A comparison of bilaterally wrist-worn accelerometers on measures of free-living physical activity in adolescents

Abstract

With the advancement of technology, the use of accelerometry-based wrist-worn activity monitors (AM) to measure free-living physical activity (PA) has become extremely popular. Whereas the influence of contralateral wrist AM placement has been evaluated in adults, no research has investigated the influence of wrist-worn AM location (dominant [D] versus non-dominant [ND] or left [LW] versus right [RW]) on measurements of free-living PA in adolescents. Purpose: The purpose of the current study was to evaluate the influence of wrist-worn AM wear location (D versus ND and LW versus RW) and AM processing algorithm (single (1R) versus double (2R) regression) on measures of free-living PA in adolescents 12-17 years of age. Methods: Eight boys (Mean±SD: 14.3±1.9 yrs, 53.9±13.6 kg, 160.5±14.5 cm, 20.6±2.3 body mass index [BMI]) and 12 girls (13.9±1.6 yrs, 56.1±9.9 kg, 160.3±7.4 cm, 21.4±2.6 BMI) participated in the study. Each subject wore two AMs for 7 consecutive days (24 hr/day), with one AM attached to the dorsal side of each wrist using locking plastic wrist bands. Subjects were instructed to engage in their usual daily activities. After downloading the raw AM data to a computer, previously validated algorithms (1R and 2R) were used to transform and summarize the data into counts (counts/day), activity energy expenditure (AEE; kcals/day), and duration of time (T; minutes/day) for 1, 5, and 10 minute bout durations within light (L) and moderate-to-vigorous (MV) intensity categories for AMs (D, ND and for LW, RW). Dependent variables (CNT, AEE, T) were compared using two-factor multivariate repeated measures ANOVA (alpha=0.05) for data processed by both algorithms separately. Scheffe's post-hoc analyses (alpha=0.01) were performed for all pairwise comparisons when indicated by the ANOVA. Results: No significant differences existed between mean values for dependent variables for any bout duration within L or MVPA intensity between monitors (D versus ND or between LW versus RW) or processing algorithm. Conclusion: These data indicate that wrist-worn AM location has no significant influence on PA outcome variables in adolescent under free-living conditions. Researchers can now allow participants the freedom to choose which wrist they would prefer the monitor be placed when being assessed.