Abstract

Background: The possibility of carrying out screening, with acceptable accuracy, of a child's bone mass status based on a physical fitness test can advance the concept of health-related physical fitness. In addition, the relevance of the applicability of this type of screening in educational environments is mainly due to the difficulty of direct assessments of bone health indicators. This study aimed to propose cut-off points for physical fitness tests based on children's bone health indicators.Methods: This is a two-phase cross-sectional study. Phase-1: 160 children (6-11 years-old) performed the 20-m sprint test (20-mST) and the 2 kg medicine ball throw test (2 kgMBTT). Areal bone mineral density (aBMD) and content was assessed by DXA. The area under the ROC curve greater than 70% was considered valid. Phase-2: It was carried out a secondary analysis in a sample with 8,750 Brazilians (6-11 years-old). The percentile values (identified in phase-1) were used to identify the values of the cut-off points in the unit of measurement of the tests. The validation of the cut-off points found was by odds ratio values and p <= 0.05.Results: Phase 1: The areas under the ROC curve were 0.710, 0.712 (boys and girls-20-mST), 0.703, and 0.806 (boys and girls-2 kgMBTT) with total spine and pelvis aBMD as the outcome. Phase 2: From percentile values, we find valid cut-off points in the Brazilian sample (OR > 3.00; p < 0.001) for boys and girls. Values ranged between 5.22 s-4.00 s to 20-mST and between 125.0 cm-160.0 cm to 2 kgMBTT. Conclusion. The 20-mST and the 2 kgMBTT presented sufficient accuracy for the screening of children aged between 6 and 11 years with greater chances of having low aBMD in the total spine and pelvis, with valid cut-off points.