363
American Journal of Epidemiology
Copyright © 2001 by The Johns Hopkins University School of Hygiene and Public Health
All rights reserved
Vol. 153, No. 4
Printed in U.S.A.
Neighborhood Socioeconomic Status and Mortality Bosma et al.
Neighborhood Socioeconomic Status and All-Cause Mortality
Hans Bosma, H. Dike van de Mheen, Gerard J. J. M. Borsboom, and Johan P. Mackenbach
This study sought to determine the contribution of neighborhood socioeconomic status to all-cause mortality
and to explore its correlates. As part of the longitudinal “Gezondheid en LevensOmstandigheden Bevolking en
omstreken” (GLOBE) study in the Netherlands, 8,506 randomly selected men and women aged 15–74 years
from 86 neighborhoods in the city of Eindhoven reported on their socioeconomic status in the 1991 baseline
survey. During the 6-year follow-up, 487 persons died. Neighborhood socioeconomic status was derived from
individual reports on socioeconomic status. Its effect on mortality was stringently controlled for four individual-
level socioeconomic indicators. Persons living in a neighborhood with a high percentage of unemployed/disabled
or poor persons had a higher mortality risk than did those living in a neighborhood with a low percentage of
unemployed/disabled or poor persons. This was independent of individual socioeconomic characteristics,
including individual unemployment/disability or reports of severe financial problems. Educational and
occupational neighborhood indicators were similarly, but less strongly, related to mortality. The prevalence of
poor housing conditions, social disintegration, and unhealthy psychologic profiles and behaviors was higher in
neighborhoods with a low socioeconomic status. Contextual effects of socioeconomic status may thus be due to
one or more of these specific circumstances. The findings indicate potential public health benefits of modifying
socioeconomic characteristics of areas. Am J Epidemiol 2001;153:363–71.
mortality; social class; social environment; unemployment
Received for publication February 3, 1999, and accepted for pub-
lication March 6, 2000.
Abbreviation: GLOBE study, Gezondheid en LevensOmstandig-
heden Bevolking Eindhoven en omstreken (Dutch acronym for
health and living conditions of the population of Eindhoven and its
surroundings).
From the Erasmus University Rotterdam, Department of Public
Health, 3000 DR Rotterdam, the Netherlands.
Correspondence to Prof. Dr. Johan P. Mackenbach, Erasmus
University Rotterdam, Department of Public Health, P. O. Box 1738,
3000 DR Rotterdam, the Netherlands.
Recently, there has been much interest in the adverse
health consequences of living in districts, regions, wards, or
neighborhoods characterized by poor socioeconomic condi-
tions (1–21). Although some studies reported negative find-
ings (22–25), there is increasing support for the hypothesis
that living in such areas has negative effects on physical and
mental health for both people with a high socioeconomic
status and those with a low socioeconomic status (26–32).
This may have important consequences for public health ini-
tiatives since effects of area socioeconomic status—net of
individual socioeconomic status–suggest that health policies
should focus not only on individuals, but also on the socio-
economic environment in which people live (26–32).
However, only a few studies stringently adjusted for equiv-
alent measures of socioeconomic status on the individual
level. Hence, there is still the possibility that the adverse
effect of living in an area with a low socioeconomic status
only holds for the low-class inhabitants and not for their bet-
ter-off neighbors in the same area.
Dutch longitudinal data (Gezondheid en Levens
Omstandigheden Bevolking Eindhoven en omstreken
(Dutch acronym for health and living conditions of the pop-
ulation of Eindhoven and its surroundings) (GLOBE) study
(33)) on 8,506 men and women living in 86 neighborhoods
in the city of Eindhoven (191,000 inhabitants in 1991) were
used to determine whether living in a neighborhood with a
low socioeconomic status is related to 6-year all-cause mor-
tality. In the analyses, we stringently controlled for individ-
ual socioeconomic status by adjusting each neighborhood
socioeconomic indicator, aggregated from individual reports
(e.g., percent of blue-collar workers in neighborhood), for
four individual socioeconomic indicators, including the
equivalent individual report (e.g., being a blue-collar worker
or not). To obtain more information on the mechanisms
underlying the association between neighborhood socioeco-
nomic status and mortality (4, 28), we further examined
whether people living in varying socioeconomic neighbor-
hoods differed in more specific housing, social, psycho-
logic, and behavioral characteristics.
MATERIALS AND METHODS
Study population
Data were collected within the framework of the GLOBE
study (33). A postal survey was conducted in 1991 among
27,070 noninstitutionalized inhabitants (aged 15–74 years)
of Eindhoven and a number of surrounding municipalities,
all in the southeastern part of the Netherlands. Stratified by
at Institute of Social Studies on November 9, 2015http://aje.oxfordjournals.org/Downloaded from
364 Bosma et al.
Am J Epidemiol Vol. 153, No. 4, 2001
age and zip code, the sample was randomly drawn from the
municipal population registries. The response rate was 70.1
percent, which resulted in a study population of 18,973
respondents. The response rates were not substantially dif-
ferent by age, sex, marital status, level of urbanization, or
social class (as indicated by zip code) (33).
In Eindhoven, the major city in the study, 8,506 GLOBE
participants who reported on their current or last occupation
(younger people, mainly students, were thus excluded)
could be attributed to 86 administrative neighborhoods
(areas) with relatively homogeneous housing. On average,
there were 99 respondents in a neighborhood (range,
5–386). Eindhoven has a total of 106 areas, but some have
only few or no inhabitants. In 1991, the 86 neighborhoods
contained 191,000 inhabitants (all ages), with an average of
2,221 inhabitants per neighborhood.
Municipal population registers provided information on
all-cause mortality during the follow-up period until mid-
1997. These have a virtually complete coverage of the pop-
ulation (34). During the 6-year follow-up period, 487 men
and women died (6 percent).
Individual and neighborhood socioeconomic status
Similar measures were used to indicate socioeconomic
status on the individual and neighborhood levels. To indi-
cate the socioeconomic status of the neighborhood, we
aggregated individual reports on socioeconomic status to
the neighborhood level. An advantage of this approach is
that the effect of neighborhood socioeconomic status on
mortality can be stringently controlled for equivalent mea-
sures on the individual level. This allows an accurate exam-
ination of genuine contextual effects and, to a large extent,
excludes the possibility that any adverse effect of poor
neighborhoods is fully based on poorer people living in
poorer neighborhoods.
Four indicators of individual socioeconomic status were
used: last attained educational level of respondent, current
or last occupational level of household breadwinner (35),
being disabled or unemployed, and presence of severe finan-
cial problems in the household. Being or not being disabled
or unemployed was based on a question asking for the sub-
jects’ main activity for which they received income or spe-
cial social security benefits. Disability implied long-term
disability. Severe financial problems were self-reported by a
single item about whether or not the household had many
problems with making ends meet.
Neighborhood socioeconomic status was indexed by four
equivalent indicators based upon aggregated individual
GLOBE data: the percent of subjects reporting primary
schooling only (range, 0–44 percent), the percent of subjects
reporting that they were unskilled manual workers (range,
0–39), the percent of subjects reporting that they were
unemployed or disabled (range, 0–28), and the percent of
subjects reporting severe financial problems (range, 0–15).
Both continuous percentage scores and quartiles of neigh-
borhoods based on the percentage in the aggregated data
were used; each quartile contained about 22 neighborhoods
(86 neighborhoods total).
Disability and unemployment were combined because
both groups share the characteristic of being without paid
work and receiving social security benefits. Moreover, they
were combined because it is likely that many disabled per-
sons would have been assigned to the unemployed group in
countries other than the Netherlands. In the Netherlands, the
relatively generous disability benefit scheme has been used
on a large scale by employers to give the unemployed a rea-
sonable level of income compensation. Our findings will not
be confounded by prevalent disease (as the main cause of
disability) because the effect of the neighborhood percent
reported as unemployed or disabled will be controlled for
individual disability or unemployment and indicators of
prevalent disease.
Correlates of neighborhood socioeconomic status
Twelve individual characteristics were explored for their
association with neighborhood socioeconomic status. These
could be classified into four groups: housing conditions, and
social, psychologic, and behavioral factors. The three hous-
ing conditions were based on self-reported cold or draft in
the house (yes, no), condensation (yes, no), or moisture and
damp in the house (yes, no). The three social characteristics
were based on self-reported perception of vandalism in
neighborhood (yes, no), social difficulties with family mem-
bers and neighbors (subjects in the most adverse quintile of
the sum of eight five-point items asking for difficulties in
contacts with eight separate groups of important others,
including neighbors vs. the others), and noise pollution from
neighbors (yes, no). The three psychologic factors were
based on self-reported low control (subjects in the most
adverse quintile of the sum of 11 five-point items asking for
an external locus of control vs. the others), passive coping
(subjects in the most adverse quintile of the sum of eight
five-point items asking for an inactive style of coping vs. the
others), and depressive coping (subjects in the most adverse
quintile of the sum of seven five-point items asking for a
depressive reaction pattern when faced with problems vs.
the others). The three behavioral factors were based on self-
reported current smoking (yes, no), physically inactivity in
leisure time (subjects reporting no time spent on sports, gar-
dening, walking, or cycling during leisure time vs. the oth-
ers), and excessive alcohol consumption (subjects drinking
more than six alcoholic beverages on 3 or more days a week
or more than four beverages on 5 or more days a week vs.
the others). More information on the measurement of these
characteristics can be found elsewhere (36, 37).
Data analysis
Each separate neighborhood socioeconomic indicator was
related to all-cause mortality. The effect of individual
socioeconomic status was also determined. The effect of
neighborhood socioeconomic status (e.g., percent of sub-
jects reporting to be unskilled manual workers) was esti-
mated as both unadjusted and adjusted for its individual-
level equivalent socioeconomic measure (e.g., occupational
level). Any residual confounding by socioeconomic status
at Institute of Social Studies on November 9, 2015http://aje.oxfordjournals.org/Downloaded from
Neighborhood Socioeconomic Status and Mortality 365
Am J Epidemiol Vol. 153, No. 4, 2001
on the individual level was taken into account by controlling
for all four individual-level indicators of socioeconomic sta-
tus. Age, sex, and baseline health status were controlled for
in these analyses. Neighborhood socioeconomic status had
similar effects on mortality in men and women and in the
young and old (there were no statistically significant inter-
actions with sex and age). Therefore, no sex- or age-specific
analyses were performed. Baseline health status was indi-
cated by two dummy variables indicating whether or not the
respondent reported any less (e.g., hypertension, migraine)
or more (e.g., heart disease, cancer) severe chronic condi-
tions in a 23-item checklist.
We further explored the associations between neighbor-
hood socioeconomic status and the adverse housing, social,
psychologic, and behavioral conditions. Neighborhood
socioeconomic status was therefore related to individual
reports of any of these conditions, adjusting for all four indi-
vidual-level indicators of socioeconomic status, age, sex,
and baseline health status. Since the information on these
conditions was available for only a subsample that was
extensively interviewed (n 2,726) (33), detailed multi-
variate analyses determining the contribution of reported
adverse conditions to the neighborhood-mortality associa-
tion were precluded.
Because individuals (level one) were nested within neigh-
borhoods (level two), the analyses with death or individual
reports of problems as the outcome were done with multi-
level logistic regression using the MLN program (38, 39).
The probability of dying of the ith individual in the jth
neighborhood was modeled as follows:
where α is the constant; β
1
to β
p
are the regression coeffi-
cients of the individual-level variables x
1
to x
p
; γ
1
is the
regression coefficient of neighborhood socioeconomic sta-
tus z
1
, and µ
j
is the neighborhood level residual (random
variable).
RESULTS
Table 1 shows that both low individual socioeconomic sta-
tus and low neighborhood socioeconomic status were related
to mortality. For example, subjects with only primary school-
ing had a 1.65 higher risk of dying during follow-up com-
pared with their highly educated counterparts. A similar odds
ratio was obtained for subjects living in a neighborhood in
which a high percentage of residents had only primary
schooling compared with those living in a neighborhood
with a low percentage having only primary schooling (odds
ratio 1.73). Subjects who were self-employed and those
with severe financial problems had particularly elevated
odds of dying (odds ratio 2.37 and 2.16, respectively) as
did subjects living in neighborhoods in which a high per-
centage were unemployed or disabled or were reporting
severe financial problems (odds ratio 1.78 and 1.47,
respectively).
The multilevel analysis showed significant variability in
mortality risks among neighborhoods (χ
2
4.3 (1 df), p
log1π
ij
>11 π
ij
22 α β
1
x
1ij
p β
p
x
pij
γ
1
z
1j
u
j
0.04). When the neighborhood socioeconomic status and con-
founders were taken into account, no between-neighborhood
variation was left. There were no statistically significant
cross-level interactions between individual and neighborhood
socioeconomic statuses. This indicates that low neighborhood
socioeconomic status was related to increased risks of mor-
tality in both subjects with a high and those with a low socio-
economic status. Finally, findings were similar across neigh-
borhoods differing in the number of respondents, i.e., the
results are probably not biased by the fact that the socioeco-
nomic status of some neighborhoods was derived from data
obtained from few respondents only.
The independence of neighborhood socioeconomic sta-
tus and individual socioeconomic status is shown in figure
1 for the educational indicators. Living in a poorly educated
neighborhood increased probabilities of dying for both
highly and poorly educated individuals. Similarly, a low
individual educational level increased probabilities of
dying within both highly and poorly educated neighbor-
hoods. For example, 3.6 percent of the subjects with a low
educational level living in a neighborhood with many
respondents with only primary schooling were estimated to
die within the follow-up period compared with 2.4 percent
of their poorly educated counterparts living in highly edu-
cated neighborhoods. The 1.7 percent deaths in highly edu-
cated subjects living in highly educated neighborhoods
compared with the 3.6 percent deaths in poorly educated
subjects living in poorly educated neighborhoods indicates
the cumulative effects of individual and neighborhood
socioeconomic statuses.
The neighborhood effects adjusted for the individual-
level equivalent measure of socioeconomic status is shown
in table 2, model 1. Except for the neighborhood occupa-
tional level (the percent of unskilled workers), effects were
statistically significant for subjects in the socioeconomically
most adverse areas. For example, the risk of mortality for
subjects living in neighborhoods with a high percentage of
subjects who reported only primary schooling was 1.50
times higher than the risk for their counterparts in highly
educated neighborhoods. The decrease in odds ratios com-
pared with the unadjusted model in table 1 is the result of
subjects in neighborhoods with a low socioeconomic status
having a higher risk of having a low socioeconomic status
themselves.
When we adjusted for all four individual socioeconomic
indicators in model 2, the odds ratios of neighborhood
socioeconomic status decreased further, causing only one
odds ratio (for the percent reporting as unemployed or dis-
abled) to remain statistically significant. The other odds
ratios were generally in the expected direction, however.
Moreover, using the continuous percentage score for neigh-
borhood socioeconomic status revealed that the presence of
severe financial problems in a neighborhood was also statis-
tically significantly related to mortality. The odds ratios
were 1.32 and 1.60 for a 10 percent increase in the neigh-
borhood percent of subjects reporting that they were unem-
ployed or disabled and those reporting severe financial
problems, respectively. These effects were independent of
individual educational and occupational levels, whether or
at Institute of Social Studies on November 9, 2015http://aje.oxfordjournals.org/Downloaded from
366 Bosma et al.
Am J Epidemiol Vol. 153, No. 4, 2001
not subjects were disabled or unemployed themselves and
whether or not subjects reported severe financial problems
themselves.
The odds ratios in table 3 show that neighborhood
socioeconomic status (percent reporting that they were
unemployed or disabled, in quartiles) was related in the
TABLE 1. Number of respondents and odds ratios of mortality and 95 percent confidence intervals by
individual and neighborhood socioeconomic status, adjusted for age, sex, and baseline health status
(n = 8,506; 487 deaths), longitudinal GLOBE study, Eindhoven, the Netherlands, 1991–1997
Individual educational level*
1 High
2
3
4 Low
Neighborhood educational level
1 Living in best-off quartile
2
3
4 Living in worst-off quartile
% score (range: 0–44)†
Individual occupational level‡
1 High
2
3 Self-employed
4
5 Low
Neighborhood occupational level
1 Living in best-off quartile
2
3
4 Living in worst-off quartile
% score (range: 0–39)†
Individual unemployment or disability
1 No
2 Yes
Neighborhood unemployment or disability
1 Living in best-off quartile
2
3
4 Living in worst-off quartile
% score (range, 0–28)†
Individual severe financial problems
1 No
2 Yes
Neighborhood severe financial problems
1 Living in best-off quartile
2
3
4 Living in worst-off quartile
% score (range, 0–15)†
No.
1,596
1,787
3,326
1,797
1,026
2,947
2,204
2,329
988
3,581
360
2,037
1,540
962
2,179
2,791
2,574
7,465
1,041
1,610
1,986
2,876
2,034
8,178
328
1,466
2,314
2,993
1,733
* 1: university and higher vocational education; 2: intermediate/higher general and intermediate vocational edu-
cation; 3: lower general and lower vocational education; and 4: primary education only.
† Odds ratios for the percent scores indicate increase in odds when there is a 10-point increase in the percent
score.
‡ 1: higher-grade professionals; 2: lower grade professionals or routine nonmanual workers; 3: self-employed
workers; 4: skilled manual workers; and 5: unskilled manual workers.
1.00
1.20
1.30
1.65
1.00
1.37
1.31
1.73
1.15
1.00
1.53
2.37
1.82
1.82
1.00
1.06
1.48
1.46
1.14
1.00
1.87
1.00
1.38
1.46
1.78
1.53
1.00
2.16
1.00
1.05
1.26
1.47
2.03
0.83, 1.72
0.94, 1.79
1.19, 2.28
0.93, 2.02
0.89, 1.93
1.18, 2.54
1.04, 1.27
1.07, 2.17
1.43, 3.94
1.27, 2.61
1.24, 2.68
0.72, 1.56
1.04, 2.11
1.03, 2.09
1.02, 1.27
1.44, 2.44
0.97, 1.97
1.04, 2.04
1.26, 2.52
1.23, 1.90
1.42, 3.28
0.76, 1.44
0.94, 1.68
1.07, 2.02
1.40, 2.93
Odds
ratio
95%
confidence
interval
at Institute of Social Studies on November 9, 2015http://aje.oxfordjournals.org/Downloaded from
Neighborhood Socioeconomic Status and Mortality 367
Am J Epidemiol Vol. 153, No. 4, 2001
FIGURE 1. Percent deceased during follow-up by individual and neighborhood educational level. Estimated for men aged 49 years without
baseline diseases (n = 6,506 deaths), longitudinal Globe study, Eindhoven, the Netherlands, 1991–1997.
predicted direction to all housing, social, psychologic, and
behavioral factors. For example, individuals living in
neighborhoods with a high percent of subjects reporting
that they were unemployed or disabled more often
reported that they experienced cold or draft in their houses
(odds ratio 2.29), that they experienced vandalism in
the neighborhood (odds ratio 2.05), that they used more
passive instead of active coping (odds ratio 1.46), and
that they more often did not engage in physical activity
(odds ratio 1.48). The associations were independent of
whether or not the subjects themselves had a high educa-
tional or occupational level, whether or not they were
unemployed or disabled, and whether or not they had
severe financial problems. There was no clear association
between neighborhood socioeconomic status and exces-
sive alcohol consumption. Similar associations were found
with the other indicators of neighborhood socioeconomic
status.
DISCUSSION
Our findings indicate that particular indicators of neigh-
borhood socioeconomic status are related to longevity in
Dutch men and women in an urban setting. After the strin-
gent control for individual socioeconomic status, the
neighborhood percentage of unemployed or disabled per-
sons and the percentage who reported severe financial
problems continued to affect mortality risks. The educa-
tional and occupational indicators of neighborhood socio-
economic status were also related to mortality, but less
strongly, and their effects were no longer statistically sig-
nificant after the stringent individual-level control. The
effects of neighborhood socioeconomic status are likely to
be genuine contextual effects because we used similar
measures on the individual level (individual reports) and
the neighborhood level (aggregated individual reports).
The association between neighborhood socioeconomic sta-
tus and mortality was not only controlled for the individual-
level socioeconomic equivalent, but also for three other
indicators of individual socioeconomic status. The higher
mortality risk for subjects living in neighborhoods with
many people who were unemployed or disabled or who
reported severe financial problems thus not only holds for
subjects who are unemployed, disabled, or poor them-
selves, but also for their employed, not disabled, and rich
neighbors living in the same area.
at Institute of Social Studies on November 9, 2015http://aje.oxfordjournals.org/Downloaded from
