Join us on March 2nd for an ActiGraph webinar:
Wearable Data Gone Awry: Cautionary Tales from the Clinical Research Trenches.Register Now
Equating Self-Report Tools To Objective Measures Of MVPA: Calibration Of The Physical Activity Questionnaire
- Presented on May 30, 2013
The Physical Activity Questionnaires (PAQ-Children and PAQ-Adolescents) are widely used to assess youth physical activity patterns but a limitation is that they do not provide information about time spent in moderate-to-vigorous physical activity (MVPA).
Purpose The purpose of this study was to calibrate in- and out-of-school activity items of the PAQ using temporally matched data from an objective physical activity monitor.
Methods A total of 195 school-aged children wore an Actigraph monitor and completed either the PAQ-C or PAQ-A one week later. Actigraph data were processed using age-specific cutpoints and split into aggregates of MVPA spent during recess, PE, lunch, after-school, evening, and weekend. Each aggregate was matched with a respective item from the PAQ’s (e.g. recess item) and standard calibration procedures were used to compute a prediction equation for each period of time. Predicted scores were combined to reflect in-school, out-of-school and total week MVPA and crossvalidated with an independent sample.
Results Regression b-weights were significant for all in-school and out-ofschool periods, indicating that individual items could significantly predict %MVPA during different periods of the week (b-weights ranged from 1.3%-4.9% MVPA). Proportion of variance explained by each item was higher for the after-school (r2=.35) and recess items (r2= .34) and ranged from .16 to .24 for the remaining items. The cross-validation results revealed that aggregates of both in-school (Mdiff = 3.8±3.1; p = .23) and out-of-school activity items (Mdiff= -15.6±21.2; p = .47) could predict synchronized minutes of accelerometer-recorded MVPA.
Conclusions Minutes of MVPA can be accurately estimated in groups of individuals during in- and out-of-school periods with the PAQ. Similar calibration procedures can be employed to increase the utility of other self-report tools for physical activity research. This work has been partially funded by a grant from European Social Fund/Fundacao para a Ciencia e Tecnologia (SFRH / BD / 60396 / 2009).