A comparison of physical activity from Actigraph GT3X+ accelerometers worn on the dominant and non-dominant wrist

The purpose of this study was to evaluate the agreement between several activity measures using raw acceleration data from accelerometers worn concurrently on the dominant and non-dominant wrist. Fifty-five adults (31.9 ± 9.7 years, 26 males) wore two ActiGraph GT3X+ monitors continuously for 1 day, one on their non-dominant wrist and the other on their dominant wrist. Paired t-tests were undertaken with sequential Holm-Bonferroni corrections to compare wear time, moderate-vigorous physical activity (MVPA), time spent in 10-min bouts of MVPA (MVPA10min) and the average magnitude of dynamic wrist acceleration (ENMO). Level of agreement between outcome variables from the wrists were examined using Intraclass Correlation Coefficients (ICC, single measures, absolute agreement) with 95% confidence intervals and limits of agreement (LoA). Time spent across acceleration levels in 40 mg resolution was also examined. There were no significant differences between the non-dominant and dominant wrist for ENMO, wear time, MVPA or MVPA10min. Agreement between wrists were strong for most outcomes (ICC≥0.92) including wear time, ENMO, MVPA, MVPA10min and the distribution of time across acceleration levels. Agreement was strong in the low acceleration bands (ICC = 0.970 and 0.922) with a mean bias of 3.08 minutes (LoA -55.18 to 61.34) and -5.43 (LoA -43.47 to 32.62). In summary, ENMO, MVPA, MVPA10min, wear time and the distribution of time across acceleration levels compared well at the group level. The LOA from the two lowest acceleration levels suggest further work over a longer monitoring period is needed to determine whether outputs from each wrist are comparable.