Table 1 Validity of measuring HR with the Apple Watch during and in recovery from walking, jogging, and running.
| Exercise Mode | Left Wrist | Right Wrist |
|---|---|---|
| Walking | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 95 (14) | 95 (14) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 94 (13) | 95 (14) |
| Standardised mean bias (90% CI) | −0.03 (−0.11 to 0.06) | 0.01 (−0.09 to 0.11) |
| Standardised typical error of the estimate (90% CI) | 0.23 (0.18 to 0.32) | 0.26 (0.21 to 0.36) |
| Correlation coefficient (90% CI) vs criterion (Polar) | 0.97 (0.94 to 0.99) | 0.97 (0.93 to 0.98) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 0 (−6 to 6) | 0 (−8 to 8) |
| In recovery from walking | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 83 (14) | 83 (14) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 88 (14) | 89 (13) |
| Standardised mean bias (90%CI) | 0.32 (0.15 to 0.49) | 0.37 (0.22 to 0.52) |
| Standardised typical error of the estimate (90% CI) | 0.46 (0.36 to 0.63) | 0.41 (0.32 to 0.56) |
| Correlation coefficient (90% CI) vs criterion (Polar) | 0.89 (0.78 to 0.95) | 0.92 (0.83 to 0.96) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 5 (−8 to 18) | 5 (−7 to 17) |
| Jogging | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 133 (15) | 133 (15) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 132 (16) | 133 (16) |
| Standardised mean bias (90% CI) | −0.03 (−0.18 to 0.11) | 0.01 (−0.15 to 0.16) |
| Standardised typical error of the estimate (90% CI) | 0.37 (0.29 to 0.50) | 0.40 (0.32 to 0.55) |
| Correlation coefficient (90% CI) vs criterion (Polar) | 0.93 (0.86 to 0.97) | 0.92 (0.84 to 0.96) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 1 (−19 to 21) | 1 (−19 to 21) |
| In recovery from jogging | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 118 (19) | 118 (19) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 131 (19) | 130 (21) |
| Standardised mean bias (90% CI) | 0.59 (0.32 to 0.86) | 0.53 (0.30 to 0.77) |
| Standardised typical error of the estimate (90% CI) | 0.72 (0.57 to 0.98) | 0.61 (0.49 to 0.84) |
| Correlation coefficient (90% CI) vs criterion (Polar) | 0.71 (0.47 to 0.86) | 0.80 (0.61 to 0.90) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 12 (−18 to 42) | 11 (−16 to 38) |
| Running | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 155 (17) | 155 (17) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 157 (18) | 157 (18) |
| Standardised mean bias (90% CI) | 0.11 (−0.12 to 0.35) | 0.11 (−0.09 to 0.31) |
| Standardised typical error of the estimate (90% CI) | 0.61 (0.48 to 0.83) | 0.53 (0.42 to 0.72) |
| Correlation coefficient (90% CI) vs criterion (Polar) | 0.81 (0.62 to 0.91) | 0.86 (0.72 to 0.93) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 2 (−21 to 25) | 2 (−18 to 22) |
| In recovery from running | ||
| Mean (SD) of criterion (Polar) (beats.min −1 ) | 144 (22) | 144 (22) |
| Mean (SD) of practical (Apple Watch) (beats.min −1 ) | 156 (19) | 155 (20) |
| Standardised mean bias (90% CI) | 0.50 (0.25 to 0.75) | 0.44 (0.19 to 0.69) |
| Standardised typical error of the estimate (90% CI) | 0.67 (0.54 to 0.92) | 0.67 (0.53 to 0.92) |
| Correlation coefficient (90%CI) vs criterion (Polar) | 0.76 (0.54 to 0.88) | 0.75 (0.53 to 0.88) |
| Mean bias (95% limits of agreement) (beats.min −1 ) | 12 (−18 to 42) | 10 (−20 to 40) |
CI: Confidence interval.