Comparison of two accelerometers during 8th grade physical education

The NL-1000 (NL; New Lifestyles Inc., Lee's Summit, Missouri) is a low cost ($48) accelerometer that promises feasible and practical physical activity (PA) assessment both within and outside of school. The NL outputs include accumulated time in moderate-to-vigorous physical activity (MVPA) and steps. Purpose: This study compared MVPA outputs between the NL and the more commonly used and expensive ($388) ActiGraph (AG; Actigraph, LLC, Fort Walton Beach, Florida) accelerometer in 8th graders during Physical Education (PE). Methods: Participants (N=169, males=88) wore both accelerometers attached to an elastic waist belt during a gender segregated 35 minute PE class. MVPA time detected by the NL was recorded from the NL screen display at class end. AG data were downloaded and time in MVPA was derived using child-appropriate validated thresholds. Means and standard deviations were calculated for minutes in MVPA for both accelerometers. Paired t-tests, effect size (ES), and Bland-Altman plots were used to compare outputs. Results: Significant differences between MVPA outputs existed for males (NL=14.2±3.3, AG=15.7±4.2, p<.001) and females (NL=5.9±2.0, AG=4.7±2.0, p<.001) during PE. ES of mean differences for males and females were 0.77 and 0.84, respectively. Bland-Altman comparisons displayed evidence of systematic bias between NL and AG MVPA outputs for males (r=.397, p<.001) but not for females. Conclusion: Small but significant differences were observed between mean NL and AG MVPA outputs for males (1.5 min) and females (1.2 min) in a 35 minute PE class. Differences may be due to instrument specific intensity thresholds or sampling lengths (i.e., epochs).