You are here

  • Home
  • Archive
  • Volume 68, Issue 4
  • Socioeconomic disadvantage and disease-specific mortality in Asia: systematic review with meta-analysis of population-based cohort studies

Article Text

Article menu

Download PDFPDF

Review
Socioeconomic disadvantage and disease-specific mortality in Asia: systematic review with meta-analysis of population-based cohort studies
  1. Prin Vathesatogkit1,2,
  2. G David Batty3,
  3. Mark Woodward1,4
  1. 1The George Institute for Global Health, University of Sydney, Sydney, New South Wales, Australia
  2. 2Department of Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
  3. 3Department of Epidemiology and Public Health, University College London, London, UK
  4. 4Department of Epidemiology, Johns Hopkins University, Baltimore, Maryland, USA
  1. Correspondence to Dr Prin Vathesatogkit, The George Institute, Level 10, King George V building, Missenden Road, Camperdown, NSW 2050, Australia; prinv{at}hotmail.co.uk

Abstract

Background That socioeconomic deprivation has shown a correlation with disease-specific mortality in Western societies is well documented. However, it is unclear whether these findings are also apparent in Asian societies. Accordingly, we conducted the first systematic review and meta-analysis of studies from Asia that have reported on the association between socioeconomic position and adult mortality risk.

Methods Relevant studies were identified through an electronic search of databases. Studies were included if they had published quantitative estimates of the association between socioeconomic status (SES) and mortality in a general population-based sample. The pooled multiple-adjusted relative risks for mortality in the lowest SES group relative to the highest SES group were studied. Random effects meta-analyses were computed.

Results A total of 29 cohort studies from 10 Asian countries were identified, comprising 1 370 023 individuals and 71 818 total deaths. The three markers of SES most widely used (education, income, occupation) were inversely related to mortality outcomes under consideration (all-causes, cardiovascular disease, cancer). For instance, the pooled relative risks of low education compared with high education were: 1.40 (95% CI 1.29 to 1.52) for all-cause mortality, 1.66 (1.23 to 2.25) for cardiovascular mortality and 1.16 (1.07 to 1.27) for cancer mortality. There was some evidence that the age of cohort members at study induction, the gross national product of the country from which the cohort was drawn and geographical region modified the association between SES and mortality.

Conclusions Concordant with findings from Western societies, socioeconomic disadvantage is associated with total and cause-specific mortality in Asia.

https://doi.org/10.1136/jech-2013-203053

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Socioeconomic inequalities in health are well documented; with the exception of very few outcomes, poorer health is more common in the disadvantaged.1 Reducing these differentials has become a priority for many governments, including the UK2 ,3 and USA,4 and for WHO.5 ,6 To date, most of the evidence for socioeconomic inequalities in health comes from western societies. By contrast, there is a paucity of data on the impact of socioeconomic status (SES) on adult mortality in Asia.

    Using the 2010 U.S. Census Bureau definition of the Asian,7 we undertook a systematic review of observational cohorts examining the relation of SES with disease-specific mortality in Asia, including a meta-analysis of the aggregated evidence. To our knowledge, this is the first study of this kind using purely Asian data.

    Methods

    We conducted a systematic review of the literature following the Meta-Analysis Of Observational Studies in Epidemiology guidelines8 and the Preferred Reporting Items for Systematic reviews and Meta-Analysis statement.9 The primary focus of this systematic review was the association between socioeconomic factors and adult mortality in Asian populations.

    Definition of Asian

    ‘Asian’ refers to a person having origins in any of the original peoples of the Far East, South-East Asia or the Indian subcontinent.7

    Search strategy

    We used a four-stage approach for identifying pertinent publications. First, we performed an electronic search of a wide range of databases from January 1966 to 31 May 2013: Pubmed, Embase, CINAHL, Social Science Research Network and the Cochrane library. The following search terms were used: (social class [MeSH] or socioeconomic factors [MeSH] or education [MeSH] or income [MeSH] or occupation [MeSH]) and mortality [MeSH] and cohort [all field] and Asia [all field], and by country. There was no language restriction. To ascertain relevance, we read the titles and abstracts of all of the articles identified. The full texts of the articles selected were read to determine if they met our criteria for inclusion. Second, we scrutinised the reference lists of the identified articles for further publications, including other publications from the same study. Third, we searched our own files. Finally, we contacted some experts in the field to ascertain if they could add to our list of retrieved publications.

    Data extraction

    Articles were considered to be relevant if they reported findings from prospective Asian cohort studies that included estimates of relative risks (RRs) (including HRs and rate ratios) for all-cause, cardiovascular disease (CVD) or cancer mortality in a general population according to different levels of SES. SES measures accepted were education, income (including economic hardship, household economic status and dimensions of the subject's dwelling) and occupation (limited to manual vs non-manual and blue vs white collar occupation). Articles were excluded if they were performed in hospital patients or based outside Asia. Articles were also excluded if they gave the RR without an associated metric of uncertainty, such as a 95% CI. In the case of multiple publications from the same or overlapping cohorts, we selected the article with the largest number of person-years of follow-up. If two studies tied in this regard we took the article with the most exhaustive level of adjustment for potential confounders.

    In the event that a study only presented RRs for a mortality outcome that was not our primary interest (eg, communicable disease or injury), reported RRs from coronary heart disease and all stroke, instead of total CVD, or reported cancer mortality, as avoidable, partly avoidable and non-avoidable, separately, we selected the suboutcome based on its highest quantitative contribution to the total, composite outcome.

    Statistical analysis

    The primary aim was to estimate pooled multiple-adjusted (taking the maximum adjustment set) RRs for mortality in the lowest SES group relative to the highest SES group studied. We used a random effects model (ie, allowing the true value from each study to vary) with inverse variance weighting. Our analyses were subdivided by pairing SES measures and mortality as: education–all-cause mortality, education–CVD, education–cancer; income–all-cause mortality, income–CVD, income–cancer; and occupation–all-cause mortality, occupation–CVD, occupation–cancer. Where sufficient data were available to give reliable results, additional analyses were conducted within subgroups. Subgroups analysed were mean age of study population (<65 vs ≥65 years), World Development Indicators10 (high- vs middle- and low-income countries), the Gini index of income inequality11 (0.2–0.29, 0.3–0.39, ≥0.4), geographical regions (East Asia vs South Asia vs South-East Asia), years of recruitment (before 1995 vs after 1995) and urbanicity (whether the majority of the study population lived in urban or rural areas). The Gini index is a number between 0 and 1, where 0 corresponds with perfect equality (where everyone has the same income) and 1 corresponds with perfect inequality (where one person has all the income—and everyone else has zero income). A priori, we excluded studies based only in the most elderly (75 years and over) when analysing the Gini index. Furthermore, to address the cohort effect,12 meta-regression was performed for the effects of year and duration of study on the associations between education–all-cause mortality and income–all-cause mortality.

    Statistical heterogeneity within studies was assessed using Cochran's Q test for heterogeneity and Higgins's I2 statistic.13 I2 is interpreted as the percentage of variability between studies due to heterogeneity, rather than chance. We undertook several additional sensitivity analyses to assess the robustness of our results and to explore potential sources of heterogeneity by (i) repeating the primary analysis after excluding studies that did not give total, composite outcomes, as described above; (ii) further excluding unadjusted and solely age–sex-adjusted estimates; (iii) further removing large cohort studies that might dominate the overall effects; (iv) analysing only studies that primarily focused on SES; and (v) comparing results from studies that split SES into only two groups and studies which used more than two groups, where we had compared the lowest with the highest groups.

    We tested for publication bias by inspecting funnel plots and performing Begg and Mazumdar rank correlation tests14 for each of the outcomes. Subgroup RRs from the same study were pooled before plotting onto funnel plots. A non-parametric ‘trim and fill’ method was performed if there was any asymmetry in a funnel plot.15 These meta-analyses were carried out using Stata, V.11.

    Results

    Study characteristics

    We identified 1956 article citations for screening of titles, abstracts and keywords (figure 1). From these, we retrieved 431 full articles for further consideration. We further identified 48 articles from reference lists. Overall, we identified 45 relevant publications based on 29 different cohort studies16–60 from which to extract data (see online supplementary table S1). Geographically, 20 cohorts were from East Asia, six were from South Asia and three cohorts were from South-East Asia (see online supplementary appendix 1). Demographic data of countries included in the analysis were provided in online supplementary appendix 2. Of the 45 publications, 29 aimed primarily to study the association between SES and mortality; the rest focused on other topics but still provided data on SES. Results were available from some of the 29 cohorts on more than one of the three SES measures of interest.

    Figure 1
    Figure 1

    Selection of publications.

    Three studies only provided results from a subset outcome rather than one of our three mortality outcomes of interest. These were studies from Vietnam58 (where we used their results on non-communicable diseases (NCDs) to represent total mortality when NCDs actually only accounted for 64% of all-cause mortality), Japan39 (where we used stroke to represent all CVD mortality—66% coverage) and Korea47 (where we used ‘avoidable’ cancers to represent all cancer mortality—76% coverage). Finally, data for 1 370 023 individuals were available for the primary analysis for all-cause mortality, 851 961 for CVD mortality and 884 474 for cancer mortality. There were 71 818 total deaths (5%), 10 790 CVD deaths (1.3%) and 7859 cancer death (0.9%). Mean duration of follow-up varied from 1 to 22 years.

    Pooled estimates of association

    Most studies demonstrated a significantly elevated risk of premature mortality in the lowest SES group relative to the highest (figure 2). The RRs for the association between low education and mortality varied from 0.99 to 4.62 across studies. Overall, those who had the lowest level of education experienced a 40% increased risk of total mortality compared with the highest level (95% CI 29% to 52%); the corresponding result for income was almost identical: 41% (24% to 61%). Being a manual or blue collar worker marginally increased risk of all-cause mortality compared with a non-manual or white collar worker, RR 1.14 (1.01 to 1.30), p value=0.04.

    Figure 2
    Figure 2

    Pooled relative risks and 95% CIs for all-cause mortality in those who had lower SES with higher SES serving as the reference (stratified by SES measure). SES, socioeconomic status; NCDs, non-communicable diseases; Adjustments: 1, age; 2, sex; 3, marital status; 4, smoking; 5, alcohol; 6, SES (ethnicity/education/wealth/income/financial capability index/household economic status/household size/household assets/type of home/number of household residents/occupation/job status/job type); 7, geographic (public health centre area/area of residence); 8, social network and support (number of meeting close relatives/having reliable friends/participation in activities/going to any religious services/enjoying good fellowship with neighbours/religion/mother tongue); 9, menopausal status; 10, physical activity (exercise/frequency of sports); 11, nutritious information (vegetable tendency/western meal tendency/energy intake/fruit intake/vegetable intake); 12, body anthropometry (BMI/height); 13, screening tests (SBP/cholesterol/glucose); 14, chronic diseases (including HTN/DM/physical/mental health); 15, others (CES-D/randomisation group/hearing impairment/doctor's visit).

    Level of education was associated with both CVD (figure 3) and cancer (figure 4) mortality. Pooled RRs comparing the extremes were 1.66 (1.23 to 2.25) and 1.16 (1.07 to 1.27) (both p value <0.001, respectively). Level of income was associated with an increased risk of cancer mortality (RR comparing the extremes 1.35 (1.08 to 1.68), p value <0.001) but not CVD mortality (RR 1.45 (0.88 to 2.36)). Occupation had no effect on either CVD or cancer mortality: the pooled RRs were 1.07 (0.86 to 1.33) and 0.99 (0.83 to 1.17), respectively. There was a high degree of heterogeneity (I2 statistics ranging from 74% to 92%) in the estimates between studies for all-cause and CVD mortality and, except when studying associations with occupation, for CVD mortality. There was no important heterogeneity when cancer mortality was the endpoint of interest.

    Figure 3
    Figure 3

    Pooled relative risks and 95% CIs for CVD mortality in those who had lower SES with higher SES serving as the reference (stratified by SES measure). SES, socioeconomic status; NCDs, non-communicable diseases; Adjustments: 1, age; 2, sex; 3, marital status; 4, smoking; 5, alcohol; 6, SES (ethnicity/education/wealth/income/financial capability index/household economic status/household size/household assets/type of home/number of household residents/occupation/job status/job type); 7, geographic (public health centre area/area of residence); 8, social network and support (number of meeting close relatives/having reliable friends/participation in activities/going to any religious services/enjoying good fellowship with neighbours/religion/mother tongue); 9, menopausal status; 10, physical activity (exercise/frequency of sports); 11, nutritious information (vegetable tendency/western meal tendency/energy intake/fruit intake/vegetable intake); 12, body anthropometry (BMI/height); 13, screening tests (SBP/cholesterol/glucose); 14, chronic diseases (including HTN/DM/physical/mental health); 15, others (CES-D/randomisation group/hearing impairment/doctor's visit).

    Figure 4
    Figure 4

    Pooled relative risks and 95% CIs for cancer mortality in those who had lower SES with higher SES serving as the reference (stratified by SES measure). SES, socioeconomic status; NCDs, non-communicable diseases; Adjustments: 1, age; 2, sex; 3, marital status; 4, smoking; 5, alcohol; 6, SES (ethnicity/education/wealth/income/financial capability index/household economic status/household size/household assets/type of home/number of household residents/occupation/job status/job type); 7, geographic (public health centre area/area of residence); 8, social network and support (number of meeting close relatives/having reliable friends/participation in activities/going to any religious services/enjoying good fellowship with neighbours/religion/mother tongue); 9, menopausal status; 10, physical activity (exercise/frequency of sports); 11, nutritious information (vegetable tendency/western meal tendency/energy intake/fruit intake/vegetable intake); 12, body anthropometry (BMI/height); 13, screening tests (SBP/cholesterol/glucose); 14, chronic diseases (including HTN/DM/physical/mental health); 15, others (CES-D/randomisation group/hearing impairment/doctor's visit).

    Additional analyses

    Six additional subgroup analyses according to study characteristics are provided in table 1. The impact of low education on mortality was modified by the mean age of the studied population. Studies in older adults gave a lower pooled estimate (RR 1.16; 1.13 to 1.20) than studies in the younger age group (RR 1.62; 1.44 to 1.81), p value for difference <0.0001. The mortality risk of having a more basic education was stronger in the middle- and low-income countries compared with high-income countries, p value <0.01. Studies in countries with a Gini index ≥0.4 found a higher pooled estimate (RR 1.76; 1.24 to 2.52) compared with studies in countries with a Gini index <0.3 (RR 1.23; 1.16 to 1.30). Comparing study regions, the effects of education on mortality were highest in South-East Asia (RR 2.72; 2.05 to 3.59), followed by South Asia (RR 1.47; 1.31 to 1.65) and East Asia (RR 1.25; 1.16 to 1.35), p value <0.001. Year of study and whether studies were based in urban or rural population did not modify the effect of education on mortality but the Gini index did (see online supplementary appendices 3 and 4). Differences by age were also apparent in studies that related income to the risk of all-cause mortality, such that an association of lower magnitude was seen in older groups (1.17; 0.96 to 1.44) in comparison with the younger (1.50; 1.30 to 1.74) (p value for difference <0.01). Study regions, urbanicity and country's income levels did not affect the association between income and mortality. However, studies conducted before 1995 showed a higher impact of income on mortality (RR 1.66; 1.28 to 2.15) than studies conducted since 1995 (RR 1.28; 1.14 to 1.44), p value <0.01. In meta-regression, neither year nor duration of study modified the relationship between education and mortality.

    View this table:
    Table 1

    Pooled relative risks for education and income in relation to all-cause mortality according to study or country characteristics

    Sensitivity analyses

    The association of education and income with all-cause mortality remained statistically significant at conventional levels after removing those studies which only reported unadjusted and age–sex-adjusted estimates and estimates from the study58 where we used deaths due to NCDs to substitute for total mortality: RR; 95% CI 1.36; 1.24 to 1.50 for education and 1.55; 1.30 to 1.86 for income. Omission of three large studies from Korea47 and India31 ,33 had little effect on the estimates for all-cause mortality (RR; 95% CI 1.37; 1.27 to 1.49 for education and 1.42; 1.17 to 1.72 for income). Analyses restricted to studies that focused on SES also gave a broadly similar set of results: the pooled RR for education and all-cause mortality was 1.38 (1.26 to 1.50), and for income and all-cause mortality was 1.44 (1.23 to 1.68). Only the association between occupation and all-cause mortality became non-significant in this context (1.09 (0.97 to 1.23), p value 0.14). In every other instance, similar RRs and their significances were observed. Analyses of studies that used binary data for education and income gave smaller RRs for the relationship with all-cause mortality (1.20 (1.12 to 1.37) and 1.28 (1.15 to 1.43), respectively, p value <0.0001) compared with studies that compared the lowest and highest SES groups (1.59 (1.41 to 1.81) and 1.49 (1.22 to 1.81) for education and income, respectively, p value <0.0001). Overall, none of these analyses removed the heterogeneity observed previously.

    Publication bias

    Begg's rank correlation test indicated no evidence of publication bias among each pair of SES measure and mortality outcome (p value >0.05). However, visual inspection of funnel plots (see online supplementary appendices 5–7) showed some asymmetry among studies between education and all-cause mortality. However, the non-parametric ‘trim and fill’ method for studies examining the relationship between education and all-cause mortality gave a pooled estimate of 1.28 (1.17 to 1.41, p value <0.0001).

    Discussion

    Although there are reasons to anticipate that the findings of research on SES and adult mortality in high-income settings may not necessarily be generalisable to other contexts, our meta-analysis of 30 associations from 29 prospective Asian cohorts provides evidence that the impact of low SES, in terms of education and income, on mortality can also be seen across Asia. However, within this general result, there was a suggestion that the effects of SES are strongest in South-East Asia and weakest in East Asia.

    Comparison with other studies

    Our finding of a strong inverse relationship between education and mortality accords with cohorts drawn from Western countries.61–70 High educational attainment improves health directly, but also indirectly through work, economic conditions, social-psychological resources and health behaviour. The similarity of direction of the problem of low education found in Asian population and the Westerners should suggest these effects of education on health as universal. The magnitude of the impact of education on mortality found in our study is slightly larger than the findings from three studies conducted in western European countries 61 ,62 ,71 and two large American cohorts,63 but smaller than the effect size found in four eastern European countries.64 The size of these associations seems to vary across regions, depending on socioeconomic backgrounds. For instance, the impact of education on mortality appears to be lower in western European compared with eastern European populations, northern European compared with southern European populations and less in American whites compared with African Americans.61 ,62 ,64–69 ,71

    It has been proposed that the level of economic inequality within a given population is a key determinant of magnitude of impact of SES on health.70 ,72–75 This is not surprising as a higher disparity will lead to a greater contrast between the lowest and highest groups of social status. Previous results from Europe,11 ,67 ,74 Latin America75 and the USA70 ,76 support this hypothesis. Our finding that the largest RRs for education occur within South-East Asia, and the lower RRs in South Asia and East Asia, respectively, also supports this hypothesis. This is because South-East Asian countries collectively have the highest disparity in income, as measured by the Gini index of income inequality,11 among the three subregions studied. The recent rapid modernisation in South-East Asia may be a factor in increasing social disparity. For example, Singapore's Gini index is the second largest among all developed economies11 and the substantial health disparity there has been commented upon.53 ,77

    We observed a reduced effect of education on mortality in the elderly. In general, western studies show that socioeconomic gradients in mortality persist in the elderly, but smaller on a relative scale when compared with younger populations.78 Education is determined early in adulthood and may not any more accurately describe the socioeconomic position of ageing people, where unmeasured community influences turn out to be highly important, particularly in the Asian society. Poor health behaviours observed in low education79–84 could also be another explanation in that premature deaths in the young are more prevalent in low- and middle-income countries than high-income countries.85

    The association between income and mortality observed in this meta-analysis shows a mixed pattern. Low income is related to all-cause and cancer mortality, but not CVD mortality. Western literature has shown a positive association between income and longevity. In Asia, the average income as measured by gross domestic product (GDP) per capita in the year 2000 is diverse, ranging from USD 347 in Bangladesh to USD 33 657 in Japan. It has been argued that it is an individual's relative, rather than absolute, income that is important for health, with a low relative income being a health hazard.86 ,87 A study in the USA found that the association between income and mortality is non-linear and the income–mortality gradient exists mainly at lower income levels.88 Similar to education, our analysis on income shows that the ranking of effects of income on mortality does not follow the country's GDP, as classified as high- or middle- to low-income countries, so much as the country's Gini index. Income is a crucial determinant factor for the affordability of treatment in fast growing Asian countries, where money is the key to access modern medicine. An example for the importance of affordability is a connection between income and mortality from chronic diseases, such as cancer and kidney disease. Similar differential effects with age were found for income as for education.

    The effect size of occupation on mortality, however, is weaker than education and income, and only reaches statistical significance for all-cause mortality. Historically, occupation has been a common measure in studies of social class inequalities in Europe. Much of the comparative research on all-cause and cause-specific mortality, comparing between manual and non-manual occupations, found that mortality is lower in high status occupations.89–91 Although there have been several classifications of occupation available, our study is limited to manual versus non-manual occupation and white collar versus blue collar, because these were the measures most often provided. It is quite unclear what occupational rank should be used in non-industrialised countries, where a higher proportion of the population works in the agricultural sector or is self-employed. This, perhaps, explains the lack of association for occupation found in this review.

    Strengths and limitations

    One of the strengths of this study is that we use only prospective data. This provides more reliable effects of SES on mortality and avoids the weaknesses of a cross-sectional design, including the migration effect that is frequently found in an early phase of epidemiological transition. The use of random effects methodology also provides more robust results than a fixed effect model, and may be justified, post hoc, by the high degree of heterogeneity observed in this meta-analysis. The various sensitivity analyses that showed broadly similar results, also suggested that our results are robust. Nevertheless, this study has several limitations. First, comparisons between SES classes are not uniform in both the number of levels of SES and the definition of SES. Half of the studies give binary data of SES, while the others use more than two levels of SES. As expected, given that it provides a more extreme comparison, the latter gave the larger estimate of the effect of SES on mortality. Second, most of the studies have already made an adjustment for SES. This might lessen the stand-alone effect of SES; however, this proves an independent effect of SES on mortality. Third, we observe a publication bias in studies on education and all-cause mortality. However, our test for publication bias by ‘trim and fill’ method gives a similar significant association. Fourth, education, income and occupation are only three of many SES measures used in studies from Asia. A selection of an appropriate SES for a society is a fundamental part for a cause-effect study, especially in Asian countries that have different social backgrounds such as the caste system in India. However, these three measurements are the core of SES and have been used in common in several studies in Asia. Fifth, some major countries are missing from this analysis, such as Indonesia, the world's fourth most populous country, and several South-East Asian countries. We noticed from our search that there are ongoing researches in these countries but results are not available yet. Last, only six cohort studies used nationally representative samples; the rest selected samples from specific areas (towns, cities or a workforce). Our study might not represent all Asian population but this is the most aggregated evidence that we have at the present.

    Mechanisms

    Explanations for how socioeconomic disadvantage might lead to CVD and other chronic diseases include access to resources (eg, education and income), environmental exposures (eg, housing conditions) and, most obviously, health-related behaviours (eg, tobacco smoking, physical activity and diet) and their physiological correlates (eg, obesity and raised levels of blood pressure and high cholesterol).6 ,92 ,93 That controlling for preventable behavioural and physiological risk factors attenuates but fails to eliminate socioeconomic gradients in health, particularly CVD,94 ,95 raises the possibility that as yet unmeasured psychological factors need to be considered (such as cognition, stress or personality type).96–100

    Conclusions

    In conclusion, socioeconomic inequality, as measured by education and income, has a negative impact on adult mortality in Asian population. The effect sizes lay between the impacts found in western European countries and eastern European countries. A higher disparity was observed in South-East Asia where there was a surge in modernisation during 1990s. Our result should strengthen the knowledge and understanding of the effect of SES on mortality in this part of the world.

    What is already known on this subject

    • Most of the evidence for socioeconomic inequalities in health comes from western societies; by contrast, there is a paucity of data on the impact of socioeconomic status on adult mortality in Asia.

    What this study adds

    • Socioeconomic inequality has a negative impact on adult mortality in Asian population.

    • The ranking of effects of income and education on mortality does not follow the country's gross national product so much as the country's inequality index.

    • The classic occupational rank might have a limited role in determining health outcomes in non-industrialised, agricultural societies.

    References

    1. ↵
      1. MacIntyre S
      . Social inequalities and health in the contemporary world: comparative overview. In: Strickland S, Shetty P, eds. Human biology and social inequality (39th symposium volume of the Society for the Study of Human Biology). Cambridge: Cambridge University Press, 1998:20–35.
    2. ↵
      Chief Medical Officer. Health in Scotland 2004. Edinburgh: Scottish Executive, 2005.
    3. ↵
      Department of Health. Choosing health; making healthy choices easier. London: The Stationery Office, 2004.
    4. ↵
      US Department of Health and Human Services. Healthy people 2010. Washington, DC: US Government Printing Office, 2000.
    5. ↵
      1. Marmot M
      . Social determinants of health inequalities. Lancet 2005;365:1099–104.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Wilkinson R,
      2. Marmot M
      . Social determinants of health: the solid facts. Copenhagen: World Health Organisation, 2003.
    7. ↵
      1. Elizabeth MH,
      2. Sonya R,
      3. Myoung OK,
      4. et al
      . The Asian population: 2010. United States Census Bureau, 2010.
    8. ↵
      1. Stroup DF,
      2. Berlin JA,
      3. Morton SC,
      4. et al
      . Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–12.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Moher D,
      2. Liberati A,
      3. Tetzlaff J,
      4. et al
      . Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535.
      OpenUrlFREE Full Text
    10. ↵
      World Development Indicators. World Bank, 2011:460.
    11. ↵
      Gini Index. World Development Indicators. World Bank, Development Research Group. 2011. http://data.worldbank.org/indicator/SI.POV.GINI (accessed Sep 2013).
    12. ↵
      1. Schaie K
      . Historical time and cohort effects. In: McCluskey KA, Reese HW, eds. Life-span development psychology: historical and generational effects. New York: Academic Press, 1984:1–15.
    13. ↵
      1. Higgins JP,
      2. Thompson SG,
      3. Deeks JJ,
      4. et al
      . Measuring inconsistency in meta-analyses. BMJ 2003;327:557–60.
      OpenUrlFREE Full Text
    14. ↵
      1. Begg CB,
      2. Mazumdar M
      . Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Steichen TJ
      . Meta-analysis in Stata: an updated collection from the stata journal. College Station, Texas: A Stata Press Publication, StataCorp LP, 2009:165–77.
    16. ↵
      1. Rahman OM
      . The effect of spouses on the mortality of older people in rural Bangladesh. Health Transition Rev 1997;7:1–12.
      OpenUrl
    17. ↵
      1. Mostafa G,
      2. Van Ginneken J
      . Trends in and determinants of mortality in the elderly population of Matlab, Bangladesh. Soc Sci Med 2000;50:763–71.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Hosegood V,
      2. Campbell OM
      . Body mass index, height, weight, arm circumference, and mortality in rural Bangladeshi women: a 19-y longitudinal study. Am J Clin Nutr 2003;77:341–7.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Hurt LS,
      2. Ronsmans C,
      3. Saha S
      . Effects of education and other socioeconomic factors on middle age mortality in rural Bangladesh. J Epidemiol Community Health 2004;58:315–20.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Pierce BL,
      2. Kalra T,
      3. Argos M,
      4. et al
      . A prospective study of body mass index and mortality in Bangladesh. Int J Epidemiol 2010;39:1037–45.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Yu ES,
      2. Kean YM,
      3. Slymen DJ,
      4. et al
      . Self-perceived health and 5-year mortality risks among the elderly in Shanghai, China. Am J Epidemiol 1998;147:880–90.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Liang J,
      2. McCarthy JF,
      3. Jain A,
      4. et al
      . Socioeconomic gradient in old age mortality in Wuhan, China. J Gerontol B Psychol Sci Soc Sci 2000;55:S222–33.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Nechuta S,
      2. Shu XO,
      3. Li HL,
      4. et al
      . Combined impact of lifestyle-related factors on total and cause-specific mortality among Chinese women: prospective cohort study. PLoS Med7 e1000339.
    24. ↵
      1. Yi Z,
      2. Gu D,
      3. Land KC
      . The association of childhood socioeconomic conditions with healthy longevity at the oldest-old ages in China. Demography 2007;44:497–518.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Zhu H,
      2. Xie Y
      . Socioeconomic differentials in mortality among the oldest old in China. Res Aging 2007;29:125–43.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Huang C,
      2. Elo IT
      . Mortality of the oldest old Chinese: the role of early-life nutritional status, socio-economic conditions, and sibling sex-composition. Popul Stud 2009;63:7–20.
      OpenUrlCrossRef
    27. ↵
      1. Wen M,
      2. Gu D
      . The effects of childhood, adult, and community socioeconomic conditions on health and mortality among older adults in China. Demography 2011;48:153–81.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Ferri CP,
      2. Acosta D,
      3. Guerra M,
      4. et al
      . Socioeconomic factors and all cause and cause-specific mortality among older people in Latin America, India, and China: a population-based cohort study. PLoS Med 2012;9:e1001179.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Ho SC
      . Health and social predictors of mortality in an elderly Chinese cohort. Am J Epidemiol 1991;133:907–21.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Woo J,
      2. Ho SC,
      3. Yu AL
      . The influence of income on morbidity, mortality and dependency in elderly Hong Kong Chinese. Arch Gerontol Geriatr 2000;30:55–61.
      OpenUrlCrossRefPubMed
    31. ↵
      1. Pednekar MS,
      2. Gupta R,
      3. Gupta PC
      . Illiteracy, low educational status, and cardiovascular mortality in India. BMC Public Health 2011;11:567.
      OpenUrlCrossRefPubMed
    32. ↵
      1. Sauvaget C,
      2. Ramadas K,
      3. Thomas G,
      4. et al
      . Body mass index, weight change and mortality risk in a prospective study in India. Int J Epidemiol 2008;37:990–1004.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Sauvaget C,
      2. Ramadas K,
      3. Fayette JM,
      4. et al
      . Socio-economic factors & longevity in a cohort of Kerala State, India. Indian J Med Res 2011;133:479–86.
      OpenUrlPubMed
    34. ↵
      1. Mohan VR,
      2. Muliyil J
      . Mortality patterns and the effect of socioeconomic factors on mortality in rural Tamil Nadu, south India: a community-based cohort study. TransR Soc Trop Med Hyg 2009;103:801–6.
      OpenUrlCrossRef
    35. ↵
      1. Jotheeswaran AT,
      2. Williams JD,
      3. Prince MJ
      . Predictors of mortality among elderly people living in a south Indian urban community; a 10/66 Dementia Research Group prospective population-based cohort study. BMC Public Health 2010;10:366.
      OpenUrlCrossRefPubMed
    36. ↵
      1. Sugisawa H,
      2. Liang J,
      3. Liu X
      . Social networks, social support, and mortality among older people in Japan. J Gerontol 1994;49:S3–13.
      OpenUrlAbstract
    37. ↵
      1. Liang J,
      2. Bennett J,
      3. Krause N,
      4. et al
      . Old age mortality in Japan: does the socioeconomic gradient interact with gender and age? J Gerontol Soc Sci 2002;57B:S294–307.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Fujino Y,
      2. Tamakoshi A,
      3. Iso H,
      4. et al
      . A nationwide cohort study of educational background and major causes of death among the elderly population in Japan. Prev Med 2005;40:444–51.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. Fujino Y
      . Occupational factors and mortality in the Japan Collaborative Cohort Study for Evaluation of Cancer (JACC). Asian Pac J Cancer Prev 2007;8(Suppl):97–104.
      OpenUrlPubMedWeb of Science
    40. ↵
      1. Ito S,
      2. Takachi R,
      3. Inoue M,
      4. et al
      . Education in relation to incidence of and mortality from cancer and cardiovascular disease in Japan. Eur J Public Health 2008;18:466–72.
      OpenUrlAbstract/FREE Full Text
    41. ↵
      1. Ishizaki T,
      2. Kai I,
      3. Imanaka Y
      . Self-rated health and social role as predictors for 6-year total mortality among a non-disabled older Japanese population. Arch Gerontol Geriatr 2006;42:91–9.
      OpenUrlCrossRefPubMedWeb of Science
    42. ↵
      1. Hirokawa K,
      2. Tsutusmi A,
      3. Kayaba K
      . Impacts of educational level and employment status on mortality for Japanese women and men: the Jichi Medical School cohort study. Eur J Epidemiol 2006;21:641–51.
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      1. Tsutsumi A,
      2. Kayaba K,
      3. Hirokawa K,
      4. et al
      . Psychosocial job characteristics and risk of mortality in a Japanese community-based working population: the Jichi Medical School Cohort Study. Soc Sci Med 2006;63:1276–88.
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      1. Iwasaki M,
      2. Otani T,
      3. Ohta A,
      4. et al
      . Rural-urban differences in sociodemographic, social network and lifestyle factors related to mortality of middle-aged Japanese men from the Komo-Ise cohort study. J Epidemiol 2002;12:93–104.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Aida J,
      2. Kondo K,
      3. Hirai H,
      4. et al
      . Assessing the association between all-cause mortality and multiple aspects of individual social capital among the older Japanese. BMC Public Health 2011;11:499.
      OpenUrlCrossRefPubMed
    46. ↵
      1. Song YM,
      2. Ferrer RL,
      3. Cho SI,
      4. et al
      . Socioeconomic status and cardiovascular disease among men: the Korean national health service prospective cohort study. Am J Public Health 2006;96:152–9.
      OpenUrlCrossRefPubMedWeb of Science
    47. ↵
      1. Song YM,
      2. Byeon JJ
      . Excess mortality from avoidable and non-avoidable causes in men of low socioeconomic status: a prospective study in Korea. J Epidemiol Community Health 2000;54:166–72.
      OpenUrlAbstract/FREE Full Text
    48. ↵
      1. Khang YH,
      2. Lynch JW,
      3. Jung-Choi K,
      4. et al
      . Explaining age-specific inequalities in mortality from all causes, cardiovascular disease and ischaemic heart disease among South Korean male public servants: relative and absolute perspectives. Heart 2008;94:75–82.
      OpenUrlAbstract/FREE Full Text
    49. ↵
      1. Khang YH,
      2. Kim HR
      . Relationship of education, occupation, and income with mortality in a representative longitudinal study of South Korea. Eur J Epidemiol 2005;20:217–20.
      OpenUrlCrossRefPubMedWeb of Science
    50. ↵
      1. Khang YH,
      2. Kim HR
      . [Socioeconomic mortality inequality in Korea: mortality follow-up of the 1998 National Health and Nutrition Examination Survey (NHANES) data]. J Prev Med Public Health 2006;39:115–22.
      OpenUrlPubMed
    51. ↵
      1. Khang YH
      . Relationship between childhood socio-economic position and mortality risk in adult males of the Korea Labour and Income Panel Study (KLIPS). Public Health 2006;120:724–31.
      OpenUrlCrossRefPubMedWeb of Science
    52. ↵
      1. Suh GH
      . Predictors of mortality in an aging community-based cohort in Korea. Psychogeriatrics 2006;6:10–8.
      OpenUrlCrossRef
    53. ↵
      1. Ma S,
      2. Cutter J,
      3. Tan CE,
      4. et al
      . Associations of diabetes mellitus and ethnicity with mortality in a multiethnic Asian population: data from the 1992 Singapore National Health Survey. Am J Epidemiol 2003;158:543–52.
      OpenUrlAbstract/FREE Full Text
    54. ↵
      1. Liu X,
      2. Hermalin AI,
      3. Chuang YL
      . The effect of education on mortality among older Taiwanese and its pathways. J Gerontol B Psychol Sci Soc Sci 1998;53:S71–82.
      OpenUrlPubMedWeb of Science
    55. ↵
      1. Lee Y,
      2. Lin RS,
      3. Sung FC,
      4. et al
      . Chin-Shan Community Cardiovascular Cohort in Taiwan-baseline data and five-year follow-up morbidity and mortality. J Clin Epidemiol 2000;53:838–46.
      OpenUrlCrossRefPubMedWeb of Science
    56. ↵
      1. Vathesatogkit P,
      2. Yamwong S,
      3. Woodward M,
      4. et al
      . Socioeconomic status and cardiovascular disease mortality in Thailand: The electricity generating authority of Thailand study. J Am Coll Cardiol 2010;55(Suppl 1):A57.E543.
      OpenUrl
    57. ↵
      1. Minh HV,
      2. Byass P,
      3. Wall S
      . Mortality from cardiovascular diseases in Bavi District, Vietnam. Scand J Public Health Suppl 2003;62:26–31.
      OpenUrlPubMed
    58. ↵
      1. Huong DL,
      2. Van Minh H,
      3. Janlert U,
      4. et al
      . Socio-economic status inequality and major causes of death in adults: a 5-year follow-up study in rural Vietnam. Public Health 2006;120:497–504.
      OpenUrlCrossRefPubMedWeb of Science
    59. ↵
      1. Minh HV,
      2. Huong DL,
      3. Wall S,
      4. et al
      . Cardiovascular disease mortality and its association with socioeconomic status: findings from a population-based cohort study in rural Vietnam, 1999–2003. Prev Chronic Dis 2006;3:1–11.
      OpenUrl
    60. ↵
      1. Granlund D,
      2. Chuc NT,
      3. Phuc HD,
      4. et al
      . Inequality in mortality in Vietnam during a period of rapid transition. Soc Sci Med 2010;70:232–9.
      OpenUrlCrossRefPubMed
    61. ↵
      1. Huisman M,
      2. Kunst AE,
      3. Bopp M,
      4. 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.
      OpenUrlCrossRefPubMedWeb of Science
    62. ↵
      1. Gallo V,
      2. Mackenbach JP,
      3. Ezzati M,
      4. et al
      . Social inequalities and mortality in Europe–results from a large multi-national cohort. PLoS ONE 2012;7:e39013.
      OpenUrlCrossRefPubMed
    63. ↵
      1. Steenland K,
      2. Henley J,
      3. 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.
      OpenUrlAbstract/FREE Full Text
    64. ↵
      1. Leinsalu M,
      2. Stirbu I,
      3. Vagero D,
      4. et al
      . Educational inequalities in mortality in four Eastern European countries: divergence in trends during the post-communist transition from 1990 to 2000. Int J Epidemiol 2009;38:512–25.
      OpenUrlAbstract/FREE Full Text
    65. ↵
      1. Mackenbach JP,
      2. Cavelaars AE,
      3. Kunst AE,
      4. et al
      . Socioeconomic inequalities in cardiovascular disease mortality; an international study. Eur Heart J 2000;21:1141–51.
      OpenUrlAbstract/FREE Full Text
    66. ↵
      1. Mackenbach JP,
      2. Kunst AE,
      3. Cavelaars AE,
      4. et al
      . Socioeconomic inequalities in morbidity and mortality in western Europe. The EU working group on socioeconomic inequalities in health. Lancet 1997;349:1655–9.
      OpenUrlCrossRefPubMedWeb of Science
    67. ↵
      1. Mackenbach JP,
      2. Stirbu I,
      3. Roskam AJ,
      4. et al
      . Socioeconomic inequalities in health in 22 European countries. N Engl J Med 2008;358:2468–81.
      OpenUrlCrossRefPubMedWeb of Science
    68. ↵
      1. Williams D,
      2. Chiquita C
      . US socioeconomic and racial differences in health: patterns and explanations. Annu Rev Sociol 1995;21:249–386.
      OpenUrl
    69. ↵
      1. Crimmins EM,
      2. Preston SH,
      3. Cohen B
      . International differences in mortality at older ages: dimensions and sources. National Research Council (US) Panel on Understanding Divergent Trends in Longevity in High-Income Countries The National Academies Press, 2011.
    70. ↵
      1. Lochner K,
      2. Pamuk E,
      3. Makuc D,
      4. et al
      . State-level income inequality and individual mortality risk: a prospective, multilevel study. Am J Public Health 2001;91:385–91.
      OpenUrlCrossRefPubMedWeb of Science
    71. ↵
      1. Huisman M,
      2. Kunst AE,
      3. Andersen O,
      4. et al
      . Socioeconomic inequalities in mortality among elderly people in 11 European populations. J Epidemiol Community Health 2004;58:468–75.
      OpenUrlAbstract/FREE Full Text
    72. ↵
      1. Babones SJ
      . Income inequality and population health: correlation and causality. Soc Sci Med 2008;66:1614–26.
      OpenUrlCrossRefPubMedWeb of Science
    73. ↵
      1. Karlsson M,
      2. Nilsson T,
      3. Lyttkens CH,
      4. et al
      . Income inequality and health: importance of a cross-country perspective. Soc Sci Med 2010;70:875–85.
      OpenUrlCrossRefPubMed
    74. ↵
      1. De Vogli R,
      2. Mistry R,
      3. Gnesotto R,
      4. et al
      . Has the relation between income inequality and life expectancy disappeared? Evidence from Italy and top industrialised countries. J Epidemiol Community Health 2005;59:158–62.
      OpenUrlAbstract/FREE Full Text
    75. ↵
      1. Biggs B,
      2. King L,
      3. Basu S,
      4. et al
      . Is wealthier always healthier? The impact of national income level, inequality, and poverty on public health in Latin America. Soc Sci Med 2010;71:266–73.
      OpenUrlCrossRefPubMed
    76. ↵
      1. Noss A
      . Household income for states: 2008–2009. US Census Bureau, US Department of Commerce, 2010.
    77. ↵
      1. Mak KH,
      2. Chia KS,
      3. Kark JD,
      4. et al
      . Ethnic differences in acute myocardial infarction in Singapore. Eur Heart J 2003;24:151–60.
      OpenUrlAbstract/FREE Full Text
    78. ↵
      1. Guilley E,
      2. Bopp M,
      3. Faeh D,
      4. et al
      . Socioeconomic gradients in mortality in the oldest old: a review. Arch Gerontol Geriatr 2010;51:e37–40.
      OpenUrlCrossRefPubMed
    79. ↵
      1. Siegrist J,
      2. Bernhardt R,
      3. Feng ZC,
      4. et al
      . Socioeconomic differences in cardiovascular risk factors in China. Int J Epidemiol 1990;19:905–10.
      OpenUrlAbstract/FREE Full Text
    80. ↵
      1. Kim KH,
      2. Shin HR,
      3. Nakama H
      . Health consciousness in relation to education in Korea--focusing on seven preventable health risk factors. Asia Pac J Public Health 1994;7:3–9.
      OpenUrlAbstract/FREE Full Text
    81. ↵
      1. Nishi N,
      2. Makino K,
      3. Fukuda H,
      4. et al
      . Effects of socioeconomic indicators on coronary risk factors, self-rated health and psychological well-being among urban Japanese civil servants. Soc Sci Med 2004;58:1159–70.
      OpenUrlCrossRefPubMedWeb of Science
    82. ↵
      1. Fukuda Y,
      2. Nakamura K,
      3. Takano T
      . Accumulation of health risk behaviours is associated with lower socioeconomic status and women's urban residence: a multilevel analysis in Japan. BMC Public Health 2005;5:53.
      OpenUrlCrossRefPubMed
    83. ↵
      1. Gupta R
      . Smoking, educational status and health inequity in India. Indian J Med Res 2006;124:15–22.
      OpenUrlPubMedWeb of Science
    84. ↵
      1. Gupta R,
      2. Deedwania PC,
      3. Sharma K,
      4. et al
      . Association of educational, occupational and socioeconomic status with cardiovascular risk factors in Asian Indians: a cross-sectional study. PloS one 2012;7:e44098.
      OpenUrlCrossRefPubMed
    85. ↵
      World Health Organization. Global status report on noncommunicable diseases 2010. WHO Library Cataloguing-in-Publication Data, 2011.
    86. ↵
      1. Wilkinson RG
      . Socioeconomic determinants of health. Health inequalities: relative or absolute material standards? BMJ 1997;314:591–5.
      OpenUrlFREE Full Text
    87. ↵
      1. Elstad JI,
      2. Dahl E,
      3. Hofoss D
      . Associations between relative income and mortality in Norway: a register-based study. Eur J Public Health 2006;16:640–4.
      OpenUrlAbstract/FREE Full Text
    88. ↵
      1. Rehkopf DH,
      2. Berkman LF,
      3. Coull B,
      4. et al
      . The non-linear risk of mortality by income level in a healthy population: US National Health and Nutrition Examination Survey mortality follow-up cohort, 1988–2001. BMC Public Health 2008;8:383.
      OpenUrlCrossRefPubMed
    89. ↵
      1. Kunst AE,
      2. Groenhof F,
      3. Mackenbach JP,
      4. et al
      . 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;316:1636–42.
      OpenUrlAbstract/FREE Full Text
    90. ↵
      1. Marmot MG,
      2. McDowall ME
      . Mortality decline and widening social inequalities. Lancet 1986;2:274–6.
      OpenUrlCrossRefPubMedWeb of Science
    91. ↵
      1. Marmot MG,
      2. Smith GD,
      3. Stansfeld S,
      4. et al
      . Health inequalities among British civil servants: the Whitehall II study. Lancet 1991;337:1387–93.
      OpenUrlCrossRefPubMedWeb of Science
    92. ↵
      1. Acheson D
      . Independent inquiry into inequalities in health: report. London: TSO, 1998.
    93. ↵
      1. Gordon D,
      2. Shaw M,
      3. Dorling D,
      4. et al
      . Inequalities in health. Bristol: The Policy Press, 1999.
    94. ↵
      1. Yarnell JW
      . The PRIME study: classical risk factors do not explain the severalfold differences in risk of coronary heart disease between France and Northern Ireland. Prospective Epidemiological Study of Myocardial Infarction. QJM 1998;91:667–76.
      OpenUrlAbstract/FREE Full Text
    95. ↵
      1. Marmot MG,
      2. Shipley MJ,
      3. Rose G
      . Inequalities in death–specific explanations of a general pattern? Lancet 1984;1:1003–6.
      OpenUrlCrossRefPubMedWeb of Science
    96. ↵
      1. Neisser U,
      2. Boodoo G,
      3. Bouchard TJ Jr.
      et al. Intelligence: knowns and unknowns. Am Psychol 1996;51:77–101.
      OpenUrlCrossRefWeb of Science
    97. ↵
      1. Hemmingsson T,
      2. Essen J,
      3. Melin B,
      4. et al
      . The association between cognitive ability measured at ages 18–20 and coronary heart disease in middle age among men: A prospective study using the Swedish 1969 conscription cohort. Soc Sci Med 2007;65:1410–19.
      OpenUrlCrossRefPubMedWeb of Science
    98. ↵
      1. Batty GD,
      2. Mortensen EL,
      3. Nybo Andersen AM,
      4. et al
      . Childhood intelligence in relation to adult coronary heart disease and stroke risk: evidence from a Danish birth cohort study. Paediatr Perinat Epidemiol 2005;19:452–9.
      OpenUrlCrossRefPubMedWeb of Science
    99. ↵
      1. Russ TC,
      2. Stamatakis E,
      3. Hamer M,
      4. et al
      . Association between psychological distress and mortality: individual participant pooled analysis of 10 prospective cohort studies. BMJ 2012;345:e4933.
      OpenUrlAbstract/FREE Full Text
    100. ↵
      1. Kivimaki M,
      2. Nyberg ST,
      3. Batty GD,
      4. et al
      . Job strain as a risk factor for coronary heart disease: a collaborative meta-analysis of individual participant data. Lancet 2012;380:1491–7.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Contributors PV took part in the literature search, data analyses and interpretation of data, wrote the manuscript and drafted the tables and figures. GDB did the literature search and critically revised the manuscript for important intellectual content. MW supervised PV did the statistical analyses and critically revised the manuscript, tables and figures. All authors read and approved the submitted version of the manuscript.

    • Funding Mahidol University, Bangkok, Thailand.

    • Competing interests PV receives a PhD scholarship grant from Mahidol University, Bangkok, Thailand. GDB was a Wellcome Trust Fellow during the preparation of this manuscript. MW has a research fellowship from the National Health and Medical Research Council of Australia.

    • Provenance and peer review Not commissioned; externally peer reviewed.