Table 2.

Odds Ratios (ORs) with 95% Confidence Intervals (95% CI) Versus Being a Light Consumer, for Each Step Decrease on the Stanine Scale for Total Alcohol Intake Per Week in Groups Using Multinomial Logistic Regression Analysis. Results Are Presented Both for the Full Cohort at Time of Conscription (A) and for the Subpopulation at Conscription and at Follow-Up (B)

Exposure: IQ test results in stanines as a continuous variable	Moderate consumers versus light consumers		High consumers versus light consumers		Abstainers versus light consumers
	OR	95% CI	OR	95% CI	OR	95% CI
(A) Full cohort (cross-sectional) crude	1.07	1.06–1.08	1.20	1.17–1.23	1.02	1.00–1.04
Adjusted
Childhood SEP	1.08	1.07–1.09	1.22	1.19–1.25	1.02	1.00–1.04
Psychiatric symptoms	1.06	1.04–1.07	1.13	1.10–1.16	1.01	0.99–1.03
Emotional stability	1.05	1.04–1.06	1.13	1.10–1.06	1.01	0.99–1.03
Father's alcohol habits	1.07	1.06–1.08	1.20	1.17–1.23	1.03	1.01–1.05
Full modela	1.06	1.04–1.07	1.12	1.09–1.15	1.00	0.98–1.02

(B) Subpopulation			Moderate and high consumers versus light consumers		Abstainers versus light consumers
At age 18 (cross to sectional)			1.38	1.07–1.77	0.99	0.73–1.34
At middle age (follow to up)			1.22	1.01–1.48	1.14	0.88–1.47

^aAdjusted for childhood socioeconomic position (SEP), psychiatric symptoms, emotional stability, and father's alcohol habits.