1. Introduction
Precision medicine has a racial health disparities problem. The precision medicine approach to health care is characterized as incorporating information about patients’ genes and environment to replace “one size fits all” medicine with “the right treatment, for the right patient, at the right time,” and aggressive efforts are underway to recruit communities of color into participating in precision medicine research (Lee et al. Reference Lee, Fullerton and McMahon2022; Martschenko and Young Reference Martschenko and Young2022). The reason to participate, the proponents argue, is because doing so will help combat the unjust health inequities that have plagued communities of color for centuries (Cooke Bailey et al. Reference Cooke Bailey, Bush and Crawford2020; Mapes et al. Reference Mapes, Foster and Kusnoor2020).
This precision medicine approach, however, has been around for a quarter century now, and the economics of it are daunting (Parker Reference Parker and Bydon2024). Precision medicines are among the most expensive health interventions on the planet (Green et al. Reference Green, Prainsack and Sabatello2023). Gene therapies for rare diseases that cost millions of dollars. Targeted therapies for cancers that average more than $150,000/year (Fleck Reference Fleck2022). These prices have led to growing alarm among clinicians about financial toxicity for their patients—heartbreaking situations in which an effective treatment may be available but accessing it will bring financial ruin (Zafar and Abernathy Reference Yousuf and Abernathy2013; Ramsey et al. Reference Ramsey, Bansal and Fedorenko2016; Gilligan et al. Reference Gilligan, Alberts and Roe2018).
The tension here is that communities of color often struggle to access basic health care, so it’s hard to see how handing their health data over to precision medicine research projects will translate into addressing health disparities if the outcome will be exorbitantly expensive treatments. Consider, as just one example of this tension, the case of sickle cell, a blood disease that tends to affect more Black people than White and is a major priority for efforts aimed at combating health disparities in the United States. Proponents of precision medicine hailed the arrival of Casgevy and Lyfgenia—gene therapies for sickle cell disease—when they were approved by the US Food and Drug Administration (FDA) in 2023. But Casgevy was priced at $2.2 million upon its release, and Lyfgenia topped the scales at $3.1 million (Sheridan Reference Sheridan2023; Rueda et al. Reference Rueda, Beriain and Montoliu2024; Scheffer and Tessema Reference Scheffer and Tessema2024). Patients with sickle cell often struggle to reliably access hydroxyurea, an effective treatment that was approved by the FDA in 1998 and costs about $50/month (Treadwell et al. Reference Treadwell, Du and Bhasin2022). Precision medicine’s health disparities problem in a nutshell is this: If a standard therapy for sickle cell can be hard to get at $50/month, why should communities of color put their health data in service of precision medicine research projects that deliver health interventions with price tags of two to three million dollars?
In this article I’ll consider a solution that has been proposed, ultimately concluding that the proposal will not resolve precision medicine’s health disparities problem and, in fact, will more likely make efforts to combat health disparities worse off. Section 2 introduces the proposal, one that involves reorienting precision medicine away from genetics and toward a focus on social and environmental determinants of health. In section 3 I provide a brief tour of the history and contemporary landscape of precision medicine, revealing why this reorientation is so unlikely to succeed; this is because precision medicine always has been a genomic enterprise, remains fundamentally genomic in practice, and shows no signs of pivoting away from this genomic essence. That’s not to say that attention to something like social and environmental determinants of health can’t get incorporated. But, as I’ll show in section 4, that incorporation works in a very specific way in precision medicine. Precision medicine has been genomic medicine for so long that conducting precision medicine research now means doing science like geneticists; it means morphing a research practice into the mold of genomics, a process I call “methodological genomification.”Footnote 1 The methodological genomification of health disparities research is not something that we should welcome, and I use a recent example involving precision medicine research on type 2 diabetes mellitus to illustrate why. If what we care about is addressing unjust health disparities, I conclude in section 5, there are better routes for research—ones that don’t require forcing health disparities researchers to do science like the geneticists.
2. Social and environmental determinants of health to the rescue?
Precision medicine, as mentioned, standardly gets described as tailoring health care by attending to the unique genes and environment of each patient. Recently, a number of philosophers, legal scholars, and clinical researchers have suggested the way around precision medicine’s health disparities problem is to pivot attention from the genetics side of precision medicine to the environmental side. The really expensive precision medicines—gene therapies, pharmacogenetic pharmaceuticals, targeted cancer treatments—have derived from molecular-genetic research. So, the thought goes, why not try directing precision medicine research toward the social and environmental determinants of health?
Dayna Bowen Matthew, in her reflections on the “threats to precision medicine equity” was an early advocate of this move, writing,
[Precision medicine] researchers should follow the example set by historical epidemiologists, sociologists, philosophers, and hosts of physicians who have reasoned that racial disparities in health status and outcomes derive from interplay between complex social forces such as economic disadvantage, institutional and interpersonal discrimination, structural barriers to healthy life choices; unequal access to healthy food, education, work and housing environments; disparate access to water and sanitation; and, indeed, non-genetic transmission of biological differences among the races. (Matthew Reference Matthew2019, 635)
Anna Jabloner and Alexis Walker offer similar guidance in their warnings about the “pitfalls of genomic data diversity”:
But if ensuring equitable benefit, or even addressing health inequality, is really the goal, the analytic insights from the social, human, and cultural sciences— precisely those identifying capitalist biomedicine, the enduring effects of history, and the structural impacts of oppression even in the absence of direct interpersonal discrimination—are essential to designing successful interventions. Bridging the gulf between social and medical scientists would mean that the analytic contributions of the social and human sciences in understanding and overcoming inequality are not skirted in efforts to address racial health inequality and are instead brought to the core of the scientific effort. (Jabloner and Walker Reference Jabloner and Walker2023, 12)
And Ilaria Galasso echoes this sentiment, counseling,
A critical research focus on the social and environmental determinants of health, if translated into sociopolitical interventions, would be to the direct benefit of those most exposed to those determinants and most at risk of (upstream as well as downstream) exclusion, to the general benefit of precision medicine inclusivity and ultimately of health—and precision medicine—equity. (Galasso Reference Galasso2024, 83; emphasis in original)
I am sympathetic to the intuition these authors share.Footnote 2 Decades of research have shown clearly that the major causes of health disparities are to be found in the social, built, physical, and chemical environments that different racial groups have been and continue to be forced to navigate, not in the genetic differences between those groups (Valles Reference Valles2021). So if you want to do something about racial health disparities, then it makes sense to focus attention on the environmental domain where we have very good reason to think the causes of the problem operate. What’s more, these environmental factors of health more often fall under the purview of public health, where addressing the health threats are the responsibility of governmental collective action; whereas the genetic products of precision medicine have traditionally been generated by the private sector, where corporations motivated by profit dictate access with the prices that they set (Valles Reference Valles2018). So if you’re worried about cost barriers contributing to health disparities, then it again makes sense to shift attention toward research insights that are more likely to be absorbed by communities and less likely to fall onto the shoulders of individual payors.
I’m sympathetic to the intuition, but I’m also skeptical that it can be realized. That’s because, to pull it off, precision medicine would need to be radically overhauled in terms of how the science is practiced, and there isn’t good reason to expect that this will happen. The more likely outcome is that it will be the health disparities research that gets radically overhauled, and not for the better. To appreciate that reality, some familiarity with how precision medicine became what it is today is in order.
3. Precision medicine is genomic medicine
Precision medicine routinely gets packaged as if it attends to the genetic and environmental contributions to health and illness, but when you look closer at how precision medicine operates in practice, it becomes clear that this packaging is misleading. “Precision medicine” is just marketing lingo for genomic medicine (Tabery Reference Tabery2024). This can be seen both historically, in terms of how precision medicine took its current shape, and contemporarily, in terms of how it exists out in the world today.
Historically, the story of precision medicine began in pharmacogenetics (Tabery Reference Tabery2023; Allen Reference Allen2025). Medical geneticists, in the 1990s, were trying to sell the benefits of uniting genetics and pharmacology to drug companies, investors, and the public, arguing that incorporating genetic information about patients and research participants into the drug discovery, development, and prescribing processes would lead to more drug approvals (good for drug companies) and fewer adverse drug reactions (good for patients). This genomic revolution in health care, the proponents claimed, would swap out the one-pill-fits-all approach to pharmacology with one that delivered the right drug, to the right patient, and at the right time (Marshall Reference Marshall1997a, Reference Marshall1998). What’s more, they promised that genetics would allow for personalizing medicine based on each patient’s unique genome (Marshall Reference Marshall1997b). “Personalized medicine,” in turn, became the branding of the genomic revolution throughout the first decade of the twenty-first century (Hedgecoe Reference Hedgecoe2004; Fuller Reference Fuller2017; Prainsack Reference Prainsack2017).
Eventually, though, many in the medical genetics community became worried that the language of personalization misled how the science and clinical practice worked because genetics didn’t in fact generate individualized treatments; it just grouped patient populations based on their shared genomic profiles (Prainsack Reference Prainsack2014; Juengst et al. Reference Juengst, McGowan and Fishman2016). An effort at rebranding occurred around 2010, where the new idea was that genetics made medicine more precise (Christensen et al. Reference Christensen, Grossman and Hwang2009; Committee on a Framework for Development of a New Taxonomy of Disease 2011). “Precision medicine” thus became the second packaging for the purported genomic revolution in health care (Plutynski Reference Plutynski, Beneduce and Bertolaso2022; Chin-Yee Reference Chin-Yee, Schramme and Walker2024). All the catchphrases from pharmacogenetics—one pill fits all; right drug, for the right patient, at the right time—came along for the ride, but the explicitly pharmacological terms like “pill” and “drug” were swapped out for more generic terms, resulting in the new notion that “one size fits all” medicine would be replaced with the “right treatment, for the right patient, at the right time.” The genomic essence of it all, however, never disappeared. In short, when we look backward, the history shows us that “precision medicine” was just an intentional rebranding of “personalized medicine,” which was just an earlier effort to make genomic medicine more attractive to nongeneticists.
That’s what the past has to say. But what about the present? Does precision medicine look less like basic marketing lingo for genomic medicine when we survey the current scientific and clinical landscapes? No. In fact, it affirms this reality. Go look around yourself. Check out a book on precision medicine from your university library (McCarthy and Mendelsohn Reference McCarthy and Mendelsohn2017). Or find an informational video designed to explain what precision medicine is (Trianni and Tsai Reference Trianni and Tsai2015). Or visit a hospital where precision medicine services are offered (UCLA Health 2023). Or sign up to participate in a precision medicine research project (National Institutes of Health 2025). Here’s what you’ll find: primers on whole-genome sequencing, patient testimonials about their experiences with genetic testing for cancers, histories of the Human Genome Project, covers and logos infused with DNA’s iconic double-helix. Now, these books, videos, clinical services, and research endeavors routinely start off by saying something about how precision medicine attends to the genetic and environmental contributions to health and disease. But then the actual examples, success stories, and services are genetic. The present thus delivers the same verdict as the past: Precision medicine is genomic medicine. This should make us leery of the thought that the science can be reformed in such a way that the social and environmental determinants of health can take center stage in that research, despite the lip service that the environment gets.
4. The methodological genomification of health disparities research
I’ve argued so far that precision medicine always has been and remains a genomic endeavor. But proponents of the proposal to shift precision medicine toward the social and environmental determinants of health could reply: Sure, that may be how things have gone, but our proposal is for a future where precision medicine looks different.
As it turns out, we don’t need to imagine what this future might look like; we can take a look at a place where it’s supposedly starting to unfold. That’s because the proponents of the proposal have pointed to a precision medicine initiative where they see it being realized: The US National Institutes of Health’s (NIH) All of Us Research Program (“All of Us”) (National Institutes of Health 2025). All of Us was announced by President Barack Obama in his 2015 State of the Union Address, and it publicly launched across the United States in 2018 (National Institutes of Health 2018a). It is advertised by the NIH as an effort on the part of the US federal government to recruit 1,000,000 diverse Americans into a longitudinal study that collects information about their genetics and environment; the stated goal is to use this diverse data to prevent health disparities by bringing precision medicine to the masses.
Matthew, who I mentioned earlier as one of the first to call for the move toward focusing precision medicine research on social and environmental determinants of health, writes approvingly of All of Us: “Efforts to address these disparities are hopeful. For example, the National Institutes of Health (NIH) has launched the ‘All of Us’ program to collect health, genomic, and behavior data from one million Americans and to specifically oversample and build trust with communities that have historically been underrepresented in research” (Matthew Reference Matthew2019, 637). And Galasso, as another instance of this embrace, praises,
By meaningfully including social determinants of health in the research—as All of Us does through the sociodemographic surveys it distributes to participants, including a recent survey specifically on the social determinants of health—precision medicine can catalyze knowledge and attention to inform and incentivize socio-political reforms and interventions to reduce social inequalities and inequities. (Galasso Reference Galasso2024, 82)
When we look beyond the All of Us advertising, however, and examine what this precision medicine research looks like in practice, it becomes less clear how “meaningful” the incorporation of social and environmental determinants of health is. Take the case of type 2 diabetes mellitus. In 2024, the NIH announced that it was making a major investment in studying the environmental contributions to diabetes with All of Us, focusing on diabetes in part because of the tremendous health disparities that exist for that chronic disease (National Institutes of Health 2024).
I’ll reveal in a moment how All of Us is planning to conduct the diabetes research. But before I do that, it will help to start by considering an example of environmental research focused on diabetes that has nothing to do with precision medicine. This will provide a baseline to see how scientists who standardly study the environmental contributions to health disparities in diabetes conduct this research, so that we can then compare that baseline to what it looks like when All of Us does the precision medicine version. Diabetes Care, in 2023, published an article that served as a review of research on how the neighborhood environment contributes to health disparities in diabetes (Mujahid et al. Reference Mujahid, Sai Ramya Maddali, Oo, Benjamin and Lewis2023). Note, to begin, that the article was just focused on the neighborhood environment; not the school environment, home environment, or work environment. Just the neighborhood. The authors detailed the great variety of data about neighborhoods that are collected and studied by the environmental health disparities research community: data about greenspaces (e.g., tree canopy cover, parks, green satisfaction among residents); walkability (e.g., street conditions, sidewalk accessibility, land use, population density, traffic patterns); food options and prevalence (e.g., grocery stores, convenient stores, fast food restaurants); the history of the geographic space (e.g., redlining, segregation); proximity to toxic pollutants (e.g., landfills, highways, factories), about housing (e.g., different types, affordability, security); and transportation options, senses of social cohesion, noise levels, and community aesthetics. Getting all that data isn’t easy for environmental health researchers. It takes time; a number of different scientific specialists guiding the data collection; collecting data over and over again; building trust with the people in the neighborhoods; and money. But the motivation for all that investment is that it’s necessary because collecting data on all those varied factors is an essential part of seriously investigating the contributions of the neighborhood environment to diabetes. Type 2 diabetes mellitus is a chronic disease that results from the intersection of many different features of life—nutrition, activity, stress, and sleep, to name a few. To understand how the environment contributes to diabetes outcomes, the researchers need to get out there in the world and study the environment in all its ever-changing complexity so as to investigate how different groups are distributed in different neighborhoods with different exposures and different resources that make diabetes easier or harder to avoid.
Now back to All of Us. Here are the data sources that All of Us will be using to assess the environmental contributions to diabetes (and note that these are the sum total of all environmental contributions to diabetes, not just the neighborhood environment): (1) blood collected from participants when they first signed up to join the study, (2) several surveys that the participants completed on a phone app, and (3) remote access to the participants’ electronic health records (National Institutes of Health 2024). That’s it. Why would the NIH think that they’re studying the environment without stepping foot in the environment? The answer is that this is what it looks like when health disparities research gets methodologically genomified.
Precision medicine has been genomic medicine for so long that, to do precision medicine research today, a researcher needs to do science like geneticists. Doing science like geneticists came to mean something very specific over the last two decades. Coming out of the Human Genome Project in 2003, geneticists were hopeful that they’d find major genes for all sorts of health traits: diabetes, cardiovascular disease, dementia, asthma, allergies, depression, you name it (Collins Reference Collins2003). But that’s not what they discovered. Instead, they learned that common, complex traits are subtly influenced by thousands of variants in the human genome, each making a miniscule contribution to the risk of developing the traits (Maher Reference Maher2008; Schaffner Reference Schaffner2016; Matthews and Turkheimer Reference Matthews and Turkheimer2022). That genomic reality forced the geneticists to alter how they conducted their research. To keep searching for smaller and smaller genetic effects, the scientific effort shifted toward recruiting bigger and bigger research cohorts—hundreds of participants, then thousands, then hundreds-of-thousands, now millions (Vogt et al. Reference Vogt, Green and Ekstrom2019; Abdellaoui et al. Reference Abdellaoui, Yengo and Verweij2023). That expansion in research participants was possible in part because, at the same time, the technological costs of genotyping and sequencing genetic data plummeted dramatically too, and so—with just a bit of blood collected once from participants—geneticists could quickly, easily, and cheaply get enormous amounts of genetic data from enormous numbers of people, and then go looking for those miniscule genetic effects (National Human Genome Research Institute 2023).
This methodological approach to doing science (e.g., the data collected, the way it is collected, the assumptions about how causes operate, the priorities placed on sample size) can be seen in precision medicine research projects, and All of Us is a perfect example. Again, it advertises itself as if it was designed to collect health data about genes and environments from research participants. But the project was conceived in the NIH’s National Human Genome Research Institute by the head of the Human Genome Project (Tabery Reference Tabery2023). A Genomics Working Group was convened in 2017 to guide decisions about how to design All of Us and what data to collect; no environmental working group was convened to provide that formative guidance (Genomics Working Group 2017). The goal of recruiting 1,000,000 participants was explicitly chosen to facilitate the search for extremely small genetic effects using just a bit of blood from those who signed up, with no consideration of how that might shape environmental research (The All of Us Research Program Investigators 2019). A series of Genome Centers were funded to sequence the All of Us participants’ DNA; no environmental center was created in anticipation of the program’s rollout (National Institutes of Health 2018b).
When All of Us says that it will study the social and environmental determinants of health for diabetes, then, it means that the health disparities research needs to follow the methodological path set in place by the geneticists. The health outcomes are measured using the exact same source that the geneticists use—the electronic health records, where the assumption is that any diabetes outcomes worth tracking will appear in the participants’ medical charts. The source of biological data that is analyzed is the exact same source that the geneticists analyze—the blood, where the assumption is that blood collected once from participants will carry markers of any environmental exposure the participants encountered in their life. The only new data source is the handful of short surveys that participants are asked to fill out about things like alcohol use, health care access, and social life.
All of Us, according to Matthew, should make us hopeful about precision medicine’s ability to pivot toward studying the social and environmental determinants of health and their impact on health disparities. This closer look at how that All of Us research is being implemented, however, should sound an alarm about the dangers of methodologically genomifying that research. Direct measurements of the environment are replaced with biosamples meant to serve as distant proxies for the environment. Repeated collections of dynamic environmental data are replaced with static data sources collected only once. Careful, research-intensive investigations of specific populations in specific places at specific times are replaced with low-cost queries of 1,000,000 people disparately spread out across the country.
There are multiple reasons why we should be wary of this methodological genomification. First, there’s a risk of what C. Thi Nguyen calls “value capture” (Nguyen Reference Nguyen2024). Value capture occurs when an individual or community adopts externally sourced values as their own, regardless of whether they work well in the new domain. It’s particularly pernicious in scenarios in which some new device or metric seemingly allows for quick and easy measurements of complex phenomena, such as when fitness and activity endeavors get replaced by FitBit step counts. In environmental health research, this has taken the form of increasing attention to the “exposome”—an explicit attempt to reimagine the sum total of all environmental exposures that a person experiences over the course of their life as if it is a single thing (like a genome) that should be as cheap, easy, and quick to measure as a genome is to sequence (Wild Reference Wild2012).
Second, methodological genomification gives the illusion of studying the social and environmental determinants of health without conducting the research in a meaningful way. The fact that All of Us has already fooled several scholars who are calling for greater investment into studying the social and environmental determinants of health should be evidence enough that this illusion can be quite convincing. It inclines advocates of studying health disparities to think that precision medicine can be reformed so as to prioritize attending to the social and environmental determinants of health when in fact that research is being methodologically genomified and morphed into something that barely resembles time-tested approaches to doing that science.
5. Conclusion
Precision medicine is genomic medicine. In fact, it has been genomic medicine for so long that, even when environmental health research is incorporated into precision medicine, it must be methodologically genomified. That means we should be wary of the suggestion that precision medicine can be reformed in such a way that it can effectively combat racial health disparities by reorienting around the social and environmental determinants of health.
That’s not to say that we should abandon attention to the social and environmental determinants of health. Quite the contrary. If what we care about is addressing racial health disparities, then that’s exactly where we should be investing research resources. But that doesn’t mean giving into the illusion that precision medicine will do this. It means investing in the types of research that social and environmental health scientists have long known interrogates the causes of health and health disparities, uninhibited by any pressures of methodological genomification.
Acknowledgments
I am grateful to Benjamin Chin-Yee and Zinhle Mncube for organizing the “Precision Medicine and (In)Justice” symposium at PSA 2024, as well as the coparticipants (Jonathan Fuller, Sara Green, Anya Plutynski, and Olivia Spalletta) and audience members for constructive feedback. I also benefited from the input of Chris Cosans, Lindley Darden, Sandra Herbert, Eric Juengst, and Kenneth Schaffner.
Funding statement
None to declare.
Declarations
None to declare.
Open access
