Stability and convergent validity of three physical activity assessments.

Abstract

The purpose of this study was to examine the stability and convergent validity of heart rate (HR) monitoring, Caltrac accelerometer, and physical activity recall (PAR) in sixth-grade girls during normal weekday activities. 46 sixth-grade girls (age 12 +/- 0.6) wore HR monitors and Caltrac accelerometers for 3 d during school, after school, and evenings. We also obtained a PAR for each day. Data were compared on the basis of kcal x h(-1). Two days' worth of data were analyzed for each participant. Intraclass correlation coefficients obtained by use of repeated measures ANOVA revealed that HR monitoring (r = 0.99) and PAR (r = 0.98) were extremely stable across 2 d, whereas Caltrac was moderately stable (r = 0.76). Pearson correlations between techniques were HR versus PAR, r = 0.50 (P < 0.01), HR versus Caltrac, r = 0.28, and Caltrac versus PAR, r = 0.76 (P < 0.01). Methods comparison plots showed poor individual subject agreement between all three types of assessment. HR and PAR were stable across 2 d. PAR underestimated caloric expenditure by approximately 14%. Caltrac showed the least utility in both reliability and validity.