PII-025 - THE POTENTIAL OF SMARTWATCH-DERIVED CARDIAC MEASUREMENTS AS BIOMARKERS OF THERAPEUTIC RESPONSE IN CHILDREN WITH DISRUPTIVE BEHAVIOR DISORDER

K. Delaney¹, M. Saliba², J. Shekunov², M. Romanowicz², P. Croarkin², A. Athreya²; ¹Mayo Clinic, Rochester, MN, United States of America, ²Mayo Clinic, Rochester, MN, United States.

Background: Ubiquitous commercial-grade smartwatches provide measurements such as resting heart rate (RHR) and heart rate variability (HRV) that are known to be indicative of autonomic nervous system functioning. This study aimed to examine the associations of these biomarkers with therapeutic response in children diagnosed with a disruptive behavior disorder and treated with behavioral therapy in conjunction with appropriate stimulant medication for children with comorbid attention deficit/hyperactivity disorder (ADHD).

Methods: This study utilized data from the Mayo Clinic PISTACHIo clinical trial, in which 50 children with disruptive behavior were treated with 12 sessions of Parent-Child-Interaction-Therapy while wearing a smartwatch. Child behavior severity was measured using the Eyberg Child Behavioral Inventory (ECBI), and RHR (bpm) was collected using Garmin VivoSmart 4 watches. HRV was calculated as the SDANN, or standard deviation of the average NN intervals, in milliseconds. A linear regression was conducted to determine the relationship between percent change in either RHR or HRV measures and percent change in ECBI scores from Pre to Post assessments. Among the covariates included were child sex, age, parent demographics, baselines for ECBI scores, and cardiac measurements. The interaction between cardiac measurements and stimulant use was assessed, controlling for the interaction between ADHD diagnosis and stimulant usage.

Results: Reductions in RHR were significantly associated with improvements in child behavior (i.e. decreased ECBI severity; Estimate = 5.14, SE = 2.06, p = 0.03), and stimulant use did not significantly impact this relationship (p = 0.40). Changes in HRV was not significantly associated with a change in ECBI in the whole sample (Estimate = 0.57, SE = 0.32, p = 0.10), but the interaction between HRV change and stimulant use was significant (estimate = -0.75, SE = 0.33, p = 0.042), demonstrating a positive relationship between HRV reduction and ECBI improvement in the subgroup taking stimulants.

Conclusion: This work suggests that changes in RHR may be promising indicators of therapeutic response in pediatric disruptive behavior disorders. In addition, this study demonstrates the need for including pharmacological exposures (e.g. stimulant use) when interpreting HRV changes as a treatment response biomarker.