Table 2.

Episodic drinking in relation to job loss experience and duration, both during (September 2008–December 2013) and before (January 2004–August 2008) the post-2008 economic crisis in The Netherlands

	Association before economic crisis OR (95% CI)	Association during economic crisis OR (95% CI)	p interaction
All
Job loss experience
Never	Ref	Ref	Ref
Former	0.96 (0.65; 1.41)	1.15 (0.85; 1.56)	0.461
Current	0.87 (0.53; 1.44)	1.24 (0.87; 1.77)	0.265
Job loss duration
0 months	Ref	Ref	Ref
1–6 months	1.14 (0.76; 1.73)	0.98 (0.70; 1.37)	0.569
>6 months	0.73 (0.47; 1.15)	1.40 (1.01; 1.94)	0.023
Men
Job loss experience
Never	Ref	Ref	Ref
Former	1.03 (0.67; 1.58)	1.14 (0.82; 1.59)	0.704
Current	0.90 (0.53; 1.55)	1.34 (0.91; 1.98)	0.244
Job loss duration
0 months	Ref	Ref	Ref
1–6 months	1.25 (0.79; 2.00)	1.02 (0.72; 1.47)	0.501
>6 months	0.77 (0.48; 1.25)	1.43 (1.00; 2.04)	0.042
Women
Job loss experience
Never	Ref	Ref	Ref
Former	0.65 (0.24; 1.80)	1.12 (0.52; 2.39)	0.409
Current	0.61 (0.14; 2.60)	0.70 (0.25; 1.97)	0.875
Job loss duration
0 months	Ref	Ref	Ref
1–6 months	0.75 (0.27; 2.08)	0.72 (0.28; 1.82)	0.946
>6 months	0.47 (0.11; 1.99)	1.19 (0.53; 2.66)	0.274

Each of the six blocks (from “ALL/job loss experience to WOMEN/job loss duration”) is based on a separate regression model. Each model included period of survey, either one of the unemployment variables, and the corresponding interaction term. Reported associations were estimated conditioned on crisis, i.e., one model run holding crisis constant before crisis and one holding crisis constant during crisis. Age, sex, education, country of origin, marital status, and household composition were included as control variables

Bold ORs and p-values are statistically significant (<0.05)