Discovering Digital Biomarkers of Panic Attack Risk in Consumer Wearables Data

Area of Science:

• Psychiatry
• Digital Health
• Wearable Technology

Background:

• Panic attacks significantly impair quality of life, often occurring unexpectedly.
• Current prediction methods rely on self-reported mood, lacking objective measures.
• Digital biomarkers from wearables offer a novel avenue for predicting panic attacks.

Purpose of the Study:

• To explore the predictive capability of digital biomarkers for next-day panic attacks.
• To investigate objective physiological, behavioral, and environmental data for panic attack prediction.
• To assess the feasibility of using consumer wearable sensors for panic attack forecasting.

Main Methods:

• Recruited 38 individuals with frequent panic attacks across the US.
• Collected daily self-reported panic attack data for 28 days.
• Utilized data from Apple Watches, analyzing ambient noise and resting heart rate via Mixed Regressions.

Main Results:

• Higher ambient noise levels were associated with increased likelihood of next-day panic attacks.
• Elevated resting heart rate was also indicative of a higher risk for experiencing a panic attack the following day.
• Preliminary findings suggest digital biomarkers can predict panic attacks.

Conclusions:

• Objective data from consumer wearables may predict imminent panic attacks.
• This predictive capability can inform treatment decisions and preventative strategies.
• Digital biomarkers open new possibilities for managing and preventing panic attacks.