Skip to main content
. 2021 Jan 28;16(1):e0245875. doi: 10.1371/journal.pone.0245875

Table 5. Between-group differences in outcomes over time.

Outcome Measure Model 1 Model 2 Model 3
β (95% CI) p-value β (95% CI) p-value β (95% CI) p-value
BMI z-score
 Usual care group reference reference reference
 Exercise group -0.12 (-0.20 to -0.047) 0.002 -0.11 (-0.20 to -0.023) 0.013 -0.11 (-0.18 to -0.038) 0.003
 Nutritional group -0.057 (-0.13 to 0.021) 0.15 -0.060 (-0.15 to 0.030) 0.19 -0.056 (-0.13 to 0.016) 0.129
%BMIp95th
 Usual care group reference reference reference
 Exercise group -1.03 (-1.05 to -1.01) 0.001 -1.03 (-1.05 to -1.01) 0.007 -1.03 (-1.04 to -1.01) 0.001
 Nutritional group -1.01 (-1.03 to 1.01) 0.17 -1.01 (-1.03 to 1.01) 0.19 -1.01 (-1.03 to 1.00) 0.13
Waist circumference, cm
 Usual care group reference reference reference
 Exercise group -1.90 (-4.03 to 0.23) 0.080 -1.30 (-3.86 to 1.25) 0.32 -1.73 (-3.68 to 0.22) 0.083
 Nutritional group -2.37 (-4.54 to -0.21) 0.032 -3.47 (-6.06 to -0.89) 0.009 -3.71 (-5.68 to -1.74) <0.001
Adiponectin, μg/mL
 Usual care group reference reference reference
 Exercise group 1.28 (1.08 to 1.51) 0.005 1.31 (1.08 to 1.58) 0.005 1.27 (1.10 to 1.47) 0.001
 Nutritional group -1.00 (-1.19 to 1.18) 0.97 -1.08 (-1.30 to 1.12) 0.44 -1.08 (-1.25 to 1.07) 0.30

Abbreviations: BMI, body mass index; %BMIp95th, percentage of the 95th percentile of age- and sex-specific body mass index.

Model 1: group × time interaction effects adjusted for age and sex in the mixed effects linear regression models (random intercept: individual).

Model 2: group × time interaction effects adjusted for age, sex, parental obesity, parental education, monthly household income, living with both parents, and sleep time in the mixed effects linear regression models (random intercept: individual).

Model 3: group × time interaction effects adjusted for age, sex, parental obesity, parental education, monthly household income, living with both parents, sleep time, and baseline values in the mixed effects linear regression models (random intercept: individual).