Reliability and validity of a questionnaire for physical activity assessment in South American children and adolescents: The SAYCARE Study

Abstract

ABSTRACT: The objective of this article is to test the reliability and validity of the new and innovative physical activity (PA) questionnaire. Methods: Subsamples from the South American Youth/Child Cardiovascular and Environment Study (SAYCARE) study were included to examine its reliability (children: n 5 161; adolescents: n 5 177) and validity (children: n 5 82; adolescents: n 5 60). The questionnaire consists of three dimensions of PA (lei-sure, active commuting, and school) performed during the last week. To assess its validity, the subjects wore accelerometers for at least 3 days and 8 h/d (at least one weekend day). The reliability was analyzed by correlation coefficients. In addition, Bland-Altman analysis and a multilevel regression were applied to estimate the measurement bias, limits of agreement, and influence of contextual variables. Results: In children, the questionnaire showed consistent reliability (q 5 0.56) and moderate validity (q 5 0.46), and the contextual variable variance explained 43.0% with 222.9 min/d bias. In adolescents, the reliability was higher (q 5 0.76) and the validity was almost excellent (q 5 0.88), with 66.7% of the variance explained by city level with 16.0 min/d PA bias. Conclusions: The SAYCARE PA questionnaire shows acceptable (in children) to strong (in adolescents) reliability and strong validity in the measurement of PA in the pediatric population from low- to middle-income countries.