RESUMO
Background: An appropriate measurement of physical activity (PA) in children is useful, since inactivity is associated to obesity, cardiovascular and metabolic risk. Aim: To assess the reliability of the INTA questionnaire of PA, to compare the derived PA score with accelerometry and to assess its ability to identify excessively inactive children. Material and methods: One hundred eighty children aged 8 to 13 years answered an interviewer-administered questionnaire about their usual PA, consisting in 5 items (recumbent, seated, walking, playing outdoor, sports). The answers were converted to a PA score with a 0-10 points scale. Reliability was tested in 87 children by test/retest conducted 3-5 days apart. The PA score was compared with 3-day accelerometry in 77 of 93 children (35 obese and 42 non obese). Receiver operating characteristic (ROC) curves were used to determine the optimal cut-point for identify an excessively sedentary child. Results: The test/retest reliability of the questionnaire was 0.69 to 0.93 (Lin coefficient). Accelerometry was significantly associated with PA score (RHO: 0.60, p =0.008), outdoor plays (RHO: 0.37, p =0.0009) and practicing of sports (RHO: 0.33, p =0.003). Obese children were less active than non obese children, according both to PA score and to accelerometry. The optimal cut-point for classifying a child as too sedentary was a score of 5 (sensitivity =0.89). Conclusions: The INTA-test is a valuable instrument for measuring usual PA in clinical practice and is easy to administer.