|
 |
|
|||||||
|
|||||||||
© 2003 American Public Health Association
RESEARCH AND PRACTICE |
The author is with the National Cancer Institute, Division of Cancer Control and Population Sciences, National Institutes of Health, Bethesda, Md.
Correspondence: Requests for reprints should be sent to Gopal K. Singh, PhD, MS, MSc, National Cancer Institute, Division of Cancer Control and Population Sciences, 6116 Executive Blvd, Suite 504, MSC8316, Bethesda, MD 20892-8316 (e-mail: gopal_singh{at}nih.gov).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Objectives. This study examined age-, sex-, and race-specific gradients in US mortality by area deprivation between 1969 and 1998.
Methods. A census-based area deprivation index was linked to county mortality data.
Results. Area deprivation gradients in US mortality increased substantially during 1969 through 1998. The gradients were steepest for men and women aged 25 to 44 years and those younger than 25 years, with higher mortality rates observed in more deprived areas. Although area gradients were less pronounced for women in each age group, they rose sharply for women aged 25 to 44 and 45 to 64 years.
Conclusions. Areal inequalities in mortality widened because of slower mortality declines in more deprived areas. Future research needs to examine population-level social, behavioral, and medical care factors that may account for the increasing gradient.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Studies involving individual social class data have shown increasing socioeconomic inequalities in US infant and adult mortality rates.1–3 However, these studies have compared social class inequalities in mortality at only 2 distant time points (e.g., 1960 and 1986). This limitation is primarily because of limited availability of socioeconomic information in US mortality statistics, which generally include only educational attainment and usual occupation/industry of the decedent.4–9 Moreover, analyses of socioeconomic differentials in mortality are hampered by incomplete and poorly reported socioeconomic data on death certificates as well as by the lack of relevant denominator data.5,7,9,10
Whereas US mortality statistics are frequently provided by age, sex, race, and cause of death, temporal analyses of socioeconomic differentials in mortality are less common.1–3,5,6,9,10 Similarly, although a substantial number of ecological studies have examined the cross-sectional association between areal social conditions and US mortality,8,11–20 temporal analyses of mortality differentials in relation to area-based deprivation or inequality measures remain scarce.9,21–28 Area-based composite deprivation indices have been used extensively in analyzing and monitoring health and mortality differentials in Europe, Australia, and New Zealand.29–38 Despite the lack of a consensus deprivation index in the United States, it is possible to construct a comprehensive, composite census-based socioeconomic index that, when linked to mortality data at an aggregate geographic level (e.g., county), could allow the monitoring of population health inequalities across time and space.8,9,21
In this article, I use census tract data to describe a composite area–based deprivation index for the United States. By linking the index to national mortality data, I examine the extent to which differentials in all-cause mortality rates by area deprivation have changed over time. Specifically, I use the areal index to stratify all 3097 US counties into 5 area deprivation groups and examine trends in areal gradients in mortality between 1969 and 1998 for men and women of all ages as well as for those in specific age groups (less than 25 years, 25–44 years, 45–64 years, and 65 years or older).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Constructing an Index of Area Deprivation
Community socioeconomic measures describe important aspects of social organization, structure, stratification, or environment, such as socioeconomic deprivation, economic inequality, resource availability, and opportunity structure.8,9,21,39–41 Although single measures representing an area’s educational and occupational composition, income and employment distributions, or housing conditions can be used to classify communities, a composite index consisting of several key indicators drawn from these domains would more accurately reflect the multidimensional characterization of a community’s socioeconomic position.8,9,21 Such a composite index should have greater validity, robustness, and explanatory power than single areal measures in documenting the extent of social disparities in health and mortality.
In constructing an index, I considered 21 socioeconomic indicators that may be viewed as approximating the material and social conditions and relative socioeconomic disadvantage in a given community. Indicators were selected on the basis of their theoretical relevance and on the basis of previous empirical research.8,9,21,29,37–40 These indicators, drawn from the 1990 census, included educational distribution (percentage of the population with less than 9 years and with 12 or more years of education), median family income, income disparity, occupational composition, unemployment rate, family poverty rate, percentage of the population below 150% of the poverty rate, single-parent household rate, home ownership rate, median home value, median gross rent, median monthly mortgage, and household crowding. Other indicators were percentages of households without access to a telephone, plumbing, or motor vehicles; English language proficiency; divorce rate; percentage of urban population; and percentage of immigrant population.21,42,43 Factor analysis and principal-components analysis were used in index construction.44,45
The initial factor analysis provided 2 factors that respectively accounted for 43% and 17% of the variance in the data. Seventeen of the indicators were clustered and had considerably larger loadings (> 0.45) on the first than on the second factor. However, 3 indicators—English language proficiency, percentage of urban population, and percentage of immigrant population—had much smaller loadings (< 0.25) on the first factor but larger loadings on the second factor. Divorce rate did not load highly on either factor. Whereas the first factor clearly indicated a theoretically and empirically meaningful clustering of the given indicators, the second factor, with only a few substantial loadings, did not lend itself to any obvious theoretical interpretation. In the final phase of the index construction, the 17 indicators were factor analyzed with a singlefactor solution. Table 1
contains these results.
|
The factor loadings for the census tract deprivation index ranged from 0.92 for percentage of population below 150% of the poverty rate to 0.45 for percentage of households without access to plumbing (Table 1
). The factor score coefficients were used to weight the 17 indicators comprising the index. Poverty, income, and education had the largest relative weights in generating the 1990 index. The index accounted for 52% of the variance in the data. The factor scale was transformed into a standardized index by arbitrarily setting the index mean and standard deviation at 100 and 20, respectively. The tract index scores were averaged to allow computation of index scores for each of the 3097 US counties. The 1990 county index scores ranged from a low of 70.22 to a high of 160.32. Higher index scores denote higher levels of deprivation.
The reliability coefficient (
) for the census tract index was 0.95, indicating a high degree of internal consistency among the indicators comprising the index.46 The reliability of the tract index was further evaluated by factor analyzing the 17 variables for different random subsamples of the US population. The factor structure matrix containing the factor loadings for the different subsamples remained essentially unchanged, indicating a high degree of index reliability for various subsets of the US population.21
The validity of the 1990 deprivation index was tested by comparing factor loadings for the same set of 17 indicators computed at the census tract, zip code, and county levels (Table 1). The factor loadings for the 3 geographic levels were generally similar in magnitude and relative importance. The percentages of explained variance and reliability coefficients were almost identical for the 1990 census tract and county indices.
The predictive validity of the 1990 deprivation index was checked by examining its correlation with a variety of county-level health outcomes for the period 1990 through 1996. The weighted correlations of the index with health outcomes were in the expected direction. The correlations with infant mortality rate and low birthweight rate were 0.48 and 0.46, respectively, and correlations with age-adjusted mortality rates from various cause-of-death categories were as follows: all causes combined, 0.58; heart disease, 0.45; stroke, 0.24; all cancers, 0.20; lung cancer, 0.27; breast cancer, -0.19; cervical cancer, 0.51; melanoma, -0.20; diabetes, 0.44; chronic obstructive pulmonary disease, 0.14; cirrhosis, 0.25; unintentional injury, 0.66; suicide, 0.27; and homicide, 0.39.
Table 1 also presents the factor loadings associated with 15 indicators comprising the 1970 county deprivation index. Census tract data for all variables and county data for mortgage and home ownership rates were not available for the 1970 census.43 The 1990 index was highly correlated with the 1970 index (r = 0.87). The quintile classifications of the 1970 and 1990 deprivation indices also corresponded closely with each other (
= 0.88).
Computing Annual Rates and Modeling Areal Gradients Over Time
The weighted population quintile distribution of the 1990 deprivation index, which classified all US counties into 5 equal population groups, was used to analyze time trends. The groups thus created ranged from representing the least-deprived to the mostdeprived areas. National mortality data files were used to obtain age-, sex-, race-, and county-specific deaths recorded from 1969 through 1998.4,5 Age-, sex-, race-, and county-specific census population estimates from 1969 to 1998 served as denominators for computing rates.47,48 Each of the 3097 counties was assigned 1 of the 5 areal deprivation categories. In the case of Alaska and Hawaii, state- rather than county-level data were used. Mortality rates for each deprivation category were age adjusted via the direct method, using the age composition of the 1970 US population as a standard and 5-year age-specific death rates.
Log-linear models were used to estimate annual exponential rates of declines in mortality rates.1 Poisson regression models were fitted to age-, sex-, race-, and county-specific death counts and populations to estimate areal gradients in mortality for 15 time periods of 2 years each.49 Areal gradients (relative mortality risks) were estimated for men and women separately after adjustment for age and race (coded White, Black, or other). In all Poisson models, the least-deprived area was selected as the reference category. There was no statistically significant interaction between race and area deprivation.
All models, estimated via the SAS GENMOD procedure, showed reasonable fit, as determined by the likelihood ratio statistic or deviance.50 In all of the models, 95% confidence intervals were adjusted for overdispersion. Trend tests were conducted by the use of 
2 statistics derived through Poisson models that included age, race, and area deprivation (coded as a continuous variable).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
The descriptive socioeconomic data presented in Table 2
indicate the relative stability of the area deprivation groups between 1970 and 1990. The relative educational standing of each deprivation group remained similar during 1970 to 1990. Median family incomes were 40% to 48% lower in the most-deprived than in the least-deprived area during this period. Median home values were 58% to 69% lower, white-collar employment rates were 27% to 29% lower, and the poverty rate was at least 3 times greater in the most-deprived than in the least-deprived area. More deprived areas had substantially higher proportions of Black and rural residents. Moreover, 62% of the population in the most-deprived areas was located in the South, in contrast to only 19% of the population in the least-deprived areas.
|
Figure 1
shows increasing areal gradients in mortality for both men and women over the past 3 decades. The areal classification based on 20% of counties (rather than 20% of the population) in each quintile produced almost identical gradients. Not only did more deprived areas have higher mortality rates than less-deprived areas during each year examined, but the gradient generally increased over time, especially for men. Although mortality rates for all deprived groups declined during 1969 to 1998, the more-deprived groups experienced slower mortality declines. Mortality among men in the least- through the most-deprived groups declined at average annual rates of 1.64%, 1.51%, 1.40%, 1.34%, and 1.13%, respectively, between 1969 and 1998. The corresponding annual rates of decline among women were 1.23%, 1.19%, 1.04%, 1.08%, and 0.97%.
|
Figure 1
also contains separate trends for Whites and Blacks. Consistently increasing areal gradients in mortality were observed among White men and women. Among Black men, areal gradients were somewhat less consistent but were as pronounced as those for White men, and they showed an increasing trend. Areal gradients were also less consistent for Black women and remained generally stable over time. Age- and sex-specific areal gradients in mortality were computed with the 1970 deprivation index as well (data not shown). Temporal trends were generally similar to those based on the 1990 index. However, areal gradients based on the 1970 index were somewhat less consistent than those based on the 1990 index.
Figure 2 presents age- and race-adjusted relative mortality risks in 5 deprivation groups (based on the 1990 index) derived from Poisson regression analyses. During 1969 through 1998, there was a positive and generally increasing gradient in mortality by area deprivation for both men and women. The increase in areal gradients was particularly marked between 1985 and 1998. The gradients were steepest for men and women aged 25–44 years, followed by those aged younger than 25 years, those aged 45 to 66 years, and those aged 65 years or older. In 1969–1970, mortality rates among men aged younger than 25 years and men aged 25 to 44 years, respectively, were 40% and 56% greater in the most-deprived than in the least-deprived area. The corresponding differentials were 46% and 67% in 1989–1990 and 55% and 77% in 1997–1998.
|
The gradients were less pronounced for women than for men in each age group. However, the mortality differentials between deprivation groups rose sharply during 1969 to 1998 for women aged 25 to 44 and aged 45 to 64 years. In 1969–1970, mortality rates among women aged 25 to 44 and aged 45 to 64 years, respectively, were 32% and 9% greater in the most-deprived than in the least-deprived area. In 1989–1990, the corresponding differentials for women in these age groups were 49% and 21%; in 1997–1998, the differentials were 67% and 29%. Areal gradients in mortality among the elderly, although considerably smaller than those for the other age groups, increased consistently in the 1990s.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
This study involved the use of a composite area–based deprivation index to analyze temporal trends in the extent of inequalities in US mortality during 1969 through 1998 among men and women in different age groups. The present analysis extended an earlier study that focused exclusively on the 25- to 64-year age group in its examination of temporal area socioeconomic inequalities in US all-cause and cardiovascular mortality.21 The findings of the present study are also consistent with investigations showing increasing inequalities in mortality by single areal socioeconomic measures.24–28
An important limitation of the study relates to the use of the 1990 deprivation index to analyze areal inequalities in mortality from 1969 to 1998. Ideally, to allow for temporal sequencing between area deprivation and mortality, a deprivation index defined at the earliest decennial time point (i.e., 1970) was preferable. However, the 1970 and 1990 indices were highly correlated, and use of the 1970 index produced mortality trends similar to those based on the 1990 index. The small degree of areal misclassification that may arise from using the 1990 index is therefore unlikely to significantly affect the general trend of increasing areal inequalities in mortality.9,21
Because of the lack of census tract or block group geocodes, it is not possible to analyze national mortality data at smaller geographic levels.9,21 Although there is a substantial degree of intracounty heterogeneity in sociodemographic conditions, it is unclear whether temporal mortality trends would differ if area deprivation were to be linked to tract-level mortality data. Nevertheless, it is advantageous to use temporal county data. Although census tracts are socioeconomically homogeneous geographic units with an average population of 4000, they are subject to change in every decennial census. Counties, on the other hand, not only are more stable sociopolitical and geographic entities, but also provide an appropriate socioeconomic, political, and community context within which many social and public health policies are formulated and implemented.9,21
Areal inequalities in US mortality have widened because of slower mortality declines among residents of more deprived areas. Although the relative standing of deprivation groups remained fairly stable during the study period, increasing inequalities in absolute deprivation between areas can be noted in Table 2, which may have contributed to increasing inequalities in mortality.21 It would be useful to examine the magnitude of temporal social inequalities in major causes of death (such as cardiovascular disease, cancer, injuries, and infectious diseases) to better understand the specific social, behavioral, and health care mechanisms involved.21,22 Currently, ecological studies of socioenvironmental, behavioral, and health care disparities by area deprivation are lacking in the United States. Future research needs to examine the roles of population distributions of smoking and alcohol use rates, diet, obesity, physical inactivity, environmental pollution, and accessibility and use of health services in explaining the increasing areal gradients shown here.8
Census-based deprivation indices could serve as an important, cost-effective analytic tool for documenting social inequalities in health and for monitoring trends in the extent of inequality over time.9,21 In the absence of routinely collected individual social class data, evaluation of health and mortality data through the use of deprivation indices holds much promise for the public health community’s efforts to reduce health disparities. Caution should be exercised, however, when comparing areal variations in mortality with individual-level socioeconomic differentials.9,14,21,24,31,51 Equating differentials at the 2 levels may lead to an ecological bias. This study analyzed areal variations in mortality as a function of an ecological variable, area deprivation. Although areal deprivation patterns in mortality by age, race, and sex are consistent with those at the individual level, the individual socioeconomic effects are generally larger than those at the area level, and temporal trends in individual socioeconomic inequalities in mortality may differ as well.2,3,14,21,31,52–55
|
Footnotes |
|---|
Note. The views expressed in this article are the author’s and not necessarily those of the National Cancer Institute.
Ethical clearance was not needed because this study used only secondary data and public-use vital statistics and census data. No human participants were contacted as part of this research.
Accepted for publication June 29, 2002.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
1. Singh GK, Yu SM. Infant mortality in the United States: trends, differentials, and projections, 1950 through 2010. Am J Public Health.1995;85:957–964.
2. Pappas G, Queen S, Hadden W, Fisher G. The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med.1993;329:103–109.
3. Feldman JJ, Makuc DM, Kleinman JC, Cornoni-Huntley J. National trends in educational differentials in mortality. Am J Epidemiol.1989;129:919–933.
4. Vital Statistics of the United States, 1970–1997, Volume 2: Mortality, Parts A and B. Hyattsville, Md: National Center for Health Statistics; 1999.
5. Murphy SL. Deaths: final data for 1998. Natl Vital Stat Rep.2000;48(11):1–108.[Medline]
6. Singh GK, Kochanek KD, MacDorman MF. Advance report of final mortality statistics, 1994. Month Vital Stat Rep.1996;45(suppl 3):1–80.
7. Hoyert DL, Singh GK, Rosenberg HM. Sources of data on socioeconomic differential mortality in the United States. J Off Stat.1995;11:233–260.
8. Singh GK, Wilkinson AV, Song FF, et al. Health and Social Factors in Kansas: A Data and Chartbook, 1997–98. Lawrence, Kan: Allen Press; 1998.
9. Singh GK, Miller BA, Hankey BF, Feuer EJ, Pickle LW. Changing area socioeconomic patterns in US cancer mortality, 1950–1998: part I—all cancers among men. J Natl Cancer Inst.2002;94:904–915.
10. Health, United States, 1998, With Socioeconomic and Health Chartbook. Hyattsville, Md: National Center for Health Statistics; 1998.
11. Kitagawa EM, Hauser PM. Differential Mortality in the United States: A Study in Socioeconomic Epidemiology. Cambridge, Mass: Harvard University Press; 1973.
12. Guest AM, Almgren G, Hussey JM. The ecology of race and socioeconomic distress: infant and working-age mortality in Chicago. Demography.1998;35:23–34.[Web of Science][Medline]
13. Singh GK, Yu SM. US childhood mortality, 1950 through 1993: trends and socioeconomic differentials. Am J Public Health.1996;86:505–512.
14. Davey Smith G, Neaton JD, Wentworth D, Stamler R, Stamler J. Socioeconomic differentials in mortality risk among men screened for the Multiple Risk Factor Intervention Trial, I: white men. Am J Public Health.1996;86:486–496.
15. McLaughlin DK, Stokes CS. Income inequality and mortality in US counties. Am J Public Health.2002;92:99–104.
16. Cubbin C, Pickle LW, Fingerhut L. Social context and geographic patterns of homicide among US black and white males. Am J Public Health.2000;90:579–587.
17. Logue EE, Jarjoura D. Modeling heart disease mortality with census tract rates and social class mixtures. Soc Sci Med.1990;31:545–550.
18. Barnett E, Casper ML, Halverson JA, et al. Men and Heart Disease: An Atlas of Racial and Ethnic Disparities in Mortality. Morgantown, WV: West Virginia University; 2001.
19. Casper ML, Barnett E, Halverson JA, et al. Women and Heart Disease: An Atlas of Racial and Ethnic Disparities in Mortality. Morgantown, WV: West Virginia University; 2000.
20. Geller AC, Miller DR, Lew RA, Clapp RW, Wenneker MB, Koh HK. Cutaneous melanoma mortality among the socioeconomically disadvantaged in Massachusetts. Am J Public Health.1996;86:538–544.
21. Singh GK, Siahpush M. Increasing inequalities in all-cause and cardiovascular mortality among US adults aged 25–64 years by area socioeconomic status, 1969–1998. Int J Epidemiol.2002;31:600–613.
22. Singh GK, Miller BA, Hankey BF. Changing area socioeconomic patterns in US cancer mortality, 1950–1998: part II—lung and colorectal cancers. J Natl Cancer Inst. 2002;94:916–925.
23. Wagener DK, Schatzkin A. Temporal trends in the socioeconomic gradient for breast cancer mortality among US women. Am J Public Health.1995;85:1003–1006.
24. Armstrong D, Barnett E, Casper M, Wing S. Community occupational structure, medical and economic resources, and coronary mortality among US blacks and whites, 1980–1988. Ann Epidemiol.1998;8:184–191.[Web of Science][Medline]
25. Wing S, Casper M, Hayes CG, et al. Changing association between community occupational structure and ischaemic heart disease mortality in the United States. Lancet.1987;2(8567):1067–1070.[Web of Science][Medline]
26. Barnett E, Armstrong D, Casper M. Evidence of increasing coronary heart disease mortality among black men of lower social class. Ann Epidemiol.1999;9:464–471.[Web of Science][Medline]
27. Tyroler HA, Wing S, Knowles M. Increasing inequality in coronary heart disease mortality in relation to educational achievement: profile of places of residence, United States, 1962 to 1987. Ann Epidemiol.1993;3(suppl):S51–S54.
28. DiLiberti JH. The relationship between social stratification and all-cause mortality among children in the United States, 1968–1992. Pediatrics.2000;105:1–6.
29. Carstairs V. Deprivation indices: their interpretation and use in relation to health. J Epidemiol Community Health.1995;49(suppl 2):S3–S8.
30. Carstairs V, Morris R. Deprivation: explaining differences in mortality between Scotland and England and Wales. BMJ.1989;299:886–889.
31. Davey Smith G, Hart C, Watt G, Hole D, Hawthorne V. Individual social class, area-based deprivation, cardiovascular disease risk factors, and mortality: the Renfrew and Paisley Study. J Epidemiol Community Health.1998;52:399–405.[Abstract]
32. Eames M, Ben-Sholomo Y, Marmot MG. Social deprivation and premature mortality: regional comparison across England. BMJ.1993;307:1097–1102.
33. McLoone P, Boddy FA. Deprivation and mortality in Scotland, 1981 and 1991. BMJ.1994;309:1465–1470.
34. Morrison A, Stone DH, Redpath A, Campbell H, Norrie J. Trend analysis of socioeconomic differentials in deaths from injury in childhood in Scotland, 1981–95. BMJ.1999;318:567–568.
35. Sloggett A, Joshi H. Higher mortality in deprived areas: community or personal disadvantage? BMJ.1994;309:1470–1474.
36. Benach J, Yasui Y. Geographical patterns of excess mortality in Spain explained by two indices of deprivation. J Epidemiol Community Health.1999;53:423–431.[Abstract]
37. Salmond C, Crampton P, Sutton F. NZDep91: a New Zealand index of deprivation. Aust N Z J Public Health.1998;22:835–837.[Web of Science][Medline]
38. Enough to Make You Sick: How Income and Environment Affect Health. Canberra, Australian Capital Territory, Australia: Australian Government Publishing Service; 1992.
39. Berkman LF, Macintyre S. The measurement of social class in health studies: old measures and new formulations. In: Kogevinas M, Pearce N, Susser M, Boffetta P, eds. Soc Inequalities Cancer. Lyon, France: International Agency for Research on Cancer; 1997:51–64.
40. Krieger N, Williams DR, Moss NE. Measuring social class in US public health research: concepts, methodologies, and guidelines. Annu Rev Public Health.1997;18:341–378.[Web of Science][Medline]
41. Link BG, Phelan JC. Understanding sociodemographic differences in health: the role of fundamental social causes. Am J Public Health.1996;86:471–473.
42. Census of Population and Housing, 1990: Summary Tape File 3A on CD-ROM. Washington, DC: US Bureau of the Census; 1992.
43. The Area Resource File (ARF): Public Use File Technical Documentation. Rockville, Md: Health Resources and Services Administration; 1996.
44. Kim J, Mueller C. Factor Analysis: Statistical Methods and Practical Issues. London, England: Sage Publications; 1978. Sage University Paper Series on Quantitative Applications in the Social Sciences, No. 14.
45. SAS/STAT User’s Guide, Version 8, Volume 1: The FACTOR Procedure. Cary, NC: SAS Institute Inc; 1999.
46. Carmines EG, Zeller RA. Reliability and Validity Assessment. London, England: Sage Publications; 1979. Sage University Paper Series on Quantitative Applications in the Social Sciences, No. 7.
47. Sink L. Estimates of the Population of Counties by Age, Sex, Race and Hispanic Origin: 1990 to 1998. Washington, DC: US Bureau of the Census; 1999.
48. Hollmann FW. United States Population Estimates, by Age, Sex, Race and Hispanic Origin: 1980 to 1988. Washington, DC: US Bureau of the Census; 1990. Current Population Reports, Series P-25, No. 1045.
49. Agresti A. An Introduction to Categorical Data Analysis. New York, NY: John Wiley & Sons Inc; 1996.
50. SAS/STAT Software: Changes and Enhancements Through Release 6.12: The GENMOD Procedure. Cary, NC: SAS Institute Inc; 1997.
51. MacRae K. Socioeconomic deprivation and health and ecological fallacy. BMJ.1994;309:1478–1479.
52. Ben-Shlomo Y, Davey Smith G. Commentary: socioeconomic position should be measured accurately. BMJ.1999;318:844–845.
53. Singh GK, Siahpush M. All-cause and causespecific mortality of immigrants and native born in the United States. Am J Public Health.2001;91:392–399.
54. Sorlie PD, Backlund E, Keller JB. US mortality by economic, demographic, and social characteristics: the National Longitudinal Mortality Study. Am J Public Health.1995;85:949–956.
55. Jefferys M. Social inequalities in health—do they diminish with age? Am J Public Health.1996;86:474–475.
This article has been cited by other articles:
|
|
 |
|
  R Ocana-Riola, C Saurina, A Fernandez-Ajuria, A Lertxundi, C Sanchez-Cantalejo, M Saez, M Ruiz-Ramos, M A Barcelo, J C March, J M Martinez, et al. Area deprivation and mortality in the provincial capital cities of Andalusia and Catalonia (Spain) J Epidemiol Community Health, February 1, 2008; 62(2): 147 - 152. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M Kelly, M. Schootman, E. A Baker, E. K Barnidge, and A. Lemes The association of sidewalk walkability and physical disorder with area-level race and poverty J Epidemiol Community Health, November 1, 2007; 61(11): 978 - 983. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Satcher, G. E. Fryer Jr., J. McCann, A. Troutman, S. H. Woolf, and G. Rust What If We Were Equal? A Comparison Of The Black-White Mortality Gap In 1960 And 2000 Health Aff., March 1, 2005; 24(2): 459 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Boyle Population geography: migration and inequalities in mortality and morbidity Progress in Human Geography, December 1, 2004; 28(6): 767 - 776. [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |