The European Journal of Public Health Advance Access originally published online on April 26, 2006
The European Journal of Public Health 2006 16(6):633-639; doi:10.1093/eurpub/ckl026
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health inequalities |
Neighbourhood income and anxiety: a study based on random samples of the Swedish population
Jonas Lofors, Vania Ramírez-León and Kristina SundquistKarolinska Institute, Center for Family Medicine, Sweden
Correspondence: Dr Kristina Sundquist, Karolinska Institute, Center for Family Medicine, Alfred Nobels allé 12, SE-141 83 Huddinge, Sweden, tel: +46 8 524 887 08, fax: +46 8 524 887 06, e-mail: Kristina.sundquist{at}klinvet.ki.se
Received November 1, 2005, accepted January 31, 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Background: Few studies have investigated the association between the neighbourhood characteristics and the vast panorama of mental disorders. This study examined whether there is an association between neighbourhood income and anxiety, a common mental disorder. Methods: A national random sample of the entire Swedish population was used, consisting of 30 884 men and women aged 25–64 years. The sample was obtained from pooled data during the period 1995–2002 from the Swedish Annual Level of Living Survey. Small area market statistics were used in order to define neighbourhoods. The proportion of individuals with incomes in the lowest national income quartile was calculated for each neighbourhood. The distribution was then divided into quartiles. A log binomial model was applied in the estimation of prevalence ratios. Four models were calculated with stepwise inclusion of the variables. Model 4 was adjusted for all the individual variables, i.e. age, gender, marital status, immigrant status, social network, housing tenure, employment status, and income. Results: In neighbourhoods with the highest proportions of individuals with low income the prevalence ratio of anxiety was 1.33 (95% confidence interval 1.24–1.42). The association demonstrated between neighbourhood income and anxiety decreased after stepwise inclusion of the individual variables and disappeared after all the individual variables were accounted for. Conclusion: Compositional explanations, rather than contextual explanations, lie behind the association between neighbourhood income and anxiety, a common mental disorder. However, we do not exclude the possibility that there is a contextual effect on severe mental disorders or among children with behavioural problems.
Keywords: anxiety, mental disorders, neighbourhood, socio-economic status
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The association between the neighbourhood environment and the individual health has received increased interest in recent research. During the last decade a number of studies have investigated the association between neighbourhood characteristics and different health outcomes.1–8 However, relatively few studies have investigated the association between the neighbourhood characteristics and the vast panorama of mental disorders. A recent study from Sweden found that individuals living in the socio-economically most disadvantaged neighbourhoods exhibited a significantly higher risk of being hospitalized for mental disorder than individuals living in the socio-economically most advantaged neighbourhoods, after adjustment for individual demographic and socio-economic characteristics.9 A study of the general adult population in the USA found that neighbourhood disadvantage was associated with higher rates of major depression and substance abuse, after controlling for individual characteristics.10 A Dutch study of incident cases of schizophrenia found that the neighbourhood environment modified the individual risk of schizophrenia.11 Another Dutch study found that neighbourhood deprivation was associated with both incidence and severity of non-psychotic, non-organic disorders, after adjustment for individual characteristics.12 In contrast, a British study found no neighbourhood variation in the prevalence of common mental disorders.13 Another study of common mental disorders did find an association between neighbourhood deprivation and common mental disorders in urban areas. However, this association was purely a result of a concentration of people with low socio-economic status in such areas, i.e. the compositional effect.14
In studies of the association between neighbourhood characteristics and health it is important to differentiate between the compositional and the contextual effect. The compositional effect is related to individual characteristics attributed to each individual, such as age, gender, marital status, and socio-economic status. It is well known that certain individual characteristics are associated with health, e.g. individuals with low socio-economic status are more likely to suffer from mental disorders.15 This will result in higher prevalence of mental disorders in neighbourhoods with a high proportion of individuals with low socio-economic status, irrespective of neighbourhood characteristics. The contextual effect, on the other hand, is related to the character of the neighbourhood in which the individuals live. This means that if individuals live in a context, i.e. a neighbourhood, that is likely to induce mental disorders, the prevalence of mental disorders in the neighbourhood will rise, irrespective of each individual's characteristics.6 Therefore, in order to differentiate between the compositional and contextual effect it is necessary to include several potential individual confounders in studies of the association between neighbourhood socio-economic characteristics, and mental disorders.
This cross-sectional study of 30 884 men and women aged 25–64 years was based on pooled data from a national random sample of the entire Swedish population. The data were obtained from the Swedish Annual Level of Living Survey (SALLS), which contains several individual variables concerning demographic, socio-economic, social and health indicators. Neighbourhood characteristics were defined on the basis of proportions of people with low income at small area level, i.e. neighbourhood income. The outcome self-reported anxiety was selected because anxiety disorders represent a most prevalent mental disorder in the population, tend to be chronic, and can be as disabling as somatic disorders. In addition, individuals with anxiety place a high strain upon the whole health care system.
The first aim of this study was to examine whether there is an association between neighbourhood income and anxiety. The second aim was to examine whether this hypothesized association remains after adjustment for several individual demographic, socio-economic and social characteristics, i.e. age, gender, marital status, immigrant status, social networks, housing tenure, employment status, and income.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
This cross-sectional study was based on a national random sample of the entire Swedish population, consisting of 15 659 women and 15 225 men aged 25–64 years. The sample was obtained from pooled data during the period 1995–2002 from the SALLS. The data in SALLS are widely used for various purposes, such as research and planning of welfare and health care policies. SALLS has been conducted by Statistics Sweden (the Swedish Government's statistic bureau) since 1975 and consists of interviews performed at home by trained interviewers. The questions include 11 areas of investigation covering the respondents' demographic and socio-economic characteristics, health status, and social environment. The response rate was
80% on average during the studied years.
Small area market statistics (SAMS) were used in order to define neighbourhoods. SAMS are small geographic units with boundaries defined by homogeneous types of buildings. Each SAMS consists of 2000 people in Stockholm and 1000 people in the rest of Sweden. The whole of Sweden consists of 9667 SAMS. The participants in the study population were represented in 7094 of these SAMS. The home addresses of the participants had been previously geocoded, which allowed us to identify the SAMS neighbourhood in which the participants lived. In total 239 participants could not be linked to their neighbourhood and they were categorized in the SAMS neighbourhoods with the highest neighbourhood income in order to avoid an overestimation of the relative risks.
Outcome variable
Anxiety was based on the question ‘Do you suffer from nervousness, uneasiness, or anxiety?’ The response alternatives were dichotomized as follows (1): Yes, severe problems and (2) Yes, slight problems versus (3) No.
Response alternative (1) included 3.9% of the respondents and response alternatives (2) and (3) included 14.5 and 81.6% of the respondents, respectively.
Explanatory variables
Neighbourhood income
Data used to calculate neighbourhood income were obtained from a national database for the entire Swedish adult population, containing annual data for each individual on, for example, socio-economic characteristics. This implies that the measurement of neighbourhood income was derived from another data source than the studied sample. The individuals included in the study had been previously geocoded to their neighbourhood of living, i.e. SAMS, which was used as a basis for the neighbourhood variable. The proportion of individuals with incomes in the lowest national income quartile was calculated for each neighbourhood, i.e. the SAMS area. The distribution was then divided into quartiles. Quartile 1 represented the richest neighbourhoods, i.e. the neighbourhoods with the lowest proportion of people with low income and quartile 4 represented the poorest neighbourhoods, i.e. the neighbourhoods with the highest proportion of people with low income. The income variable was based on annual disposable family income and took into account the number of people in the family. A weighted system was used whereby small children were given lower consumption weights than adolescents and adults. For the calculation of the neighbourhood variable we used data on women and men aged 25–64 from December 1997 because these age groups represent the socio-economically active part of the population.
Individual variables
Gender: female or male.
Age was categorized by dividing the respondents into the following groups: 25–34, 35–44, 45–54, and 55–64 years of age at the time of the interview.
Marital status was classified in two categories: living alone and married/cohabiting.
Immigrant status was classified into four groups. The first group consisted of Swedish-born people. The second group consisted of people born in Finland, the largest immigrant group in Sweden. The classification in the third and fourth group was based on the most probable reason for migration. The third group consisted of people born in countries with mainly labour immigration to Sweden (e.g. Denmark, Germany, Greece, Italy, Norway, Portugal, Spain, the UK, the USA, and Yugoslavia). The fourth group consisted of people born in countries, which are often referred to as refugee countries (e.g. Chile, Czechoslovakia, Estonia, Ethiopia, Hungary, Iran, Iraq, Lebanon, Romania, and the Soviet Union).
Social networks was based on the following five items (1): Having at least one close friend (no = 1 point) (2), Meeting friends and other acquaintances (seldom or never = 1 point) (3), Exchanging favours with neighbours (seldom or never = 1 point) (4), Casual interaction with neighbours (seldom or never = 1 point), and (5) Parents alive (no = 1 point).
Housing tenure was classified into two categories (1): ownership and2 renting.
Employment status was classified into two categories (1): employed and (2) non-employed.
Income was defined as annual individual disposable income. The women and men were divided into four equal-sized groups according to income level.
Many psychiatric disorders decline with increasing age and socio-economic status, which is why we included age and income in the analysis.16 The variable marital status was included because of its relationship with mental health.17 Finally, the importance of immigrant status for mental health was recently confirmed by the Swedish National Institute of Public Health18 and in a longitudinal study of different migrant groups in Sweden.19
Statistical analysis
A log binomial model20 was applied in the estimation of prevalence ratios (PR) with 95% confidence intervals (95% CIs). Four models, adjusted for age and gender, were calculated. Model 1 included the neighbourhood variable. Models 2–4 were calculated after stepwise inclusion of the individual variables. In Model 2 marital status and immigrant status were added and in Model 3 social networks and housing tenure were added. Model 4, i.e. the full model, included all the explanatory variables. Women and men were analysed together because of similar risk patterns.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Table 1 shows the distribution of the sample, in percentages, by neighbourhood income and the individual variables. Quartile 1 represents neighbourhoods with the highest neighbourhood income, i.e. neighbourhoods with the lowest proportion of individuals with low income. The highest proportion of refugees was found in quartile 4, i.e. neighbourhoods with the lowest neighbourhood income. Approximately twice as many non-employed individuals and individuals renting their home lived in quartile 4 compared with quartile 1.
|
Table 2 shows the prevalence of anxiety, in percentages, by neighbourhood income and the individual variables. The highest total prevalence of anxiety was found in quartile 4. All the individual demographic and socio-economic variables seemed to be associated with the prevalence of anxiety. For example, the prevalence of anxiety was especially high among refugees, non-employed individuals, individuals with a poor social network and individuals renting their home.
|
Table 3 shows the prevalence ratios of anxiety after stepwise inclusion of the independent variables in four different models. In Model 1, adjusted for age and gender, there was an apparent gradient; when neighbourhood income decreased the risk of anxiety increased. In quartile 4 the prevalence ratio of anxiety was 1.33 (95% CI 1.24–1.42), i.e. the risk was increased by 33% in neighbourhoods with the lowest neighbourhood income. The increased risk in quartile 4 remained significant in Model 2, adding marital status and immigrant status, as well as in Model 3, when social networks and housing tenure also were included. However, in Model 4, i.e. the full model, the increased risk of anxiety in quartile 4 no longer remained significant, after adjustment for all the independent variables.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
This study shows that the demonstrated association between neighbourhood income and anxiety decreased after stepwise inclusion of the individual variables and disappeared in the final model, after accounting for all the individual variables, i.e. age, gender, marital status, immigrant status, income, employment status, housing tenure, and social networks. Our study indicates that compositional explanations, rather than contextual explanations, lie behind the association between neighbourhood income and anxiety, which represents a common mental disorder.
Our study is in agreement with a study from the UK where no neighbourhood variation was found in the prevalence of common mental disorders, assessed by a General Health Questionnaire.13 However, in another study by the same authors there was an association between neighbourhood socio-economic characteristics and common mental disorders, but only among those who were economically inactive and hence more likely to spend more time at home.21 A Dutch study14 found that mental disorders in general cumulated in deprived urban areas, but mainly as a result of a concentration of people with low socio-economic status in these areas, i.e. the results were explained by compositional factors. Another Dutch study,22 which examined the neighbourhood effect on child behaviour problems, found that living in a more deprived neighbourhood was associated with higher levels of child problem behaviour, after adjustment for parental socio-economic status. The results could therefore be explained by contextual factors, which is in contrast to our study. However, our study examined the prevalence of anxiety among individuals aged 25–64 years, based on an annual survey that only includes adult participants. Another study from the UK,23 based on a sample of people aged 75 years and older, found no association between living in the most socio-economically deprived areas and anxiety, which disagrees with our study. Our study found an association between neighbourhood income and anxiety before adjustment for all the individual variables. However, the study from the UK included only elderly people, while our study was based on the economically active people.
It is possible that the contextual effect in some of the previous studies of the association between neighbourhood characteristics and mental disorders could be explained by residual confounding. For example, residual confounding could be present if only a few individual variables are adjusted for in the final models. Residual confounding could also constitute a problem if the individual variables are inadequately or insufficiently measured. Our study included in total eight individual potential confounders that were collected by well-trained interviewers in the participants' homes, i.e. age, gender, marital status, immigrant status, social network, housing tenure, employment status, and income.
Some individual characteristics are more likely to be associated with increased prevalence of anxiety, which could explain why the contextual effect in our study disappeared after adjustment for the individual variables. If we look at the results for the individual variables in our study the association with anxiety is obvious, which is in part consistent with data from the Epidemiologic Catchment Area (ECA) study in the United States. According to that study, women, those who are separated or divorced, and those in low socio-economic groups have higher rates of anxiety than men, those who are married, and those in high socio-economic groups.24
However, we do not exclude the possibility that there is a contextual effect on mental health. Our study included the outcome variable anxiety, which represents a common mental disorder. In addition, only individuals aged 25–64 years were included in the sample. It is possible that there is a neighbourhood effect on severe mental disorders or among children and adolescents with behavioural problems.
Possible features in socio-economically deprived neighbourhoods with a negative impact on people's mental health could include violent and non-violent crime, litter, broken glass, unleashed dogs, and abandoned cars. This kind of features could make people feel unsafe and increase negative feelings. In socio-economically deprived neighbourhoods people also suffer from poor social networks, leisure-time passiveness, and alienation.25–28 In addition, the association between mood disorders and poor social networks has been shown in several previous studies.29–32 Thus, poor social networks might act as one possible mediator between neighbourhood socio-economic characteristics and mental disorders. When we accounted for all the individual variables, including social networks, the association demonstrated between neighbourhood income and anxiety disappeared.
Strengths and limitations
There are several strengths in our study. First, the sample size is large and representative of the entire Swedish population. Second, SAMS neighbourhoods are relatively small (1000–2000 people) and, regarding type of buildings, homogenous. Third, our neighbourhood measure was based on the entire Swedish population and not the studied sample. This implies that the measurement of neighbourhood income was derived from another data source than the individual-level socio-economic characteristic in order to avoid sampling variation between neighbourhoods when estimating the proportions of individuals with low income in each neighbourhood. Fourth, the reliability of the survey questions, collected in face-to-face interviews by well-trained interviewers, was high when a sample of the participants were re-interviewed (test–retest method).33 Fifth, the character of the question regarding anxiety is very personal but was included in the interview in a set of health-related questions, which improves the possibility of obtaining a reliable answer.
There are also some limitations to our study. First, response bias may have occurred if the non-respondents (20%) differed from the respondents with respect to the outcome measure. However, the non-response rate was relatively low compared to many other similar types of surveys. Second, residual confounding likely exists because individual socio-economic status cannot be measured precisely and completely.34 However, besides income we included housing tenure and employment status, variables that both are related to the actual standard of living. Third, the measure of mental health, i.e. anxiety, is self-reported, which implies a risk of over- or underestimation of the actual prevalence of anxiety. However, we have no reason to believe that this bias would differ between neighbourhoods. Fourth, we used a traditional contextual model in the analysis, since it was not possible to use more complex multilevel modelling. Although the number of SAMS was large enough for each group, the number of individuals per SAMS was too small to allow a correct calculation of the variance measure. Nevertheless, contextual models can be adequate when examining the association between group-level variables and individual-level variables, if appropriate adjustment for residual correlation is taken into account.6 Several previous studies have used a similar analytical approach.5,35–37 Finally, we had no data of the physical environment in the neighbourhood. Poor physical environments could make people feel unsafe and increase feelings of anxiety.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The association demonstrated between neighbourhood income and anxiety disappeared in the final model, after all the individual variables were accounted for, i.e. age, gender, marital status, immigrant status, income, employment status, housing tenure, and social networks. Our study indicates that compositional explanations, rather than contextual explanations, lies behind the association between neighbourhood income and anxiety, which represents a common mental disorder. However, we do not exclude the possibility that there is a contextual effect on severe mental disorders or among children and adolescents with behavioural problems.
Key points
|
|
Acknowledgments |
|---|
The authors wish to thank Sanna Sundquist at the University of California, San Diego, for technical assistance. This work was supported by grants from the National Institutes of Health (1 R01 HL71084-01), the Swedish Research Council, the Swedish Council for Working Life and Social Research, the Knut and Alice Wallenberg Foundation, the Karolinska Institute, and the Stockholm County Council.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
1 Sundquist K, Winkleby M, Ahlen H, Johansson SE. Neighborhood socioeconomic environment and incidence of coronary heart disease: a follow-up study of 25,319 women and men in Sweden. Am J Epidemiol 2004; 159: 655–62.
2 Sundquist K, Malmstrom M, Johansson SE. Neighbourhood deprivation and incidence of coronary heart disease: a multilevel study of 2.6 million women and men in Sweden. J Epidemiol Community Health 2004; 58: 71–7.
3 Ellaway A, Anderson A, Macintyre S. Does area of residence affect body size and shape?. Int J Obes 1997; 21: 304–8.[CrossRef][ISI][Medline]
4 Kleinschmidt I, Hills M, Elliott P. Smoking behaviour can be predicted by neighbourhood deprivation measures. J Epidemiol Community Health. 1995;49(Suppl 2):S72–7.
5 Winkleby M, Cubbin C. Influence of individual and neighbourhood socioeconomic status on mortality among black, Mexican-American, and white women and men in the United States. J Epidemiol Community Health 2003; 57: 444–52.
6 Kawachi I, Berkman LF. Neighborhoods and health. New York: Oxford University Press, 2003.
7 Sundquist J, Malmstrom M, Johansson SE. Cardiovascular risk factors and the neighbourhood environment: a multilevel analysis. Int J Epidemiol 1999; 28: 841–5.
8 Yen IH, Kaplan GA. Poverty area residence and changes in physical activity level: evidence from the Alameda County Study. Am J Public Health 1998; 88: 1709–12.
9 Sundquist K, Ahlen H. Neighbourhood Income and Incidence of Psychiatric Hospital Admissions: A Multilevel Follow-up Study of 4.5 Million Women and Men in Sweden. Health & Place, 2003, In press.
10 Silver E, Mulvey EP, Swanson JW Neighborhood structural characteristics and mental disorder: Faris and Dunham revisited. Soc Sci Med 2002; 55: 1457–70.[CrossRef][ISI][Medline]
11 van Os J, Driessen G, Gunther N, Delespaul P
Neighbourhood variation in incidence of schizophrenia. Evidence for person-environment interaction. Br J Psychiatry 2000; 176: 243–8.
12 Driessen G, Gunther N, Van Os J Shared social environment and psychiatric disorder: a multilevel analysis of individual and ecological effects. Soc Psychiatry Psychiatr Epidemiol 1998; 33: 606–12.[CrossRef][ISI][Medline]
13 Weich S, Holt G, Twigg L, et al. Geographic variation in the prevalence of common mental disorders in Britain: a multilevel investigation. Am J Epidemiol 2003; 157: 730–7.
14 Reijneveld SA, Schene AH. Higher prevalence of mental disorders in socioeconomically deprived urban areas in the Netherlands: community or personal disadvantage?. J Epidemiol Community Health 1998; 52: 2–7.[Abstract]
15 Sundquist K, Frank G, Sundquist J. Urbanisation and incidence of psychosis and depression: Follow-up study of 4.4 million women and men in Sweden. Br J Psychiatry 2004; 184: 293–8.
16 Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry 1994; 51: 8–19.[Abstract]
17 Carroll LJ, Cassidy JD, Cote P. Factors associated with the onset of an episode of depressive symptoms in the general population. J Clin Epidemiol 2003; 56: 651–8.[CrossRef][ISI][Medline]
18 Sundquist J, Johansson S-E. Födelselandets betydelse – En rapport om hälsan hos olika invandrargrupper i Sverige. Stockholm: Statens Folkhälsoinstitut, 2002.
19 Westman J, Hasselstrom J, Johansson SE, Sundquist J. The influences of place of birth and socioeconomic factors on attempted suicide in a defined population of 4.5 million people. Arch Gen Psychiatry 2003; 60: 409–14.
20 Skov T, Deddens J, Petersen MR, Endahl L. Prevalence proportion ratios: estimation and hypothesis testing. Int J Epidemiol 1998; 27: 91–5.
21 Weich S, Twigg L, Holt G, et al. Contextual risk factors for the common mental disorders in Britain: a multilevel investigation of the effects of place. J Epidemiol Community Health 2003; 57: 616–21.
22 Kalff AC, Kroes M, Vles JS, et al. Neighbourhood level and individual level SES effects on child problem behaviour: a multilevel analysis. J Epidemiol Community Health 2001; 55: 246–50.
23 Walters K, Breeze E, Wilkinson P, et al. Local area deprivation and urban-rural differences in anxiety and depression among people older than 75 years in Britain. Am J Public Health 2004; 94: 1768–74.
24 Regier DA, Narrow WE, Rae DS. The epidemiology of anxiety disorders: the Epidemiologic Catchment Area (ECA) experience. J Psychiatr Res 1990;24 (Suppl 2):3–14.
25 Dalgard O. Bomiljø og psykisk helse (Environment and mental health) (dissertation) In Norwegian with English summary. Oslo: University of Oslo, 1980.
26 Lindberg G, Lindén A-L. Housing market segmentation in Swedish local authorities. Immigrants and Swedes; Young and old. Scand Housing Planning Res 1986; 3: 233–48.
27 Tonboe JC. Colonization of urban civil society. Acta Sociol 1988; 31: 303–18.
28 Sundquist J, Rosén U. The influence of social surroundings on consultation of private care, emergency department, hospital out-patient departments, and primary health care. 1384 Consultations mad by a defined population in a residential area. Eur J Public Health 1993; 3: 188–92.
29 Amann G. Social network and social support deficits in depressed patients: a result of distorted perception?. Eur Arch Psychiatry Clin Neurosci 1991; 241: 49–56.[CrossRef][ISI][Medline]
30 Cheng AT, Chen TH, Chen CC, Jenkins R. Psychosocial and psychiatric risk factors for suicide. Case-control psychological autopsy study. Br J Psychiatry 2000; 177: 360–5.
31 Hirschfeld RM, Montgomery SA, Keller MB, et al. Social functioning in depression: a review. J Clin Psychiatry 2000; 61: 268–75.[ISI][Medline]
32 Johnsson Fridell E, Ojehagen A, Traskman-Bendz L. A 5-year follow-up study of suicide attempts. Acta Psychiatr Scand 1996; 93: 151–7.[ISI][Medline]
33 Statistics Sweden. Re-interviews in the Survey of Living Conditions (ULF). Appendix No 12. Stockholm: Statistics Sweden, 1989.
34 Kaufman JS, Cooper RS, McGee DL. Socioeconomic status and health in blacks and whites: The problem of residual confounding and the resiliency of race. Epidemiology 1997; 8: 621–8.[CrossRef][ISI][Medline]
35 Cubbin C, Hadden WC, Winkleby MA. Neighborhood context and cardiovascular disease risk factors: the contribution of material deprivation. Ethn Dis 2001; 11: 687–700.[Medline]
36 Diez Roux AV, Merkin SS, Arnett D, et al. Neighborhood of residence and incidence of coronary heart disease. N Engl J Med 2001; 345: 99–106.
37 Soobader MJ, LeClere FB. Aggregation and the measurement of income inequality: effects on morbidity. Soc Sci Med 1999; 48: 733–44.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||