. Author manuscript; available in PMC: 2015 Oct 1.

Published in final edited form as: Alcohol Clin Exp Res. 2014 Sep 24;38(10):2541–2549. doi: 10.1111/acer.12524

Table 2.

Cox proportional hazards regression models of adolescent drinking milestones

	Drinking milestones in ever-drinkers before age 18 (N=1,550)
	Models of first intoxication		Models of first DSM-5 symptom

	Hazard ratio (95% CI)	χ² p value	Hazard ratio (95% CI)	χ² p value
Univariate model set
rs1229984^a	0.72 (0.56−0.91)	.006	0.69 (0.50−0.94)	.02
peer drinking^b	1.89 (1.70−2.10)	<.0001	2.27 (1.98−2.60)	<.0001
Multivariate model set
rs 1229984	0.76 (0.61−0.96)	.02	0.73 (0.54−0.97)	.03
peer drinking	1.88 (1.69−2.09)	<.0001	2.26 (1.97−2.60)	<.0001
Interaction model set
rs 1229984	0.56 (0.41−0.76)	.0002	0.45 (0.26−0.77)	.004
peer drinking	1.81 (1.62−2.01)	<.0001	2.17 (1.88−2.50)	<.0001
rs1229984*peer drinking	2.10 (1.32–3.32)	.002	2.29 (1.21–4.30)	.01
Examination of GxE term in interaction model set
None/few best friends drink (GA/AA vs GG)	0.56 (0.41–0.76)	.0002	0.45 (0.26–0.77)	.004
Most/all best friends drink (GA/AA vs GG)	1.16 (0.82–1.65)	.39	1.03 (0.73–1.45)	.87

Reference ADH1B rs1229984 genotype GG was compared to GA/AA;

Reference peer drinking status none/few best friends drink was compared to most/all best friends drink; All models adjusted for gender and ethnicity.