Accelerometer Validation of the Global Physical Activity Questionnaire Within an Urban Community

Purpose This study aimed to test the validity of the Global Physical Activity Questionnaire (GPAQ) against accelerometer data, as feasible, cost-effective instruments are required for the surveillance of physical activity (PA). The questionnaire has been previously assessed in terms of reproducibility and validity however, validation produced poor results and further replication of the work was called for.

Methods Participants (n=101) were randomly selected, stratified by PA level and gender, from 1000 people who completed a household survey in Northern Ireland and consented to participate in further research. Participants wore an Actigraph GT3X accelerometer for 7-days and filled in GPAQ. To test the reliability of measurements, the protocol was repeated for a sub-sample 1 month later.

Results 95 participants completed the study (44 females, 51 males; mean age 44 years, range 17-77). There was moderate agreement between GPAQ and accelerometer MVPA minutes/day (Spearman’s rho 0.48, p=0.01). Sub-group analyses showed similar moderate correlations of 0.43 and 0.50 for females (p<0.05) and males (p<0.01) respectively. The absolute mean difference of both instruments for MVPA minutes/day was 89 (SD±49), (76 (SD±35) MVPA minutes/day for females; 105 (SD±58) for males). Bland and Altman limits of agreement were -133 to 135 MVPA minutes/day. Analysis of test-retest correlations is ongoing.

Conclusions Results showed that the agreement of GPAQ with objective measurement of PA, using an accelerometer, showed a moderate correlation. Therefore, GPAQ provides an acceptable and cost-effective tool to use when measuring PA in the community or large population studies where objective instruments are not feasible.