*3.1. Calibration Study*

Figure 1 presents the relationship between Speedabs and the accelerometer output. It demonstrates that the three age-groups fall approximately on the same regression line for hip data, while children show higher values compared to the other age-groups for the thigh placement. These results confirm our assumptions of similar mechanical work captured with the hip placement and less similarity with the thigh placement.

Figure 2 present the relationship between Speedabs and VO2net and between Speedeq and VO2net, respectively. It shows that while children have higher VO2net than the other age-groups for the same Speedabs, there was approximately one common regression line with the Speedeq. VO2net may therefore be considered a measure of similar effort in different age-groups according to our assumption.

Figure 3 display the relationship between Speedabs and MET and between Speedeq and MET, respectively. The MET-values differ between the three age-groups for both Speedabs and Speedeq, which confirms the assumption that MET is not a measure of effort comparable between age-groups.

**Figure 1.** Relationship between absolute speed (Speedabs) and the accelerometer measure of physical activity (PA) intensity (mg) for (**A**) hip and (**B**) thigh placement, in children 9–11, adolescents 14–16 and adults 23–44 years old.

**Figure 2.** Relationship between (**A**) Speedabs or (**B**) equivalent speed (Speedeq) and net oxygen consumption (VO2net), in children 9–11, adolescents 14–16 and adults 23–44 years old.

**Figure 3.** Relationship between (**A**) Speedabs or (**B**) Speedeq and Metabolic Equivalent of Task (MET), in children 9–11, adolescents 14–16 and adults 23–44 years old.

Figure 4 presents the relationship between MET and the corresponding VO2net. It is the starting-point to determine the VO2net cut-points for LPA (1.5 METs), MPA (3.0 METs), VPA (6.0 METs) and VVPA (9.0 METs) based on adult data. If the same MET-values had been applied to the children and adolescent groups, considerable higher VO2net-values (and effort) had been used for accelerometer calibration in these age-groups.

**Figure 4.** Definition of the common VO2net cut-points representing light PA (LPA), moderate PA (MPA), vigorous PA (VPA) and very vigorous PA (VVPA) based on adult MET values, in children 9–11, adolescents 14–16 and adults 23–44 years old.

The next step was to apply the common VO2net cut-points to all three age-groups to determine the corresponding accelerometer mg values (Figure 5). The three age-groups achieved different regression lines, with the lowest accelerometer cut-points observed in children. Hence, when moving with the same effort, the three age-groups move at different Speedabs and consequently do different mechanical work. Still, there were some overlapping between the age-groups.

**Figure 5.** Calibration of accelerometer measures of absolute PA intensity (mg) for (**A**) hip and (**B**) thigh placement in children 9–11, adolescents 14–16 and adults 23–44 years old, using the common VO2net cut-points defined and presented in Figure 4.

In contrast, when MET-based cut-points were applied, the adults achieved the lowest accelerometer values (Figure 6). Table 1 provides the numerical values of the VO2net-based and the MET-based cut-points for LPA, MPA, VPA and VVPA for the two body placements in the three age-groups, expressed in mg.

**Figure 6.** Calibration of the accelerometer measure of absolute PA intensity (mg) for (**A**) hip and (**B**) thigh placement in children 9–11, adolescents 14–16 and adults 23–44 years old, using the standard MET cut-points (1.5, 3.0, 6.0, 9.0) in all three groups.


**Table 1.** Accelerometer cut-points (mg) by body placement, age-group and calibration method.

Children 9–11, adolescents 14–16 and adults 23–44 years old. VO2net cut-points (mL·min−1·kg<sup>−</sup>1): LPA 1.1, MPA 5.5, VPA 14.2 and VVPA 22.9. MET cut-points (METs): LPA 1.5, MPA 3.0, VPA 6.0 and VVPA 9.0. Explained variation (R2) is presented for the fit of the regression line between VO2net/MET and accelerometer mg.
