Table 3.
Path coefficients, direct and indirect effects from the path model examining parental knowledge and peer substance use as indirect pathways linking externalizing polygenic scores and adolescent externalizing behaviors among participants of European ancestry
| Paths | B | SE | β | p |
|---|---|---|---|---|
| EXT PGS → T1 EXT | .14 | .03 | .17 | <.001 |
| EXT PGS → T1 par knowledge | −.07 | .02 | −.12 | .004 |
| EXT PGS → T1 peer sub use | .04 | .01 | .13 | <.001 |
| EXT PGS → T2 EXT | .14 | .03 | .14 | <.001 |
| EXT PGS → T2 par knowledge | −.04 | .02 | −.06 | .047 |
| EXT PGS → T2 peer sub use | .06 | .02 | .11 | <.001 |
| T1 EXT → T2 EXT | .51 | .10 | .40 | <.001 |
| T1 EXT → T2 par knowledge | −.05 | .06 | −.07 | .421 |
| T1 EXT → T2 peer sub use | .06 | .05 | .09 | .226 |
| T1 par monitor → T2 EXT | −.16 | .09 | −.09 | .070 |
| T1 par monitor → T2 par knowledge | .43 | .05 | .43 | <.001 |
| T1 par monitor → T2 peer sub use | −.09 | .04 | −.10 | .020 |
| T1 peer sub use → T2 EXT | .42 | .16 | .13 | <.001 |
| T1 peer sub use → T2 par knowledge | .11 | .09 | .06 | .263 |
| T1 peer sub use → T2 peer sub use | .63 | .09 | .40 | <.001 |
| Indirect Effects | B | 95% CI | β | p |
| EXT PGS → T1 EXT → T2 EXT | .070 | [.030, .123] | .068 | .003 |
| EXT PGS → T1 par knowledge → T2 EXT | .011 | [.000, .032] | .011 | .145 |
| EXT PGS → T1 peer sub use → T2 EXT | .018 | [.004, .039] | .017 | .043 |
Note. Abbreviations: EXT PGS = externalizing polygenic score. T1 = Time 1. T2 = Time 2. EXT = externalizing behaviors. Par knowledge = parental knowledge. Peer sub use = peer substance use. CI = confidence intervals. The EXT PGS included in the analyses was regressed on the ten genetic ancestry principal components. Age and sex were included as covariates. Statistically significant coefficients/effects are bolded.