HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: AJPH.2004.054700v1 95/11/2035    most recent Submit a response Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Nusselder, W. J. Articles by Kunst, A. E. Search for Related Content PubMed PubMed Citation Articles by Nusselder, W. J. Articles by Kunst, A. E. Related Collections Disability Socioeconomic Factors Mortality Surveys

Wilma J. Nusselder, PhD, Caspar W.N. Looman, MSc, Johan P. Mackenbach, MD, PhD, Martijn Huisman, PhD, Herman van Oyen, MD, DrPH, MPH, DTM&H, Patrick Deboosere, MA, Sylvie Gadeyne, MA and Anton E. Kunst, PhD

Wilma J. Nusselder, Anton E. Kunst, Johan P. Mackenbach, Martijn Huisman, and Caspar W.N. Looman are with the Department of Public Health, Erasmus MC, Rotterdam, The Netherlands. Sylvie Gadeyne and Patrick Deboosere are with the Interface Demography, Centrum voor Sociologie, VUB, Brussels, Belgium. Herman van Oyen is with the Unit of Epidemiology, Scientific Institute for Public Health, Brussels.

Correspondence: Requests for reprints should be sent to Dr. Wilma J. Nusselder, Erasmus MC, Department of Public Health, PO Box 1738, 3000 DR Rotterdam, The Netherlands (e-mail: w.nusselder{at}erasmusmc.nl).

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 

Objectives. We examined the contribution that specific diseases, as causes of both death and disability, make to educational disparities in disability-free life expectancy (DFLE).

Methods. We used disability data from the Belgian Health Interview Survey (1997) and mortality data from the National Mortality Follow-Up Study (1991–1996) to assess education-related disparities in DFLE and to partition these differences into additive contributions of specific diseases.

Results. The DFLE advantage of higher-educated compared with lower-educated persons was 8.0 years for men and 5.9 years for women. Arthritis (men, 1.3 years; women, 2.2 years), back complaints (men, 2.1 years), heart disease/stroke (men, 1.5 years; women, 1.6 years), asthma/chronic obstructive pulmonary disease (COPD) (men, 1.2 years; women, 1.5 years), and "other diseases" (men, 2.4 years) contributed the most to this difference.

Conclusions. Disabling diseases, such as arthritis, back complaints, and asthma/COPD, contribute substantially to differences in DFLE by education. Public health policy aiming to reduce existing disparities in the DFLE and to improve population health should not only focus on fatal diseases but also on these nonfatal diseases.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 
Socioeconomic differences in health have been reported to be substantial and persistent. Mortality and morbidity, including the prevalence of reported chronic conditions and disability, is higher in the lower socioeconomic groups than among persons with a higher socioeconomic status.^1,2 Together, these disadvantages in mortality and morbidity have been shown to create large differences in population health. For a variety of definitions of health, for different indicators of socioeconomic status, and on the basis of both cross-sectional and longitudinal data, studies on socioeconomic differences in health expectancy have consistently shown that persons in the lower socioeconomic groups spend fewer years free of disability or in good health. Moreover, despite their shorter total life expectancy, these persons live longer with disability or ill health than persons in higher socioeconomic groups.^3–16

Elimination of inequalities in population health is a primary goal of health politics.^1,17 Greatest success is likely to be achieved by targeting diseases that have the greatest impact on inequalities in health. Some prior studies have examined the contribution of specific diseases to socioeconomic health differences. Mortality rates among persons with lower socioeconomic status were shown to be higher for almost all causes of death,^1,18,19 but the contribution of specific causes to differences in total mortality has been found to vary between countries.^19,20 Only 3 studies^21–23 assessed the contribution of specific causes to disparities in life expectancy, showing largest contributions for ischemic heart diseases, other cardiovascular diseases, cancers, and respiratory diseases. A major limitation of these studies, however, is that they included only the fatal consequences of diseases. Socioeconomic differences in nonfatal health outcomes have been taken into account in studies on health expectancy. Although socioeconomic differences in health expectancy have shown to be even more pronounced than in life expectancy,^3–16 none of these studies has examined the contribution of specific diseases to these differences.

We extended prior studies on the contribution of specific diseases to socioeconomic health differences in disability-free life expectancy (DFLE). On the basis of Belgian data, we used a new method²⁴ to examine the contribution of specific diseases to inequalities in health expectancy measures. Our study assessed the contribution that 7 disease groups make to educational differences in DFLE.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 
We used Belgian data because of the availability of information by level of education both on disability and chronic disease from the large Health Interview Survey and on mortality from the National Mortality Follow-Up Study.

Mortality
Data on the number of deaths by age, gender, level of education, and underlying cause of death for the period 1991–1996 were obtained from the National Mortality Follow-Up Study.^8,25 This study was based on a linkage of the 1991 census with the National Register (1991–1996). In the analyses, we included persons aged 30 and over, yielding 27 635 thousand person-years at risk and 486 thousand deaths in the period of 1991–1996.

Disability
Cross-sectional data on long-term disability by gender, age, and level of education were obtained from Belgian Health Interview Survey 1997 (HIS).²⁶ This survey was based on a stratified, multistage sample from the National Register of the entire population of Belgium, without restriction to nationality or age. Of 11568 households that were initially selected for an interview, 3546 were not eligible or could not be contacted for different reasons. The percentage of households agreeing to participate after being contacted was 58.1% (n=4664).²⁷ Within the participating households, 10339 subjects were selected for a face-to-face interview, of which 10221 subjects were interviewed. The probability of nonparticipation was higher among the smaller households, younger persons, women, in the Brussels region, and among persons living in institutions. In this study, we included only persons aged 30 and over (n = 6632) but excluded 171 persons with incomplete information on disability (n=25), chronic diseases (n=48), or level of education (n=98), yielding 6461 subjects in the analyses (3143 men and 3318 women). We used normalized weights (with a mean of 1) to take into account the complex sampling design and differential nonresponse.²⁸

Long-term disability was measured with functional limitations in mobility included in the short form health survey (SF 36).²⁹ Functional limitations occur at an early stage in the disablement process and are seen as precursors of activities of daily living (ADL) disability occurring at a later stage.³⁰ Persons were considered to be disabled if they indicated that they had 1 or more moderate or severe limitations in lifting groceries, climbing 1 or more flights of stairs, bending, kneeling, stooping, or walking 1 block or longer distances.

Definition of Disease Groups and Level of Education
We used self-reported data on the presence of disability and chronic diseases in the HIS to estimate disability by cause (see Statistical Analysis). The presence or absence of chronic diseases was assessed on the basis of a structured list ("checklist"), comprising a broad number of somatic diseases, and a category of "other diseases." The list did not cover all chronic conditions; for example, injuries and dementia were not included. Seven disease groups were compiled from the original chronic conditions included in the checklist of chronic conditions: asthma/chronic obstructive pulmonary disease (COPD), heart disease/stroke, diabetes mellitus, back complaints, arthritis, cancer, and 1 group including all other diseases. Disability that could not be attributed to these disease groups was included in "background disability." Deaths by cause, classified according to the International Classification of Diseases, Ninth Revision (ICD-9),³¹ were grouped into the same disease groups, assuming no mortality from back complaints and arthritis (see footnote to Table 4). Persons with an unknown cause of death (n=7404) were included in the group "other causes of death." Distributing these deaths proportionally across the known causes did not alter the results.

Statistical Analysis
Disability data by cause of disease were not available and had to be estimated from individual information on the presence or absence of disability, the presence or absence of various disease groups, age, and level of education. We used a multivariate additive regression model,³² which is described in more detail elsewhere.²⁴ The method takes into account that subjects without a reported disease can be disabled (this risk is referred to as "background") and that subjects can have >1 disease (comorbidity). The main reason for using this method is that additive cause-specific disability prevalence can be obtained in the presence of competing causes of disability. The regression model is specified as follows:

(1)

(2)

where is the estimated probability that the person has at least moderate disability, e is the base of the natural logarithm and the linear predictor. The latter is defined as the sum of the background rate by gender, 5-year age group, and education (_gae) and the cause-specific rates of disability (ß_dgae, labeled as "disabling impact") for the diseases (d ) that are present in the respondent (given by the dummy variable _d ). Background disability was handled as a cause that is prevalent for everyone, to reflect the fact that disability also occurs in cases that cannot be attributed to reported diseases. The disabling impact is estimated as the product of an age pattern (equal for each disease, gender, and level of education) and a disease effect, which varies by gender, level of education, and disease. The age pattern of the disabling impact (according to 3 broad age groups) allows the disabling impact to increase by age. The disabling impacts obtained from the model may differ by disease, gender, age group and level of education.

Disability prevalence by cause estimated from the regression model (Generalized Linear Interactive Modeling 4 [GLIM]; NAG Ltd., Oxford, UK) depends on the prevalence of the disease and the disabling impact of the disease (Table 2).

The contribution of specific diseases to educational differences in DFLE was estimated by using a decomposition tool to partition the differences in health expectancy into additive contributions of causes.²⁴ This technique is based on the Sullivan method and is an extension of the Arriaga method for total life expectancy.^36,37 The tool assesses the difference in health expectancy because of smaller (higher) total mortality rates and/or disability prevalence (by age) from a given cause, relative to a reference year/group. First, the difference in the number of person-years with(out) disability (by age) is decomposed into 2 parts: the first part reflecting the smaller (larger) number of person-years lived ("mortality effect") and the second part reflecting the smaller (higher) prevalence of disability ("disability effect"). Second, these differences in age-specific mortality rates and disability prevalence are further decomposed by cause.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 
Table 1 summarizes the prevalence of cause-specific disability by age, gender, and level of education. Important causes of disability are arthritis, heart disease/stroke, "other diseases," back complaints (younger ages), and asthma/COPD. For most diseases, cause-specific disability is higher among the lower educated, except for back complaints (women) and "background." Differences in the prevalence of cause-specific disability are the result of differences in the prevalence of the disease and in the disabling impact. Most diseases were more prevalent among lower-educated persons, the major exception being back complaints in women (data not shown). Also the disabling impact is higher among the lower educated, in particular for cancer, arthritis, heart disease/stroke, back complaints (only men), and asthma/COPD (only women) (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 
This is the first study that has assessed the overall contribution that various diseases make to socioeconomic differences in population health. Using DFLE in addition to life expectancy offered us the possibility to study the contribution of specific diseases as causes of both death and disability to socioeconomic differences in health. We showed that the causes contributing most to the disadvantage in DFLE of lower-educated persons include the following: back complaints (men) and arthritis (women), "other diseases" (men), followed by heart disease/stroke and asthma/COPD.

Our results support the conclusion that mortality from heart disease/stroke, "other diseases," cancer, and asthma/COPD among lower socioeconomic groups contributes most to the lower life expectancy of this group.^21–23 Our study confirms that educational inequalities in DFLE exceed inequalities in life expectancy³ and showed that this also holds for specific causes. Higher mortality from heart disease/stroke, asthma/COPD, and "other diseases" reduced the DFLE of the lower educated. On top of this, the higher prevalence of disability from these causes further increased their disadvantage in DFLE. In addition to this contribution from diseases that are both fatal and disabling, higher disability prevalence caused by nonfatal diseases increased the DFLE disadvantage of the lower educated further. For most diseases, higher disability prevalence reflected a combination of higher disease prevalence and higher disabling impact. This is in line with existing information pointing at the higher prevalence of chronic diseases, including cardiovascular diseases, respiratory diseases, musculoskeletal diseases,^38–42 and the less favorable course of chronic diseases and of long-term disabilities⁴³ among lower socioeconomic groups.

A number of limitations of the data and methods need to be considered in evaluating the results. Relying on respondents’ self-reports of morbidity may have biased the results, in particular when differences exist in reporting behavior between lower and higher educated persons. Information on the effect of socioeconomic status on reporting of disability is very limited. However, the only study⁴⁴ that has assessed the potential effect of education on self-reporting of disability did not find such an effect. Moreover, performance-based measures confirm that lower-educated persons are more disabled than higher-educated persons.^45,46 With respect to the accuracy of self-reports of chronic diseases, most uncertainty relates to arthritis.^47–49 Although persons with pain or stiffness may have attributed their complaints to arthritis without having consulted their general practitioner, no differences in overreporting by level of education were found.⁴⁷ Underreporting of arthritis has been shown to be higher among persons with lower education,⁴⁷ and thus, our estimate of the contribution of arthritis to educational differences in DFLE might be conservative. Although for other diseases it is not established whether education has an effect on (under)reporting,^47,50,51 if such an effect would be present, it might be limited because underreporting is less likely to occur in persons with disability.^47,48 So, although the causes of disability should not be considered as precisely and clinically diagnosed diseases, and confirmation of our results with clinical data on diseases is warranted, we have no reasons to expect that our overall conclusions are seriously biased.

We used statistical associations between self-reported disability and diseases to attribute disability to diseases (and to background), assuming that diseases and conditions that caused disability were still present and reported in the survey. Violation of this assumption, occurring in situations where disability is caused by (1) prior conditions (such as accidents), (2) diseases still present but not mentioned as separate entities in the checklist (such as dementia), and/or (3) diseases included in the checklist, but not reported, will have caused an overestimation of "background" at the expense of "other diseases" or a specific disease group. We cannot rule out that violation of the assumption occurs more often among lower-educated persons, because this population is more often involved in accidents and has more health problems,¹ and might be more likely to underreport chronic disease.⁵⁰ However, it is noteworthy that most prevalent disabling diseases were included in the checklist. Moreover, we found that disability from background was higher in higher-educated persons (at older ages). This is a puzzling finding but might reflect—more important than misclassification—that higher-educated persons have more disability as a result of frailty or "old age," which could relate to less mortality selection in that group.

Although we used a similar classification of education, the percentage of lower-educated persons was 69% in the mortality follow-up of the census compared with 46% in the HIS. There is evidence to suggest that there is some differential misclassification of individuals according to their educational level, especially in the survey. This misclassification implies that differences in DFLE between persons with lower and higher education might be larger than presented in our study.

Socioeconomic differences in institutionalization—in combination with the underrepresentation of this population segment in the HIS—might have caused an underestimation of the educational differences in DFLE, and more importantly of the contribution of diseases that are prevalent among institutionalized persons, such as stroke and dementia (included in "other diseases"). However, because mortality risks of persons living in institutions are very high, the effect will be only small.

The calculation of DFLE and the decomposition of differences in DFLE are both based on the Sullivan method, which is the standard method for health expectancy calculations on a routine basis. Although the Sullivan method generally provides a good measure of the current health composition of a population (group),^52,53 and thus of educational differences, it is not based on transition rates. Consequently, the decomposition analysis quantifies to what extent differences in disability prevalence and total mortality (in each age group) from each cause contribute to differences in DFLE but does not show which underlying health dynamics contribute most to these differences.

The results of this study may depend upon the disability measure used. Prior work on socioeconomic differences in DFLE has found that the overall pattern of lower DFLE among lower socioeconomic groups is present across all disability measures⁷ but that the size of difference in DFLE varies between disability measures.^3,5 Because some diseases were found to affect specific disabilities,⁵⁴ the choice of the items included in the disability indicator might affect the association between specific diseases and disability. In an explorative analysis using a different disability indicator that included sensory and mobility limitations as well as restrictions in ADL, we found that, although the contributions of specific diseases to the difference in DFLE differed, the overall picture pointing at the large impact of nonfatal diseases was the same.

Our study provides important information for policymakers and researchers, because it documents the large role of nonfatal diseases such as back complaints and arthritis in socioeconomic differences in population health. The large contribution of these diseases remained undetected in prior studies, simply because only fatal consequences of diseases were taken into account. Our results show that, although higher mortality from fatal diseases in lower-educated persons reduces their DFLE, the major contribution of these diseases is through their higher prevalence of disability in persons with lower education. Although years lost to mortality and years lived with disability cannot be weighted equally, and reducing mortality inequalities should stay high on the priority list, the burden of disabling diseases, responsible for health inequalities among survivors, should not be ignored. Next to reducing inequalities in the onset and course of fatal diseases, a major challenge lies in reducing the onset and disabling impact of nonfatal diseases among lower socioeconomic groups.

	Acknowledgments

 
This study was funded by the European Union Fifth Framework Program on Quality of Life and Management of Living Resources (contract QLK6-CT-1999-02161).

Human Participant Protection
No protocol approval was needed for this study.

	Footnotes

 
Peer Reviewed

Contributors
W. J. Nusselder, A. E. Kunst, and J. P. Mackenbach originated and designed the study. M. Huisman, S. Gadeyne, P. Deboosere, and H. van Oyen delivered and processed the empirical data. W. J. Nusselder and C. W. N. Looman developed and applied the decomposition technique. W. J. Nusselder and A. E. Kunst wrote the article. All of the authors interpreted the results and commented on drafts of the article.

Accepted for publication February 15, 2005.

	References

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION References

 
1. Mackenbach JP, Bakker J, Kunst AE, Dideerichsen F. Socioeconomic differences in health: an overview. In: Mackenbach JP, Bakker J, eds. Reducing Inequalities in Health. London, England: Routledge; 2002:3–24.

2. Amaducci L, Maggi S, Langlois J, et al. Education and the risk of physical disability and mortality among men and women aged 65 to 84: the Italian Longitudinal Study on Aging. J Gerontol A Biol Sci Med Sci. 1998;53:M484–M490.[Abstract]

3. Cambois E, Robine JM, Hayward M. Social inequalities in disability-free life expectancy in the French male population. Demography. 2001;38:513–524.[ISI][Medline]

4. Crimmins EM, Hayward MD, Saito Y. Differentials in active life expectancy in the older population of the United States. J Gerontol B Psychol Sci Soc Sci. 1996; 51:S111–S120.[Abstract]

5. Doblhammer G, Kytir J. Social inequalities in disability-free and healthy life expectancy in Austria. Wien Klin Wochenschr. 1998;110:393–396.[ISI][Medline]

6. Sihvonen AP, Kunst AE, Lahelma E, Valkonen T, Mackenbach JP. Socioeconomic inequalities in health expectancy in Finland and Norway in the late 1980s. Soc Sci Med. 1998;47:303–315.

7. Valkonen T, Sihvonen AP, Lahelma E. Health expectancy by level of education in Finland. Soc Sci Med. 1997;44:801–8.

8. Bossuyt N, Gadeyne S, Deboosere P, Van Oyen H. Socioeconomic inequalities in health expectancy in Belgium. Public Health. 2004;118:3–10.[CrossRef][ISI][Medline]

9. Wilkins R, Adams OB. Health expectancy in Canada, late 1970s: demographic, regional, and social dimensions. Am J Public Health. 1983;73:1073–1080.[Abstract/Free Full Text]

10. Bebbington A. Regional and social variations in disability-free life expectancy in Great Britain. In: Robine JM, Mathers CD, Bone MR, Romieu I, eds. Calculation of Health Expectancies: Harmonization, Consensus Achieved and Future Perspectives. Montpellier, France: Colloque INSERM/John Libbey Eurotext; 1993:175–191.

11. Guralnik JM, Land KC, Blazer D, Fillenbaum GG, Branch LG. Educational status and active life expectancy among older blacks and whites. N Engl J Med. 1993;329:110–116.[Abstract/Free Full Text]

12. Martinez-Sanchez E, Gutierrez-Fisac JL, Gispert R, Regidor E. Educational differences in health expectancy in Madrid and Barcelona. Health Policy. 2001;55:227–231.[CrossRef][ISI][Medline]

13. Melzer D, McWilliams B, Brayne C, Johnson T, Bond J. Socioeconomic status and the expectation of disability in old age: estimates for England. J Epidemiol Community Health. 2000;54:286–292.[Abstract/Free Full Text]

14. Kaprio J, Sarna S, Fogelholm M, Koskenvuo M. Total and occupationally active life expectancies in relation to social class and marital status in men classified as healthy at 20 in Finland. J Epidemiol Community Health. 1996;50:653–660.[Abstract/Free Full Text]

15. Bronnum-Hansen H, Andersen O, Kjoller M, Rasmussen NK. Social gradient in life expectancy and health expectancy in Denmark. Soz Praventivmed. 2004;49:36–41.[ISI][Medline]

16. Crimmins EM, Cambois E. Social inequalities in health expectancies. In: Robine JM, Jagger C, Mathers CD, Crimmins EM, Suzman RM, eds. Determining Health Expectancies. Chichester, England: John Wiley & Sons; 2003:111–125.

17. World Health Organization Regional Office for Europe. Health21:An Introduction to the Health for All Policy Framework for the WHO European Region. Copenhagen, Denmark: World Health Organization; 1998.

18. Steenl andand K, Henley J, Thun M. All-cause and cause-specific death rates by educational status for two million people in two American Cancer Society cohorts, 1959–1996. Am J Epidemiol. 2002;156:11–21.[Abstract/Free Full Text]

19. Kunst AE, Groenhof F, Mackenbach JP, Health EW. Occupational class and cause specific mortality in middle aged men in 11 European countries: comparison of population-based studies. EU Working Group on Socioeconomic Inequalities in Health. BMJ. 1998;316:1636–1642.[Abstract/Free Full Text]

20. Huisman M, Kunst A, Bopp M, et al. Educational inequalities in cause-specific mortality in middle-aged and older men and women in eight western European populations. Lancet. 2005;365:493–500.[ISI][Medline]

21. Wong MD, Shapiro MF, Boscardin WJ, Ettner SL. Contribution of major diseases to disparities in mortality. N Engl J Med. 2002;347:1585–1592.[Abstract/Free Full Text]

22. Shkolnikov VM, Valkonen T, Begun A, Andreev EM. Measuring intergroup-inequalities in length of life. Genus. 2001;52:33–62.

23. Valkonen T. Socio-economic mortality differences in Europa. In: Beets G, Van den Brekel H, Cliquet R, Dooghe G, de Jong-Gierveld J, eds. Population and Familiy in the Low Countries 1993: Late Fertility and Other Current Issues. Lisse, The Netherlands: Swets & Zeitlinger; 1994:127–150.

24. Nusselder WJ, Looman CW. Decomposition of differences in health expectancy. Demography. 2004;41:315–334.[CrossRef][ISI][Medline]

25. Gadeyne S, Deboosere P. Socio-economische ongelijkheid in sterfte op middelbare leeftijd. Statistics Belgium Working Paper. Brussels, Belgium: Statbel; March 2002. Report 6.

26. Health Interview Survey 1997. Brussels, Belgium: Scientific Institute of Public Health, Epidemiology Unit; 1997. Available at: http://www.iph.fgov.be/epidemio/epien. Accessed September 7, 2004.

27. Burzykowski T, Molenberghs G, Tafforeau J, Van Oyen H, Demarest S, Bellamammer L. Missing data in the Health Interview Survey 1997 in Belgium. Arch Public Health. 1999;57:107–129.

28. Snedecor GW, Cochran WG. Statistical Methods. Iowa City, IA: The Iowa University Press; 1967.

29. Ware JEJr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473–483.[ISI][Medline]

30. Robine JM, Jagger C, Romieu I. Setting a Coherent Set of Health Indicators. Montpellier, France: European Commission; 2002.

31. International Classification of Diseases, Ninth Revision. Geneva, Switzerland: World Health Organization; 1980.

32. Clayton D, Hills N. Statistical Models in Epidemiology. Oxford, UK: Oxford University Press; 1993.

33. Sullivan DF. A single index of mortality and morbidity. HSMHA Health Reports. 1971;86:347–354.[ISI][Medline]

34. Sullivan DF. Disability Components for an Index of Health. Washington, DC: Government Printing Office; 1971. PHS publication 1000, series 2, report 42.

35. Namboodiri K, Suchindran CM. Life Tables and Their Applications. Orlando, Fl: Academic Press; 1987.

36. Arriaga EE. Measuring and explaining the change in life expectancies. Demography. 1984;21:83–96.[ISI][Medline]

37. Arriaga EE. Changing trends in mortality decline during the last decades. In: Ruzicka L, Wunsch G, Kane P, eds. Differential Mortality. Methodological Issues and Biosocial Factors. Oxford, UK: Clarendon Press; 1989:105–129.

38. Gutzwiller F, La Vecchia C, Levi F, Negri E, Wietlisbach V. Education, disease prevalence and health service utilization in the Swiss National Health Survey "SOMIPOPS." Prev Med. 1989;18:452–459.[CrossRef][ISI][Medline]

39. La Vecchia C, Negri E, Pagano R, Decarli A. Education, prevalence of disease, and frequency of health care utilisation. The 1983 Italian National Health Survey. J Epidemiol Community Health. 1987; 41:161–165.[Abstract/Free Full Text]

40. Pincus T, Callahan LF, Burkhauser RV. Most chronic diseases are reported more frequently by individuals with fewer than 12 years of formal education in the age 18–64 United States population. J Chronic Dis. 1987;40:865–874.[CrossRef][ISI][Medline]

41. Mackenbach JP. Socio-economic health differences in The Netherlands: a review of recent empirical findings. Soc Sci Med. 1992;34:213–26.

42. Sturm R, Gresenz CR. Relations of income inequality and family income to chronic medical conditions and mental health disorders: national survey. BMJ. 2002;324:20–23.[Abstract/Free Full Text]

43. van der Meer JB, Mackenbach JP. Course of health status among chronically ill persons: differentials according to level of education. J Clin Epidemiol. 1998; 51:171–179.[CrossRef][ISI][Medline]

44. Daltroy LH, Larson MG, Eaton HM, Phillips CB, Liang MH. Discrepancies between self-reported and observed physical function in the elderly: the influence of response shift and other factors. Soc Sci Med. 1999; 48:1549–1561.

45. Parker MG, Thorslund M, Lundberg O. Physical function and social class among Swedish oldest old. J Gerontol. 1994;49:S196–S201.[ISI][Medline]

46. Berkman LF, Seeman TE, Albert M, et al. High, usual and impaired functioning in community-dwelling older men and women: findings from the MacArthur Foundation Research Network on Successful Aging. J Clin Epidemiol. 1993;46:1129–1140.[CrossRef][ISI][Medline]

47. Kriegsman DM, Penninx BW, van Eijk JT, Boeke AJ, Dee.g., DJ. Self-reports and general practitioner information on the presence of chronic diseases in community dwelling elderly. A study on the accuracy of patients’ self-reports and on determinants of inaccuracy. J Clin Epidemiol. 1996;49:1407–1417.[CrossRef][ISI][Medline]

48. Metzger MH, Goldberg M, Chastang JF, Leclerc A, Zins M. Factors associated with self-reporting of chronic health problems in the French GAZEL cohort. J Clin Epidemiol. 2002;55:48–59.[CrossRef][ISI][Medline]

49. Simpson CF, Boyd CM, Carlson MC, Griswold ME, Guralnik JM, Fried LP. Agreement between self-report of disease diagnoses and medical record validation in disabled older women: factors that modify agreement. J Am Geriatr Soc. 2004;52:123–127.[CrossRef][ISI][Medline]

50. Mackenbach JP, Looman CW, van der Meer JB. Differences in the misreporting of chronic conditions, by level of education: the effect on inequalities in prevalence rates. Am J Public Health. 1996;86:706–711.[Abstract/Free Full Text]

51. Wu SC, Li CY, Ke DS. The agreement between self-reporting and clinical diagnosis for selected medical conditions among the elderly in Taiwan. Public Health. 2000;114:137–142.[CrossRef][ISI][Medline]

52. Crimmins EM. What can we expect from summary indicators of population health? In: Murray CJL, Salomon JA, Mathers CD, Lopez AD, eds. Summary Measures of Population Health. Geneva, Switzerland: World Health Organization; 2002:213–219.

53. Mathers CD, Robine JM. How good is Sullivan’s method for monitoring changes in population health expectancies? J Epidemiol Community Health. 1997;51:80–86.[Abstract/Free Full Text]

54. Fried LP, Guralnik JM. Disability in older adults: evidence regarding significance, etiology, and risk. J Am Geriatr Soc. 1997;45:92–100.[ISI][Medline]

This article has been cited by other articles:

				P. Sainio, T. Martelin, S. Koskinen, and M. Heliovaara Educational differences in mobility: the contribution of physical workload, obesity, smoking and chronic conditions J Epidemiol Community Health, May 1, 2007; 61(5): 401 - 408. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: AJPH.2004.054700v1 95/11/2035    most recent Submit a response Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Nusselder, W. J. Articles by Kunst, A. E. Search for Related Content PubMed PubMed Citation Articles by Nusselder, W. J. Articles by Kunst, A. E. Related Collections Disability Socioeconomic Factors Mortality Surveys