Quantifying physical activity in physical education via pedometry: a further analysis of steps/min guidelines.

BACKGROUND The validity of common pedometer steps/min guidelines for 1st-12th grade physical education physical activity (PA) recommendations (ie, 33% and 50% PA) was investigated. METHODS Data sets from previous research, where physical education PA was quantified via pedometry, were combined. Participants (1st-12th grade, N = 1152) with concurrent steps/min and observed %PA scores were included. Data were analyzed using correlation, regression, and receiver-operator-characteristic (ROC) statistics. Alpha was set at .05. RESULTS Overall, by gender and school level group (ie, 1st-6th, 7th-12th) PA outcome measures were strongly correlated and significant (r = .85-.92). Steps/min₂, lesson time₃, stature₄, and BMI₅ were significant predictors (r₁₂.₃₄₅=.91) of %PA₁. Steps/min accounted for 85.4% of the variance for %PA; however, the other predictors only accounted for an additional 0.5%. ROC analyses indicated that steps/min was an excellent discriminator (AUC ≥ .90) of %PA guideline achievement. Steps/min values of 60.6 and 82.2 were the most accurate cut points overall for the 33% and 50% PA guidelines, respectively. Steps/min cut points for gender and school level demonstrated agreement with the overall steps/min cut points. CONCLUSIONS These findings support the contention that common steps/min guidelines can be applied in the surveillance of physical education PA.