Genetic lineage can predict disease risk more accurately than race

By Monisha Ravisetti
April 29, 2021
Genetic ancestry is important in assessing disease risk, but it's important to steer clear of racism. (Unsplash/Laura Furhman)

Genetic ancestry is important in assessing disease risk, but it's important to steer clear of racism. (Unsplash/Laura Furhman)

Considering race and ethnic background when assessing disease burden is believed to walk a fine line between informed care and systemic racism, but arguing that heritage can weigh heavily on one's propensity for illness, researchers are advocating for a refined way of incorporating genetic ancestry without teetering toward the side of prejudice.

A study published April 15 in Cell demonstrates a specialized method of investigating genetic ties to health outcomes by comparing tens of thousands of de-identified DNA samples and electronic health records. The catalog provided clear associations between hereditary patterns and certain illnesses, the authors say, illustrating the utility of such factors in predicting health conditions when defined carefully.

"There's a long history of using race and ethnicity in research because we know it's very true that we see differences in diseases and race," said study author Eimar E. Kenny, founding director of Icahn School of Medicine's Institute for Genomic Health and associate professor of medicine and genetics at Mount Sinai.

"On the other hand," she continued, "there's an acknowledgement that using race and ethnicity in medicine can cause some harms — it can even cause racism."

Kenny emphasized that present methods that attach individuals to generalized ethnic groups, such as Black or Hispanic, without pinpointing exact ancestry, paint an inaccurate picture of relevant disease burden.

"Asthma is a really good example," she told Fastinform. "If you were just to consider Hispanic/Latino as one monolith, you would miss the fact that the rates of asthma are highest in Puerto Rican populations compared to anybody else, and lowest in Mexican populations compared to anybody else."

While acknowledging that these associations aren't always the case, Kenny said that genetic pieces of information can be invaluable when predicting which patients are prone to which diseases.

"We sort of demonstrated, using research data, that this layer of information really ought to be taken into account when we think about disease risk at an individual level or a population level," she said.

The researchers evaluated more than 36,000 diverse DNA samples given by volunteers to an established biobank at Mount Sinai. Samples were linked to the respective individual's electronic health record collected by the institution, meaning each sample was associated with disease data. All information was de-identified.

The result was a comprehensive database of genetic lineage and disease burden.

"It's using data that health systems capture operationally to function and to do business," Kenny said, "and repurposing that to tell us about health outcomes."

She continued, "For example, if you have a medical billing code for Type 2 diabetes, that's giving a piece of information that you might actually have Type 2 diabetes."

The team drew the correlations by using machine-learning mechanisms that indicated some genetic traits found across sub-populations have a strong relationship with the likelihood of developing particular diseases. The asthma example was one of the notable findings.

"Genetic ancestry is telling us about our human history and our lineages," Kenny said, "every linear trend tracing back to our ancestors in sub-Saharan Africa and the types of movements and migrations of humans around the planet, back and forth."

The work draws an updated picture of demographic connections and history for any human on Earth and from every source of the family tree. That's an upgrade, Kenny says, from the standard model of ticking a box about race that oversimplifies one's identity.

"A lot of people, faced with a question about their race and ethnicity, either don't opt to complete the questionnaire or tick multiple boxes," Kenny said. "It's an issue that is very important to think about, because obviously people's self-identity can be very complicated."

In the same vein, the researchers acknowledged their study does not explore biological sex differences or mixed-race individuals, as the team corrected for these factors. 

"We did see that about 6% to 7% of participants in our study appeared to have more than one group that they could necessarily fit in," she said. "It's something that we're looking to the future for."

Kenny further explained that there are two things to think about regarding genetic ancestry's application in the clinic. First is determining the arenas of clinical care where this information could be applied. She suggests that the most obvious would be specialties of medical genetics but also subspecialties such as cardiology, immunology, neurology and oncology, which all study inheritable diseases. 

The second critical factor, Kenny says, is understanding what regulations must be in place to ensure unprejudiced use of data that indicates heritage. She notes that health systems are getting to a point where people can be associated with a particular race from MRI imaging, let alone genetic information.

"I'm stressing that we don't just immediately put this into clinical care — we really have to do the next step of implementation research," Kenny said. "And from that … start to also work with policymakers and with regulators."

She continued, "It requires the same level of scrutiny and the same level of ethical upholding of standards in health systems as applies to other any other patient data."

The study, "Toward a fine-scale population health monitoring system," published April 15 in Cell, was authored by Gillian M. Belbin, Sinead Cullina, Stephane Wenric, Noam D. Beckmann, Arden Moscati, Emily R. Soper, Benjamin S. Glicksberg, Denis Torre, Ariella Cohain, Steve Ellis, CBIPM Genomics Team, Erwin P. Bottinger, Judy H. Cho, Ruth J.F. Loos and Eimear E. Kenny, Icahn School of Medicine at Mount Sinai; Genevieve L. Wojcik and Elena P. Sorokin, Stanford University; Ruhollah Shemirani and Jose-Luis Ambite, University of Southern California; Danny S. Park and Noura S. Abul-Husn, Icahn School of Medicine at Mount Sinai and University of California, San Francisco; Adam Auton, Albert Einstein College of Medicine; Noah A. Zaitlen, University of California, Los Angeles; Christopher R. Gignoux, University of Colorado Anschutz Medical Campus; and Regeneron Genetics Center.

Saving
We use cookies to improve your experience on our site and to show you relevant advertising.