Validity of the Block Kids Physical Activity Screener for Moderate to Vigorous Physical Activity
- Presented on May 29, 2014
Purpose: Objectively measuring physical activity in children is often impractical; thus self-report methods are frequently used but may not be valid for determination of physical activity intensity. This study examined the accuracy and validity of the Block Kids Physical Activity Screener (PAS) for estimating moderate to vigorous physical activity (MVPA) in schoolchildren, and evaluated differences by sex, race, and weight status.
Methods: Boston area schoolchildren (n=101; 4th-8th grade) were recruited, demographic information was collected by parent report, and measured height and weight were used to determine BMI z-score. Minutes of daily MVPA were objectively measured using accelerometers (ACC; Actigraph GT3X) worn on the hip for 7 days and subsequently measured using the PAS, which assesses the frequency and duration of team-based leisure activities, sports, and extracurricular exercise over the prior week. Estimates of intraclass correlation coefﬁcients (ICC), standard error of measurements (SEM), and Bland-Altman plots were used to assess the level of agreement and variability between ACC- and PAS-measured MVPA of all participants and by duration of MVPA, sex, race, and weight status
Results: Of the 101 enrolled, 64 completed the study with at least 4 valid ACC days (mean (SD): age, 11.7 (1.0) yrs; BMIz-score, 0.7 (1.1); 42.2% overweight/obese; 43.8% female; 43.8% Caucasian). Mean durations of ACC- and PAS-measured MVPA were 91.1 and 97.5 min/day, respectively (P=0.59). Overall, the level of agreement between ACC- and PAS-measured MVPA was fair (ICC=0.30; 95% CI=0.06-0.51) with a moderate amount of measurement error (SEM=56.1 min), which increased with increasing levels of MVPA. Children over-reported more if their MVPA exceeded 120 min (SEM=99.2 vs 30.3 min). Overall, the Bland-Altman plot showed a moderate amount of random variability and no signiﬁ cant bias, with no observable differences by sex, race, and weight status.
Conclusion: The PAS showed fair to moderate agreement with ACC-measured MVPA, with no observational differences by sex, race, and weight status. Increased over-reporting was observed for measured MVPA greater than 120 minutes; thus further research should assess factors influencing over-reporting, which may have implications for improvement of self-report methods.