Table 2.
Boys | Girls | |||
---|---|---|---|---|
Baseline child and household correlates | N | Externalizing SCBE T-score, mean ± SD | N | Externalizing SCBE T-score, mean ± SD |
Child’s age | ||||
3 y | 92 | 51.2 ± 9.2 | 99 | 52.1 ± 9.9 |
4 y or older | 156 | 52.0 ± 9.6 | 171 | 52.2 ± 9.2 |
P valueb | 0.40 | 0.80 | ||
Child’s race/ethnicity | ||||
White, non-Hispanic | 126 | 52.3 ± 8.3 | 139 | 51.3 ± 8.8 |
Hispanic or not White | 123 | 51.1 ± 10.4 | 132 | 53.2 ± 10.1 |
P value | 0.26 | 0.16 | ||
Overweight/obese status | ||||
Not overweight/obese | 150 | 52.5 ± 9.5 | 194 | 53.3 ± 9.2 |
Overweight/obese | 98 | 50.5 ± 9.2 | 76 | 49.4 ± 9.7 |
P value | 0.11 | 0.002 | ||
Physical activity, hours/wk | ||||
< 7 | 28 | 52.1 ± 11.1 | 26 | 54.4 ± 8.5 |
7 to <14 | 54 | 50.3 ± 8.9 | 70 | 51.4 ± 10.2 |
14 to <24 | 74 | 52.6 ± 9.7 | 79 | 51.4 ± 8.6 |
≥ 24 | 73 | 52.2 ± 8.9 | 63 | 52.3 ± 10.3 |
P value | 0.94 | 0.84 | ||
Screen time, hours/day | ||||
< 1.5 | 53 | 51.8 ± 10.1 | 62 | 51.2 ± 9.7 |
1.5 to <2.5 | 66 | 52.2 ± 8.7 | 82 | 52.0 ± 9.5 |
2.5 to <3.5 | 57 | 51.8 ± 9.8 | 56 | 51.1 ± 9.5 |
≥ 3.5 | 56 | 51.4 ± 9.4 | 47 | 54.3 ± 9.5 |
P value | 0.66 | 0.13 | ||
Parent/guardian education level | ||||
Did not graduate HS | 36 | 53.8 ± 9.1 | 42 | 54.6 ± 9.8 |
HS graduate/GED | 81 | 51.7 ± 9.4 | 92 | 50.2 ± 9.4 |
Post-HS education | 134 | 51.1 ± 9.5 | 135 | 52.8 ± 9.1 |
P value | 0.45 | 0.50 | ||
Household income-to-needs ratio | ||||
Quartile 1, <0.47 | 46 | 53.5 ± 10.5 | 66 | 49.6 ± 9.0 |
Quartile 2, ≥0.47 to <0.76 | 55 | 51.6 ± 10.3 | 58 | 54.3 ± 8.8 |
Quartile 3, ≥0.76 to <1.12 | 56 | 52.2 ± 9.2 | 55 | 52.5 ± 10.8 |
Quartile 4, ≥1.12 | 52 | 51.4 ± 8.3 | 55 | 53.2 ± 9.4 |
P value | 0.91 | 0.05 | ||
Parent/guardian CES-D score | ||||
< 16, not depressed | 168 | 51.9 ± 9.6 | 163 | 52.5 ± 9.5 |
≥ 16, depressed | 65 | 51.6 ± 9.3 | 80 | 51.1 ± 9.8 |
P value | 0.35 | 0.34 |
aHigher T-scores on the SCBE indicate lower levels of externalizing behavior
bFrom a linear mixed effects model with externalizing SCBE T-score as the outcome and a continuous variable for each ordinal sociodemographic predictor. For nominal categorical variables, a type III F test was used. A random intercept was specified to account for clustering by classroom