Table 2.

Results of negative binomial regression model: coefficients and P-values, years 2004 and 2015

	2004 N = 25 016		2015 N = 43 916
	Coefficient	P-value	Coefficient	P-value
Age	0.022	0.000	0.020	<0.001
Gendera
Female	0.122	0.000	0.064	<0.001
Marital statusa
Couple	0.050	0.006	0.023	0.175
Working statusa
Retired	0.317	0.000	0.330	<0.001
Unemployed	0.202	0.000	0.203	0.001
Disabled	0.709	0.000	0.723	<0.001
Homeworker	0.264	0.000	0.299	<0.001
Other	0.349	0.000	0.411	<0.001
Educationa
Lower secondary	–0.045	0.050	–0.055	0.010
Upper secondary	–0.074	0.001	–0.088	<0.001
Tertiary	–0.058	0.028	–0.133	<0.001
Net wortha
2nd quintile	–0.063	0.006	–0.036	0.079
3rd quintile	–0.083	0.001	–0.091	<0.001
4th quintile	–0.080	0.001	–0.114	<0.001
5th quintile	–0.150	0.000	–0.179	<0.001
Incomea
2nd quintile	0.026	0.252	–0.011	0.596
3rd quintile	0.014	0.558	–0.015	0.500
4th quintile	0.015	0.565	–0.028	0.287
5th quintile	–0.035	0.232	–0.046	0.118
Smoking statusa	0.107	0.000	0.109	<0.001
Physical inactivitya	0.135	0.000	0.195	<0.001
Obesitya	0.294	0.000	0.348	<0.001
Countrya
Austria	–0.254	<0.001	0.130	<0.001
Germany	–0.060	0.031	0.311	<0.001
Sweden	–0.045	0.088	0.019	0.462
Spain	0.025	0.342	0.179	<0.001
France	–0.060	0.015	0.184	<0.001
Denmark	0.022	0.468	0.206	<0.001
Greece	–0.141	<0.001	0.107	<0.001
Switzerland	–0.338	<0.001	–0.035	0.267
Belgium	0.003	0.915	0.283	<0.001
LR test H0:α = 0b		<0.001		<0.001

^aReference categories: ‘Italy’ for country, ‘male’ for gender, ‘not-in-couple’ for marital status, ‘employed’ for working status, ‘primary’ for education, ‘1st quintile’ for income and net worth, ‘no’ for smoking status, physical inactivity and obesity.

^bα is the overdispersion parameter for negative binomial regression model.

Source: Share survey release 6.0.0, 2004 and 2015.