Our study covered 310 million person-years in people aged from 0 to 95+, with 3,161,490 deaths that occurred in Belgium between 1991 and 2020. Deprivation was defined by deciles of housing deprivation indices 1991, 2001, and 2011 on the level of statistical sector (Fig. 1). During the study period, ASMRs decreased for men and women in all deprivation deciles. Reduction in absolute ASMR was greatest for males in the most deprived decile (-406 per 100,000 person-years) and females in the least deprived deciles (-293 per 100,000 person-years), whereas relative reductions were greatest for males (-31%) and females (-25%) in the least deprived deciles (Table 2). Comparing the period 2011–2020 with 1991–2000, the gap between the most and least deprived deciles has narrowed for men, but widened for women.
Table 2
Trends in mortality inequalities according to housing deprivation deciles
| 1991–20001 | 2001–2010 | 2011–2020 | Change (%)4 |
Age-standardised mortality rates per 100 000 person-years and 95% CI |
Male | | | | |
Most deprived2 | 1649 [1634–1664] | 1495 [1484–1506] | 1243 [1231–1254] | -24.62 |
Least deprived3 | 1193 [1181–1207] | 923 [911–935] | 826 [815–837] | -30.76 |
Difference | 456 | 572 | 417 | -8.55 |
Female | | | | |
Most deprived | 1402 [1389–1415] | 1271 [1262–1281] | 1140 [1130–1150] | -18.69 |
Least deprived | 1156 [1144–1170] | 1027 [1013–1041] | 863 [851–875] | -25.34 |
Difference | 246 | 244 | 277 | 12.60 |
Mortality attributable to housing inequality (%) |
Male | 21.0 [20.12–21.85] | 26.5 [25.13–27.99] | 19.4 [18.33–20.47] | -7.6 |
Female | 12.6 [11.61–13.51] | 11.27 [9.48–13.02] | 15.5 [14.47–16.75] | 23.2 |
Mean number of Potential Years of Life Lost attributable to housing inequality |
Male | 1.50 | 1.58 | 0.92 | -38.7 |
Female | 0.61 | 0.69 | 0.46 | -24.6 |
Mortality is adjusted to the Belgian population structure in 2019. |
1 Housing deprivation index 1991 applied for years 1991–2000, index 2001 for 2001–2010, and index 2011 for 2011–2020 |
2 Most deprived refers to the most deprived decile of the housing deprivation indices 1991, 2001, 2011. |
3 Least deprived refers to the least deprived decile of the housing deprivation index 1991, 2001, 2011. |
4 Change in mortality between 1991–2000 and 2011–2020 |
The proportion of mortality attributable to housing inequality varied by time period, sex and age. Across all years under study, the PAF was higher in men than in women, increasing first sharply from 21.0% (95% CI, 20.1–21.9) to 26.5% (95% CI, 25.1–28.0) in the periods 1991–2000 and 2001–2011, then decreasing to 19.4% (95% CI, 18.3–20.5) in the period 2011–2020. In women, mortality attributable to inequality has been increasing over time from 12.6% (95% CI, 11.6–13.5) to 15.5% (95% CI, 14.5–16.8) between 1991–2000 and 2011–2020. In the most deprived deciles, the proportion of deaths attributable to housing deprivation peaked in middle-aged men and women and accounted for more than a half of all deaths occurring in the middle-aged population (Figures S1, S2 and S3 in Supplementary Information).
If the whole population of Belgium presented the same risk of mortality as the least deprived groups measured by the HDI in the three reference years 1991, 2001, 2011, 568,605 fewer deaths would have occurred between 1991 and 2020. This represents about 18% of all deaths that could be attributable to housing deprivation (Fig. 2).
The results of our life table calculation suggested that on average, men and women in the least deprived deciles live longer than in the most deprived deciles in all observed periods. For the period 1991–2000, men and women in the least deprived deciles were expected to live 78.1. and 82.8 years compared to 72.7 and 79.9 years of men and women in the most deprived deciles (Table 3). The gap in life expectancy between the least and most deprived deciles was 5.4 and 2.9 years for men and women. In the following periods of 2001–2010 and 2011–2020, the gaps between the most and least deprivation groups for men and women increased to 7.1 and 4.0 years, and consequently decreased to 5 and 3.3 years, respectively.
Table 3
Life expectancies at birth in Belgium by the most and the least housing deprivation deciles in 1991–2000, 2001–2010, and 2011–2020.
| 1991–20001 | 2001–2010 | 2011–2020 | Change in time (in years) |
Male | | | | |
Most deprived2 | 72.72 | 73.90 | 76.98 | 4.26 |
Least deprived3 | 78.07 | 80.96 | 81.96 | 3.89 |
Difference | 5.35 | 7.06 | 4.98 | |
Female | | | | |
Most deprived | 79.88 | 80.68 | 82.53 | 2.65 |
Least deprived | 82.78 | 84.65 | 85.84 | 3.06 |
Difference | 2.90 | 3.97 | 3.31 | |
Mortality is adjusted to the Belgian population structure in 2019. |
1 Housing deprivation index 1991 applied for years 1991–2000, index 2001 for 2001–2010, and index 2011 for 2011–2020 |
2 Most deprived refers to the most deprived decile of the housing deprivation indices 1991, 2001, 2011. |
3 Least deprived refers to the least deprived decile of the housing deprivation index 1991, 2001, 2011. |
4 Change in mortality between 1991–2000 and 2011–2020 |
Over the three decades, life expectancy at birth increased for all deciles, but not in the same way. Using the Arriaga method, we compared how differences in age-specific mortality contributed to the life expectancy gap between the most and least deprived deciles in each period under study. The increase in life expectancy is observed in all age groups, but it is driven mostly by declines in mortality in the age group 60–69 in men and in the age group 70–79 in women in all three decades. These age groups contribute about 25% to the difference in life expectancy between the most and least deprived deciles in 1991–2000, 2001–2010, and 2011–2020 (Table S3 in Supplementary Information).
The results of our life table modeling also revealed that premature mortality due to housing deprivation caused a mean of 4.7, 3.74 and 3.1 PYLLs per person in 1991–2000, 2001–2010 and 2011–2020. Sex-stratified results showed that mean number of PYLLs has always been greater for men than women across all deciles, and it had a declining trend in both sexes (Fig. 3). The mean PYLL for men and women was 6.1 and 3.2 in the period 1991–2000, 4.8 and 2.7 in 2001–2010, and 3.9 and 2.3 years in 2011–2020, respectively. If everyone had the mortality risk of the least deprived groups measured by the HDI 1991, 2001, and 2011, the PYLLs would decrease to a 3.6, 2.6 and 2.4 years per person. The average number of PYLLs attributable to housing deprivation is shown for each sex in Table 1.
Our results also suggested that 23%, 30% and 22% of PYLLs can be attributed to inequality in years 1991, 2001 and 2011. In both sexes, this proportion peaked in the period 2001–2010, equally with 68% in men and women.