Data from Garmin fitness trackers aid researchers in panic attack

October 9, 2025
Physiological data from a fitness tracker — in combination with questionnaires and environmental data — could be used to predict a panic attack, according to studies carried out by the Graduate Institute of Biomedical Electronics and Bioinformatics (BEBI) at National Taiwan University (NTU). Three cohort studies, ranging in duration from 1 to 3 years, used the Garmin vívosmart® series to monitor patients. The hope is that the research may help clinicians to provide more appropriate, timely and personalized treatment for panic disorder patients.
Panic Attack Prediction
“The combination of smartwatches and apps showed promising results. The data from these devices helped make it possible to predict a panic attack a few days in advance, giving people a heads-up to take action,” Dr. Chan-Hen Tsai, researcher at NTU, says about the completion of these studies. “The apps also helped users reduce their symptoms and feel more in control, demonstrating that technology can be a powerful tool in managing panic disorder and preventing attacks. In simple terms, these studies show that with the right tools, technology can help people with panic disorder by predicting panic attacks and offering support to manage their condition more effectively day-to-day.”
A panic attack is an intense form of anxiety accompanied by multiple somatic presentations. Panic attacks are known to be triggered by psychological stress or specific situations that induce agoraphobia, the fear of being unable to escape. However, few studies have predicted recurrent panic attacks using real-life data.
The first research project carried out by BEBI gathered data from 59 participants in a yearlong study. Information on environmental conditions and emotional state was fed back to the team using a mobile app. The Garmin vívosmart 4 automatically collected daily physiological metrics, including heart rate, activity and sleep. According to Dr. Tsai, Garmin fitness trackers were chosen as “promising tools” that patients can wear all day, with up to 7 days of battery life. Many of the participants gave positive feedback after learning to self-monitor their emotional and physiological states through fitness trackers and regular questionnaires.
The study concluded that, with the combination of data, it is possible to predict panic attacks up to 7 days before they happen. A prediction model could help clinicians and patients monitor, control and carry out early intervention for recurrent panic attacks in hospital settings and in patients’ daily lives1.
Further Studies and Promising Outcomes
Since the initial study, two more studies were conducted over 2 and 3 years, with one focusing on how sleep and physical activity affect people with panic disorder. This study aimed to predict symptoms up to a week in advance. Participants wore Garmin vívosmart devices to track their daily sleep, activity levels and heart rate. This collected data combined with information from questionnaires and interviews was used to power a machine learning model to predict panic attacks. Researchers were able to predict when someone might experience a panic attack in the following week with more than 92% of accuracy2. This means that, in most cases, they could successfully anticipate when symptoms were likely to get worse.
Their most recent study explored how a mobile-assisted case management program (including the prediction model) could help people with panic disorder reduce their symptoms and improve the quality of care they receive. Participants used a mobile app for extra support, in addition to their regular treatment, and stayed in touch with health care professionals remotely for more personalized treatment. As a result, participants who engaged in the mobile-assisted case management reported feeling more in control of their symptoms, more self-aware and better supported by others3.
These three studies are very helpful not only for individuals with panic disorder but also for clinicians. Using machine learning models, the researchers focused on key factors that contribute to recurring panic attacks. They identified the ideal ranges for sleep duration, exercise levels and heart rate to help prevent attacks. Specifically, they found that an average heart rate of 72-87 bpm, a maximum heart rate of 100-145 bpm and a resting heart rate of 55-60 bpm are beneficial. In addition, daily activity goals such as climbing more than nine floors, getting between 6 hours 23 minutes and 10 hours 50 minutes of total sleep, at least 50 minutes of deep sleep and keeping awake time under 53 minutes can be helpful. Using data from fitness trackers and smartphone apps, clinicians can offer personalized lifestyle adjustments to improve treatment outcomes for panic disorder.
To learn more, visit www.garmin.com/third-party-studies-overview for additional…