Reliability and validity of Web-SPAN, a web-based method for assessing weight status, diet and physical activity in youth

Background: Web-based surveys are becoming increasing popular. The present study aimed to assess the reliability and validity of the Web-Survey of Physical Activity and Nutrition (Web-SPAN) for self-report of height and weight, diet and physical activity by youth. Methods: School children aged 11-15 years (grades 7-9; n = 459) participated in the school-based research (boys, n = 225; girls, n = 233; mean age, 12.8 years). Students completed Web-SPAN (self-administered) twice and participated in on-site school assessments [height, weight, 3-day food/ pedometer record, Physical Activity Questionnaire for Older Children (PAQ-C), shuttle run]. Intraclass (ICC) and Pearson's correlation coefficients and paired samples t-tests were used to assess the test-retest reliability of Web-SPAN and to compare Web-SPAN with the on-site assessments. Results: Test-retest reliability for height (ICC = 0.90), weight (ICC = 0.98) and the PAQ-C (ICC = 0.79) were highly correlated, whereas correlations for nutrients were not as strong (ICC = 0.37-0.64). There were no differences between Web-SPAN times 1 and 2 for height and weight, although there were differences for the PAQ-C and most nutrients. Web-SPAN was strongly correlated with the on-site assessments, including height (ICC = 0.88), weight (ICC = 0.93) and the PAQ-C (ICC = 0.70). Mean differences for height and the PAQ-C were not significant, whereas mean differences for weight were significant resulting in an underestimation of being overweight/obesity prevalence (84% agreement). Correlations for nutrients were in the range 0.24-0.40; mean differences were small but generally significantly different. Correlations were weak between the web-based PAQ-C and 3-day pedometer record (r = 0.28) and 20-m shuttle run (r = 0.28). Conclusions: Web-SPAN is a time-and cost-effective method that can be used to assess the diet and physical activity status of youth in large cross-sectional studies and to assess group trends (weight status).