Agreement of physical activity metrics from ActivPAL and ActiGraph accelerometers worn concurrently among college students at a residential university

Abstract

Objective To examine agreement between concurrently worn ActivPAL (AP) and ActiGraph (AG) accelerometers in college students.Methods Thirty-six students (age 21.6 ± 2.4 years) wore AP and AG accelerometers for 4–8 days. Agreement was assessed using paired t-tests, intraclass correlation coefficients, mean absolute percentage error, Bland-Altman analyses, and ±10% equivalence testing.Results AG wear time averaged 15.5 ± 1.0 h/day. Compared with AP, AG measured 698 ± 805 fewer steps/day, 69 ± 70 more sedentary min/day, and 112 ± 57 fewer zero cadence min/day. Peak 1-min and 30-min cadence demonstrated strong agreement and met equivalence criteria. Daily steps showed agreement but did not meet equivalence; sedentary and zero cadence time demonstrated the weakest agreement.Conclusions Device agreement varied by metric, emphasizing the importance of measurement selection in college student research. Reliable step-based measurement may support research needed to develop step recommendations and physical activity promotion strategies for college students.