JMIR Publications

ABSTRACT

Background:

Measuring heart rate variability (HRV) via wearable photoplethysmography (PPG) sensors like smartwatches is gaining popularity for monitoring of many health conditions. However, missing data caused by insufficient wear compliance or signal quality can degrade the performance of health metric or algorithm calculations. Research is needed on how to best account for missing data and to measure data validity to assess the accuracy of metrics derived from wearable sensor data.

Objective:

To evaluate the influence of missing data on HRV metrics collected from smartwatches both at rest and during physical activity in real-world settings and to evaluate HRV agreement and consistency between wearable PPG and gold standard wearable electrocardiogram (ECG) sensors in real-world settings.

Methods:

Healthy participants were outfitted with a smartwatch with a PPG sensor that collected high resolution inter beat interval data to wear continuously (day and night) for up to 12 months. New data sets were created with various amount of missing data and then compared to the original (reference) data sets. 5 min windows of each HRV metric (median RR interval [RR], standard deviation of the RR interval [STDRR], root mean square error of the RR interval [RMSDRR], low frequency power [LF], high frequency power [HF], and the ratio of LF to HF power [LF/HF]) were compared between the reference dataset and the missing data sets (~10%, ~20%, ~35%, and ~60% missing data). HRV metrics calculated from the PPG sensor were compared to HRV metrics calculated from a chest-worn ECG sensor.

Results:

Median RR, STDRR, and RMSDRR remain stable at rest until at least ~35% data degradation (P < 0.05). LF, HF, and LF/HF are less resilient to missing data both at rest and during light activity and are unstable even at ~10% data degradation (P < 0.05). Median RR intervals calculated from PPG sensors had moderate agreement (intraclass correlation coefficient [ICC] = 0.585) and consistency (ICC = 0.589) and LF had moderate consistency (ICC = 0.545) with ECG sensors. Other HRV metrics demonstrated poor agreement (ICC = 0.071 – 0.472).

Conclusions:

This study describes methodology for the extraction of HRV metrics from PPG sensor data that resulted in stable and valid metrics while utilizing the least amount of available data. While smartwatches containing PPG sensors are valuable for remote monitoring of patients, future work is needed to identify best practices for using these sensors to evaluate HRV in medical settings.