Discovering Digital Biomarkers of Panic Attack Risk in Consumer Wearables Data

Area of Science:

• Psychiatry
• Digital Health
• Wearable Technology

Background:

• Panic attacks affect over 10% of US adults, significantly impairing quality of life due to their unpredictable nature.
• Current management strategies lack advance warning systems, highlighting a need for predictive tools to aid individuals in managing or preventing panic attacks.
• While self-reported mood and social media data show some predictive ability, objective physiological, behavioral, and environmental measures for panic attack prediction remain underexplored.

Approach:

• This study investigated the predictive capability of objective data from consumer wearable devices for next-day panic attacks.
• Data were collected from 38 individuals experiencing regular panic attacks across the US over 28 days.
• Participants completed daily surveys on panic attacks and provided access to data from their Apple Watches.

Key Points:

• Preliminary findings indicate that objective measures, specifically louder ambient noise and higher resting heart rate, are associated with an increased likelihood of experiencing a panic attack the following day.
• These objective physiological and environmental markers offer potential for early detection of heightened panic attack risk.
• The study demonstrates the feasibility of using passively collected data from consumer wearables for panic attack prediction.

Conclusions:

• Objective data from consumer wearables, including ambient noise and resting heart rate, show promise in predicting next-day panic attacks.
• These findings open new possibilities for developing proactive interventions and improving the management of panic disorder.
• Leveraging wearable technology could lead to personalized, preventative strategies for individuals with panic attacks.