Table 3.
Results of Longitudinal Generalized Estimating Equations for Covariates, Genotypes, Parenting, and Genotype × Parenting Interactions Predicting Conduct disorder Symptoms During Assessment Years 6–9
| Predictor |
β |
95% Confidence Interval |
z |
p < |
|---|---|---|---|---|
| Time (assessment waves) | −0.07 | (−0.18 – 0.04) | −1.21 | .23 |
| Site | −0.34 | (−0.93 – 0.25) | −1.13 | .26 |
| Cohort | −0.45 | (−0.94 – 0.05) | −1.77 | .08 |
| Child’s age in year 1 | 0.24 | (−0.09 – 0.57) | 1.42 | .16 |
| Child’s sex | −0.54 | (−1.09 – 0.02) | −1.89 | .06 |
| ADHD symptoms in year 1 | 0.08 | (0.03 – 0.14) | 3.10 | .002 |
| Family income (log) | −0.15 | (−0.35 – 0.05) | −1.47 | .15 |
| African American1 | −0.19 | (−0.78 – 0.40) | −0.63 | .53 |
| Other ethnic group1 | 0.16 | (−0.54 – 0.87) | 0.46 | .65 |
| Parent interviewer blind | −0.05 | (−0.42 – 0.32) | −0.25 | .81 |
| Youth interviewer blind | 0.07 | (−0.24 – 0.38) | 0.47 | .65 |
| 9/9 genotype2 | 4.86 | (3.40 – 6.32) | 6.52 | .0001 |
| 9/10 genotype2 | 0.91 | (−0.49 – 2.31) | 1.27 | .21 |
| Positive parenting3 | 0.02 | (−0.28 – 0.31) | 0.10 | .92 |
| Negative parenting3 | 0.12 | (−0.21 – 0.45) | 0.71 | .48 |
| Child disruptive behavior3 | 0.014 | (−0.01 – 0.02) | 0.89 | .38 |
| 9/9 × positive parenting | −0.80 | (−1.30 – −0.30) | −3.13 | .002 |
| 9/9 × negative parenting | −1.13 | (−1.71 – −0.549) | −3.79 | .0002 |
| 9/10 × positive parenting | −0.28 | (−0.75 – 0.18) | −1.19 | .24 |
| 9/10 × negative parenting | −0.04 | (−0.482 – 0.40) | −0.17 | .87 |
1
Contrast group is non-Hispanic white
2
Contrast group is 10/10 genotype
3
Observed during mother-child interaction task in year 1
Note: Conduct disorder symptoms = number of symptoms reported in each assessment wave by either the parent or the youth (combined); estimates of main effects for genotype and parenting are changed by the presence of interaction terms; all tabled tests of significance used 1 df.