Posted by Katherine J. Wu

https://www.theatlantic.com/health/archive/2025/08/nih-scientifically-justified-research/683913/?utm_source=feed

Since January, President Donald Trump’s administration has been clear about its stance on systemic racism and gender identity: Those concepts—championed by a “woke” mob, backed by Biden cronies—are made-up, irrelevant to the health of Americans, and unworthy of inclusion in research. At the National Institutes of Health, hundreds of research studies on health disparities and transgender health have been abruptly defunded; clinical trials focused on improving women’s health have been forced to halt. Online data repositories that contain gender data have been placed under review. And top agency officials who vocally supported minority representation in research have been ousted from their jobs.

These attacks have often seemed at odds with the administration’s stated goals of fighting censorship in science at the NIH and liberating public health from ideology. But its members behave as though they have no dogma of their own—just a wholehearted devotion to scientific rigor, in the form of what the nation’s leaders have repeatedly called “gold-standard science.” This pretense—that the government can obliterate entire fields of study while standing up for free inquiry—is encapsulated by what’s become a favored bit of MAHA rhetoric: All research is allowed, the administration likes to say, so long as it’s “scientifically justifiable.”

On Friday, the phrase scientifically justified appeared several times in a statement by NIH Director Jay Bhattacharya that set the agenda for his agency and ordered a review of all research to make sure that it fits with the agency’s priorities. “I have advocated for academic freedom throughout my career,” he wrote in a letter to his staff that accompanied the statement. “Scientists must be allowed to pursue their ideas free of censorship or control by others.” But his announcement went on to warn that certain kinds of data, including records of people’s race or ethnicity, may not always be worthy of inclusion in research. Only when its consideration of those factors has been “scientifically justified,” he wrote, would a project qualify for NIH support.

That message may seem unimpeachable—in keeping, even, with the priorities of the world’s largest public funder of biomedical research: NIH-backed studies should be justified in scientific terms. But the demand that Bhattacharya lays out has no formal criteria attached to it. Scientific justifiability is, to borrow Bhattacharya’s description of systemic racism, a “poorly-measured factor.” It’s imprecise at best and, at worst, a subjective appraisal of research that invites political meddling. (Neither the NIH nor the Department of Health and Human Services, which oversees it, responded to my questions about the meaning and usage of this phrase.)

Judging scientific merit has always been one of the NIH’s most essential tasks. Tens of thousands of scientists serve on panels for the agency each year, scouring applications for funding; only the most rigorous projects are selected to receive portions of the agency’s $47 billion budget—most of which goes to research outside the agency itself. All of the thousands of grants the agency has terminated this year under the Trump administration were originally vetted in this way, by subject-matter experts with deep knowledge of the underlying science. Many of the studies have been recast, in letters from the agency, as being “antithetical to the scientific inquiry,” indifferent to “biological realities,” or otherwise scientifically unjustified.

The same language from Bhattacharya’s email appears in other recent NIH documents. Last week, an official at the agency sent me a copy of a draft policy that, if published, would prohibit the collection of all data on people’s gender (as opposed to their sex) by any of the agency’s researchers and grantees, regardless of their field of study. It allows for an exception only when the consideration of gender is “scientifically justified.” The gender-data policy was uploaded to an internal portal typically reserved for agency guidance that is about to be published, but has since been removed. (Its existence was first reported by The Chronicle of Higher Education.) When reached for comment, an HHS official told The Atlantic that the policy had been shot down by NIH leadership, but declined to provide any further details on the timing of that shift, or who, exactly, had been involved in the policy’s drafting or dismissal.

Still, if any version of this policy remains under consideration at the agency, its aims would be in keeping with others that are already in place. One NIH official told me that one of the agency’s 27 institutes and centers, the National Institute for General Medical Sciences, has, since April, sent out hundreds of letters to grantees noting, “If this award involves human subjects research, information regarding study participant ‘gender’ should not be collected. Rather, ‘sex’ should be used for data collection and reporting purposes.” Payments to those researchers, the official said, have been made contingent on the scientists agreeing to those terms within two business days. “Most have accepted,” the official told me, “because they’re desperate.” (The current and former NIH officials who spoke with me for this article did so under the condition of anonymity, to be able to speak freely about how both Trump administrations have affected their work.)

Collecting data on study participants’ gender has been and remains, in many contexts, scientifically justified—at least, if one takes that to mean supported by the existing literature on the topic, Arrianna Planey, a medical geographer at the University of North Carolina, told me. Evidence shows that sex is not binary, that gender is distinct from it, and that acknowledging the distinction improves health research. In its own right, gender can influence—via a mix of physiological, behavioral, and social factors—a person’s vulnerability to conditions and situations as diverse as mental-health issues, sexual violence, cardiovascular disease, infectious diseases, and cancer.

The Trump administration has expressed some interest in gender-focused research—but in a way that isn’t justified by the existing science in the field. In March, NIH officials received a memo noting that HHS had been directed to fund research into “regret and detransition following social transition as well as chemical and surgical mutilation of children and adults.” That framing presupposes the conclusions of such studies and ignores the most pressing knowledge gaps in the field: understanding the long-term outcomes of transition on mental and physical health, and how best to tailor interventions to patients. (Bhattacharya’s Friday statement echoed this stance, specifically encouraging “research that aims to identify and treat the harms these therapies and procedures have potentially caused to minors.”)

According to the draft prohibition on collecting gender data, NIH-employed scientists would be eligible for an exception only when the scientific justification for their work is approved by Matthew Memoli, the agency’s principal deputy director. Memoli has played this role before. After Trump put out his executive order seeking to abolish government spending on DEI, Memoli—then the NIH’s acting director—told his colleagues that the agency’s research into health disparities could continue as long as it was “scientifically justifiable,” two NIH officials told me. Those officials I spoke with could not recall any instances in which NIH staff successfully lobbied for such studies to continue, and within weeks, the agency was cutting off funding from hundreds of research projects, many of them working to understand how and why different populations experience different health outcomes. (Some of those grants have since been reinstated after a federal judge ruled in June that they had been illegally canceled.)

The mixing of politics and scientific justifiability goes back even to Trump’s first term. In 2019, apparently in deference to lobbying from anti-abortion groups, the White House pressured the NIH to restrict research using human fetal tissue—prompting the agency to notify researchers that securing new funds for any projects involving the material would be much more difficult. Human fetal tissue could be used in some cases, “when scientifically justifiable.” But to meet that bar, researchers needed to argue their case in their proposals, then hope their projects passed muster with an ethics advisory board. In the end, that board rejected 13 of the 14 projects it reviewed. “They assembled a committee of people for whom nothing could be scientifically justified,” a former NIH official, who worked in grants at the time of the policy change, told me. “I remember saying at the time, ‘Why can’t they just tell us they want to ban fetal-tissue research? It would be a lot less work.’”

The NIH’s 2019 restriction on human-fetal-tissue research felt calamitous at the time, one NIH official told me. Six years later, it seems rather benign. Even prior to the change in policy, human fetal tissue was used in only a very small proportion of NIH-funded research. But broad restrictions on gathering gender data, or conducting studies that take race or ethnicity into account, could upend most research that collects information on people—amounting to a kind of health censorship of the sort that Bhattacharya has promised to purge.

The insistence that “scientifically justifiable” research will be allowed to continue feels especially unconvincing in 2025, coming from an administration that has so often and aggressively been at odds with conventional appraisals of scientific merit. Robert F. Kennedy Jr., the head of HHS, has been particularly prone to leaning on controversial, biased, and poorly conducted studies, highlighting only the results that support his notions of the truth, while ignoring or distorting others. During his confirmation hearing, he cited a deeply flawed study from a journal at the margins of the scientific literature as proof that vaccines cause autism (they don’t); in June, he called Alzheimer’s a kind of diabetes (it’s not); this month, he and his team justified cutting half a billion dollars from mRNA-vaccine research by insisting that the shots are more harmful than helpful (they’re not), even though many of the studies they cited to back their claims directly contradicted them. Kennedy, it seems, “can’t scientifically justify any of his positions,” Jake Scott, an infectious-disease physician at Stanford, who has analyzed Kennedy’s references to studies, told me.

Bhattacharya’s call for a full review of NIH research and training is predicated on an impossible, and ironic, standard. Scientists are being asked to prove the need for demographic variables that long ago justified their place in research—by an administration that has yet to show it could ever do the same.

https://www.theatlantic.com/health/archive/2025/08/nih-scientifically-justified-research/683913/?utm_source=feed