Photoplethysmographic Heart Rate Accuracy During Exercise In Temperate And Hot Environments

Increased skin blood flow (SkBF) and local sweat rate (LSR) during exercise heat stress may decrease the accuracy of heart rate (HR) from wearable devices using photoplethysmography by interfering with light reflection at the epidermis. PURPOSE: To test the hypothesis that heat stress reduces the accuracy of photoplethysmographic HR measurement during exercise. METHODS: Twenty-seven participants (52% female; mean ± SD, age = 25 ± 5 y, body mass = 76.2 ± 14.0 kg) cycled in 22 °C (TEMP) and 35 °C (HOT) at progressively increasing intensities (50%, 60%, 70%, 80% of age-predicted maximal HR) over 40 min while wearing 3 photoplethysmographic devices (Apple Watch Series 8, Garmin vivosport 5, SlateSafety BAND V2) and a criterion device (Polar H10). HR, SkBF, and LSR were recorded during rest, exercise, and recovery. Device performance was evaluated using equivalence testing (equivalence zone, ±5 bpm), and differences in mean absolute error of HR measurement (MAEHR) in TEMP vs. HOT were analyzed using a two-way (condition × intensity) analysis of variance. RESULTS: Based on 90% and 95% confidence intervals (CI), device HR was both equivalent and not different from criterion HR for Apple in both conditions at all intensities. Garmin HR was equivalent but statistically different from criterion HR only during recovery in HOT (95% CI [0.5, 2.9]). SlateSafety HR was not equivalent during HOT at rest (90% CI [-2.6, 5.3]) and very light intensity (90% CI [-6.5, 2.8]), and during TEMP at vigorous intensity (90% CI [-6.4, 2.8]) but was not different from criterion HR. MAEHR was not different between conditions for Garmin (TEMP, 1.7 ± 2.3 bpm; HOT, 1.4 ± 0.9 bpm; P = 0.13; η2G = 0.01) or SlateSafety (TEMP, 3.8 ± 7.2 bpm; HOT, 3.2 ± 12.5 bpm; P = 0.88; η2G < 0.01) but was lower during HOT for Apple (TEMP, 0.9 ± 1.5 bpm; HOT, 0.7 ± 0.4 bpm; P = 0.02; η2G = 0.04), despite greater SkBF (TEMP, 35 ± 36 AU; HOT, 67 ± 46 AU; P < 0.001) and LSR (TEMP, 0.13 ± 0.17 mg·cm-2·min-1; HOT, 0.27 ± 0.23 mg·cm-2·min-1; P < 0.001) during HOT. CONCLUSION: Heat exposure did not increase error in photoplethysmographic HR measurement, although SkBF and LSR were elevated. Supported by: This study was supported by an equipment grant from SlateSafety (Atlanta, GA).