0103 Feasibility of examining component-specific effects of yogic breathing on heart rate variability during sleep: A three-arm pilot RCT

Abstract

Introduction: Wearable devices and mind‐body interventions (MBIs) continue to receive widespread interest as tools for improving sleep. This study investigated the feasibility of using an automated electronic survey system and wearable heart rate (HR) monitor in the context of a fully remote clinical trial study to produce detailed measures of par‐ticipant adherence, daily sleep quality, and associations with physio‐ logical outcomes captured by wearable devices. Methods: Eighteen healthy participants (age 18‐30yrs, 12 female) were randomized to one of three 8‐week long interventions: slow‐ paced breathing (SPB, N=5, 24.6 ± 2.1 years, 4 female), mindfulness (M, N=6, 23.7 ± 3.7 years, 4 female), or yogic breathing (SPB+M, N=7, 24.3 ± 3.1 years). Participants completed two weeks of daily intervention‐specific training with 20‐min guided practice, and sub‐sequent tasks including experimental stress induction. Participants started a 24‐hour HR recording using a Polar H10 chest strap on the night prior. Then, participants were instructed to repeat their assigned intervention practice daily, using a guided audio similar completing a detailed practice log. HR interbeat interval data were examined with spectral analysis using full spectrograms for inspec‐ tion of timecourse and frequency‐specific patterns in both the noc‐ turnal recordings and daily practice sessions. Results: Participants completed an average of 75% of daily prac‐ tice sessions across the 8‐week intervention period (SPB: 77%, M: 65%, SPB+M: 77%). An automated procedure was developed to analyze and visualize the timecourse of HRV‐derived breathing patterns in the 754 completed practice sessions and 36 nocturnal recordings. The three groups were then successfully distinguish‐able based on breathing rates and mindfulness questionnaires. Nocturnal HR recordings demonstrated visually identifiable pat‐terns of interindividual variability and intraindividual consistency. Statistical analysis is ongoing to further characterize these patterns. Conclusion: These findings support feasibility for a fully remote, semi‐automated clinical trial study assessing component‐specific effects of these MBIs on sleep, including detailed spectral analysis of high‐quality HR data. Future studies would benefit from exam‐ining scalability of this type of study design with wearable physi‐ology in both clinical and nonclinical populations.