Congruency of Motion Sensors to Detect Change following a Sedentary Behavior Intervention

Introduction Time spent in sedentary behavior (SB) has deleterious effects on health. As a result, there is a strong scientific need to evaluate methods to assess SB.

Purpose To determine responsiveness of two motion sensors to detect change in free-living, occupational SB during an intervention to decrease sitting activity.

Methods Adults who spent > 60% of their working day sitting were recruited to participate in an intervention to reduce SB at work. SB was assessed using two accelerometers, an Actigraph GTX3 (AG) worn on a belt, at the midline of the right thigh, and an activPAL (AP) affixed to the middle of the right thigh. SB was assessed during working hours for three consecutive days (baseline) and during the same three days the week following while undergoing the intervention (post). Data from both motion sensors were time-matched to allow direct comparison. SB was determined from time spent sitting/lying by AP and time below 100 cts/min, using 60-s epochs by AG. Pearson correlation coefficients were calculated to determine strength of association between the measures. Difference between AP and AG measures of SB at baseline and post were compared with paired samples t-test. Baseline-post intervention difference scores were compared using paired samples t-tests.

Results Sixty-seven adults (45.3 ± 11.2y; 29.2 ± 7.7 kg/m2) completed the intervention. At baseline, participants spent approximately six of their eight working hours in sedentary activities. Time spent in SB as assessed by AG and AP were correlated (baseline: r=0.64, p<0.001; post: r=0.5, p<0.001). Measures of SB differed between AP and AG at baseline (AP 355.6 ± 11.0 min; AG 386.8 ± 8.2 min; p<0.001) and post assessment (AP 337.3 ± 10.8 min; AG 368.8 ± 7.7 min; p=0.002). Baseline-post AP difference scores (16.5 ± 6.8 min) were not significantly different than baseline-post AG difference scores (16.2 ± 5.6 min, p=0.56).

Conclusion Results reveal that both the AP and AG were able to assess responsiveness to change following a SB intervention to a similar degree. However, the AP and the AG significantly differed in their baseline and post assessment of time spent in SB.

This work was supported by funding source 1-UL1-RR031973.