The physiological and clinical-behavioral effects of HRV biofeedback in adolescents with autism: a pilot randomized controlled trial

Background Adolescents with autism present lower levels of cardiac vagal modulation in comparison to typically developing peers. These lower values are also associated with psychosocial and behavioral problems. Heart Rate Variability Biofeedback (HRVB) was therefore suggested as an interesting avenue for further exploration since it focusses on the up-regulation of cardiac vagal modulation and has demonstrated positive effects on mental health outcomes. However, scarce evidence was present regarding the effectiveness of HRVB in this population. It was hypothesized that HRVB would increase the level of cardiac vagal modulation in adolescents with autism which would result in positive effects on physiological outcome measures and psychosocial parameters. Furthermore, it was hypothesized that a home-based, non-supervised HRVB training would be feasible in this population. Methods A single-blind, randomized sham-controlled pilot trial was used. During the initial single-blind phase, adolescents with autism performed supervised HRVB (n=24) or a sham training (n=20) for five weeks with one guided session per week and home-based practice (20 min) on the remaining days. In the subsequent follow-up period of five weeks, half of the adolescents of each group received HRVB training at home, in a non-supervised manner, whereas the other subset did not practice during that period. A combination of physiological, cortisol and behavioral data was collected during standardized stress-provoking assessments before (T0) and after each training period (T1 and T2). Results Supervised HRVB resulted in a late increase in cardiac vagal modulation in adolescents with autism. Heart rate increased and cortisol decreased significantly immediately after the supervised HRVB training, but none of these effects remained after the follow-up period of five weeks. None of the outcome measures on psychosocial functioning and self-reported stress revealed a significant change following the supervised HRVB training. The home-based HRVB training was feasible in this population but with a significant decrease in compliance rate. However, adolescents reported significantly lower symptoms of stress after this training period, regardless of the prior training (sham or HRVB). Conclusion HRVB is feasible and effective in adolescents with autism as demonstrated by late-emerging increases in cardiac vagal modulation and less self-reported symptoms of stress. Researchers are encouraged to replicate this study with a larger sample size and to further explore the possible working mechanisms of HRVB.