Heart Rates and Step Counts: A Novel Approach to Eating Disorder Care

What potential do everyday devices, like smartwatches, have in transforming the management and treatment of eating disorders? 

In this episode, Under the Cortex explores how commonly used technology, such as heart rate monitors and step counters, can be used to understand binge-eating episodes. APS’s Özge Gürcanlı Fischer Baum interviews Qinxin Shi from the University of Utah, whose research is featured in Clinical Psychological Science. Together, they discuss the potential implications of this innovative approach for identifying and treating eating disorders. 

Send us your thoughts and questions at [email protected]. 

Unedited Transcript

[00:00:00.270] – APS’s Özge Gürcanlı Fischer Baum

Can the technology we use every day, like smartphones and smartwatches, play a role in addressing eating disorders? How can this information transform the way we approach these conditions? And what does it mean for future treatments? This is Under the Cortex. I am Özge Gürcanlı Fischer Baum with the Association for Psychological Science. In this episode, Under the Cortex examines how monitoring heart rate and step count can be used to predict the binge event. I’m joined by Qinxin Shi from the University of Utah, who has a recent article on this topic published in APS’s journal, Clinical Psychological Science. Qinxin, thank you for joining me today. Welcome to Under the Cortex.

[00:00:49.730] – Qinxin Shi

Thank you, Ozge. It’s such a pleasure to be here.

[00:00:54.140] – APS’s Özge Gürcanlı Fischer Baum

Can you please tell us a little bit about yourself and what type of psychologist are I’m a developmental researcher, developmental psychologist.

[00:01:05.610] – Qinxin Shi

I study children’s and adolescents’ mental health development as a graduate student at Texas A&M University. Then after I graduated 2020, I started focusing on the digital phenotyping and the eating disorder behavior from a dynamic system perspective at University of Utah as a postdoc researcher. I’m currently continuing my We’re focusing on the mental health development and at Children National Hospital at DC, Washington DC. To get more clinical training for adolescents with functional difficulties and how we can further utilize technology.

[00:01:45.210] – APS’s Özge Gürcanlı Fischer Baum

Yeah. Thank you for that introduction. Today, we are going to talk about your research. We are going to talk about eating disorders. What initially got you interested in this topic?

[00:01:57.330] – Qinxin Shi

Yeah, this is a great question. My original interest was not actually on eating disorder in the very beginning. I mentioned when I was a graduate student, I study depressive symptoms and delinquency, focusing on children and adolescents from pre K till a 12th grade. I particularly said there are early childhood risk and resiliency factors associated with their procuring internalizing or depressed symptoms and delinquency development. I use longitudinal data sets that track them manually from kindergarten to 12th grade. I’m really proud of my work as a graduate student. I sometimes wondered how I can apply my findings into the real-life setting. The reason I mentioned this because there was one time I was in a conference, and one high school teacher, after I shared about the early child risks and resiliency factor associated with a long-term mental health development, then she approached me and she told me, Oh, this is really cool. All the risk factors and the resiliency factors you find out in the early childhood at the elementary school exactly matches with my expectations as a school teacher. But can you tell me… I’m a high school teacher, but can you tell me how I can help my students with mental health issues, like depression, delinquency?

[00:03:29.230] – Qinxin Shi

I It’s really back at the moment because I don’t have a good answer for her. The reason is because I cannot go back in time to help the lessons struggle with mental health, to eliminate all the risk factors. It’s really hard to also provide all the protective factors. In the same time, let’s say for my study, for example, teacher warmth is a protective factor, peer acceptance is a protective factor. But that is from others. It’s hard to leverage that in a real setting. I think that’s where my interest started to thinking about how we can let the adolescents use a This is a tool when they can help themselves, when they need the help the most. I think that is where my mind are going. Then I was really fortunate to get involved into the project, which is the BEGIN Project. The short name is BEGIN, which is the Binge Heeding Genetics Initiative Project. With my mentor,, Brian Plastal, University of Utah, and also Cynthia Buhik and the entire amazing team from the UNC. That project is to answer the question when and how the mental health challenges emerge, and eventually, how we can utilize technology to provide the timely intervention to the patients who need the help the most.

[00:05:06.930] – APS’s Özge Gürcanlı Fischer Baum

This is a really great study that you did. It is about the developmental trajectory of protection There are negative factors, eating disorders, and also there is technology involved. We are going to go step by step. Can you give us an overview of what binge type eating disorders like bulimia nervosa bulimia nausea and binge eating disorder are? What are the treatment outcomes like for them?

[00:05:36.870] – Qinxin Shi

Both bulimia nausea and binge eating disorder are considered as two types of eating disorder behaviors. The difference is bulimia nausea is characterized by an episode of binge eating, but followed by the compensatory behaviors like vomiting or excessive exercise. But in contrast, binge eating disorder, not involving those compensatory actions. It’s just the terminology differences. Both of them are binge eating, but the only difference is whether you’re follow with compensatory behavior or not. So bulimia, follow with the compensatory behavior, vomiting, exercise, and fintech eating disorder does not follow with compensatory behavior. So the both two types of The eating user behavior have serious physical and emotional consequences. A lot of amazing literature already support us on such a statement. So currently, the most effective treatment is cognitive behavior therapy, which is having short term as CBT. But even with the complete of the CBT, CBT treatment, evidence show that still up to 50% of individuals with either bulimia or other binge eating disorder, continue engaging in unhealthy eating disorder behaviors, unfortunately.

[00:07:08.130] – APS’s Özge Gürcanlı Fischer Baum

Yeah, thanks for those definitions. Your study used passive data, and specifically using technology, you looked at heart rate to monitor individuals with these disorders. Could you explain how these technologies work and why they were chosen for this research?

[00:07:28.690] – Qinxin Shi

Yeah, this is such That’s a great question. You’re right on the target for the key of our study. First, I wanted to define the passive data a little bit more since some of the audience may not very familiar with the term. The word passive, of course, different from traditionally you fill in a questionnaire. The passive data, which means you do not actively involving with filling in the answer for the questions. For example, if you are wearing Apple Watch, if you’re bringing a phone with you, so it will collect your person’s heart rate and physical activity data, steps data. You don’t need to do things and they will passively collecting the data from you. That is the passive sensing data. For our study, especially for the eating disorder behavior, we particularly take the heart rate and physical activity. Sometimes we refer it as ectography, sometimes we refer it as steps data, but they all can change interchangeably. Sometimes I would say physical activity, sometimes I would say ectography, I believe sometimes I will see steps, but I just want to be clear to the… Some of the standards, they’re all the same in our study. But to the original question, why we pick heart rate and steps data?

[00:08:57.280] – Qinxin Shi

There are two main reasons. The first reason is because the empirical or theory of support of the associations between heart rate steps with eating disorder or binge eating behavior. The previous literature already show that individuals with binge eating behavior display different patterns of a heart rate reactivity and heart rate variability in response to stress. Similarly to the physical activity or the steps indicator as well, showing that patients with eating disorders or with eating disorders behaviors usually have a lower physical activity compared to the healthy peers. But overall, I don’t want to go in-depth with all the literature stuff, but I want to mention that there’s a great amount of literature support us to using heart rate and steps data as an indicator to examine situations with even disorders. Another big reason is the place of most heart rate and physical activity, like steps, are the most universally collected phenotyping by using the variable devices. For example, you always collect the calories the patients are consuming, but without have the participants actively typing in exactly what the calories the person’s intake. You don’t device to passively collect that. Even though there may be a better indicator, but because the heart rate and steps with the current advance of the technology, those two are the most commonly and the universally used type of phenotypes with the variable devices like Apple Watch.

[00:10:53.010] – Qinxin Shi

Because of this widespread availability, it really increases the potential for the findings to be generalized across diverse populations with eating disorder. Also, thinking about the dead in the heart rate and the steps data from the Apple Watch, from iPhone, they’re non-invasive, they’re minimized, the use point further, and their commonality in everyday devices. I think it’s really talk about the practical reason, and that’s another main reason we kick the heart rate and stress data as our indicator to study the associations with eating disorders.

[00:11:32.460] – APS’s Özge Gürcanlı Fischer Baum

Yeah, it is a brilliant idea, and you followed your participants over 30 days. Interestingly, you found no significant difference in the average heart rate or step count before a binge event. Could you explain why this might be?

[00:11:49.380] – Qinxin Shi

Yeah, so this is a really interesting result. If we are emphasizing on the value of the heart rate and steps data. What I mean value is thinking about a particular number. For instance, 85, 65, and before the bench events. Then if you think about the steps value, for example, the cumulative steps value right before the bench, right before the two-hour window for the bench, if you’re thinking about two values in there, we do not find differences in terms the value differences in terms of the bench window and the match non-bing window. Let’s say if you binge at 2:00 PM today, and because we select two-hour window as our study window, and so during that two-hour window for the Binge event, your hard rate is one value and steps is one value. With all the matched control period, which you do not bench, the values are not really different from each other. Which means if you really have to have a number to differentiate bench and non-benching events, we do not find that magical number.

[00:13:20.180] – APS’s Özge Gürcanlı Fischer Baum

Right. Yeah. No, that is surprising, right? Also, you looked at the stability of step counts, and it is in fact a potential indicator of approaching binge events. What do you mean by stability, and how can tracking this lead to more accurate predictions?

[00:13:43.350] – Qinxin Shi

Yeah, that’s another amazing That’s another amazing question. Actually, I earlier said, there’s no magic one particular value to help us to differentiate the binge and non-binging periods. In our study, we actually find that it is the temporal pattern of the heart rate and steps that start to becoming stabilizing when approaching the bench events that differentiate to the non-bench time window. Let me further elaborate this a little bit so you can better understand what I mean by this. First, let me talk a bit more about the concept of the stability. This go hands-with the concept of the dynamic system theory as well. Imagine we’re all leaving in this very We are a human beings. We are very complex. We are all living in this crazy world. Everything constantly happen to us, knocking us around. There is predictive events, unpredictable events. They’re all around us. I’m going to use my amazing mentor, John’s example to describe this. We all sign our signatures. Let’s say this is 6 AM in the morning, you’re fresh, you have good sleep, you sign your signature. In contrast, now it’s already 1 AM, you’re tired after a full day work, you want to go back, but you still need to sign your signature.

[00:15:30.230] – Qinxin Shi

But consider, sometimes you are really happy because you sign your amazing offer and you sign your signature. For their other events, you sign something make you really, really sad or desperate. You sign your signature. Every time, at best, you’re going to sign your signature differently because you cannot find the exact same signature. I mean, exactly, exactly the same. What I mean is you’re going to sign But overall, you can identify it is your signature. Other people can identify it is your signature. It is that stability. It is that core signature that is associated with your signature that’s always going to emerge out. No matter when, where, and under what context you sign your name, there’s always going to be a stability within your signature that’s going to emerge so people can identify. The concept of the stability in here is by using heart rate and steps as an indicator to help us better understand when the beam to event is going to approach. The heart rates and steps are followed their own algorithm. It’s a silica pattern, right? Go up and down, up and down. But it has a stability, which means when you’re running, of course, your heart rate is going to go up.

[00:17:05.700] – Qinxin Shi

Your steps go up, but you are not always running. You’re going to come down into a stability. It’s your set point. It is your comfort zone. Sometimes I’m understanding this as your comfort zone. You’re going to come back. It’s your stability. It’s not going to be off the air forever. Even you’re under stress, this thing’s on the extreme. This thing happened, this thing happened. You off your set point, but you always come back. That is concept of the stability. The reason we really care about the stability is we want to know why someone bench eating? Is the pattern, is the temporal pattern becoming… Does something break the stability? Which means it’s like a sudden shift, so they’re going to move into to a different stage. With that so, that can be used as a warning signal for capture that certain trend and it can predict when the bench is coming.

[00:18:14.180] – APS’s Özge Gürcanlı Fischer Baum

Yeah, that’s a great answer. Thank you very much. Then I have a question about the digital technologies in the future. The idea of using passive sensor data such as heart rate and step counts opens up new possibilities, right? How do you see this approach being applied to other disorders or conditions in the future?

[00:18:40.090] – Qinxin Shi

Well, I think this is another amazing question. I think It can open with many other conditions like depressive symptoms, anxiety, so where the physical patterns such as heart rate variability and even sleep determinants can use as a signal to further predict the symptoms. The type of the real-in-time monitoring can help us further intervene earlier and more effectively in a wide range of the behavior health conditions. In our paper, we really focus only on the eating disorder behavior, but I think it’s for sure can widely apply to many others.

[00:19:28.550] – APS’s Özge Gürcanlı Fischer Baum

So once we I see that stability, in fact, it can be a warning sign maybe for individuals. Finally, is there a key takeaway or final thought that you would like to leave with our listeners?

[00:19:45.190] – Qinxin Shi

Yeah, of course. I think the main message I really wanted to share is I got really excited because we are now living in the The world that technology offers us so much more opportunity to study our study psychology from a nonlinear perspective. Before, we often time assuming things are associated in a linear way, but actually it can be more complex than that. Our human being complex, our environment is really complex. Back to the by utilizing the heart rate and steps as an indicator. By using this dynamic system perspective, we can really allow the heart rate and steps holding hands and thinking about how they are dancing together. And using this dancing as a temporal pattern and see how it will change before the binge eating behavior emerge. I think it’s empowered us with right now the the utilizing of technology and intensive collected data to help us unpack the nonlinear dynamics associations among the constructs we usually study. Also in a traditional way, we cannot really control a lot of factors. We open selective variables we’re going to study. But by considered the dynamic system perspective, we’re thinking the stability, which means no matter how the environment is going to move us around, We’re always going to go back to the stability.

[00:21:32.390] – Qinxin Shi

That’s to help us to, by utilizing this theory, it helped us to control a lot of factors that we all generally cannot really control in our traditional way to study in the study concepts in a linear perspective.

[00:21:47.950] – APS’s Özge Gürcanlı Fischer Baum

Yeah. So your data gives us a perspective about how we can think about each individual’s stability point, in other words, their comfort zone, and then we can use it for more individual to visualize help and intervention. Qinxin, this was such a pleasure. Thank you very much for joining us today.

[00:22:07.030] – Qinxin Shi

Oh, thank you so much for having me. It’s such a pleasure.

[00:22:13.780] – APS’s Özge Gürcanlı Fischer Baum

This is Özge Gürcanlı Fischer Baum with APS, and I have been speaking to Qinxin Shi from the University of Utah. If you want to know more about this research, visit psychologicalscience.org. Would you like to reach us? Send us your thoughts and questions at [email protected].