Impact of accelerometer data processing decisions on the sample size, wear time and physical activity level of a large cohort study

Table 3 Physical activity and sedentary behavior estimates using data from vertical axis only and from vector magnitude assessment, Women's Health Study, 2011-2013

Assessment method:	Sedentary^a(min/d)		Light-intensity^b(min/day)		MVPA^c(min/day)
	Vertical axis	Vector magnitude	Vertical axis	Vector magnitude	Vertical axis	Vector magnitude
Detailed-log	598.7 (537.3, 656.9)	518.5. (453.4, 585.3)	277.2 (227.6, 330.7)	343.8 (283.9, 404.0)	8.9 (2.7, 21.1)	19.7 (8.0, 37.7)
Limited-log + Troiano [5]	530.1 (480.1, 578.6)	474.6 (417.0, 529.6)	290.4 (242.2, 342.7)	358.7 (300.2, 418.4)	9.6 (3.0, 22.3)	20.7 (8.7, 38.7)
Limited log + Choi [11, 19]	581.6 (521.1, 639.8)	506.0 (439.2, 570.9)	287.9 (238.7, 340.7)	357.4 (297.4, 417.6)	9.4 (3.0, 22.1)	20.6 (8.7, 38.6)

Note: All differences between vertical axis and vector magnitude (VM) were statistically significant P < 0.001. Includes all participants with at least 1 valid day (≥10 h wear-time) for particular assessment method.
^aSedentary time is defined as time during which the accelerometer registers vertical cpm <100 [21] and VM cpm <200 [18].
^bLight-intensity physical activity time is defined as time during which the accelerometer registers vertical cpm between 100 and 1951 [12] and VM cpm between 150 and 2690 [22].
^cMVPA time is defined as time during which the accelerometer registers vertical cpm > = 1952 [12] and VM cpm > = 2691 [22].

ISSN: 1471-2458