Do changes in social and economic factors lead to changes in drinking behavior in young adults? Findings from three waves of a population based panel study

Table 4 Fixed effects linear regression model of living arrangement, education/employment, income, and deprivation as exposure variables and frequency of hazardous drinking episodes per month as outcome variable

Model	1: Crude models^b		2: Full model^c
Variable	β	95% CI	β	95% CI
Living arrangement
Other multiperson household	0.49	0.00,0.97	0.29	-0.20,0.79
One person household	0.50	0.04,0.97	0.33	-0.14,0.81
Couple only or with others	-0.28	-0.70,0.13	-0.42	-0.85,0.00
Family household (in parent role)	-0.99	-1.58,-0.41	-1.17	-1.76,-0.57
Family household (in child role) (ref.)	0		0
p-value ^a		< 0.0001		< 0.0001
Education/Employment
Not in education - employed	-0.01	-0.43,0.41	-0.18	-0.65,0.30
Not in education – not employed (NEET)	0.09	-0.44,0.63	-0.09	-0.63,0.45
In education - employed	-0.08	-0.50,0.34	-0.20	-0.66,0.27
In education -not employed (ref.)	0		0
p-value ^a		0.9066		0.8668
Increase in log personal income	0.12	-0.03,0.26	0.12	-0.05,0.29
p-value ^a		0.1064		0.1512
One decile increase in area deprivation	0.06	0.01,0.11	0.06	0.00,0.12
p-value ^a		0.0277		0.0513
One unit increase in individual deprivation	0.20	0.07,0.33	0.25	0.11,0.39
p-value ^a		0.0030		0.0005

^aType III Wald tests, which for multichotomous categorical variables (e.g. labour force status) provides a statistical test of the whole construct (not just one non-referent compared to referent comparison).
^bThe crude models include each of the independent variables individually.
^cThe full model includes all independent variables.

ISSN: 1471-2458