Correspondence of parasympathetic-mediated heart rate variability derived from electrocardiogram and photoplethysmography signals in ethnically diverse adolescents

Abstract

The reliable and valid assessment of heart rate variability (HRV) is important to understanding autonomic functioning in youth. However, use of electrocardiogram (ECG) signals to derive estimates of parasympathetic-mediated HRV may present logistical barriers in applied settings. Thus, this study investigated the concordance between high-frequency HRV [HF-HRV] and root mean square of successive differences [RMSSD] derived from ECG and photoplethysmography (PPG) signals during a video baseline, resting baseline, and mental arithmetic among an ethnically diverse sample of ninety-six youths (Mage = 13.84; 51% male; 62.5% ethnic minority). Results showed moderate to almost perfect agreement between PPG- and ECG- derived HF-HRV and RMSSD for video and resting baselines (Lin's correlations ranged from 0.93 to 0.99) and limits of agreement (LoA) ranging from −0.48 to 0.58 (HF-HRV) and − 11.37 to 9.32 (RMSSD). Conversely, we found poor to moderate agreement for the mental arithmetic task (Lin's correlations ranged from 0.88 to 0.91) and LoA ranging from −0.68 + 0.94 (HF-HRV) and − 17.58 + 20.69 (RMSSD), though we did find ethnic minority youth had higher and moderate Lin's correlations (0.93 to 0.94). Overall, there was a bias towards higher HF-HRV and RMSSD values with PPG. Findings suggest that PPG-derived HF-HRV and RMSSD may be viable alternatives for ECG in baseline conditions, but tasks requiring movement or eliciting mild stress responses may result in less than perfect values and missing data patterns. It is imperative that future studies replicate these findings in other ethnically diverse youth samples and expand to younger children and applied settings.