In Belgium, in 2018, strong socio-economic inequalities were observed in many health determinants.
People with a low educational level (EL) were three times more likely to be daily smokers and twice more likely to be obese or daily consumers of sugary drinks than high educated people.
People with a low EL were also twice less likely to have a sufficient fruits/vegetables consumption or to practice enough physical activity as compared to high EL.
Between 1997 and 2018, inequalities in daily smoking increased, while no clear trend was observed for inequality in the other health determinants studied.
2.Background
Socio-economic (SE) health inequalities refer to systematic disparities in health between SE groups, most often in disfavour of those on the lower position of the social and/or economic scale. SE health inequalities have been consistently observed in industrialized societies for the whole scope of health-related indicators, ranging from health determinants to mortality [1;2]. Tackling health inequalities is a priority for the WHO [3], the European Union [4], and for Belgium [5-7]. In order to assess progress towards reducing health inequalities, it is important to measure and monitor them [8,9].
Inequalities in health determinants have been computed from the data of the Health Interview Surveys 1997-2018. More information on the methodology can be found in this annexe. The health determinants chosen were daily smoking, obesity, performing at least 150 min of moderate to vigorous physical activity per week, daily consumption of at least 5 portions of fruits and vegetables, and daily consumption of sugar-sweetened beverages. The educational level (EL), grouped in three categories, was chosen as marker of socio-economic position to examine inequalities.
Beside the prevalence rates by SE level, we also calculated the magnitude of the inequalities by computing three inequality indices:
absolute and relative difference in age-adjusted rates between the low and the high ELs,
Population Attributable Fraction (PAF), i.e. the percentage of gain in health (or health determinant) expected in the whole population if all groups experienced the health (or health determinant) of the highest educated group.
3.Results
Situation in 2018
Daily smoking presents very large inequalities. While daily smokers represented 27.5% of the low educated group in 2018, they were only 9.4% in the high educated group. This represents an absolute difference of 18.1 percentage points (ppt), and a relative risk of almost 3, meaning that people with a low EL were three times more likely to smoke daily than the high educated people. If each EL had the level of smoking of the high educated people, the prevalence of smoking would decrease by 37.5% in the population.
The absolute difference in obesity prevalence between the low and high EL was 10 ppt, with people from the low EL being 1.8 times more likely to be obese than people with a high EL. The prevalence of obesity in the whole population would reduce by 22.7% if each EL had the same percentage of obesity as the high educated group.
Physical activity also shows very large inequalities; as it is a positive indicator for which we thrive for a higher prevalence, the values of the inequality indices are inversed, that is absolute inequalities will show values inferior to zero and relative inequalities values inferior to one. In 2018, 38.5% of high educated people practiced at least 150 min of moderate to vigorous aerobic physical activity per week, versus 18.8% in the low EL, representing an absolute difference of 19.7 ppt. People with a low EL were twice less likely to be physically active than people with a high EL. Bringing the physical activity practice of all ELs to the level of the high educated group would increase the physical activity practice by 23.9% at population level.
Large inequalities were also observed in nutrition. Twice more people were meeting the target of consuming 5 portions of fruits/ vegetables per day among people with a high EL than among those with a low EL, corresponding to an absolute difference of 8.4 ppt. Bringing the consumption of fruits/vegetables of all EL to the one of people in the high EL would increase the fruits/vegetable consumption in the whole population by 33.5%. People with a low EL were twice more likely to drink daily sugar-sweetened beverages than people from the high EL, the absolute difference between the 2 groups was 15.4 ppt. If people from all EL would reduce their consumption of sugar-sweetened beverages to the level of the high EL then the overall consumption level would decrease by 31.5%.
It is noteworthy that alcohol consumption does not present the same SE pattern as most health determinants. The SE pattern of excess consumption of alcohol is unclear and inconclusive.
Socio-economic inequalities in selected health determinants, Belgium, 2018 Source: Own calculation based on Health Interview Survey [10] * statistically different from 0% for absolute difference and PAF, and statistically different from 1 for the relative difference (p<0.05)
Age-adjusted prevalence rate low EL
Age-adjusted prevalence rate high EL
Absolute difference
Relative difference
PAF
Daily smoking (%people ≥ 15)
27.5%
9.4%
18.1%*
2.9*
37.5%*
Obesity (%people ≥ 18, BMI ≥ 30)
22.0%
12.0%
10.0%*
1.8*
22.7%*
At least 150 min of physical activity per week (%people ≥ 18)
18.8%
38.5%
-19.7%*
0.5*
-23.9%*
Daily consumption of 5 portions of fruits and vegetables (%people ≥ 6)
8.2%
16.6%
-8.4%*
0.5*
-33.5%*
Daily consumption of sugar-sweetened beverages (%people all age)
29.3%
13.9%
15.4%*
2.1*
31.5%*
Trends
The age-adjusted prevalence of daily smoking is decreasing between 1997 and 2018, but almost exclusively due to a strong decrease among people with a high EL, with few changes in the other ELs. Consequently, inequalities have clearly increased over time according to all three inequality indices (absolute, relative and PAF).
The age-adjusted prevalence of obesity increased between 1997 and 2018 in all ELs. A slight non-significant increase in absolute difference was observed, with no remarkable trends in relative inequality. The PAF decreased as the share of people pertaining to the low ELs (and with a high prevalence) has decreased.
The daily consumption of sugary drinks has decreased between 2013 and 2018 for all groups. We observe a slight non-significant decrease in absolute difference; the relative difference and PAF on the other hand remained constant.
The indicators used to assess physical activity and the consumption of fruits and vegetables in the HIS 2018 were new indicators, and therefore no trend can be described.
Daily smoking
Obesity
Sugary drinks
Absolute difference
Relative difference
PAF
Prevalence of daily smoking among people aged 15 and over, by educational level, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of obesity among people aged 18 and over, by educational level, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Proportion of the population that drinks sugary drinks daily, by educational level, Belgium, 2013-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Absolute low-versus-high EL inequalities in health determinants indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Relative low-versus-high EL inequalities in health determinants indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
PAF in health determinants indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
The Percentage-point (ppt) is the arithmetic difference between two percentages, for instance with 16% in group A and 8% in group B, the difference is 8 ppt, corresponding to a relative excess of 100%.
References
Mackenbach J. Health inequalities: Europe in profile. Expert Report commissioned by the EU. Department of Health Publications; 2006.
Feinstein JS. The relationship between socioeconomic status and health : A review of the literature. The Milkbank Quarterly. 1993
WHO Commission on Social Determinants on Health. Closing the gap in a generation: health equity through action on the social determinants of health. Geneva: WHO; 2008.
Executive Agency for Health and Consumer. Second Programme of Community Action in the Field of Health 2008-2013. European Commission; 2007.
Vlaamse overheid. Vlaamse Actieplan Geestelijke Gezondheid, Strategisch plan 2017-2019. 2017.
Gouvernement wallon. Plan prévention et promotion de la santé en Wallonie. Partie 1: définition des priorités en santé. Namur; 2017.
The prevalence of anxiety and depressive disorders and of suicidal thoughts differs by socio-economic group, with a higher prevalence among people in the lowest than in the highest socio-economic level.
In 2018, socio-economic inequalities in mental health conditions were larger than for the physical health conditions with relative differences around 2.
When looking at the evolution, absolute inequalities in anxiety and depressive disorders have strongly increased between 2008 and 2013, and stayed stable at a higher level between 2013 and 2018, which is a disappointing evolution. Between 2013 and 2018, the relative inequalities in depressive disorders have also worsened.
2.Background
Socio-economic (SE) health inequalities refer to systematic disparities in health between SE groups, most often in disfavour of those on the lower position of the social and/or economic scale. SE health inequalities have been consistently observed in industrialized societies for the whole scope of health-related indicators, ranging from health determinants to mortality [1;2]. Tackling health inequalities is a priority for the WHO [3], the European Union [4], and for Belgium [5-7]. In order to assess progress towards reducing health inequalities, it is important to measure and monitor them [8,9].
Inequalities in mental health (MH) have been computed from the data of the Belgian Health Interview Surveys 1997-2018. Methodological details are given in this annexe. The inequalities are described for three indicators: the prevalence of depressive disorders (based on the PHQ-9 scale) and anxiety disorders (based on the GAD-7 scale) in the last two weeks were selected in reason of their high frequency; the prevalence of suicidal thoughts in the last twelve months was also chosen because it indicates severe mental distress, and the suicide rate is quite high in Belgium. The educational level (EL) (in three levels) was chosen as a marker of the socio-economic position to examine inequalities.
Beside the prevalence of mental health conditions by socio-economic level, we also calculated the magnitude of the inequalities by computing inequality indices, we considered :
the absolute and relative differences in age-adjusted prevalence rates between the lowest and the highest ELs,
the Population Attributable Fraction (PAF), i.e. the percentage of gain in health expected in the whole population if all groups experienced the health of the most educated group.
3.Results
Situation in 2018
In 2018, absolute inequalities in mental health conditions were larger than for physical conditions, ranging from 9.8 percentage points (ppt) for depressive disorders to 3.5 ppt for suicidal thoughts.
Relative inequalities in mental health conditions were particularly high. The lowest EL group included 2.5 times more people with depressive disorders, 2.0 times more people having suicidal thoughts, and 1.8 times more people experiencing anxiety disorders than the highest EL group. If all groups had the same proportion of people with mental health conditions as the highest EL group, then the prevalence of depressive disorders in the whole population would be reduced by 30%, anxiety disorders by 24%, and suicidal thoughts by 20%.
Socio-economic inequalities in mental health conditions, people aged 15 years and over, Health Interview Survey, Belgium, 2018 Source: Own calculation based on Health Interview Survey [10] * statistically different from 0% for absolute difference and PAF, and statistically different from 1 for the relative difference (p<0.05)
Age-adjusted prevalence rate low EL
Age-adjusted prevalence rate high EL
Absolute difference
Relative difference
PAF
Depressive disorders in the last two weeks (% people ≥ 15)
16.2%
6.4%
9.8%*
2.5*
30.2%*
Anxiety disorders in the last two weeks (% people ≥ 15)
15.3%
8.4%
6.9%*
1.8*
23.8%*
Suicidal thoughts in the last twelve months (% people ≥ 18)
6.9%
3.4%
3.5%*
2.0*
19.9%*
Trends
Trends in anxiety disorders:
The age-adjusted prevalence has increased in all ELs from 2008 to 2018.
The absolute inequalities have increased between 2008 and 2013 then stayed stable.
The relative inequalities have slightly and non-significantly decreased from 2004 to 2018.
In conclusion, there is no worsening in the inequalities related to anxiety disorders in the last 5 years. The evolution is however not satisfactory, as a decrease in absolute inequalities would be the minimal progress expected.
Trends in depressive disorders:
The age-adjusted prevalence has increased in all ELs from 2004 to 2013 and then decreased in 2018.
The absolute inequalities have increased between 2008 and 2013 then stayed stable.
The relative inequalities have decreased from 2004 to 2013 then increased in 2018. This can be attributed to a smaller proportional decrease of the prevalence of depressive disorders in the low than in the high EL.
In conclusion, while the prevalence of depressive disorders improved in all ELs during the last 5 years, the evolution of inequalities is disappointing: the absolute inequalities did not decline (what would be the minimal improvement aimed, and would correspond to a larger decrease in depressive disorders in the less than in the more advantaged people), and the relative inequalities even increased.
The age-adjusted prevalence of suicidal thoughts has increased in all ELs from 2008 to 2018. Inequality in suicidal thoughts does not show notable trends.
When looking at inequalities at population level, the Population Attributable Fraction (PAF) tends to decrease since 2004 for all indicators, which is partly due to a change in the population composition, with the low educated groups containing a decreasing share of the population over time. In the last 5 years period however, it remained stable for depressive disorders, as the change in the population composition was compensated by an increase of the relative inequalities.
Anxiety disorders
Depressive disorders
Suicidal thoughts
Prevalence of anxiety disorders in the last 2 weeks (based on the GAD-7 scale) among people aged 15 and over by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of depressive disorders in the last 2 weeks (based on the PHQ-9 scale) among people aged 15 and over by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of suicidal thoughts in the last twelve months among people aged 18 and over by educational level, 2008-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Absolute difference
Relative difference
PAF
Absolute differences in anxiety and depressive disorders among low-versus-high EL groups, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Relative differences in anxiety and depressive disorders among low-versus-high EL groups, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
PAF in anxiety and depressive disorders, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
The Percentage-point (ppt) is the arithmetic difference between two percentages, for instance with 16% in group A and 8% in group B, the difference is 8 ppt, corresponding to a relative excess of 100%.
References
Mackenbach J. Health inequalities: Europe in profile. Expert Report commissioned by the EU. Department of Health Publications; 2006.
Feinstein JS. The relationship between socioeconomic status and health : A review of the literature. The Milkbank Quarterly. 1993
WHO Commission on Social Determinants on Health. Closing the gap in a generation: health equity through action on the social determinants of health. Geneva: WHO; 2008.
Executive Agency for Health and Consumer. Second Programme of Community Action in the Field of Health 2008-2013. European Commission; 2007.
Vlaamse overheid. Vlaamse Actieplan Geestelijke Gezondheid, Strategisch plan 2017-2019. 2017.
Gouvernement wallon. Plan prévention et promotion de la santé en Wallonie. Partie 1: définition des priorités en santé. Namur; 2017.
In 2018, as observed previously, socio-economic inequalities were present for being affected by a non-communicable disease. The percentage of people reporting a chronic illness or condition was highest among people of the lowest socioeconomic level, and this percentage decreased as the socio-economic position increases. For this general indicator “reporting a chronic disease’”, the socio-economic inequalities were low (after adjustment for age).
However, the inequalities in suffering from multiple conditions together (multimorbidity) were larger, meaning that people of the lower SE are more prone to cumulate health problems.
In 2018 inequalities were also observed in many specific chronic conditions, namely for osteoarthritis, high blood pressure, urinary incontinence in people aged 65+, migraine-like headache, chronic obstructive pulmonary disease (COPD) in people aged 65+, diabetes, asthma and acute myocardial infarction (AMI) in 65+.
Over time, the inequalities in reporting a chronic condition or multimorbidity have fluctuated. They have decreased in 2018 as compared to 2013.
For most of the specific conditions also, inequalities did not increase or even tended to slightly decrease between 2013 and 2018. For diabetes and COPD the inequalities tend to have slightly decreased already since 2008; for asthma a decrease was observed since 2013.
2.Background
Socio-economic (SE) health inequalities refer to systematic disparities in health between SE groups, most often in disfavour of those on the lower position of the social and/or economic scale. SE health inequalities have been consistently observed in industrialized societies for the whole scope of health-related indicators, ranging from health determinants to mortality [1;2]. Tackling health inequalities is a priority for the WHO [3], the European Union [4], and for Belgium [5-7]. In order to assess progress towards reducing health inequalities, it is important to measure and monitor them [8,9].
Inequalities in non-communicable diseases (NCDs) have been computed from the data of the Belgian Health Interview Surveys 1997-2018. The educational level (EL) (in three categories: low, mid, high) was chosen as a marker of the socio-economic position to examine inequalities. Methodological details can be found in this annexe. Beside the prevalence of NCDs by socio-economic level, we also calculated the magnitude of the inequalities by computing three inequality indices:
The absolute difference, which is the difference between the age-adjusted prevalence rates in the low versus the high ELs,
The relative difference (Rate Ratio), which is the ratio of the age-adjusted prevalence rates in the low versus the high ELs,
The Population Attributable Fraction (PAF), i.e. the percentage of gain in health expected in the whole population if all groups experienced the health of the most educated group.
3.Results
Situation in 2018
SE inequalities for the global indicator “suffering from a chronic condition” are small: after age-adjustment, chronic conditions are reported by 31% of people of the low EL versus 27% of people in the high EL, leading to a rate difference of 3.9 percentage-point (ppt) and a rate ratio of 1.1. However, the inequalities are larger if the co-occurrence of several chronic conditions (multimorbidity) is considered, with a rate difference reaching 5.7 ppt and a rate ratio of 1.4 (meaning a 40% excess of multimorbidity in people from the lowest versus people of the highest EL).
When looking at each condition separately, we observe SE inequalities in many conditions. This is the case for osteoarthritis, high blood pressure (in people aged 65+), migraine-like headache, chronic obstructive pulmonary disease (COPD) in people aged 65+, urinary incontinency in people aged 65+, diabetes, asthma, acute myocardial infarction (AMI) in 65+.
The absolute inequalities (rate differences) were moderate, ranging from 1.4 ppt for AMI (in 65+) to 6.6 ppt for urinary incontinence (in 65+).
The relative inequalities were:
large for AMI (in 65+) and COPD (in 65+), with respectively 2.0 and 1.9 times more people suffering from AMI (65+) and COPD (65+) in the low than the high EL group;
moderate (between 1.4 and 1.6) for urinary incontinence (65+), migraine-like headache, diabetes, and asthma;
small (between 1.1 and 1.3) for osteoarthritis, and high blood pressure.
If all EL groups had the same level as the high EL group then the prevalence of AMI (65+), COPD (65+), urinary incontinence (65+), and diabetes would be respectively reduced by 37.8%, 28.6%, 21.5%, and 20.3% in the whole population.
Socio-economic inequalities in selected non-communicable diseases, people aged 15 years and over, Health Interview Survey, Belgium, 2018 Source: Own calculation based on Health Interview Survey [10] * statistically different from 0% for absolute difference and PAF, and statistically different from 1 for the relative difference (p<0.05)
% reporting urinary incontinence among people aged 65+
17.5%
10.9%
6.6%*
1.6*
21.5%*
% reporting migraine-like headache
12.1%
8.8%
3.3%*
1.4*
13.3%*
% reporting COPD among people aged 65+
11.1%
5.8%
5.4%*
1.9*
28.6%*
% reporting diabetes
7.5%
4.7%
2.9%*
1.6*
20.3%*
% reporting asthma
7.2%
4.8%
2.4%*
1.5*
16.8%*
% reporting myocardial heart infarction among people aged 65+
2.8%
1.4%
1.4%
2.0
37.8%
Trends
The percentage of people reporting one or more chronic diseases (multimorbidity) has increased since their first estimate in the HIS (2001), even after age-adjustment. The prevalence of many specific diseases has also increased (however, it decreased for COPD). The evolution of these percentages differs by SE level and by pathology, leading to changes in the SE inequalities by NCDs.
For reporting a chronic condition, the inequalities (as measured with all three inequality indices) strongly increased from 2001 to 2013, reaching in 2013 a large level of absolute inequality (10 ppt). In the last period (2013 to 2018), the inequalities in reporting a chronic condition, as measured with all three inequality indices, dropped.
For multimorbidity, a decrease of inequality was observed in 2013 and further in 2018 as compared to 2004 with all three inequality indices.
When looking at the evolution of the prevalence and inequalities by specific condition, we observe that:
Inequalities indicators for urinary incontinence (65+), IMA (65+), high blood pressure and osteoarthritis have no notable trends.
The age-adjusted prevalence of migraine-like headache was stable between 2004-2013 and increased in 2018. The rate difference significantly decreased between 2001 and 2008, but then increased in 2013 and 2018 (increase not significant), with a same evolution observed for the relative inequalities (statistically not significant). So an increase (not significant) in inequalities is observed in migraine-like headache in the last period.
The age-adjusted prevalence of COPD (65+) decreased from 2001 to 2018, with variable evolutions by EL. The inequalities in COPD prevalence (measured both with rate difference and rate ratio) decreased between 2008 and 2018.
The age-adjusted prevalence of diabetes has increased in all EL groups since 1997, with a small stagnation in 2008. Qua inequalities, the low-versus-high absolute difference in rates remained stable since 2008 at a high level; the relative difference has decreased since 2008, reflecting a smaller proportional increase of the prevalence of diabetes in the low than in high EL. So, even if the RR has decreased in 2018, the global evolution of inequality for diabetes is still disappointing: indeed, it is important to obtain a reduction in absolute inequalities, which would require a more favorable evolution of the prevalence of diabetes in the socially disadvantaged group than in the advantaged group.
The age-adjusted prevalence of asthma was stable until 2013 and increased in 2018. The inequalities in asthma have increased from 2001 to 2013 (measured as rate difference and rate ratio) and decreased in 2018.
When looking at inequalities at population level, the Population Attributable Fraction (PAF) was particularly high for COPD (65+) and diabetes in 2008 and has decreased afterwards. For all indicators, the PAF tends to decrease since 2008 or since 2013, which is partly due to a decrease of the rate ratio, and to a change in the population composition, with the low educated group representing a decreasing share of the population over time.
Chronic condition
Multimorbidity
Migraine
COPD
Diabetes
Ashtma
Prevalence of chronic condition by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of multimorbidity by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of migraine-like headache by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of COPD (65+) by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of diabetes by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Prevalence of asthma by educational level, 1997-2018, Belgium Source: Own calculation based on Health Interview Survey, Sciensano [10]
Absolute difference
Relative difference
PAF
Absolute difference in NCDs indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
Relative difference in NCDs indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
PAF in NCDs indicators, Belgium, 1997-2018 Source: Own calculation based on Health Interview Survey, Sciensano [10]
The Percentage-point (ppt) is the arithmetic difference between two percentages, for instance with 16% in group A and 8% in group B, the difference is 8 ppt, corresponding to a relative excess of 100%.
Multimorbidity
The occurrence of at least 2 of the following diseases: chronic lung disease, heart disease, hypertension, diabetes, cancer, and arthropathy.
References
Mackenbach J. Health inequalities: Europe in profile. Expert Report commissioned by the EU. Department of Health Publications; 2006.
Feinstein JS. The relationship between socioeconomic status and health : A review of the literature. The Milkbank Quarterly. 1993
WHO Commission on Social Determinants on Health. Closing the gap in a generation: health equity through action on the social determinants of health. Geneva: WHO; 2008.
Executive Agency for Health and Consumer. Second Programme of Community Action in the Field of Health 2008-2013. European Commission; 2007.
Vlaamse overheid. Vlaamse Actieplan Geestelijke Gezondheid, Strategisch plan 2017-2019. 2017.
Gouvernement wallon. Plan prévention et promotion de la santé en Wallonie. Partie 1: définition des priorités en santé. Namur; 2017.
In the period 2015-2019, a strong income gradient of mortality was observed; and this was also the case with other socio-economic indicators.
During the first wave of the COVID-19 pandemic:
among people aged 40-64 years, mortality remained stable in each income group, resulting in no change in mortality inequalities.
among people aged 65+, mortality increased in all income groups, but the increase was higher among the disadvantaged groups, leading to an increase of the inequalities by income in the 65+ group.
The contribution of specific conditions to the inequalities in the premature mortality (under 75 years) among men was largest for lung cancer, ischemic heart diseases, suicide and chronic obstructive pulmonary diseases (COPD). Among women, these conditions are ischemic heart disease, lung cancer, cerebrovascular diseases and COPD.
2.Background
Socio-economic (SE) health inequalities refer to systematic disparities in health between SE groups, most often in disfavour of social groups lower on the social scale. SE health inequalities have been consistently observed throughout industrialized societies for the whole scope of health topics, ranging from health determinants to mortality [1;2]. Tackling health inequalities is a priority for the WHO [3], the European Union [4], and for Belgium [5-7]. In order to assess progress towards reducing health inequalities, it is important to measure and monitor them [8, 9].
SE inequalities can be calculated using different socio-economic markers, for instance, income level, educational level, occupation, or a multi-dimensional score combining several SE indicators. SE inequalities can be expressed in terms of absolute difference (here, the difference in mortality rates between the lowest and the highest advantaged group) or relative difference (here, the ratio of the mortality rates of the extreme groups). Methodological details are provided in this annexe.
Besides mortality rates by socio-economic groups, we also examined excess mortality rates to assess inequalities. Excess in mortality during the COVID-19 crisis compared to a reference mortality rate (e.g. the average mortality rate over the 5 previous years) can be expressed in absolute terms (as the difference in mortality rates between the COVID-19 period and the baseline period, here 2015-2019) or in relative terms, with a “p-score” (defined as the excess mortality rates divided by the baseline mortality rate). Several studies have described recent inequalities in all-cause mortality [10–12]. Results presented here mostly originate from the study of Decoster et al [11] and highlight the changes in income inequalities in mortality during the COVID-19 crisis. Inequalities in cause-specific mortality originate from previous studies [13,14].
3.Results
Inequalities in all-cause mortality
Inequalities by income
During the period 2015-2019
A strong income gradient in mortality was observed [11] in the pre-COVID-19 period (2015-2019).
For men aged 40-64 years, the absolute inequalities gradient in mortality, measured with the slope index of inequality (SII), reached 185 per 100,000 person-year, meaning that the mortality rate in the lowest income level exceeds the mortality rates of the highest income level by 185 per 100,000 person-year. The relative inequality, measured with the Relative Index of Inequality (RII), reached 5.3, meaning that the mortality rate (in this age group) was 5.3 times higher in the lowest than in the highest income level. For women aged 40-64 years, inequalities, although slightly smaller than in men, remained quite high, with the SII reaching 93 per 100,000 and the RII 3.9.
For men aged 65+, the absolute SII reached 596 per 100,000 person-year, with a RII of 1.76. For women aged 65+, the SII reached 499 per 100,000 person-year, with a RII of 2.05.
During the first wave of the COVID-19 crisis (March-May 2020)
The impact of the first wave of COVID-19 on inequalities in mortality depends on age groups.
For men and women aged 40-64 years, the mortality rates observed during the first COVID-19 wave did not change for all income groups compared with previous years, so inequalities did not change.
On the contrary, for men and women aged 65+, mortality rates increased significantly during the COVID-19 months for all income groups. This mortality jump was unequally distributed among income groups: in absolute terms, mortality rates increased more in the lowest (excess mortality of +350 per 100,000 person-year) than in the highest income level (excess mortality +150 per 100,000 person-year). At the same time, the relative excess mortality (that is the excess mortality in the studied year divided by the baseline mortality, or “P-score”) was rather similar across the different income levels, so no significant gradient could be observed. This smoother effect of unequal mortality change on relative than on absolute inequalities is expected when mortality is on the rise: indeed the proportional change is less affected by an increase in mortality rate in groups where mortality is high than in groups with a lower mortality rate.
This led to an increase of the SII, which jumped to respectively 791 and 672 per 100,000 person-year in men and women aged 65+. The increase in the RII was more modest, passing from 1.76 to 1.86 in men and from 2.05 to 2.31 in women.
Men
Women
Mortality rates by income deciles among men aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Mortality rates by income deciles among women aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Men
Women
Absolute excess mortality by income deciles among men aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Absolute excess mortality by income deciles among women aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Men
Women
Relative excess mortality (p-scores) by income deciles among men aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Relative excess mortality (p-scores) by income deciles among women aged 65+, period 2015-2019 versus the first wave of COVID-19 crisis, Belgium Source: Decoster et al. [11]
Other socio-economic determinants
In this section, we present results highlighting the changes observed during the first wave of the COVID-19 crisis (compared to previous years) in all-cause mortality inequalities according to other SE determinants.
For the 2011-2015 period, inequalities in all-cause mortality according to several SE determinants have been calculated by Aerden et al. and summarized as inequalities in life expectancy [10].
Change in inequalities by educational level, first COVID-19 wave (March-May 2020)
According to Decoster et al. [11], educational inequalities in mortality among people younger than 65 years, similarly to income inequality, remained unchanged during the COVID-19 pandemic.
For people aged 65+, the well-known negative educational gradient in mortality became stronger during the COVID-19 crisis and the change was even more pronounced than for the income gradient. The increase in educational inequalities was observed for both absolute and relative inequalities.
Change in inequalities by a multi-dimensional indicator, first two COVID-19 waves (March-May & October-November 2020)
Bourguignon et al. [12] looked at the relationship between excess mortality among the 80+ and a multi-dimensional SE status during the first two COVID-19 waves; she came to a different conclusion for people aged 80+ than for younger people, with a larger mortality increase in the advantaged than in the disadvantaged group. This observation might partly result from a health selection effect – hypothesizing that people in the lowest SE group who have reached older age are less vulnerable, resulting consequently in a reduction or even an inversion of health inequality – and partly because people with an undetermined SE status presented the highest mortality rate and are likely to pertain to the most disadvantaged group.
Inequalities in cause-specific mortality
During the period 2011-2015
A study by Eggerickx et al. [14] presents inequalities by groups of cause of death (COD) and age in the period 2011-2015, expressed as the ratio of the probability of dying in the different SE groups (those groups are derived from a multi-dimensional SE score distributed in quartiles) as compared to the highest SE group. This is a relative measure of inequality.
For each COD group and at each age, a social gradient in the mortality risk is observed: the higher the social group, the lower the risk of dying. The relative SE inequalities by COD groups were the highest for diseases linked to the respiratory system and the circulatory system.
The relative SE inequalities in mortality decrease with age. A health selection effect – hypothesizing that people in the lowest SE group who have reached older age are less vulnerable, reducing mortality inequality – might play a role. The age pattern in absolute and relative inequalities can partly be explained by the small number of deaths at young ages in both groups, which would always yield a small absolute difference, but might yield a large relative difference.
Contribution of specific causes of deaths to the inequality in premature mortality, period 2001-2006
A previous study by Renard et al. [13] examined the contribution of specific causes of deaths to the total inequality in mortality below 75 years at the level of the population. The contribution was highest in men for lung cancer, ischemic heart diseases, suicide and chronic obstructive pulmonary diseases (COPD); in women, the contribution of ischemic heart disease, lung cancer, cerebrovascular diseases and COPD was highest. This points out the causes of deaths for which the reduction in inequality would most benefit the whole population, by reducing the global premature mortality level.
Men
Women
Ranking of the causes of deaths by their contribution to the inequalities in premature mortality, measured as Population attributable fraction, men, Belgium, 2001 Source: Own calculations based on census 2001 linked with a 5-years' mortality follow-up
Ranking of the causes of deaths by their contribution to the inequalities in premature mortality, measured as Population attributable fraction, women, Belgium, 2001 Source: Own calculations based on census 2001 linked with a 5-years' mortality follow-up
Mackenbach J. Health inequalities: Europe in profile. Expert Report commissioned by the EU. Department of Health Publications; 2006 p.
Feinstein JS. The relationship between socioeconomic status and health : A review of the literature. The Milkbank Quarterly. 1993.
WHO Commission on Social Determinants on Health. Closing the gap in a generation: health equity through action on the social determinants of health. Geneva: WHO; 2008.
Executive Agency for Health and Consumer. Second Programme of Community Action in the Field of Health 2008-2013. European Commission; 2007.
Vlaamse overheid. Vlaamse Actieplan Geestelijke Gezondheid, Strategisch plan 2017-2019. 2017.
Gouvernement wallon. Plan prévention et promotion de la santé en Wallonie. Partie 1: définition des priorités en santé. Namur; 2017.
Braveman PA. Monitoring equity in health and healthcare: a conceptual framework. JHealth PopulNutr. 2003.
Maeseneer JD, Willems S. Terugdringen Sociale Gezondheidskloof: van concept naar politieke implementatie. Gent: Ghent University; 2021.
Aerden KV, Damiens J, Moortel DD, Eggerickx T, Gourbin C, Huegaerts K, et al. Causes of health and mortality inequalities in Belgium: multiple dimensions, multiple causes.
Decoster A, Minten T, Spinnewijn J. The Income Gradient in Mortality during the Covid-19 Crisis: Evidence from Belgium. J Econ Inequal. 2021.
Bourguignon M, Damiens J, Doignon Y, Eggerickx T, Fontaine S, Lusyne P, et al. Variations spatiales et sociodémographiques de mortalité de 2020-2021 en Belgique. L’effet de la pandémie Covid-19, Document de travail 27. Louvain-la-Neuve: Centre de recherche en démographie; 2021.
Renard F, Tafforeau J, Deboosere P. Premature mortality in Belgium in 1993-2009: leading causes and 15 years changes. 2014. doi: 10.1186/2049-3258-72-34
Eggerickx T, Sanderson J-P, Vandeschrick C. Mortality in Belgium from nineteenth century to today: Quetelet Journal. 2020.