In Simple Terms
A small inherited difference can make the same medicine work well for one person but less well—or cause more side effects—for someone else. A study of about 100,000 Americans found many gene differences linked to how people respond to medicines, including some that earlier research had missed. Turning those findings into better everyday care will take reliable funding, testing, and information doctors and patients can use.
Nearly 100,000 People Offer a Broader View of Our Genes
Choosing a medicine is not just a matter of finding the right drug for a disease. A dose that suits one patient may be too low or too high for another, and two people taking the same medicine for the same reason may not get the same results. Inherited differences in genes involved in the body’s handling of medicines may help explain some of this variation. In a large study published on November 11, 2026, in Science Translational Medicine, participants from a range of backgrounds gave researchers a broader view of those differences.
The researchers analyzed genetic data from nearly 100,000 Americans taking part in the National Institutes of Health’s All of Us program. They focused on 16 genes whose products play a role in how the body responds to prescription medicines. The most striking finding was that every participant in the study group had at least one variant in one of these 16 genes.
That does not mean everyone will have a problem with their medication, or that a gene variant is itself a diagnosis or a reliable forecast of what will happen to a patient. Rather, the finding shows how common genetic variation is in these genes. It also suggests that a patient’s genetic makeup could be useful when assessing treatment, alongside their health, age, other medicines, and many additional clinical factors.
The study also found that about one in four participants had previously taken a medicine whose effects could be influenced by the genetic differences the researchers examined. The genes are connected to medicines used in a range of fields, including some selective serotonin reuptake inhibitor antidepressants, cholesterol-lowering drugs known as statins, certain cancer treatments, and clopidogrel, a blood-thinning medicine sold under the brand name Plavix.
The effects of genetic variants differ from one gene and medicine to another. Some may be linked to the way the body metabolizes a drug—that is, processes and breaks it down—which could make a standard dose unsuitable for some people. A variant might point to a possible need for a higher or lower dose, a risk of more severe or different side effects, or reduced treatment effectiveness in certain cases. The findings are not a one-size-fits-all prescription, nor a reason to change medication without medical advice. They expand the information available to help explain why patients can respond differently.
Why a More Diverse Study Changes What We Can See
Many past genetic studies relied on groups that did not adequately reflect the diversity of the population. When a database draws on a limited range of backgrounds, variants that are common in some groups may be less visible in the analysis—or may not be detected at all. The All of Us program aims to collect genetic data from one million Americans with diverse backgrounds, giving researchers a chance to study variation across a broader population.
The study reports that this diversity helped researchers find variants that had not stood out in earlier research using other databases, including the UK Biobank. That does not mean any one database can represent everyone in the world, or that comparing databases alone establishes the significance of every variant. The broader point is that expanding who is included in research makes it more likely that researchers will detect genetic patterns missing from narrower studies.
This matters for health beyond the number of variants discovered. If precision-medicine tools are developed using data that represent some groups more fully than others, their predictions may be less suitable for groups that were underrepresented. Including more diverse participants from the outset can reduce the risk of carrying that gap into later medical decisions, although it cannot, by itself, eliminate disparities in care.
Lauren Salisbury, a health policy researcher who has studied equity and pharmacogenomics and was not involved in the study, says the findings add to growing evidence that tailoring care using genetic information may benefit people from diverse backgrounds. This point carries weight because the question is not only whether genetic testing can improve a decision, but also whether its benefits will reach different groups fairly.
From a Gene Variant to a Treatment Decision
The basic idea is that some genes are linked to enzymes or other processes the body uses to handle particular medicines. A version of a gene can change how active these processes are, affecting how quickly a medicine is processed or how it acts. But moving from this biological possibility to a clinical decision takes careful interpretation. Finding a variant alone does not tell us whether a drug will fail or harm a patient, just as symptoms or a past experience do not automatically prove that genes were the cause.
Philip Empey, director of the Center for Clinical Pharmaceutical Sciences at the University of Pittsburgh and the study’s senior author, emphasizes that a poor response to a medicine or an adverse effect cannot be definitively attributed to genes without an assessment. Still, growing knowledge about how people differ—including in their genes—allows researchers and clinicians to ask better questions about why responses vary and to identify situations in which inherited factors may warrant closer attention.
This is where finding a genetic association differs from using it in practice. A doctor needs a result that can be interpreted in the context of the patient, clear guidance on what it means for a particular medicine, and information about which drugs may be affected. Several possible variants can also make interpretation more complicated than relying on a single marker. And a variant may not matter equally for every patient or in every circumstance.
There is another practical obstacle: many patients have never had their genomes sequenced—that is, had a broad reading of their genetic information. Even when genetic information is available, some doctors may not know how it relates to the medicines they prescribe, or the results may not be presented in a way that makes them easy to use when a decision is being made. Publishing a study or developing a laboratory test is therefore not enough. Care also needs systems that can provide testing, explain results, connect them to a specific treatment, and track what happens next.
It is also important not to let the promise of personalized medicine become the assumption that genes explain every experience with a drug. Genetic variation is one possible factor in a wider picture, and it can interact with the disease, other treatments, and a person’s overall health. The study helps improve the map of variants worth examining, but it does not claim that genomic information alone can settle the choice of drug or dose for every individual.
A Large Database Faces a Funding Test
The study drew on the All of Us program, established by the National Institutes of Health with the goal of collecting data from one million Americans with diverse backgrounds. But cuts in funding have affected the program’s ability to keep expanding its participant base and providing its services. The program relies on two main sources of support: the 21st Century Cures Act, passed in 2016, and annual appropriations approved by Congress.
The funding stream tied to the 2016 law is scheduled to end in fiscal year 2027. At the time this report was prepared, the government was operating under a temporary budget known as a continuing resolution, while the program expected to receive $122 million once a final budget was passed. That would be an 80% decrease from the program’s funding peak in 2023. The figures make clear that continuing the project at its previous scale and ambition is not guaranteed.
Joshua Denny, who served as the program’s chief executive until the end of October before moving into a leadership role at the University of Chicago, said reduced resources had forced major cuts to participant recruitment and outreach. These included scaling back enrollment efforts at academic medical centers, federally qualified health centers, and mobile units. Such channels can help reach people who may not have easy access to studies centered only at major institutions or medical centers.
The program also stopped providing genetic test results to participants because of funding concerns. That decision illustrates how tight budgets can affect both data collection and the direct benefits participants receive. All of Us is not just a research effort; its design also involves participants engaging with their own information. Long-term funding is therefore part of the debate about what the project can deliver to science and to individuals.
The effects of these cuts also extend to efforts to recruit children. The program began enrolling children in 2023 and had registered nearly 5,000 of them, but their data were not available to researchers when Sara Van Driest, the program’s director of pediatrics, spoke. In her view, enrolling children on a large scale will require stable—and even increasing—funding in the years ahead, so that a database can be built to help shape the future of precision care for them.
Potential Benefits Depend on Getting the Knowledge Into Clinics
The study points to important possibilities, but it also highlights the chain of steps between a research finding and a patient. Researchers must collect data from diverse groups, analyze genetic differences, establish what those differences mean for the use of specific medicines, and then make testing, interpretation, and guidance available in healthcare settings. If any link in this chain fails, the knowledge may remain in scientific papers without changing a patient’s experience.
Broader representation in studies remains central because prediction tools built on incomplete data may not work equally well for everyone. But diversity alone is not a complete solution. It improves the chances of detecting genetic differences, while reducing disparities also requires access to appropriate testing and care, training for healthcare providers, and information that can be understood and used. Cuts to the All of Us program may limit its ability to complete recruitment and make the benefits of its research more widely available.
Patients should not take the findings as a reason to seek genetic testing or change a medication dose on their own. The study does not prescribe an alternative treatment for each person, nor does it say that any particular variant must explain a side effect or a weak response. Instead, it supports developing a medical approach that takes individual differences seriously and combines genetic information with clinical assessment, rather than assuming that the average response applies equally to everyone.
The Bottom Line
An analysis of nearly 100,000 Americans found that all participants carried at least one variant in one of 16 genes linked to how the body responds to medicines, and that one in four had taken a drug whose effects could be influenced by those variants. The diversity of the participants helped reveal variants that had not stood out in earlier databases, underscoring the value of genetic research that includes a broader range of people. But clinical benefits still depend on testing, doctors’ knowledge, and careful interpretation of multiple variants. Meanwhile, the program behind the data faces steep financial pressure: expected funding of $122 million, 80% below its 2023 peak, alongside challenges in recruitment, sharing results, and including children. Bringing personalized medicine closer to the clinic will require both diverse data and stable funding.