Table 1.

Wearable devices dataset characteristics.

	Train set (N $=$ 3115)	Test set (N $=$ 779)
Gender distribution:
Female	1035 (33%)	260 (33%)
Male	2080 (67%)	519 (67%)
Age distribution:
$\le$ 20 years	13 (0.4%)	3 (0.4%)
20–30 years	351 (11.3%)	72 (9.2%)
30–40 years	1057 (33.9%)	247 (31.7%)
40–50 years	961 (30.9%)	269 (34.5%)
50–60 years	556 (17.8%)	148 (19.0%)
$>60$ years	177 (5.7%)	40 (5.1%)
Body mass index categories:
Underweight (BMI $<18.5$)	55 (1.8%)	5 (0.6%)
Normal weight (18.5 $\le$ BMI $<25.0$)	1258 (40.4%)	321 (41.2%)
Overweight (25.0 $\le$ BMI $<30.0$)	1179 (37.8%)	287 (36.8%)
Obese (BMI $\ge$ 30.0)	623 (20.0%)	166 (21.3%)
Devices:
Apple Watch	2912 (93.5%)	726 (93.2%)
Garmin	203 (6.5%)	53 (6.8%)
Number of observed days (mean ± SD)	287±149	288±149
Reference VO2 max (ml/kg/min, mean ± SD)	36.16±6.66	36.18±6.76
Number of VO2 max measurements (mean ± SD)	8.69±5.43	8.96±5.53