# Table 3 Multivariate linear regression analyses to predict physical activity in children with juvenile idiopathic arthritis and controls

B 95 % CI lower 95 % CI upper P
PAL
Reference 1.75 1.67 1.83 <0.01
Controls 0.10 0.03 0.18 0.01
Age centered 10 years 0.04 0.02 0.07 <0.01
BMI centered 17 kg/m2 −0.01 −0.02 −0.00 0.04
Gender −0.07 −0.13 −0.01 0.02
JIA season −0.14 −0.23 −0.05 <0.01
MVPA
Reference 1.70 1.27 2.13 <0.01
Controls 0.41 −0.02 0.83 0.06
Age centered 10 years 0.20 0.07 0.33 <0.01
BMI centered 17 kg/m2 −0.02 −0.08 0.04 0.50
Gender −0.12 −0.43 0.20 0.47
JIA season −0.50 −1.00 0.01 0.06
Sedentary time
Reference 18.86 18.36 19.37 <0.01
Controls −0.59 −1.09 −0.09 0.02
Age centered 10 years −0.13 −0.28 0.03 0.10
BMI centered 17 kg/m2 0.15 0.08 0.21 <0.01
Gender −0.02 −0.38 0.35 0.92
JIA season 0.78 0.18 1.37 0.01
1. The regression equation for PAL is as follows:
2. $$\begin{array}{l}\mathrm{PAL}\kern0.5em =\kern0.5em \mathrm{reference} + 0.10\ *\ \mathrm{control} + 0.04\ *\mathrm{age}\ \left(\mathrm{centered}\ 10\right) + -0.01\ *\ \mathrm{B}\mathrm{M}\mathrm{I}\ \left(\mathrm{centered}\ 17\right) + -0.07\ \\ {}*\ \mathrm{gender} + -0.14\ *\ \mathrm{season}\end{array}$$
3. The reference for this equation is a 10 year old boy with JIA, a BMI of 17 kg/m2 of which the data was collected in the summer. So a healthy girl (no JIA) of 8 years old, a BMI of 20 has a predicted PAL of (1.75 + 0,10 * 1 + 0.04 * (8–10) + −0.01 * (20–17) + −0.07 * 1 = 1.73
4. JIA juvenile idiopathic arthritis, BMI body mass index, PAL physical activity level, MVPA moderate to vigorous activity expressed in hours/day. Sedentary time expressed in hours/day, CI confidence interval of B. Reference category: Boy of 10 years, with a BMI of 17, with JIA, who filled in the diary in the summer period