Drug development is among the slowest, most failure-prone processes in modern science, with . Today, artificial intelligence methods have accelerated the first step ā€” plucking promising molecules out of endless possibilities ā€” but countless challenges remain. A successful drug must be not only safe and effective, but also able to bypass the bodyā€™s defenses and reach the right target.

Most drug candidates fail such optimizations. Thatā€™s where a new research institute at the ĢģĆĄÓ°ŹÓ“«Ć½ has focused its attention. Housed in the UW School of Pharmacy, the brings together experts in artificial intelligence, drug discovery, pharmacology, data science and biotechnology to ease the bottleneck between promising molecules and successful drugs.Ģż

The Institute opened in July 2026 and is led by three UW faculty members: , an associate professor of medicinal chemistry who oversees the Instituteā€™s AI-enabled molecular design; , the Milo Gibaldi Chair of Pharmaceutics and expert in drug metabolism and disposition; and , a professor of psychiatry and behavioral science in the UW School of Medicine, who leads drug discovery, translation and commercialization efforts.

UW News spoke with the three co-directors about why drug candidates fail, how AI is speeding drug development and how I2D3 hopes to help get drugs to market more quickly.

What separates a promising molecule from a full-fledged drug? What properties need to be considered, and how can a developer work toward them?

Gaurav Bhardwaj: It really depends on the disease indication you are targeting and the therapeutic modality. Letā€™s say you have a promising molecule that interacts with the disease-causing protein. Delivery becomes equally important ā€” do we need an orally delivered drug? Do we need to cross the blood-brain barrier? If the disease requires daily dosing, then injectable or IV methods arenā€™t optimal. If itā€™s delivered orally, then the molecule needs to be able to get across the gut barrier, and also needs to be stable enough that it doesnā€™t get chewed up by the body. It also needs to stay in the body for a reasonable time. A successful drug molecule has to meet all these and more criteria, and ultimately all these criteria are encoded by the sequence and structure of the molecule.

The Institute is devoted to aspects of drug development that are often overlooked. What problem do you see the Institute being able to help solve?Ģż

GB: Traditional drug discovery and development is a trial-and-error-based process. Either you find a useful molecule in nature and spend years optimizing it for human use, or you create many random combinations of molecules and hope that one of them has the function you need. Both of these approaches are highly unsuccessful, which has created a bottleneck.

Now the field is also focusing on an idea called rational drug design. It started long before AI but is now becoming even more common. People are using AI methods to design new molecules. However, a lot of that work has focused on the first step ā€” finding a molecule that binds to a specific protein, or has a specific function in the body. Thatā€™s still not a drug, itā€™s just more candidates.

The bottleneck has now shifted. Itā€™s no longer finding that first molecule, but now, how do you add all the other drug-like properties? Thatā€™s what the Institute is trying to do. Letā€™s build the models that ultimately make molecules that are going to be successful all the way through the drug development pipeline.Ģż

Marco Pravetoni: I see our work also as accelerating discovery. I work on substance use disorders, and my lab develops vaccines, antibodies and next-generation antibody-like molecules that target drugs in the body. With these new tools, instead of working to design 10 antibody candidates in a lab, we could design 1,000 or more, and then we can accumulate enough data to reduce any risks, so that what we bring to clinical trials is more likely to be successful. AI can do a lot of that.

How can you make it more likely that a drug candidate succeeds in trials?Ģż

Nina Isoherranen: Part of it is predicting whatā€™s going to happen to a drug in humans before itā€™s ever given to humans. That should increase the success rate and eliminate the waste of doing a lot of unsuccessful trials.Ģż

We can also build machine learning and AI approaches to predict drug disposition in an individual person. What we talk about today are ā€˜digital twins,ā€™ which refers to a computational model of the individual patient and their characteristics. For example, how does your kidney function? What is your body mass index? And so forth. Then we generate a digital version of you. We can then predict how a certain drug would behave in your body and build the best strategy.Ģż

Thereā€™s also an access-to-treatment question here. Pregnancy is a great example ā€” we often donā€™t know how drugs work in pregnant women because weā€™ve never done trials. To be safe, we say that pregnant people shouldnā€™t take those drugs, but that means they donā€™t have access to a potentially hugely beneficial medication. If we can use AI and machine learning to predict how pregnant people respond to medications and how their bodies handle drugs differently from nonpregnant people we can make more medications accessible

Now with AI and machine learning, I think we can get to a place where we can truly sample the full space of possibilities.Ģż

How can the methods youā€™re building help with these individualized treatments?Ģż

NI: We know that drugs behave differently in different people. Even if we give them the exact same drugs and concentrations, people may still have different responses because of factors inherent to our bodies.

During drug development the candidate drug needs to be studied to see responses in different populations. Before you get a drug approved, you need to understand how liver disease, for example, is going to change exposure to that drug and whether you need to change the dosing. Thereā€™s a lot of guidance on drug interactions. Pharmacists manage drug interactions all the time, but it gets very complicated when you combine multiple patient factors. Now, if we have good predictive tools, we can predict whatā€™s going to happen without having to do trials.Ģż

The ultimate goal here is to be able to predict, using model computational tools, whatā€™s going to happen in individual humans before you ever give them a drug. Whatā€™s the right dose? The right timing?Ģż

UW has established itself as a leader in these fields already. Iā€™m thinking especially of the UW Medicine , whose director, , recently won the Nobel Prize in Chemistry. How does I2D3 fit into the broader UW ecosystem?Ģż

MP: IPD is a world leader in designing novel proteins, and the UW also has outstanding capabilities in clinical testing and implementation through the . However, there remains a critical translational space between discovery and clinical application ā€”Ģż one that focuses on the pharmaceutical development needed to turn promising innovations into viable therapeutic products. Thatā€™s where I2D3 can play a leading role.

For example, when researchers at IPD develop a new protein, I2D3 can partner with them early to address formulation, manufacturability, stability, delivery, and other key pharmaceutical considerations that are essential for advancing a discovery toward the clinic and ultimately the marketplace. I2D3 would serve as a core translational partner, helping bridge the gap between innovation and implementation.

IPD brings unmatched strengths in protein design, ITHS provides expertise in clinical translation, and I2D3 contributes the drug development and pharmaceutical sciences capabilities needed to move discoveries across the translational continuum. Together, these organizations can create a powerful and highly integrated ecosystem.

For more information, visit . To reach the researchers, contact Alden Woods at acwoods@uw.edu.

Over the past five years, the national conversation around health and nutrition has become ā€™ Most prominently, Health and Human Services Secretary Robert F. Kennedy Jr.ā€™s blames these foods for a host of chronic health issues and has to remove some UPFs from the food supply.Ģż

But thereā€™s a glaring problem: Nobody can agree on how, exactly, to define ultra-processed foods. The lack of a clear definition has stymied legislative and regulatory efforts to curb UPF consumption, and caused confusion for people evaluating their own diets.Ģż

This spring, a panel of 14 nutrition, food science, policy and legal experts gathered to create a more practical and operational definition. The panelā€™s final report, , suggests an ingredient-based approach to identify ultra-processed foods, while also recommending a series of policies to reduce peopleā€™s exposure to them.Ģż

The panel was co-chaired by , clinical professor emeritus of health systems and population health at the UW and executive director of the nonprofit group . UW News sat down with Krieger to discuss the new definition, the debate around ultra-processed foods and how people can limit their consumption.

How have we traditionally defined ultra-processed foods, and where does that definition fall short?Ģż

Jim Krieger: The ultra-processed food concept was developed by who’s a physician and epidemiologist at the University of Sao Paolo in Brazil. He was trying to understand an increase in obesity and chronic disease rates, particularly in kids and young adults, and noticed some important changes in the diet that weren’t fully explained by just the usual nutrient profiling ā€” like fat, sugar, salt. Monteiro came up with this concept of ultra-processed foods. The categorization system is called , which means ā€˜newā€™ in Portuguese, and classifies foods across a spectrum beginning with unprocessed ingredients and ending in ultra-processed. Ultra-processed foods are the ones that are the most highly processed industrial products, basically.Ģż

The Nova definition is geared toward research, to really examine the effects of ultra-processed foods on a range of health outcomes. Using it, numerous studies have found ultra-processed foods to be associated with a whole host of health problems, like diabetes and heart disease and depression. Thatā€™s the basis to say, well, thereā€™s probably sufficient evidence to figure out what we can do to reduce exposure to ultra-processed foods by reducing sales and consumption. That requires policy, and to have policy you need to have a definition of ultra-processed food thatā€™s suitable for regulation or legislation, and thatā€™s where the rub comes. The definition for research doesnā€™t really work in a policy context, because the Nova research definition uses multiple factors, including ingredients, processing techniques, and other factors to identify products as ultraprocessed. Applying Nova can require individual-level review of ambiguous products by skilled nutrition experts that may not be feasible in policy contexts.

You co-chaired a panel of 14 experts who came up with a new definition. What did that work look like, and whatā€™s your new definition?

One criticism of the movement against UPFs has been that some foods that are technically ultra-processed are actually quite nutritious. Iā€™m thinking of products like yogurts, whole-grain breads and tofu. How does your definition account for that?

JK: You want your definition to be sensitive enough to pick up most UPFs, but also specific ā€” that is, not capture foods that are not truly UPFs or even those that can be part of a healthy diet. The way we addressed that was the FDA, a couple years ago, developed criteria for what they call a . If a company wants to say its products are healthy and put that on the package, it has to meet .Ģż

We decided that even if a product is ultra-processed, if it meets the FDAā€™s criteria for a ā€˜Healthyā€™ food, then it should be exempted from policy. That cuts out edge cases ā€” healthier foods that are also UPFs.

Your final report also dives into policy and makes recommendations for lawmakers to consider. What is the current status of ultra-processed food policy across the U.S.?Ģż

JK: Over the last couple of years there has been a flurry of activity, particularly at the state legislative level. Some states say theyā€™re getting rid of ultra-processed foods in school meals, for example, but they have a somewhat random list of ingredients or additives they donā€™t like.Ģż

A few states have tried to take a more evidence-based approach. The best example of that is California, where they passed . They used the same kind of ingredient-list approach that we recommend, simplified a little bit. However, for a food to be a UPF under Californiaā€™s definition, it must also be high in fat, sugar or salt, which raises a problem ā€” about 35% of all ultra-processed foods do not have those levels. Proposed legislation in states like Pennsylvania have avoided this problem by sticking with the Nova-based definition, as recommended by our expert panel, rather than adding on fat, sugar and salt criteria.ĢżĢż

Thereā€™s also been a huge amount of movement in a couple of countries, especially in Latin America. Furthest along is , where the government just issued a regulation to require ultra-processed food labels on packages. Theyā€™re basically using the Nova definition as well.

There are policies moving now, which is why we felt it was important to say, use a good definition of ultra-processed for what youā€™re doing, and then think about certain policy ideas as the best bets for doing something about the problem. For us, those ideas include requiring labels on packages identifying a product as ultra-processed and removing UPFs from food served in schools, childcare and in government facilities.Ģż

Many of the factors that lead people to choose ultra-processed foods are systemic. As an example, food deserts leave some people without easy access to affordable fresh foods. How did the panel consider those factors in making your policy recommendations?

JK: Increasing access to healthy foods gets into another set of policies that are well-described and, to varying extents, are being put into place. Our panel focused specifically on UPF policies. We did assess whether any of our policy recommendations would have unintended consequences of making food less available or affordable for people with low incomes and then we figured out strategies for mitigating those effects.Ģż

A great example would be if you tax even a subset of ultra-processed foods, thatā€™s going to make them less affordable, and thatā€™s a challenge. We recommend that the tax revenues raised from these policies go toward vouchers or incentives for people with lower incomes that they can cash in for fruits and vegetables at a more affordable price.Ģż

We also rejected some policy ideas. One was restricting the use of SNAP (the Supplemental Nutrition Assistance Program, commonly known as food stamps) benefits for purchasing ultra-processed foods. We felt that would have too much of a negative impact on food affordability for people using SNAP, so we did not recommend that.Ģż

What comes next for this research?Ģż

JK: Our next step is to share this definition with as many policymakers as we can, and then offer them technical support if they want to figure out how to use this in legislation or regulation. The second thing is if any legislators, policymakers or advocates want to pursue policies, weā€™re available to provide technical support and assistance.Ģż

Thereā€™s also a bunch of additional research that would be useful. We have some of those recommendations at the end of the report, but as an example, thereā€™s controversy right now over different subgroups of UPFs and whether some are more harmful than others. The research thatā€™s been done so far has been somewhat confusing and flawed. Thereā€™s also research on how, exactly, ultra-processed foods cause all these problems. There are a lot of great, interesting hypotheses, but I wouldnā€™t say any of them are definitive.Ģż

As an expert in this field, how do you approach ultra-processed foods in your own life, and how would you recommend people think about reducing their own consumption?Ģż

JK: Right now, about 60% of the calories that Americans consume are ultra-processed foods. Starting there, small and incremental steps are great. You canā€™t totally change your diet overnight, so the bottom-line message is to think about small things you can do.Ģż

The first challenge is identifying UPFs, which is where we started this conversation. If there are ingredients in food that you donā€™t have in your kitchen, itā€™s likely going to be a UPF, and if itā€™s a long ingredient list that looks like a chemistry lab, itā€™s probably a UPF. Then you can consider what the food looks like. If itā€™s super bright, like Froot Loops or Doritos, that probably means itā€™s a UPF.Ģż

But then once youā€™ve identified ultra-processed foods, what do you do? Letā€™s think about one thing you want to do. Say you drink a lot of sweetened beverages. You can think of what you can swap in that works for you, like sparkling water that doesnā€™t have a lot of additives, or coffee or tea.

As for me, I donā€™t eat many ultra-processed foods. When I started doing this work, when my kids were younger, I fed them all sorts of UPFs. I certainly wouldnā€™t do that now and wish I had known better. But if I want to have something thatā€™s ultra-processed, thatā€™s fine, itā€™s not going to kill you to have just a little bit. As long as your , then thatā€™s great.Ģż

The panel was co-chaired by Krieger and Lindsey Smith Taillie of the University of North Carolina at Chapel Hill and convened by Mary Story and Megan Elsener Lott of Duke University. A full list of panel members is included in the panelā€™s This work was funded by Healthy Eating Research and the Robert Wood Johnson Foundation.Ģż

For more information or to contact Krieger, email Alden Woods at acwoods@uw.edu.Ģż

The human mouth is full of wonders. Itā€™s home to hundreds of species of bacteria, fungi, viruses and protozoa, which work in delicate harmony to maintain our oral health. Sometimes, though, this complex system ā€” known as the oral microbiome ā€” can fall out of balance. Anaerobic bacteria build up on the tongue and in the little pockets between our teeth and gums. There, they break down organic matter and spew out a foul odor.

To rebalance the oral microbiome and cure chronic halitosis, researchers at the ĢģĆĄÓ°ŹÓ“«Ć½ are embarking on a first-of-its-kind experiment. These clinical trials transplant bacteria and other minuscule critters from healthy donors into patients with halitosis. If successful, the healthy microbiota will crowd out the bad and patientsā€™ bad breath will improve.

ā€œWe know the oral microbiome can get out of whack. The question is, can you rebalance it? That is the hypothesis weā€™re proposing,ā€ said , a UW professor of restorative dentistry and co-lead of the project.Ģż

The experimental procedures build off recent breakthroughs in , commonly known as stool transplants, which have become a go-to treatment for gastrointestinal infections and bacterial imbalances.Ģż

The research team has so far completed four transplants, with preliminary evaluations underway. Theyā€™re seeking pairs of participants ā€” a patient with chronic halitosis and a donor, ideally an intimate partner, family member or trusted friend ā€” to undergo these relatively simple procedures.

To start, researchers complete a full periodontal exam of the donor to ensure their microbiome is healthy. Then they collect bacteria from the donor and suspend it in a small volume of saline. At the same time, recipients undergo a deep cleaning to remove the harmful bacteria and disrupt the ā€” the thin, sticky layer of microorganisms that lines surfaces in the mouth. Recipients rinse with the donor solution, and researchers inject a concentrated version into the gumline. Ninety days after the transplant, participants self-report whether their breath has improved.Ģż

ā€œWhat weā€™re trying to do is severely disrupt the original bacteria, and then we bring in the new guys to take hold and establish a new biofilm,ā€ said co-lead , a research scientist and affiliate faculty member in the UW School of Dentistry. ā€œIf we bring enough of the new bacteria and they outcompete the ones that we disrupted, the healthy ones will take over. Itā€™s a numbers game.ā€Ģż

This research was funded by the Dean and Margaret Spencer Clinical Research Fund. Co-investigators include professor of clinical practice and professor , both of the UW School of Dentistry. For more information, to reach the researchers or to inquire about participating, contact Pozhitkov at pozhit@uw.edu or Wee at awe@uw.edu.

Micropantries ā€” commonly called ā€œlittle free pantriesā€Ģż ā€” and community fridges are a frequent sight throughout Seattle and the greater Puget Sound region. One estimate suggests that they supply around 4 million pounds of food per year to neighbors in need in the Seattle area, more than the stateā€™s largest food bank. The curbside cupboards are a decentralized, community-driven effort to fight food insecurity and reduce food waste at the neighborhood level, but their ad hoc nature limits their dependability ā€” users donā€™t know when food is available without repeatedly checking, and donors donā€™t know what foods are needed most.

Now, anyone who interacts with micropantries or community fridges in the Seattle area can try out an experimental app, made by ĢģĆĄÓ°ŹÓ“«Ć½ researchers, that brings a suite of new features to the micropantry network. , maps many local pantries across the region. The app also gives each pantry an activity feed where users can share food theyā€™ve donated, report on stock levels, add requests to a wish list, post photos and leave other notes. The research team also retrofitted some pantries with sensors that anonymously auto-report their usage and stock levels to the app in real time.

ā€œThis is an effort to document and quantify the phenomenon of micropantries,ā€ said , a senior research scientist at the UW . ā€œLots of micropantries and community fridges popped up around the time of the COVID-19 pandemic, and I was curious about who uses them and how they are used.ā€

The team was cognizant of privacy concerns and designed the smart pantry tech accordingly.

ā€œPutting cameras in the pantries could give us a lot of information about what specific foods are moving through the system, but that may also deter users who are concerned about privacy,ā€ said , a UW doctoral student in the Paul G. Allen School of Computer Science & Engineering who designed and built the sensor suite. ā€œInstead, we settled on simpler sensors that measure weight and interactions like opening the door to measure stock levels while preserving everyoneā€™s anonymity.ā€

The researchers hope that neighbors will find new ways to connect and help one another through these tools. A user might see that stock levels are low in a nearby pantry, for example, and decide to add some food. Another user might request certain foods to accommodate their dietary restrictions.Ģż

The sensor-equipped pantries are a small subset of the dozens of pantries throughout Seattle, but in addition to providing some neighborhoods with enhanced food tracking, they will generate aggregate data that will help Dalla Chiaraā€™s team study donor and usage behavior. Dalla Chiara also plans to survey donors to learn more about what motivates people to provide food to pantries.

ā€œWe know that there is a lot of food insecurity in Seattle and in the United States in general,ā€ Dalla Chiara said. ā€œBut we know that there is also a lot of food waste ā€” lots of people have a surplus of food. And we want to see how grassroots efforts like micropantries can address both food insecurity and waste at the same time.ā€

Dalla Chiara and his team recently completed a refit on a cold, sleeting March day at a pantry owned by Saint Paulā€™s Episcopal Church near Seattle Center. The church keeps the pantry regularly stocked, and rector Stephen Crippen is curious about the data the new system will produce.

ā€œIt puts numbers on what weā€™re actually accomplishing,ā€ Crippen said. ā€œIt helps us get in touch with whatā€™s going on on this street.ā€

The research team is also working with local businesses and nonprofits to encourage and track food distribution throughout the pantry network. In April, Seattle-based recycling startup ran a nonperishable food drive across Seattle and delivered 25,000 pounds of food to the ; from there, volunteers from the Cascade Bicycle Clubā€™s distributed the food to micropantries around the city by bike, giving the network an infusion of both food and usage data. The and the nonprofit helped support the projectā€™s community fridges effort.

Dalla Chiara recognizes that there are other grassroots online, and he doesnā€™t want his app to replace those services. Nor does he expect the smart pantry network to remain in service indefinitely ā€” it costs about $150 to retrofit each pantry with sensors, and all that tech will be difficult to maintain after the study concludes in October of this year. At its core, the project is an effort to learn about micropantry usage and explore how technology might encourage sharing of resources and mutual aid systems.

ā€œWeā€™re trying to measure and quantify goodwill,ā€ Dalla Chiara said. ā€œBehind each little free pantry there is a whole system of behaviors ā€” people trying to help one another. If we can understand that system better, we can support it better.ā€

Other UW collaborators include , professor of civil and environmental engineering and director of the Urban Freight Lab; , assistant teaching professor of environmental and occupational health sciences; , assistant professor of food systems, nutrition and health; and , assistant professor in the Allen School.

For more information, contact Dalla Chiara at giacomod@uw.edu.

In the three-and-a-half years since the U.S. Supreme Court overturned the constitutional right to an abortion in Dobbs v. Jackson Womenā€™s Health Organization, the fragmented state of abortion access has put medical professionals in a precarious position. Many states have tightened abortion restrictions, with some enacting criminal penalties up to in for physicians who perform abortions. Medical schools have

New research led in part by the ĢģĆĄÓ°ŹÓ“«Ć½ found that the new restrictions are not only affecting the current medical workforce ā€” they may be shaping the next generation of physicians. The study, , found that applications to medical residency programs in states that enacted new abortion restrictions dropped sharply following the Dobbs ruling.

The decrease occurred among both male and female applicants. Applications to specialties related to reproductive health ā€” obstetrics and gynecology, family medicine, internal medicine and emergency medicine ā€” saw the largest decreases.

The new study builds on that had shown decreased application rates to residency programs in states with abortion restrictions by applying causal methodologies to understand the impact of the Supreme Court decision and isolating results from male and female applicants.

ā€œThis research provides important empirical evidence about how state-level policy changes following Dobbs may influence decisions made by medical trainees about where to pursue their graduate medical education,ā€ said co-author , a UW professor of health economics and director of the Comparative Health Outcomes, Policy, and Economics (CHOICE) Institute.

To learn more about the research, UW News sat down the paperā€™s three authors: Basu; lead author , assistant professor of medicine at the University of North Carolina at Chapel Hill; and co-author , assistant clinical professor of internal medicine at the University of Arizona. Both Ganguly and Morenz completed their internal medicine residencies at the UW School of Medicine.

The medical residency match process is quite different from traditional higher-ed applications. Can you explain how that works, and how it relates to your studyā€™s findings?

Dr. Anna Morenz: Applicants may apply to as many programs as they want, with some applying to dozens of programs. At the end of interviews, theyā€™ll rank those programs based on their preferred landing spots. The programs, in turn, will rank all the applicants that they received. A computer algorithm then matches everyone with the goal of filling all the residency slots, and itā€™s very good at that. We know that . So programs are still filling their residency slots even in states with restrictions.

What concerns us about these findings is that there’s an early signal of people avoiding applications to these states. That has potential implications for the quality of the applicants to restricted states, which could not be assessed in our data. There’s typically a high likelihood that people stay where they train for their residency, but if you landed in a restricted state that was low on your rank list, you may be more likely to complete your training and then leave to a non-restricted state. We aim to look at this very important question in projects to come.

Anisha and Anna, youā€™re both practicing primary care physicians. How big a part of a physicianā€™s training is abortion and other pregnancy-related care?Ģż

Dr. Anisha Ganguly: Itā€™s not a big part of our training traditionally, though there has been a movement to integrate more abortion care into primary care residencies. Thatā€™s more the case in family medicine rather than internal medicine, because medication abortion has now become the most common means for abortion care. As internists, we commonly diagnose pregnancies and care for women with medical conditions as they consider family planning.

AM: I do think it’s important to note that a huge percentage of primary care physicians are trained in family medicine. And family medicine physicians are trained in delivery of babies, management of prenatal care, miscarriage management, contraception and abortion. Anisha and I trained in internal medicine, and there is increasing interest to include medication abortion training in internal medicine, as it is fully within our scope of practice.

The effects of the Dobbs decision have been well-documented, and previous work on this topic highlighted changes in OB/GYN residency applications. Whatā€™s new in your study specifically?Ģż

Anirban Basu: We had a much longer pre-period than previous studies. We looked back to 2019 to see what had been happening to application rates in these two kinds of states ā€” those that eventually restricted abortion access and those that didnā€™t ā€” and we showed that these rates had been moving similarly until the ruling. That gives a little more weight to the evidence to say the change is due to the ruling.Ģż

The second big thing is that previous studies did not distinguish whether men and women were changing their behavior similarly. I think thatā€™s a very important finding in our study, that male applicants are changing their behavior at an even higher rate.Ģż

AG: I agree that the gender stratification was an important contribution. The other stratified analysis that we explored was about how specialty type may be driving some of the effects that we saw. A lot of people can reason that OB/GYN applicants would be affected by this directly, and there’s a lot of literature to support that. But what we’re showing is that it’s not just the OB/GYN workforce that’s going to be impacted. It’s the primary care workforce and the emergency medicine workforce.Ģż

We’re hoping that message spreads a little more broadly. This is not just about women’s health. It’s about the future of primary care and the person who’s going to save you from your heart attack in the future.

Letā€™s talk a little more about that gender stratification. You found that male applicants changed their application preferences at a greater rate than female applicants, which looks like a surprising result. Whatā€™s going on there?Ģż

AG: When we generated our original hypotheses, we thought we were going to see increased effects among women applying to residency, but we actually ended up seeing that there were long-term disparities that existed pre-Dobbs between restricted and non-restricted states. This was likely because of the and other state-level laws that were affecting women’s behavior. What we’re seeing is that women had been reading the tea leaves about access to reproductive health care prior to the Dobbs decision, but the decision did unmask a wider problem that drove a lot of new behavior among men.

One of the messages that we are getting from this paper is this is an ā€œall of usā€ problem. It’s not just about women physicians. It’s about men who are also making choices about their professional autonomy and also about access to reproductive health care for their families. Women have been and will be considering their personal access to care and autonomy before this decision, but perhaps these state restrictions after Dobbs may have newly increased awareness among men.Ģż

Among all these shifts, you found one group whose application rates didnā€™t change significantly: people applying to highly competitive medical specialities. What do you think explains that stickiness?Ģż

AG: Anna and I had brainstormed about this being a potential effect modifier, because people who are applying in highly competitive specialties like orthopedic surgery or dermatology apply very broadly and don’t get to exercise a lot of choice about where to go. Whereas for large specialities like internal medicine, family medicine or pediatrics, there are a lot of programs in a lot of places, so applicants have more options. In those cases, state-level policies like abortion restrictions can factor more into peopleā€™s decision-making.

At an institutional level, what changes could be made to address these trends?Ģż

AG: Institutions can make choices to mitigate some of these effects by supporting candidates with access to reproductive care within the scope of the restrictions that exist. Other industries are building in travel benefits for women who may need to travel to access these services.Ģż

It’s not this aspect of a decision alone that shapes a residency applicantā€™s choice to go to a specific place or program. But there are other things that institutions can do to make trainees, particularly women, feel supported and valued. If you’re existing in an environment where state policies make women feel a lack of autonomy, then there are workforce policies that can be in place to bolster that sense of autonomy. That could take the shape of parental leave policies, lactation policies, other things that institutions can do to make women feel like, even if this part of your voice has been taken away, we’ll help you with the rest.

AB: One policy that has a long history of literature supporting it is financial incentives. Physicians do respond to financial incentives, but in many cases those incentives need to be quite steep to get people to change their decisions.Ģż

AM: The other option is training opportunities. A lot of programs in states that had laws or restrictions that preceded the Dobbs decision would set up partnerships with organizations in another state where they could send their trainees to learn about pregnancy termination and miscarriage management. Thatā€™s a burden on residency programs and residents both. You have to set up housing and travel agreements. But thatā€™s another key thing that programs need to keep in mind in order to recruit applicants.Ģż

For more information or to contact the researchers, contact Alden Woods at acwoods@uw.edu.

Over five decades, the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) has become known as the nationā€™s. Low-income families receiving WIC benefits ā€” which provides nutritious food in designated categories, nutrition education and access to other social services ā€” have .Ģż

But many families who are income eligible to participate in WIC arenā€™t receiving those benefits. Research has found that households who donā€™t use the full amount of their nutrition benefits are more likely to drop from the program.Ģż

New research by the ĢģĆĄÓ°ŹÓ“«Ć½ has found that households who redeem more of their benefits in the most popular food categories are more likely to remain in the program long-term. Better understanding these patterns could help WIC agencies identify families who might need a little extra encouragement to stay enrolled.

The study was .Ģż

ā€œFinding ways to identify kids and families that are at risk of dropping out of the program is of high importance,ā€ said , a UW assistant professor of health systems and population health and first author of the study. ā€œThatā€™s basically what weā€™ve identified ā€” a way to flag families who may be at risk of dropping off.ā€

WIC provides families with food benefits in , with fruits and vegetables and eggs as the most popular. In partnership with (PHFE WIC), a Southern California WIC agency with a large research and evaluation division, researchers analyzed redemption data from 188,000 participating infants and children 0-3 years old, between the years 2019 and 2023.Ģż

Among those children, higher redemption of fruits and vegetables, eggs, whole milk and infant formula was associated with lower risk of their household discontinuing WIC participation.Ģż

The risk of discontinuation decreased in a somewhat linear fashion as redemption rates increased.

Chaparro hopes that local WIC agencies will build on these findings and seek new ways to engage families at risk of dropping off. All WIC providers must offer nutrition education, which could be an opportunity to target households with lower redemption rates in popular categories.Ģż

The findings come just over a year after the U.S. Department of Agriculture, which oversees WIC, . Among other changes, the 2024 rule significantly increased benefits for fresh fruits and vegetables, which has proven popular.

ā€œThe expansion of fruit and vegetable benefits for WIC families has been among the most important policy changes of the last decade,ā€ said , director of research and evaluation at PHFE WIC and co-author of the study. ā€œFamilies want more fruits and vegetables, and this research demonstrates that their inclusion in the WIC food package is essential for longer-term engagement in the program.ā€

of the University of Tennessee and PHFE WIC is the corresponding author. This study was funded by .

When uncontrolled wildfires moved from the foothills above Los Angeles into the densely populated urban areas below in January 2025, evacuation ensued and a thick layer of toxic smoke spread across the region. Air quality plummeted. Local hospitals braced for a surge,.ĢżĢż

Research led by the ĢģĆĄÓ°ŹÓ“«Ć½ and Kaiser Permanente Southern California sheds new light on how the Los Angeles fires affected peopleā€™s health, and how people navigated the health care system during an emergency. In the rapid study, published , researchers analyzed the health records of 3.7 million Kaiser Permanente members of all ages living in the region. They found that health care visits did rise above normal levels, especially virtual services.ĢżĢż

In the week after the fires ignited, Kaiser Permanente members made 42% more virtual visits for respiratory symptoms than expected. Those living near a burn zone or within Los Angeles County also made 44% and 40% more virtual cardiovascular visits, respectively, than expected.Ģż

In-person outpatient visits for respiratory symptoms also increased substantially. Members who lived near a burn zone or within Los Angeles County made 27% and 31% more virtual cardiovascular visits, respectively, than expected.Ģż

Extrapolating to all insured residents of the county, the researchers estimated an excess of 15,792 cardiovascular virtual visits, 18,489 respiratory virtual visits and 27,903 respiratory outpatient visits in the first week of the fires.Ģż

The results suggest that people may rely more heavily on virtual health care during climate-related emergencies, and that providers should better prioritize virtual and telehealth services as they prepare for future crises.Ģż

ā€œWe saw over 6,241 excess cardiorespiratory virtual visits in the week following the fire ignition. This represents a substantial increase in care,ā€ said, a UW associate professor of environmental and occupational health sciences and of epidemiology who led the research. ā€œWhile the fires clearly impacted health, virtual care likely enhanced the ability of providers to meet the health care needs of people experiencing an ongoing climate disaster.ā€Ģż

In collaboration with Kaiser Permanente Southern California, an integrated health care system with millions of members across the region, researchers analyzed health records of people who were highly or moderately exposed to wildfires. They defined high exposure as living within about 12 miles (20 kilometers) of a burn zone, and moderate exposure as living within Los Angeles County but farther than 12 miles during the time of the fires.ĢżĢż

Researchers looked back three years to estimate how many health care visits to expect in the weeks following Jan. 7 ā€” the first day of the fires ā€” under typical conditions. They then estimated how many people sought care in the first week of the fires, when smoke levels were highest, evacuations took place, and Los Angeles County public schools were closed.

In addition to the spike in cardiovascular and respiratory visits, researchers found a sharp increase in the number of visits for injuries and neuropsychiatric symptoms. On Jan. 7, outpatient injury visits were 18% higher than expected among highly exposed members, and virtual injury visits were 26% and 18% higher than expected among highly and moderately exposed groups, respectively. Among those same groups, outpatient neuropsychiatric visits rose 31% and 28% above expectations, respectively.

While both groups made significantly more visits than expected, proximity to the fires mattered. When researchers zoomed in on respiratory-related virtual visits, they found that minimally exposed members made 31% more visits, moderately exposed members made 36% more, and those living in highly exposed areas made 42% more.ĢżĢż

ā€œWhile healthcare systems often plan to increase the number of hospital beds available or clinic staffing during an emergency, this work highlights the importance of considering virtual care capacity,ā€ said, a UW doctoral student of epidemiology and co-author on the study. ā€œThis may be particularly true for climate disasters like wildfires, during which people are advised to stay indoors or when people must evacuate ā€” motivating them to seek care online if at all possible. As climate disasters increase in frequency and intensity, it is essential that health care systems know how to prepare for a sudden and dramatic surge in health care utilization.ā€Ģż

Other authors on this study are , and of Kaiser Permanente Southern California; of the University of California, Berkeley; of Kaiser Permanente Hawaii; and of Columbia University; and of the Scripps Institution of Oceanography at UC San Diego; and of the Scripps Institution and the University of Rennes in France.

This research was funded by the National Institute on Aging and the National Institute for Environmental Health Sciences.Ģż

For more information or to reach the research team, contact Alden Woods at acwoods@uw.edu.

About . Birth before 37 weeks can lead to a cascade of health risks, both immediate and long-term, making prevention a vital tool for improving public health over generations.Ģż

In recent years, researchers have identified a potential link between wildfire smoke ā€” one of the fastest-growing sources of air pollution in the United States ā€” and preterm birth, but no study has been big or broad enough to draw definitive conclusions. A new study led by the ĢģĆĄÓ°ŹÓ“«Ć½ makes an important contribution, analyzing data from more than 20,000 births to find that pregnant people who are exposed to wildfire smoke are more likely to give birth prematurely.

ā€œPreventing preterm birth really pays off with lasting benefits for future health,ā€ said lead author , a UW postdoctoral researcher in environmental and occupational health sciences. ā€œItā€™s also something of a mystery. We donā€™t always understand why babies are born preterm, but we know that air pollution contributes to preterm births, and it makes sense that wildfire smoke would as well. This study underscores that wildfire smoke is inseparable from maternal and infant health.ā€

In the study, ,Ģżresearchers used data from the , a federal research project focused on how a wide range of environmental factors affect childrenā€™s health. The sample included 20,034 births from 2006-2020 across the contiguous United States.

Researchers estimated participantsā€™ average daily exposure to fine particulate matter, or PM2.5, generated by wildfire smoke, and the total number of days they were exposed to any amount of smoke. They estimated the intensity of smoke exposure by how frequently participants were exposed to wildfire PM2.5 levels above certain thresholds.

They found that pregnant people exposed to more intense wildfire smoke were more likely to give birth prematurely. In mid-pregnancy, exposure to any smoke was associated with an elevated risk of preterm birth, with that risk peaking around the 21st week of gestation. In late pregnancy, elevated risk was most closely associated with exposure to high concentrations of wildfire PM2.5, above 10 micrograms per cubic meter.

ā€œThe second trimester is a period of pregnancy with the richest and most intense growth of the placenta, which itself is such an important part of fetal health, growth and development,ā€ said co-author , a UW professor of environmental and occupational health sciences and of pediatrics in the UW School of Medicine. ā€œSo it may be that the wildfire smoke particles are really interfering with placental health. Some of them are so tiny that after inhalation they can actually get into the bloodstream and get delivered directly into the placenta or fetus.ā€Ģż

The link was strongest and most precise in the Western U.S., where people were exposed to the highest concentrations of wildfire PM2.5 and the greatest number of high-intensity smoke days. Here, the odds of preterm birth increased with each additional microgram per cubic meter of average wildfire PM2.5.

Itā€™s possible those results were more precise simply because the West experiences more wildfire smoke on average, making the exposure model perform better, Sherris said. But there may be other factors behind the regional differences.Ģż

The composition of wildfire smoke is different across the country. In the West, smoke tends to come from fires nearby, while in places like the Midwest, smoke has typically drifted in from faraway fires. and reacts with sunlight and airborne chemicals, which could have affected the results. Researchers also noted that external factors like co-occurring heat or housing quality may have effects that arenā€™t fully understood.Ģż

Researchers hope that future studies will examine the exact mechanisms by which wildfire smoke might trigger preterm birth. But in the meantime, Sherris said, evidence for a link is now strong enough to take action.Ģż

ā€œThere are a couple avenues for change,ā€ Sherris said. ā€œFirst, people already get a lot of public health messaging and information throughout pregnancy, so thereā€™s an opportunity to work with clinicians to provide tools for pregnant people to protect themselves during smoke events. Public health agenciesā€™ messaging about wildfire smoke could also be tailored to pregnant people and highlight them as a vulnerable group.ā€

Co-authors include , doctoral student of environmental and occupational health sciences at the UW; , clinical associate professor of environmental and occupational health sciences at the UW; , professor of biostatistics at the UW; , associate professor of environmental and occupational health sciences and of epidemiology at the UW; , postdoctoral fellow of epidemiology at the UW; and , assistant professor of environmental and occupational health sciences at the UW. A full list of co-authors is included with the paper.

This research was funded by the Environmental influences on Child Health Outcomes (ECHO) program at the National Institutes of Health under multiple awards. A full list of ECHO funding awards is included with the paper.Ģż

For more information or to contact the researchers, email Alden Woods at acwoods@uw.edu.

Targeted drug delivery is a powerful and promising area of medicine. Therapies that pinpoint the exact areas of the body where theyā€™re needed ā€” and nowhere theyā€™re not ā€” can reduce the medicine dosage and avoid potentially harmful ā€œoff targetā€ effects elsewhere in the body. A targeted immunotherapy, for example, might seek out cancerous tissues and activate immune cells to fight the disease only in those tissues.

The tricky part is making a therapy truly ā€œsmart,” where the medicine can move freely through the body and decide which areas to target.

Researchers at the ĢģĆĄÓ°ŹÓ“«Ć½ took a significant step toward that goal by designing proteins with autonomous decision-making capabilities. In a proof-of-principles study in Nature Chemical Biology, researchers demonstrated that by adding smart tail structures to therapeutic proteins, they could control the proteinsā€™ localization based on the presence of specific environmental cues. These protein tails fold themselves into preprogrammed shapes that define how they react to different combinations of cues. In addition, the experiment showed that the smart protein tails could be attached to a carrier material for delivery to living cells.

Advances in synthetic biology also allowed the researchers to manufacture these proteins cheaply and in a matter of days instead of months.

ā€œWeā€™ve been thinking about these concepts for some time but have struggled with ways to increase and automate production,ā€ said senior author , a UW professor of chemical engineering and bioengineering. ā€œWeā€™ve now finally figured out how to produce these systems faster, at scale and with dramatically enhanced logical complexity. We are excited about how these will lead to more sophisticated and scalable disease-honing therapies.ā€

The concept of programmable biomaterials isnā€™t new. Scientists have developed numerous strategies to make systems responsive to individual cues ā€” such as pH levels or the presence of specific enzymes ā€” that are associated with a particular disease or area of the body. But itā€™s rare to find one cue, or ā€œbiomarker,ā€ thatā€™s unique to one spot, so a material that hones in on just one biomarker might act on a few unintended places in addition to the target.

One solution to this problem is to seek out a combination of biomarkers. There might be many areas of the body with particular enzyme or pH levels, but there are likely fewer areas with both of those factors. In theory, the more biomarkers a material can identify, the more finely targeted drug delivery can be.

In 2018, DeForestā€™s lab created a new class of materials that responded to multiple biomarkers using Boolean logic, a concept traditionally used in computer programming.

ā€œWe realized that we could program how therapeutics were released based simply on how they were connected to a carrier material,ā€ DeForest said. ā€œFor example, if we linked a therapeutic cargo to a material via two degradable groups connected in series ā€” that is, each after the other ā€” it would be released if either group was degraded, acting as an OR gate. When the degradable groups were instead connected in parallel ā€” that is, each on a different half of a cycle ā€” both groups had to be degraded for cargo release, functioning as an AND gate. Excitingly, by combining these basic gates we could readily create advanced logical circuits.ā€

It was a big step forward, but it wasn’t scalable ā€” the team built these large and complex logic-responsive materials manually through traditional organic chemistry.

But over the next several years, the related field of synthetic biology advanced by leaps and bounds.

ā€œThe field has developed exciting new protein-based tools that can allow researchers to form permanent bonds between proteins,ā€ said co-first author , a UW doctoral student of bioengineering. ā€œIt opened doors for new protein structures that were previously unachievable, which made more complex logical operations possible.ā€

Additionally, it became practical to use living cells as factories to produce these complex proteins, allowing scientists to design custom DNA blueprints for new proteins, insert the DNA into bacteria or other host cells, and then collect the proteins with the desired structure directly from the cells.

With these new tools, DeForest and his team streamlined and improved many steps of the process at once. They designed and produced proteins with tails that spontaneously fold into more bespoke shapes, creating complex ā€œcircuitsā€ that can respond to up to five different biomarkers. These new proteins can attach to various carriers ā€” hydrogels, tiny beads or living cells ā€” for delivery to a cell, or theoretically a disease site. The team even loaded up one carrier with three different proteins, each programmed to deliver their unique cargo based on different sets of environmental cues.

ā€œWe were so excited about the results,ā€ DeForest said. ā€œUsing the old process, it would take months to synthesize just a few milligrams of each of these materials. Now it takes us a couple of weeks to go from construct design to product. It’s been a complete game changer for us.ā€

ā€œThe skyā€™s the limit. You can create delayed and independent delivery of many different components in one treatment,ā€ Ross said. ā€œAnd I think we could create much, much larger logical circuits that a protein can be responsive to. Weā€™re at the point now that the technology is outpacing what weā€™ve seriously considered in terms of applications, which is a great place to be.ā€

The researchers will now continue searching for more biomarkers that proteins could target. They also hope to start collaborating with other labs at the UW and beyond to build and deploy real-world therapies.

The team outlined other uses for the technology as well. The same tools could manufacture therapies within a single cell and direct them to specific regions, a sort of microcosm of how the process works in the body. DeForest also envisions diagnostic tools like blood tests that could, say, turn a certain color when a complex set of cues within the blood sample are present.

DeForest thinks the first practical applications are likely to be cancer treatments, but with more research, the possibilities feel endless.

ā€œThe dream is to be able to pick any arbitrary location inside of the body ā€” down to individual cells ā€” and program a material to go and act there,ā€ he said. ā€œThatā€™s a tall order, but with these technologies weā€™re getting closer. With the right combination of biomarkers, these materials will just get more and more precise.ā€

Co-authors include , a former UW undergraduate student of chemical engineering; , a UW undergraduate student of bioengineering; and , a UW doctoral student of chemical engineering.

This research was funded by the National Science Foundation and the National Institutes of Health.

For more information, contact DeForest at profcole@uw.edu.Ģż

In the 10 years since the federal government , studies have suggested the policy has wide-ranging benefits. Students in participating schools , are and .

Now, as cuts to food assistance programs threaten to slash access to universal school meals, a new study led by the ĢģĆĄÓ°ŹÓ“«Ć½ finds another potential benefit to the programs: Students in participating schools were less likely to have high blood pressure, suggesting that universal free meals might be a powerful tool for improving public health.Ģż

ā€œHigh blood pressure is an important public health problem that isnā€™t studied as much on a population level as obesity,ā€ said , a UW postdoctoral researcher of health systems and population health and lead author of the study. ā€œWe have evidence that CEP increases participation in school meals, and we also have evidence that school meals are more nutritious than meals that kids obtain elsewhere. This is a public health policy that is delivering nutritious meals to children who may not have previously had access.ā€

For the study, , researchers linked two datasets that rarely interact. They obtained medical records of patients ages 4-18 from community health organizations, and used patientsā€™ addresses to identify the school they attended. The data encompassed 155,778 young people attending 1,052 schools, mostly in California and Oregon.

Researchers estimated the percentage of students with high blood pressure before and after schools opted into universal free meals, and compared those results against eligible schools that had not yet participated in the program. They also tracked studentsā€™ average systolic and diastolic blood pressure readings. All data were aggregated at the school level.Ģż

They found that school participation in the CEP was associated with a 2.71% decrease in the proportion of students with high blood pressure, corresponding to a 10.8% net drop over five years. School participation in CEP was also associated with a decrease in studentsā€™ average diastolic blood pressure.Ģż

ā€œIn previous work on the health impacts of universal free school meals, our team found that adoption of free meals is associated with decreases in and , which are closely linked to risk of high blood pressure,ā€ said , a professor of health, society and behavior at the University of California Irvineā€™s Joe C. Wen School of Population & Public Health and senior author of the study. Jones-Smith conducted much of this research while on faculty at the UW School of Public Health. ā€œSo in addition to directly affecting blood pressure through provision of healthier meals, a second pathway by which providing universal free meals might impact blood pressure is through their impact on lowering risk for high BMI.ā€Ģż

Improved nutrition of school meals may have helped drive the decrease, researchers said. The 2010 law that established the CEP also created stronger nutritional requirements for school meals. As a result, those meals now more closely resemble the , which to be an effective tool for managing hypertension.Ģż

Despite the evidence supporting the DASH dietā€™s effectiveness, public health officials previously lacked an effective mechanism to encourage people with high blood pressure to follow its recommendations. ā€œWe know there are a lot of barriers to people eating this diet,ā€ Localio said, but the combination of universal free meals and increased nutritional standards likely helped students overcome those barriers.

The study also contradicts the common misperception that universal free meals mostly benefit wealthier students, because students from low-income families would already receive free meals. The study sample consists primarily of low-income patients, with 85% of included students enrolled in public health insurance such as Medicaid.

ā€œThere is a perception that providing universally free school meals will only improve outcomes for students of relatively higher-income families, but our findings suggest that there are benefits for lower-income children as well,ā€ Jones-Smith said. ā€œPotential mechanisms for this include decreasing the income-related stigma around eating school lunch by providing it free to all students and eliminating the time and paperwork burden of individually applying, thus decreasing barriers to participation in school meals.ā€

These findings come at an uncertain time for universal free meals. A school is eligible to participate in the CEP if . In this way, recent cuts to the Supplemental Nutrition Assistance Program (SNAP), the nationā€™s largest food assistance program, may affect schoolsā€™ access to the program.

ā€œWeā€™re in a contentious time for public health, but it seems like thereā€™s bipartisan support for healthy school meals,ā€ Localio said. ā€œThereā€™s legislation being considered in a number of states to expand universal free meals, and these findings could inform that decision-making. Cutting funding to school meals would not promote childrenā€™s health.ā€Ģż

Co-authors on the study include , research professor emeritus of health systems and population health at the UW; , teaching professor of economics at the UW; Wyatt Benksen and Aileen Ochoa of OCHIN; and , associate professor of nursing at the UW. This study was funded by the Eunice Kennedy Shriver National Institute of Child Health & Human Development.Ģż

For more information or to contact the researchers, email Alden Woods at acwoods@uw.edu.