Propriedades psicométricas de questionários de atividade física na adolescência: revisão sistemática

The aim of this review was to retrieve and identify studies that evaluated the psychometric properties (validity, reliability and responsiveness) of questionnaires used to measure physical activity (PA) in adolescents. It is a systematic review carried out according to the PRISMA protocol. Searches were performed in: MEDLINE (via Ovid), CINAHL, LILACS, Cochrane, SciELO, PsycINFO, Web of Science and SPORTDiscus, using specific terms. The risk of bias was assessed through the COSMIN Checklist. Twenty-six studies were included in the review, of which only five were performed in Brazil. The quality of the studies ranged from poor to good. Of the 25 instruments that were evaluated for test-retest reliability, 11 presented intraclass correlation coefficients ≥ 0.70, including the Physical Activity Questionnaire for Adolescents (QAFA). Regarding validity, only one instrument of the 32 evaluated for this property presented coefficient correlation ≥ 0.70 and five presented coefficients of moderate magnitude (0.56 a 0.69), two of them from Brazil. However, both were compared to subjective criterion measures. The responsiveness of the questionnaires was not assessed in any of the studies. The PA questionnaires included in this review had better reliability indexes than validity and were not investigated for their responsiveness. These results reinforce the need to carry out new studies with more adequate methodological quality that investigate mainly the validity and responsiveness of these instruments.