relative sodoeconomic mortality differences (SEMD) per data-set were calculated using inequality indices developed by Pamuk. The results show a decrease of ...
Socioeconomic inequalities in perinatal and infant mortality from 1854 to 1990 in Amsterdam, the Netherlands HENDR1KE VAN DE MHEEN, SIJMEN A. REIJNEVELD, JOHAN P. MACKENBACH *
Key words: trends, infant mortality, perinatal mortality, socioeconomic status
vj^ocioeconomic inequalities in health have received a lot of attention in the last decade, starting with the publication of the Black Report.1 In particular in Great Britain and Scandinavia a great deal of research has been done on this subject.2-3 The results show that in lower socioeconomic groups health problems are more frequent than in higher socioeconomic groups. One of the targets of the World Health Organization is to reduce these health inequalities by 25% by the year 2000.4 Trends in time are an important issue in research on socioeconomic inequalities in health. Is the health gap between lower and higher socioeconomic groups closing or widening? Most evidence which may give the answer to this question comes from mortality data. Some authors have demonstrated that inequality in adult mortality has increased since the early 1950s.5"11 Unfortunately, in the Netherlands data on trends in adult mortality by socio' economic status are scarce. Results from a geographical study suggest that in the Netherlands too the relationship between socioeconomic level and adult mortality has become more negative since 1950.12
n'. S.A. Reijneveld3. If. Macfcenbach1 1 Department of Public Health. Erasmus University, Rotterdam, The Netherlands 2 Department of Epidemiology. Municipal Health Service. Amsterdam, The Netherlands Correspondence: K van de Mheen, Department of Public Health, Erasmus University, P.O. Box 1738, 3000 DR Rotterdam. The Netherlands. Ul.+31 10 4087714, fax +31 10 4366831
This paper focuses on socioeconomic differences in infant and perinatal mortality. Perinatal mortality is defined as stillbirths plus deaths in the first week (per 1,000 births, i.e. stillbirths at 28 or more weeks of gestation and live births). Infant mortality is defined as deaths under 1 year of age per 1,000 live births.13 Mortality differences between children from different socioeconomic groups are generally regarded as unacceptable inequalities.14 Because children have no choice in their living and environmental conditions, these differences are even more unjust than in an adult population. Many studies show that socioeconomic differences in infant and perinatal mortality exist.15>16Studies on infant and perinatal mortality in various Western European countries and the United States have demonstrated that the inverse relationship between social class and perinatal, neonatal and postnatal mortality did not narrow from approximately 1910 to the mid-1960s.15 Pamuk1^ has demonstrated that relative inequality in infant mortality in Great Britain has not decreased in the second half of the twentieth century. In Scandinavia, socioeconomic differences in infant mortality seem to have increased during the last 2 decades.18-19 We have investigated trends in socioeconomic mortality differences (SEMD) in infant and perinatal mortality in the Netherlands. Children in the Netherlands are very healthy judging by these indicators: infant mortality is one of the lowest in the world (63 per 1,000 live births in 1992).20 However, SEMD in infant and perinatal mortality do exist. In the Netherlands, in particular in Am-
Downloaded from http://eurpub.oxfordjournals.org/ by guest on June 12, 2013
Trends in sodoeconomic differences in infant and perinatal mortality in Amsterdam were studied for the period 1854-1990, using published and unpublished material, at the aggregate and at the individual level. Absolute and relative sodoeconomic mortality differences (SEMD) per data-set were calculated using inequality indices developed by Pamuk. The results show a decrease of the absolute differences in both Infant and perinatal mortality. For infant mortality, this is mainly due to the overall dedine of the Infant mortality rate. Relative differences in infant mortality did not decrease during the study period. This is the result of separate developments in 3 time periods. From approximately 1850 to approximately 1910 an Increase in relative differences can be seen, a trend which is reversed from approximately 1910 to the end of World War II. After World War II relative differences seem to stabilize at the same level. For perinatal mortality, for which only data from the post-World War II period are available, the decrease in the absolute differences is due both to the overall dedine of the perinatal mortality rate and to a dedine of relative differences between sodoeconomic groups. It is conduded that although SEMD in infant and perinatal mortality have dedined In an absolute sense, they still exist and that the relative position of deprived groups concerning infant mortality was not ameliorated during the study period.
Inequalities m permataljmfcmt manabty
sterdam, interesting data about infant and perinatal mortality are available over a period of approximately 150 years. In this paper we will study the trends in socioeconomic differences in infant and perinatal mortality in Amsterdam from 1854 to 1990. DATA AND METHODS Data We used published as well as unpublished data in our analysis, at both the individual and aggregate levels. Published data on the relation between socioeconomic status (SES), on the one hand, and infant mortality and perinatal mortality, on the other, are available from die second half of die nineteenth century until approximately
1980. Complementary analyses were performed on unpublished data from die last 15 yean of 3 different databases.21"23 Information on die studies used in diis analysis is given in tables i-3. 2 1 " 3 * The SES of an individual refers to his or her position in die social hierarchy; in studies on infant and perinatal mortality SES is always defined by die SES of the parents. This was based on die occupational level of die father in most of the studies used at die individual level, hi studies at die aggregate level, income or wealdi (based on income tax per earner) was used as die SES indicator. This mainly concerned die income of die father, because in die Netherlands up until 1958 income tax was based on the income of die male head of die household. For both the
Table 1 Infant mortality per 1,000 live births 1854-1990 by jocioeconomic status (SES), aggregate level, Amsterdam
Author/database
SES indicator
1854-1859
Israels"
Wealth by number of poor people receiving medical aid
1891-1894
Mortality Estimated level of total mortality level population of highest SES
Rll
SI1
4
23.1-
49.5
239
50
39.7'
79.0
239
199
6 50 11
22.7 b
32.8
161'
25.1 b
36.0
162'
145 143
94.2C
53.4
81.2 f
56.7
215
Centrale Commissie voor de Statistiek
Wealth by income tax 1892/1893
1909-1911
Sajet and Van Getderen 24 Bureau Statistiek der Getneente Amsterdam
Wealth by income tax (% tax assessment: of income >2,2OO of total number of tax assessments 1915/1916)
1919-1921
Meulenhoff" Bureau Statistiek der Gemeente ^ Amsterdam
Wealth by income tax (% tax assessments of income >2,2OO of total number of tax assessments 1915/1916)
12
75.4C
29.0
52.1
38.4
1929-1931
Meulenhoff29 Bureau Statistiek der Gemeente Amsterdam
Wealth by income tax (% tax assessments of income >3,000 of total number of tax assessments and of total number of working population 1931/1932)
15
74.1 C
19.0
35.1
25.7
1946-1949
Bureau Statistiek der Gemeente Amsterdam
Wealrh by income tax (% tax assessments of Income >3,000 of total number of tax assessments and of total number of working population 1931/1932)
29
46.6 b
10.5
27.8
22.5
1972-1978
Lau-IJierman et aL / no published data, source database: Doornbosand Nordbeck 21
Wealth index (1971) based on occupation, income, education, telephone possession
17
66.2 d
4.9
9.9
7.4
1979-1983
N o published data, source database: Van der Maas, et aL 22
Wealth index (1983) based on 16 indicators related to SES, for example, number of social benefits, crowding, number of 1-parent families
17
84.4 d
6.4
18*
76.5^
5.9
10.7 10.7
7.7
No published data, source database: Amsterdam Municipal Population Register
Mean income 1984, based on income tax
8.6
6.0
8.6
6.1
1986-1990
22 21 h
Rll: relative Index of inequality (Sll/tstimattd mortality highest SESxlOO) S1I: dope index of inequality (estimated morality lowest SES - estimated mortality highest SES) a: Rll and SU per number of buthi in SES group b: Rll and SU per number of persons in SES group c Rl] and SII per number of live births in SES group d: Rll and SU per number of 0 yean old in SES group e: Infant mortality per 1,000 births f. Infant mortality per 1,000 0-1 yean old f. Ranking neighbourhoods by mean income 1984 hu Ranking neighbourhoods by wealth index 1983
76.4
d
5.1 4.7
7.6
Downloaded from http://eurpub.oxfordjournals.org/ by guest on June 12, 2013
Year
Number of neighbourhoods
EUROPEAN JOURNAL OF PUBLIC HEALTH VOL. 6 1996 NO. 3 Methods Presenting inequalities in infant and perinatal mortality by SES group over time requires that an inequality measure takes into account mortality rates in all die SES groups and die distribution of die population over die SES groups. Moreover, it should reflect only die socioeconomic dimension of inequality in health.* 3 By using the size of die SES groups in die calculation, one measures not only the effect of decreasing SES on healdi but also die total impact of socioeconomic inequalities in health upon the healdi status of die population as a whole. Thus, the impact of changes over time in the size of die SES groups is taken into account. In addition, relative and absolute differences are important. A relative measurement presents the frequency of mortality in the lowest SES group as a percentage of the mortality rate of the highest SES group. An absolute measurement presents
individual and aggregate levels we included only studies which permitted a comparison with respect to the SES indicator and the number of SES groups. Therefore, we excluded studies in which other SES indicators like crowding or housing conditions were used. ^ Because most studies at die aggregate as well as at the individual level concerned die city of Amsterdam, we excluded studies which concerned odier cities only.-'7"39 It is unknown what aspects related to socioeconomic inequalities (for example, level of urbanization, environmental characteristics, houses, occupational composition and total level of infant and perinatal mortality) differed between cities in die period under review. Finally, only studies in which the siie of die SES groups in proportion to die total population was given could be included in this analysis, because diis information was necessary to calculate the measures of inequality we used (see below).
Year 1937-1940 1941-1943 1946-1950 1960-1963 1975-1980
Author/database Bureau Statistiek der Getneente Amsterdam Bureau Statistiek der Gemeente Amsterdam33 Bureau Statistiek der Gemeente Amsterdam Bureau Statistiek der Gemeente Amsterdam Unpublished data, source database: Doombos and Nordbeck21
Mortality level of Estimated mortality total population level of highest SES 29.5 21.2
Number of SES groups
Rll
SI1
4
86.6C
18.4
Occupation
4
44.3C
12.0
32.8
27.1
Occupation
4
no.tf
20.2
27.2
18.4
Occupation
4
61.2 C
7.0
14.8
11.5
Occupation
4
ai.tf
5.8
9.3
6.7
SES indicator Occupation
Rll: relative Index of inequality (SII/esttm»ted mortality highest SESxlOO) S1I: dope Index of inequality (estimated mortality lowest SES - estimated mortality highest SES) a: RII and SH per number of births in SES group b: Rll tnd SO* per number of persons in SES group a RII and SII per number of live births in SES group A RII and SII per number ofO yean old in SES group e: Infant mortality per 1,000 births f-. Infant mortality per 1,000 0-1 yean old g: Ranking neighbourhoods by mean income 1984 h: Ranking neighbourhoods by wealth Index 1983
Table 3 Perinatal mortality per 1,000 births 1946-1980 by socioeconomic status (SES), individual level, Amsterdam Number of SES groups
RJ1
Sll
Occupation
4
64.9"
15.4
31.0
23.8
Occupation
4
38.0*
7.9
24.5
20.8
Occupation
4
25.21
3.1
13.8
12.3
Year
Author/database
SES indicator
1946-1950
Bureau Statistiek der Gemeente Amsterdam3 Bureau Statistiek der Gemeente Amsterdam Unpublished data, source database: Doombos and Nordbeck21
1960-1963 1975-1980
RII: relative index of inequality (Sll/estimated mortality highest SESxlOO) Sll: slope Index of inequality (estimated morality lowest SES - estimated morality highest SES) a: Rll and SH per number of births In SES group b: RII and SU per number of persons in SES group c: RU and SII per number of live births in SES group ± Rll and Sll per number of 0 yean old in SES group e: Infant mortality per 1,000 blrthj f: Infant mortality per 1,0000-1 yean old g: Ranking neighbourhoods by mean income 1984 h: Ranking neighbourhoods by wealth index 1983
Mortality level of Estimated mortality total population level of highest SES
Downloaded from http://eurpub.oxfordjournals.org/ by guest on June 12, 2013
Table 2 Infant mortality per 1,000 live births 1937-1980 by jocioeconomic status (SES), individual level, Amsterdam
inequalities m pennatal]'infant mortality
log (M ; ) = a + f3 x SESj
Number of deaths ( p « 1000 births)
Obswvsd
0
0.25
0.5
0.75
1
Socioeconomic statu* (0-ttgh, 1-low)
Figure 1 Perinatal mortality by socioeconomic status, 1960-1963
from log-linearity were observed. Differences between SES groups were statistically significant (p
