any more than they would from engaging with some other consumer product.
“The question is why would someone want to participate. It’s a for-profit program, and for a for-profit company.”
Consuelo Wilkins, vice president for health equity at Vanderbilt University Medical Center
When it comes to 23andMe, “the question is why would someone want to participate. It’s a for-profit program, and for a for-profit company. Why would people who are ethnic minorities want to pay for services that they may benefit less from than people of European descent?” Wilkins said.
It’s critical for 23andMe to demonstrate the value of its spit kits to communities of color, Wilkins said. The company should address whether it plans to take steps that would make the decision to hand over genetic data to the company more valuable over time, she said. That might include developing more robust reference datasets, reinterpreting genetic variants currently classified as being of unknown significance, and expanding its ancestry data to be more precise for people who are of African and Asian descent, she said.
Recruit people of color to participate in research
The experts consulted by STAT called on 23andMe to prioritize recruiting underrepresented populations to enroll in its research studies — an endeavor that, to the company’s credit, it’s been working on for years.
In 2011, the company launched its Roots into the Future program, a genetic study of more than 10,000 African Americans. In 2016, 23andMe started its African Genetics Project, a study recruiting first- or second- generation emigrants from several countries in sub-Saharan Africa. 2018 saw the launch of 23andMe’s Global Genetics Project, which aims to expand the company’s population reference data by collecting data from individuals with recent heritage from dozens of countries in Africa, Oceania, Asia, and the Americas.
23andMe has also received funding from the NIH to try to address disparities in genetic research, such as a project to create a reference panel of information to help scientists develop studies that benefit the African American community. The NIH has also funded 23andMe’s effort to develop a new way to detect disease-causing genetic variants among ethnically diverse populations.
These projects have made a dent in the problem. But the people of color recruited through these initiatives still represent only a tiny fraction of the millions of people in 23andMe’s databases.
Joyce Tung, 23andMe’s vice president of research who’s been at the company for more than a decade, acknowledged in a statement to STAT that her company and the broader scientific community “still have a long way to go” in diversifying genetic research. “There is no silver bullet to solve this problem, and it will require prioritizing multiple efforts now and into the future in order to ensure that we help everyone benefit from the human genome,” Tung said.
Consider compensating for data
Separate from concerns about diversity, some experts and startups have questioned the idea that people should be asked to pay for genetic tests and reports about their health and ancestry from commercial companies. 23andMe, after all, benefits enormously from the genetic data that people pay $100 or more to share — and the scarcity of data from people from underrepresented groups makes their samples all the more valuable.
All of which raises the question: Should 23andMe pay people Black communities and other underrepresented populations for their data?
Cunningham thinks so. He pointed to the example of LunaDNA — a platform which offers ownership shares to people who contribute their DNA — and called for 23andMe to adopt a revenue-sharing model with communities of color.
“The current model just doesn’t work for many folks,” Cunningham said.
Not all the experts agreed. They said they worried that offering money to give up genetic data, particularly given the economic crisis caused by Covid-19, could be considered coercive or might be frowned upon by the institutional review boards that decide whether to green-light studies.
“With the current unemployment rate, how do you incentivize that in a way that would not be judged by an institutional review board … as being coercive in some sense?” said Nancy Cox, who directs the genetics institute at Vanderbilt University Medical Center.
Cox did not see an easy answer.
Recognize that responsibility goes beyond 23andMe and its competitors
Experts consulted by STAT emphasized that the blame for the whiteness of genetics — and the responsibility for diversifying it — should not lay entirely at the feet of 23andMe.
Along with academic researchers and other genetics companies, other players in the ecosystem around genetic data have a responsibility to take steps that could incentivize more equitable genetic research, the experts said.
For example: Drug companies that purchase aggregated and de-identified genetic data from companies like 23andMe could commit to prioritizing purchasing diverse datasets or to paying more for them, said Robert Green, a medical geneticist and physician at Harvard and Brigham and Women’s Hospital. Green advises several genetic testing companies and has worked closely with Cunningham and others to try to recruit diverse populations for his team’s genetic studies.
Green added: “We all have to take some responsibility at every level for this.”