Validation of the Godin-Shephard questionnaire in prepubertal girls

PURPOSE: This study was designed to provide validation of the Godin-Shephard questionnaire for the assessment of habitual activity in prepubertal children. METHODS: Twenty-nine female competitive gymnasts and 32 recreational and nongymnasts, aged 7-11 yr, underwent multiple measures of physical activity over a 15 month period. The Godin-Shephard questionnaire (G-S) was completed once, in the middle of the study period. The Physical Activity Checklist Inventory (PACI) was completed 6 times, at intervals throughout the study, for 2 weekend days and 4 weekdays. Caltrac activity monitors (Hemokinetics, Madison, WI) were worn by each subject in two waves consisting of 2 weekdays and 1 weekend day each. Self-reported activity (h/wk) was recorded by the gymnasts in daily logs and for the nongymnasts through quarterly queries regarding participation in organized activities. A rank order was assigned to each subject for each tool. Spearman rank order correlations assessed the relationship between the G-S versus each measure and combination of measures. Subjects were also divided into gymnast and recreational/nongymnast groups, and alternately into 7-9 yr old and >or= 10 yr old groups, re-ranked and re-analyzed. RESULTS: The average rank of combined PACI, Caltrac and h/wk scores showed only moderate correlation with ranked G-S scores for the overall group and for the 7-9 yr old subset, rho = 0.380 (P = 0.003) and rho = 0.422 (P = 0.043), respectively. The average rank of combined PACI and Caltrac scores also showed only moderate correlation with the ranked G-S scores for the gymnast subset, rho = 0.423 (P = 0.025). No measure or combination of measures correlated significantly with the ranked G-S scores for the >or= 10 yr old subset or the nongymnast subset. CONCLUSION: Based upon comparison with multiple single day measures of activity, the G-S did not provide a good estimate of habitual activity in preadolescent females.