Can a heart rate variability biomarker identify the presence of autism spectrum disorder in eight year old children

Autonomic nervous system (ANS) activity is altered in autism spectrum disorder (ASD). Heart rate variability (HRV) derived from electrocardiogram (ECG) has been a powerful tool to identify alterations in ANS due to a plethora of pathophysiological conditions, including psychological ones such as depression. ECG-derived HRV thus carries a yet to be explored potential to be used as a diagnostic and follow-up biomarker of ASD. However, few studies have explored this potential. In a cohort of boys (ages 8 - 11 years) with (n=18) and without ASD (n=18), we tested a set of linear and nonlinear HRV measures, including phase rectified signal averaging (PRSA), applied to a segment of ECG collected under resting conditions for their predictive properties of ASD. We identified HRV measures derived from time, frequency and geometric signal-analytical domains which are changed in ASD children relative to peers without ASD and correlate to psychometric scores (p<0.05 for each). Receiver operating curves area ranged between 0.71 - 0.74 for each HRV measure. Despite being a small cohort lacking external validation, these promising preliminary results warrant larger prospective validation studies.