The impact of housing conditions on mortality in Belgium (1991–2016)

Abstract

In the Western countries, including Belgium, life expectancy has increased over the last decades, as well as social inequalities in health and mortality. Existing research tends to approximate socioeconomic status with the educational level, occupational status or income. Housing is yet another socioeconomic factor that is much less considered when studying inequalities in mortality. Indeed, housing is a complex and multidimensional element impacting several aspects of a person’s health and well-being (physical, mental and social). The data used in this research are the result of the coupling of the population censuses of 1991, 2001 and 2011 and the National Register. They cover the entire population of Belgium over 25 years. Through life tables analysis and multivariate logistic regression models, this research contributes to existing research by setting trends in housing conditions between 1991 and 2016 in relation to social inequalities in health and mortality over this period in Belgium. It shows that housing conditions have a specific effect on the risk of death, in addition to other socio-economic characteristics. On the 2011–2015 period, once controlling for demographic and socioeconomic characteristics (education, professional status and income), a 25% higher mortality rate separates tenants from owners on the one hand, and population living in low quality housing from those living in better quality housing on the other hand. Ensuring good housing conditions seems a necessary step to reduce inequalities that should be considered in social policies.

Appendices

Appendix 1

Distribution of the housing score in the Census Population (2011). Source: National Register, 2011 Population Census—calculation by DEMO-UCL and author

Housing score	Freq.	Percentage	Cumulative percentage
0.000	3334	0.03	0.03
0.429	8512	0.07	0.1
0.500	704	0.01	0.1
0.571	698	0.01	0.11
0.600	1024	0.01	0.12
0.800	3677	0.03	0.15
0.857	6163	0.05	0.2
1.000	2558	0.02	0.22
1.143	35,658	0.29	0.51
1.200	727	0.01	0.51
1.333	50,329	0.41	0.93
1.400	6103	0.05	0.98
1.429	1121	0.01	0.98
1.571	1,38,079	1.13	2.11
1.600	3908	0.03	2.14
1.714	1,77,233	1.45	3.59
1.833	1,14,821	0.94	4.53
2.000	95,189	0.78	5.3
2.143	7,87,954	6.43	11.74
2.200	2201	0.02	11.75
2.286	1,56,200	1.27	13.03
2.333	62,356	0.51	13.54
2.400	49,978	0.41	13.95
2.571	5,19,059	4.24	18.18
2.667	1,20,224	0.98	19.16
2.714	3,24,331	2.65	21.81
2.800	1630	0.01	21.82
2.857	7,74,104	6.32	28.14
3.000	5544	0.05	28.19
3.143	2,39,492	1.95	30.14
3.167	1,11,669	0.91	31.05
3.200	2959	0.02	31.08
3.286	26,70,853	21.8	52.88
3.429	2,60,227	2.12	55
3.600	2588	0.02	55.02
3.667	60,431	0.49	55.52
3.714	17,20,333	14.04	69.56
4.000	16,89,890	13.79	83.35
99.000	7,39,405	6.04	89.39
	12,99,933	10.61	100

Appendix 2: Logistic regression models

See Tables 6, 7, and 8.

Table 6 Odds ratio and significativity on the risk of death on the 1992–1996 period

Table 7 Odds ratio and significativity on the risk of death on the 2002–2006 period

Table 8 Odds ratio and significativity on the risk of death on the 2011–2015 period