New NIH Policy on Animal Models in Research: What Does It Mean for the Migraine Field?

The National Institutes of Health has announced a plan to move away from animal models towards human-based approaches instead. Scientists who spoke to Migraine Science Collaborative say the new policy could slow the development of new headache and pain treatments, though ultimate impacts remain unclear.

In April 2025, just weeks after the U.S. Food and Drug Administration (FDA) declared it would reduce animal testing in the development of monoclonal antibodies and other drugs, the National Institutes of Health (NIH) announced it would follow suit by prioritizing the funding of studies that embrace “innovative, human-based science.”

NIH Director Jay Bhattacharya called the move “a new era in innovation” that would “fundamentally reimagine the way research is conducted” as NIH prioritizes funding for emerging human-based technologies like computational models, induced pluripotent stem cells (iPSCs), and organoid systems over traditional animal models. The NIH stated that this change was necessary because of the difficulties the biomedical research field has faced in translating findings from animal models into better understanding and treatment of human disease.

Initial responses to the announcement from press and research organizations were somewhat circumspect. But a July 10, 2025, statement explaining that NIH would “no longer issue NOFOs [Notices of Funding Opportunities] exclusively supporting animal models or limit/specify the types of models that must be used,” and may issue notices that “exclude proposals for animal use entirely,” caused more consternation (see here for example), said Andrew Russo, a migraine researcher at the University of Iowa, Iowa City, US.

“The idea of emphasizing human-relevant research and looking at human tissues to validate what genes are expressed or what kind of cells may be involved with different phenomena can absolutely help us make advances. We’ve been moving in that direction in the migraine and pain fields for some time,” he said. “Yet, it’s unclear what the new NIH policy actually means for animal work, and, frankly, it’s naïve and simplistic to believe we can eliminate all animal research without significant consequences.”

Gregory Dussor, a migraine researcher at the University of Texas at Dallas, US, agreed. He said that animal models have been integral to the development of novel migraine therapeutics.

“This is the kind of research that leads to fundamental observations that create the pathways which ultimately lead to new drugs and treatments,” he said. “We’ve seen a lot of new drugs in migraine over the past decade because of animal work. And while the models in migraine today are fairly primitive, we are missing an opportunity for a deeper understanding of the development and progression of migraine if we do not continue to challenge, improve, and optimize those models to make them more predictive and translationally valuable.”

How to show causation? How to know if a drug will work?
Currently, as Dussor shared in a previous MSC news story, there are only about 10 laboratories that focus solely on animal research in migraine. Yet Jeffrey Mogil, McGill University, Montreal, Canada, a basic science researcher who studies pain, said animal models allow researchers to do key behavioral experiments.

Today, pain scientists rely on animal models to assess behaviors that act as proxies for pain, such as mechanical or thermal hypersensitivity. Migraine researchers, too, use animal models (see review) to assess pain-like behaviors relevant to headache, such as cephalic allodynia, cutaneous allodynia, or mechanical allodynia. But the “human-based” technologies that the NIH is promoting can’t provide confirmation that a specific aspect of biology is actually responsible for the pain, or its alleviation.

“There are certain questions that simply can’t be answered without an animal model,” Mogil said. “These behavioral experiments are the last step in showing that this gene or that protein is causally responsible for pain. Without them, we can’t show causation, only correlation.”

While some researchers, including those who work for pharmaceutical companies, are looking at data science and artificial intelligence (AI) models to help them develop new treatments, Russo said those approaches, too, are limited in the kind of insights they can offer.

“AI models rely on data we’ve already generated from past research studies,” he explained. “So, it’s unclear what new information they can really give us or how to determine if what they come up with is worth pursuing.”

He added that, while there has been exciting work using iPSCs to create nociceptors from tissue donated by people who have rare forms of migraine like hemiplegic migraine, it’s unclear how translatable those results would be to common migraine or other headache conditions. And, equally as important, it is difficult to know whether these models will help researchers understand whether a particular compound would effectively treat migraine, which could delay the development of new therapeutics.

“An organoid can’t help us answer the question of whether a drug affects migraine. Physiology is not just a single organ, and migraine affects more than just one bodily system,” Russo said. “We need animal models, as well as other kinds of studies, to do the kind of cross-validation across different types of experiments to make sure that what we do ultimately translates in headache. And if we can’t do animal work, I think it’s going to really slow down the development of future treatments for migraine.”

Losing a generation of scientists
Beyond hindering the already painstaking journey to develop new medications, the NIH’s decision, as well as recent cuts to funding, may deter young investigators from pursuing careers in biomedical research altogether.

“This isn’t specific to migraine because animal models are used in all diseases,” said Dussor. “But depending on how this decision ends up being implemented, I think we’ll see people decide to work on different things. Maybe promising young scientists go the clinical route and become physicians. Maybe they still end up as researchers, but in physics or geology. It’s hard to know.”

Russo said he is afraid biomedical science will lose the next generation of scientists – especially as younger students and postdoctoral fellows question how to find a foothold in a field with limited or confusing funding opportunities.

“Students are worried and rightfully so,” he said. “The NIH was an operation that was working really, really well. But if these younger scientists can’t get funding, or if they are limited in the ways they can look at the kind of questions they want to answer, I wouldn’t blame them for looking at other opportunities.”

Let the flowers bloom
While the researchers who spoke to Migraine Science Collaborative said they believe the new policy change is short-sighted, they agreed that the NIH hasn’t clarified how it will work in practice.

The NIH released follow-up guidance explaining it is not outlawing animal models per se, but prioritizing human-based technologies and models. Grant reviewers will have the authority to determine the merit of proposed approaches and whether those methods are suited to the research topic in question. Yet it remains unclear who will have the final say on approvals for studies that involve animal models.

When Migraine Science Collaborative reached out to the NIH for more details about how their new human-based priorities might affect migraine research, they replied with the following statement:

“In July 2025, NIH announced it will no longer develop new funding opportunities focused exclusively on animal models of human disease. Rather, going forward, new funding opportunities will be designed more broadly with language that also encourages [that] various approaches be considered. This means researchers may choose any model they deem appropriate – including a combination of approaches – to answer a research question when submitting applications seeking NIH support. This strategy is intended to open the possibilities of which types of models can be submitted in response to funding opportunities, not be restrictive or prescriptive.”

While this may be somewhat of a relief to researchers concerned about a blanket prohibition against animal models, Dussor said there is still a lot of confusion about how grant reviewers will consider future grants that involve some aspect of animal research.

“Many reviewers understand the value of animal work,” he said. “How will they now evaluate a grant without important preliminary data from animal models that supports a particular idea? How do they determine whether an approach is valid when it is missing critical steps that we’ve used for many, many years? It’s going to make reviewing more difficult, and, frankly, even with what the NIH has put out, it’s still not clear what reviewers should be looking for. I think it’s going to take a while before the dust settles on what exactly all this means and how it will affect future research projects.”

After receiving the statement from NIH, Migraine Science Collaborative followed up with the agency, asking whether animal research would still be supported even if only within the context of a larger research project that includes other models. However, NIH did not respond to that question.

Mogil, for his part, said he understands NIH’s desire to encourage scientists to adopt new and exciting technologies. But he worries that, with these policy decisions, the agency is “throwing the baby out with the bathwater.”

“Every successful treatment out there started with an animal model,” he said. “Science works best when you let a thousand flowers bloom. Everyone can do the kind of work they want to do. They can put their efforts into whatever they believe is the smartest approach to answer the question they are trying to answer. And those decisions are best made from the bottom up, from the researchers doing the work, not from a policy that will likely get in the way of the very progress the NIH claims they want to see in the future.”

Kayt Sukel is a freelance writer based outside of Houston, Texas.

Image credit: Lydia Polimeni, National Institutes of Health.