UCSF home page UCSF home page About UCSF UCSF Medical Center
UCSF navigation bar
UCSF Cardiology
Transforming medicine through innovation and collaboration.
Department of Medicine

Faculty Spotlight: Dr. Kirstin Aschbacher

Behavior Change Meets Data Science

Dr. Kirstin Aschbacher works at the intersection of clinical psychology and data science, combining digital tools and insights into what makes us tick to promote positive behavioral change.

Born in Pasadena, Calif., Dr. Aschbacher grew up playing piano and guitar, and learned the basics of computer programming from her father, a mathematician. She graduated from Brown University with a bachelor’s degree in music composition and theory, and studied piano performance at the New England Conservatory of Music.

A tragic event altered her life course: when she was 20 years old, her younger sister died in a car accident. “I was introduced to what clinical psychologists could offer people, and I realized that there is this a whole psychological toolbox that can help you through life,” said Dr. Aschbacher. “In our culture, mental health is very stigmatized. Often, people only learn about this toolbox if they have a severe mental health problem or life event, but these are fundamental life skills that should be more like learning to write a check or how to maintain your car.”

The desire to make that toolbox more widely accessible fueled her professional focus on health psychology and lifestyle change. Dr. Aschbacher earned her doctoral degree in clinical psychology from a joint program of UCSF and San Diego State University, then completed a yearlong internship in behavioral medicine and psychology at the University of Washington Medical Center and Harborview Medical Center in Seattle. She then came to UCSF, completing a psychology and medicine postdoctoral fellowship, as well as another fellowship in complex systems analysis.

From the beginning of her psychology training, she was interested in mind-body science, and in San Diego she worked with a research group investigating the effects of chronic stress on cardiovascular disease. At UCSF, she built new collaborations and skills, using engineering and mathematical modeling to learn more about stress biology.

Dr. Aschbacher was recruited to the UCSF Department of Psychiatry faculty in 2012, and received a K23 career development award from the National Heart, Lung, and Blood Institute to pursue her burgeoning interest in the overlap between data science and behavioral medicine. “I believe that the new frontier of behavior science involves digital technology, because you can get data at such a larger scale and speed from wearable devices, and even test out lifestyle interventions,” she said.

Tackling Big Challenges with Big Data

While she had taught herself a number of computer programming languages during graduate school and beyond, she decided to take her skills to the next level by attending the Insight Data Fellowship in Palo Alto, a seven-week data science boot camp. “You had to have a PhD to even apply, and half the group were astrophysicists,” said Dr. Aschbacher. “I was the only psychologist there. It was an intimidating group, but I got through the program and learned a lot.”

Towards the end of the fellowship she met many industry leaders, and was recruited as a data scientist in the behavior change team at Jawbone, which manufactures the UP fitness tracker. A significant percent of UP users were actively trying to lose weight. To help these users reach their weight loss goals, Dr. Aschbacher and her colleagues tested the efficacy of various notification strategies – such as which messages users found most useful for helping them exercise more or eat less.

They also applied psychological insights to hone these messages. For example, they tried to help users self-regulate what they eat, sending empathetic messages normalizing the temptation to eat delicious but unhealthy foods and providing basic behavioral skills training. They also encouraged users to log their food intake to counteract the natural instinct to lapse at times when willpower is worn down, such as snacking late at night.

The team also developed adaptive interventions – personalizing messages based on how much a person exercised or what they ate since the previous message. For example, if the app recommended a step goal of 10,000 and a specific user fell far short of meeting that number, they tried to set a more realistic target as the next goal. “Ideally, the goal you recommend should be based on a user’s history of behavior and maybe also what they’ve done that day so far,” said Dr. Aschbacher.

Their experiments usually included between 10,000 and 100,000 participants, numbers that many academics can only dream of. “One of the biggest learnings I got out of being in industry was how scale really changes everything,” said Dr. Aschbacher. She contrasts that to her clinical training, where she might co-lead a small group at a place like the VA. “It’s a nice model, but it’s not very scalable in terms of being able to reach millions of people in all kinds of locations and deliver a cost-effective service,” she said. “And it may be that some self-regulation strategies are best learned out in the world, where people are experiencing the stressors and temptations of daily life… You want to practice applying your new skills in the moment you really need them. With digital technology, you might be able to have a ‘coach’ you can turn to when the experience is really fresh in your mind.”

Best of Both Worlds

Dr. Aschbacher was recruited back to UCSF in 2017 by Dr. Jeffrey Olgin, chief of the Division of Cardiology and one of the principal investigators of the Health eHeart Study, a novel investigation that uses smartphones, wearable devices and apps to gather big data from participants worldwide, in order to ultimately predict, understand and treat heart disease.

She is particularly interested in developing interventions to improve health using machine learning – a field of computer science in which computers can learn from finding patterns in vast amounts of data without being explicitly programmed. It’s the technology that helps fuel speech recognition, self-driving cars and many other innovations.

“Machine learning doesn’t seem to be as widespread in academia as it is in industry, but it has a lot of power and potential,” said Dr. Aschbacher. “On the other hand, one thing scientists do particularly well is to gather data in a way that is really rigorous and uses strong research designs. My dream is to combine the best of both worlds to advance lifestyle medicine.”

“Kirstin brings a unique combination of skills into the Division, including a background in psychology, industry-sharpened user experience and design, as well as proficiency in data analysis and machine learning,” said UCSF cardiologist Dr. Geoff Tison, who also leverages big data to promote cardiovascular health. “Her combination of skills is synergistic with and will be valuable for a range of research and programmatic initiatives in the Division.”

Many of Dr. Aschbacher’s research interests focus around preventive cardiology, such as using digital tools to help patients eat better, exercise more and reduce chronic stress. She has already begun to collaborate with researchers within the Division of Cardiology in areas such as how to best control high blood pressure among pregnant women, and how to support positive behavioral change among patients of the UCSF Center for Prevention of Heart and Vascular Disease.

Dr. Aschbacher commonly works with interdisciplinary teams of physicians, immunologists, microbiologists and neuroscientists, and has forged long-term collaborations with Dr. Elissa Epel and Dr. Ashley Mason in the UCSF Department of Psychiatry.

“Kirstin is an expert in the intersection of lifestyle and behavioral risk factors – including stress, diet and self-regulation – for cardiovascular disease and obesity, and focuses on digital health and big data methodologies,” said Dr. Mason. “She is an outstanding collaborator with whom I have had the good fortune to work for several years, and key collaborator of the UCSF BEE (Biology and Experience of Eating) Lab.”

Eating and Obesity

Translating good advice into practice is easier said than done. “A lot of interventions, especially in the digital wellness space, tend to invest heavily in what we call ‘psychoeducation,’” said Dr. Aschbacher. “For example, they may tell users, ‘These are the kinds of foods you should eat.’ But even with knowledge, many of us find it hard to stick with our intentions. A lot of my recent research focuses on how we measure and improve our ability to ‘self-regulate’ or control our behavior. When does self-regulation fail us, and why? Knowing this, how can we design an intervention differently?”

For example, self-regulation plays a key role in eating and obesity. In order to better understand what drives reward-based eating and cravings, Dr. Aschbacher and Dr. Mason are conducting a unique and integrative study. They are assessing how cravings are linked with the way that consuming sugar influences participants’ cognitive function and metabolism.

For example, participants took a cognitive test when they were fasting, and then again an hour after drinking a sugary beverage. It turns out that drinking a high-sugar beverage appears to affect obese and lean individuals differently in terms of their tendency to focus on food cues and to eat more sweets, even after a full meal. “We’re trying to connect the dots between what the patient is experiencing, what they are eating, and any underlying differences in the neurocircuitry or brain metabolism, so we can help people learn to self-regulate more effectively,” said Dr. Aschbacher. “People don’t want to just count their calories, they want help managing temptation and sticking with their intentions.”

Make it a Habit

Yet another area that intrigues Dr. Aschbacher is helping people make the leap from goals to habits. For example, many exercise apps are very goal-oriented. “You have a step goal for today, and another step goal tomorrow,” she said. “But the thing about goals is that they’re effortful – you have to make a conscious effort to hit them, and over time they lose their reward value. The first time you hit your goal, you feel good and the app congratulates you, but once you’ve gotten the same message for two weeks, it stops having that rewarding effect.”

One of the biggest challenges to behavior change is getting new behavior to stick. “That’s why habit formation is so important,” said Dr. Aschbacher. “The next era of apps has to focus less on goals and more on habits. A healthy habit should be intrinsically rewarding and fit into your lifestyle without always having to think about it.”

Dr. Aschbacher has cultivated her own exercise habit, which includes a morning walk, Zumba and spin classes, running, biking to work and hiking. She also takes an oil painting class once a week and is studying Mandarin. She is married to Grean Chiranakhon, a software engineering manager, and together they have a young son, Tarin.

Her ultimate goal is to build a self-regulation app that takes a different approach to wellness. “I want to change the dialogue, so it’s less about reaching goals and counting calories,” said Dr. Aschbacher. “Let’s figure out how to make the process of choosing healthy habits feel just a little bit easier and more natural. Although there are many wellness apps on the market, it is important to show, through research, that your wellness app actually impacts health. My goal in combining behavior change science with machine learning is to create change that sticks.”