Table 4.

Cross-lag model for cannabis-user subsample

Predictors	Cannabis use, age 17 β (SE)	Depressive symptoms, age 17 β (SE)	Cannabis use, age 20 β (SE)	Depressive symptoms, age 20 β (SE)	Cannabis use, age 22 β (SE)	Depressive symptoms, age 22 β (SE)
Covariates
Education, age 17	-0.171 (0.057)**	-0.216 (0.097)*
Family income, age 17	0.015 (0.005)**	0.012 (0.009)
Race	0.284 (0.164)†	0.532 (0.281)†
IQ score, age 11	0.006 (0.006)	-0.010 (0.011)
Primary caregiver depressive	0.017 (0.014)	0.026 (0.024)
Symptoms, age 12
Antisocial behavior, age 12	0.038 (0.028)	-0.021 (0.049)
Adulthood incarceration/probation	-0.286 (0.233)	-0.377 (0.404)
Tobacco use, age 17	0.315 (0.151)*	0.421 (0.262)
Alcohol use, age 17	0.144 (0.095)	0.356 (0.164)*
Cannabis use, age 17			0.008 (0.270)	-0.096 (0.179)		0.200 (0.140)
Depressive symptoms, age 17			0.408 (0.159)*	0.290 (0.106)**	-0.138 (0.190)
Cannabis use, age 20					-0.054 (0.123)	0.052 (0.055)
Depressive symptoms, age 20					0.139 (0.184)	0.482 (0.082)***
Model R2	.272***	.219**	.080	.073	.012	.290***

Notes: n = 111. β = unstandardized regression coefficient. Model fit indices: χ²(38) = 61.28, p = .010; χ²/df = 1.61, comparative fit index = .79, root mean square error of approximation = .07, standardized root mean square residual = .06.

^†p < .1;

^*p < .05;

^**p < .01;

^***p < .001.