D-23 Free Communication/Poster - Assessment and Promotion of Physical Activity in School Children: JUNE 3, 2010 1:00 PM - 6:00 PM: ROOM: Hall C
The Physical Activity Questionnaire (PAQ) is a commonly used self-report measure of physical activity in children but it has proven difficult to interpret the outcome score (a value ranging from 1 to 5).
PURPOSE: The purpose of the study was to develop and cross validate a simple calibration equation that can provide estimates of moderate-to-vigorous physical activity (MVPA) from the PAQ score. Calibration was done by relating the subjective, self-report, PAQ score to objective estimates of MVPA collected from an accelerometry-based monitor.
METHODS: Participants included 170 youth (101 female and 69 male) from two elementary schools (n = 95) and two middle schools (n=75) in a small Midwestern town (mean age = 11.13 +/- 2.43). Participants wore an Actigraph activity monitor for a 7 day period and then completed either the PAQ-C or the PAQ-A survey to assess general levels of activity performed during the week. The accelerometer data were screened for compliance and processed (using cutpoints from Ekelund) to provide estimates of average daily MVPA. The PAQ instrument was scored by averaging the data from the eight individual items (the recess item was not used from the PAQ-C). A restricted maximum likelihood regression model was used on a sample of participants (n=113) to determine the relationship between the PAQ outcome score and average daily minutes of MVPA from the Actigraph. A simple model based on PAQ score, age and gender was selected as the most parsimonious and effective solution. The model was then applied to an independent cross validation (n=57) sample and the validity was evaluated using correlation analyses and t-tests.
RESULTS: The final fitted calibration model included in the validation analysis was as follows: Predicted MVPA = [(9.5666*PAQ) + (-10.97041*age) + (11.43817*sex) + 174.50]. The equation explained 52% of the variance in average daily MVPA levels. The cross-validation analysis supported the validity of the equation. There were non-significant differences between predicted and observed estimates of MVPA for the group (t=.39, p =.70). The correlation between measured and predicted MVPA levels was also moderate and significant (r =.66).
CONCLUSIONS: The resulting PAQ calibration equation provides a useful way to estimate levels of MVPA in groups.