Skip to main content

Table 2 Agreement of CHAP-child and the ActiGraph 100 cpm cut-point with activPAL for estimating various participant-season level sedentary pattern variables (N = 127 participant-seasons from 65 participants)

From: CHAP-child: an open source method for estimating sit-to-stand transitions and sedentary bout patterns from hip accelerometers among children

  Total sedentary time (min/day) Breaks in sedentary time (num/day) Time spent in bouts ≥ 30 min (min/day) Mean bout duration (min) Usual bout duration (min) Alpha (unitless)
Descriptive statistics
 activPAL, Mean (SD) 455.1 (88.4) 82.5 (19.3) 111.5 (54.9) 5.74 (1.46) 14.68 (4.26) 1.40 (0.04)
 CHAP-child, Mean (SD) 483.1 (89.0) 81.5 (17.4) 121.8 (62.3) 6.12 (1.45) 15.13 (3.92) 1.39 (0.03)
 ActiGraph 100 cpm, Mean (SD) 364.3 (79.0) 91.1 (14.4) 28.1 (24.6) 4.02 (0.70) 6.53 (1.82) 2.13 (0.16)
Agreement between CHAP-child and activPAL
 Bias, Mean (SD) 28.1 (31.7) -1.0 (8.8) 10.3 (30.1) 0.37 (0.79) 0.46 (2.17) -0.01 (0.02)
 Mean absolute error 30.6 6.8 23.4 0.61 1.48 0.02
 Mean absolute percent error 6.7% 8.2% 21.0% 10.6% 10.1% 1.4%
 Spearman correlation 0.93 0.87 0.85 0.84 0.84 0.76
 Concordance correlation 0.89 0.88 0.86 0.83 0.85 0.72
Agreement between ActiGraph 100 cpm and activPAL
 Bias, Mean (SD) -90.7 (58.2) 8.6 (16.1) -83.3 (50.0) -1.72 (1.13) -8.15 (3.58) 0.73 (0.15)
 Mean absolute error 92.2 14.9 83.3 1.75 8.15 0.73
 Mean absolute percent error 20.3% 18.1% 74.7% 30.5% 55.5% 52.1%
 Spearman correlation 0.75 0.57 0.55 0.62 0.60 0.53
 Concordance correlation 0.48 0.49 0.14 0.24 0.10 0.01
  1. Bias ActiGraph method minus activPAL, CHAP Convolutional Neural Network Hip Accelerometer Posture classification method, CI Confidence interval, cpm Counts per minute, min Minute, num Number, SD Standard deviation