Deepening diversity

 

The study aimed to better understand how genomic variants influence the risk of forming certain diseases in different ethnic groups. The work was funded by the National Human Genome Research Institute (NHGRI) and the National Institute on Minority Health and Health Disparities (NIMHD), both parts of the U.S. National Institutes of Health (NIH).

 

According to Dr. Lucia Hindorff, program director in the Division of Genomic Medicine at NHGRI and a co-author of the paper, “Genomic studies, especially genome-wide association studies, have been an incredible way to discover new genetic variants that are associated with a wide range of human diseases. We’ve known for awhile, though, that these studies are heavily biased toward participants of European ancestry. If research studies don’t include diverse participants ... we have incomplete information about whether the benefits or harms of the research apply more broadly.”

 

Researchers looked for genomic variants in DNA associated with measures of health and disease. The team found that some genomic variants are specifically found in certain groups. Others, such as some related to the function of hemoglobin, are found in multiple groups.

 

“This study is a major step forward in providing an example of how and why studying diverse populations is scientifically important. There were 27 novel genetic variants associated with 18 different health or disease characteristics. You can make existing information more useful by narrowing down which previously known genetic variants are more likely to be associated with disease, and which may be false positive signals. There were 38 genetic variants in this category,” Hindorff adds.

 

“If we keep focusing on the populations which are easy to study, which tend to be of European ancestry, we will make existing disparities in health knowledge worse. The research shows that if you take existing European ancestry studies and add either the same number of European ancestry participants or non-European participants, you discover more new genetic variants when you add more European ancestry participants,” she continues. “However, you widen the gap in the usefulness of the information—in the specific example in the paper, the ability to explain a person’s height.”

 

The study also assessed whether known disease associations with 8,979 established genomic variants and specific diseases in European ancestry populations could be detected in the studied populations. Their findings show that the frequency of genomic variants associated with certain diseases can differ from one group to another.

 

“Two genetic associations were highlighted in the paper,” Hindorff says. “The first was a genetic variant that was associated with cigarette smoking. While the variant was absent in most populations, it was found in 17 percent of participants who self-identified as Native Hawaiian or Pacific Islander.”

 

Not finding the variant in all groups, despite large numbers of participants in each group, strengthens the argument that findings from one population cannot always be generalized to others.

 

“A second variant was in the well-known hemoglobin gene, which is known for its role in sickle cell anemia. In the PAGE study, the hemoglobin variant was found to be associated with hemoglobin A1C levels, a marker of diabetes. The novel aspect was that this association was associated with HbA1C in Hispanic/Latino participants, as well as African American participants,” Hindorff notes. “This hemoglobin variant finding shows how what we see as ancestry-specific findings may be transferable to other groups that share components of that genetic ancestry.”

 

A vast majority of human genomics research uses data based mostly on populations of white European ancestry. An earlier study has shown that among 2,500 recently published human genomics papers, only 19 percent of the individuals studied were non-European participants.

 

“We have to be better about increasing diversity at all steps of the scientific enterprise. It starts with designing studies that can address scientific questions relevant to diverse populations, including for diseases with known health disparities,” Hindorff points out. “Funding agencies can also help with providing dedicated funding support to enhance diversity in biomedical research.”

 

“Here, we often see that increasing the diversity of the workforce can help with better recruitment and retention of study participants. Once the data are collected, there are ways to improve data analysis and interpretation by making sure that tools to analyze data from diverse populations are accessible and usable,” she says. “Finally, there is the step of making the study results and data available so that additional research opportunities can be identified, and the knowledge can be used to improve health care and health care delivery for all.”

 

The Nature study, led by researchers at the Icahn School of Medicine at Mount Sinai, the Fred Hutchinson Cancer Research Center and other academic centers, is the result of work undertaken by the Population Architecture using Genomics and Epidemiology (PAGE) consortium within the last five years.

 

“One key message from the PAGE paper is that genetic diversity is clearly a spectrum, which is not how we tend to think about diversity from a social perspective. From a genetics perspective though, it’s important to think about genetic diversity more broadly—there is a scientific advantage to analyzing the data from diverse populations together, rather than analyzing it separately,” Hindorff concludes. “Diversity is a theme that is taking hold in the scientific community, and I am particularly excited for the next generation of human genomics studies. It would be wonderful if we continued to consider diversity as one of the foundational tenets of designing and implementing genomic research.”