Table 6.
Mixed logistic (binary) regression analysis | Mixed logistic (multinomial) regression analysis | ||||||||
---|---|---|---|---|---|---|---|---|---|
| |||||||||
Stunted vs. not stunted (reference category) | Underweight vs. normal weight (reference category) | Overweight/obese vs. normal weight (reference category) | |||||||
|
|||||||||
B | Estimate (95% CI) | p | B | Estimate (95% CI) | p | B | Estimate (95% CI) | p | |
|
|||||||||
Sex | |||||||||
Girls (reference) | — | — | — | ||||||
Boys | 0.26 | −0.35; 0.87 | 0.405 | −0.42 | −1.30; 0.47 | 0.357 | 0.55 | 0.04; 1.05 | 0.034 |
Age (years) | 1.37 | 1.02; 1.72 | <0.001 | 0.74 | 0.32; 1.15 | <0.001 | −0.91 | −1.22; −0.60 | <0.001 |
Socio-economic status (0–9) | −0.05 | −0.17; 0.56 | 0.332 | −0.07 | −0.22; 0.09 | 0.414 | 0.07 | −0.06; 0.19 | 0.286 |
BMI | |||||||||
Overweight/obese (reference) | — | — | — | ||||||
Normal weight | 1.55 | 0.05; 3.04 | 0.042 | — | — | — | — | — | — |
Underweight | 1.56 | −0.17; 3.29 | 0.078 | — | — | — | — | — | — |
Stunting | |||||||||
Stunted (reference) | — | — | — | ||||||
Not stunted | — | — | — | −0.25 | −1.12; 0.62 | 0.577 | 1.20 | −0.28; 2.68 | 0.111 |
Flexibility (cm) | 0.08 | 0.03; 0.13 | 0.002 | −0.02 | −0.08; 0.05 | 0.600 | 0.04 | 0.00; 0.08 | 0.035 |
Upper-body muscular strength (kg) | −0.44 | −0.56; −0.31 | <0.001 | −0.37 | −0.54; −0.20 | <0.001 | 0.32 | 0.24; 0.40 | <0.001 |
Lower-body muscular strength (cm) | −0.01 | −0.03; 0.01 | 0.297 | 0.01 | −0.01; 0.04 | 0.295 | −0.04 | −0.05; −0.03 | <0.001 |
CRF (VO 2 max, mL −1 kg −1 min −1 ) | 0.05 | −0.02; 0.11 | 0.190 | 0.03 | −0.07; 0.12 | 0.590 | −0.21 | −0.29; −0.15 | <0.001 |
| |||||||||
Overall model information | Corrected model: F(9,820)=9.9, p<0.001 Constant term: B=−15.4, Estimate (95% CI): −20.6; −10.2, T=−5.8, p<0.001 | Corrected model: F(16,820)=10.1, p<0.001. Normal vs. overweight: Constant term: B=−7.6, Estimate (95% CI): −14.9; −0.3, T=−2.0, p=0.042. Normal vs. overweight: Constant term: B=14.6, Estimate (95% CI): 9.2; 20.0, T=5.3, p<0.001 |
In the mixed logistic regression models, cases were excluded listwise from the analysis if they had missing data in one or several of the covariates or in the dependent variables. Therefore, the analyses were based on the following number of students: stunting: n=838, BMI status: n=838. B is the adjusted unstandardized estimate of the Beta coefficient. Estimate (95% CI) are the unstandardized Beta coefficients, 95% confidence interval. All p-values are calculated using mixed linear regression, adjusting for clustering of school classes. CFR, cardiorespiratory fitness; BMI, body mass index