Table 4.
Fractional logistic regressions predicting the probability of membership in each perinatal risky drinking trajectory class
| Model 1 (N=432) | Class 2 (postpartum initiation) |
Class 3 (moderate risk) |
Class 4 (high risk) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| B | SE | p | 95% CI | B | SE | p | 95% CI | B | SE | p | 95% CI | |
| Caregiver AUDIT score | .01 | .03 | .760 | [−0.05, 0.08] | .03 | .02 | .814 | [−0.06, 0.08] | .02 | .07 | .806 | [−0.11, 0.15] |
| Positive alcohol expectancies | .01 | .03 | .690 | [−0.05, 0.07] | −.15 | .03 | .203 | [−0.03, 0.13] | .10 | .07 | .157 | [−0.04, 0.24] |
| Negative alcohol expectancies | −.03 | .04 | .427 | [−0.10, 0.04] | .01 | .04 | .726 | [−0.10, 0.07] | .01 | .08 | .857 | [−0.14, 0.17] |
| Peer alcohol before pregnancy | −.11 | .13 | .374 | [−0.36, 0.14] | .05 | .04 | <.001 | [0.28, 0.83] | 1.13 | .35 | .001 | [0.45, 1.82] |
| Peer alcohol during pregnancy | .15 | .10 | .137 | [−0.05, 0.35] | −.02 | .04 | <.001 | [0.21, 0.70] | .91 | .27 | .001 | [0.39, 1.44] |
| Peer alcohol after pregnancy | .34 | .10 | .001 | [0.14, 0.53] | .55 | .14 | .021 | [0.05, 0.57] | .45 | .21 | .029 | [0.05, 0.86] |
| Model 2 (N=145) | ||||||||||||
| Coping motivations | −.07 | .11 | .500 | [−0.28, 0.14] | .07 | .09 | .454 | [−0.11, 0.24] | .16 | .12 | .197 | [−0.08, 0.40] |
| Enhancement motivations | −.20 | .10 | .057 | [−0.40, 0.01] | .06 | .10 | .567 | [−0.14, 0.25] | .23 | .15 | .130 | [−0.07, 0.53] |
Note. Reference group = Class 1 (low-risk). Model 1 includes the full sample, whereas Model 2 focuses on the subset of mothers who were already drinking (at least 5 times a year) before pregnancy. Models covaried for minority race, age of alcohol initiation, linear and quadratic effects of age at T0, and the month of pregnancy at T0.