Table 3.
Regression coefficients of the association between serum leptin levels at 16–17 years and academic performance in HS
| Intercept | Coefficient | Robust SE | R2 | |
|---|---|---|---|---|
| 9th grade | ||||
| Model 1 | 491.2*** | −0.56 | 0.32 | 0.04 |
| Model 2 | 501.9*** | −0.62* | 0.30 | 0.09 |
| 10th grade | ||||
| Model 1 | 495.6*** | −0.76 | 0.43 | 0.02 |
| Model 2 | 508.3*** | −0.82* | 0.37 | 0.14 |
| 11th grade | ||||
| Model 1 | 511.7*** | −0.47 | 0.35 | 0.04 |
| Model 2 | 522.0*** | −0.57 | 0.34 | 0.10 |
| 12th grade | ||||
| Model 1 | 527.9*** | −0.87* | 0.37 | 0.04 |
| Model 2 | 540.5*** | −0.98*** | 0.37 | 0.13 |
| HS GPA | ||||
| Model 1 | 511.2*** | −0.71* | 0.31 | 0.04 |
| Model 2 | 521.3*** | −0.80*** | 0.30 | 0.16 |
SE are robust to heteroscedasticity.
Model 1: adjusted for sex, fat mass and parental education. Model 2: also included IR, quality of diet, PA status, type of secondary education and iron supplementation in infancy.
*p<0.05; **p<0.01; ***p<0.001.
GPA, grade point average; HS, high school; PA, physical activity.