The promise — and the gap
Imagine a future where your doctor estimates your risk of a heart attack, a specific cancer, or a mental health condition not from a quick checklist but from a number derived from your DNA.
That future is closer than many people realize. Polygenic risk scores — PRS for short — are already entering some clinics. But a new analysis of 25 years of research shows that the people studied so far don't look much like the world that will use the technology.
PRS combines hundreds or thousands of small genetic variants to estimate the risk of common conditions. They've been used in research on heart disease, schizophrenia, diabetes, breast cancer, and dozens of other illnesses.
Some health systems have started piloting them as part of preventive care. The first commercial PRS tests are already on the market.
That makes the question of who is included in the underlying research more urgent than ever. A risk score built from one population may not be accurate when applied to another.
The old way versus the new way
For most of medical history, risk prediction was based on family history and lifestyle factors. Genetics played a role only in obvious single-gene conditions.
PRS changes the math. Even small genetic differences, summed across thousands of locations in the genome, can produce a score that meaningfully separates higher-risk and lower-risk individuals.
But this only works well if the underlying genetic data look like the patient being scored. Scores trained on people of European descent often perform poorly when used on people of African, Asian, or Latin American ancestry.
How the analysis worked
Imagine taking a snapshot of a research field every year for 25 years. Where did the papers come from? Who funded them? What diseases were being studied? Which groups of patients were left out?
The team did this kind of bird's-eye view by analyzing every PRS-related publication in a major scientific database between 1999 and 2024. They mapped publication growth, geographic spread, funder patterns, and how the topics evolved over time.
That kind of analysis can't tell us whether any single test is accurate. But it tells us which directions the field is moving — and which corners are being neglected.
The study snapshot
The team analyzed 10,269 publications across 2,185 sources, tracking growth rates, country and institutional contributions, funding sources, collaboration networks, and shifting research themes. They modeled the growth curve to estimate when the field is approaching maturity.
PRS research has grown explosively. Annual publications have risen at more than 21% per year since 1999, with a sharp acceleration after 2017. By 2024, more than 1,500 PRS-related papers were published in a single year.
Mathematical modeling suggests the field is approaching a turning point in 2026, where it shifts from rapid growth into a more mature, applied phase.
But growth came unevenly. The United States, China, and the United Kingdom together account for the majority of publications. A handful of elite research centers dominate output. Funding is also concentrated, with a small number of public and philanthropic sources providing roughly a quarter of all acknowledged support.
Topics shifted from foundational genetic concepts toward concrete disease prediction in mental health, heart disease and diabetes, and cancer.
This concentration matters because the people in those studies often don't reflect the people who will receive the tests.
Where this fits in the bigger picture
Genetic medicine has always struggled with diversity. Early reference databases overrepresented European populations, and PRS inherited that problem. While many researchers are now actively working to broaden the data, the catch-up is slow.
This new analysis shows that the structural issue runs deeper than just patient enrollment. Funding networks, institutional partnerships, and research priorities are all centered in a few places. Solving the diversity problem requires changing those upstream patterns too.
If you're considering a genetic risk test marketed to consumers, ask how it was developed and which populations it has been validated in. Many of these tests work much better for people of European ancestry than others.
If your doctor offers PRS as part of clinical care, ask the same questions. The accuracy and meaning of the number on the report depends heavily on whether your background was well represented in the underlying research.
The analysis looked at publication patterns, not at the underlying ancestry of the patients in those studies. So while it doesn't directly measure diversity, the geographic and institutional concentration it documents tracks closely with that broader concern. Bibliometric methods also can't tell us whether any specific clinical tool is reliable.
The field's next decade will be defined by how well it broadens its base. New initiatives are funding genome studies in Africa, Latin America, the Pacific, and South Asia. As those data come online, PRS accuracy for non-European patients should improve. The question is how quickly that progress reaches the clinic.