Subjective social status and mortality: the English Longitudinal Study of Ageing
European Journal of Epidemiology
Subjective social status and mortality: the English Longitudinal Study of Ageing
Panayotes Demakakos 0 1
Jane P. Biddulph 0 1
Cesar de Oliveira 0 1
Georgios Tsakos 0 1
Michael G. Marmot 0 1
0 & Panayotes Demakakos
1 Department of Epidemiology and Public Health, University College London , London , UK
Self-perceptions of own social position are potentially a key aspect of socioeconomic inequalities in health, but their association with mortality remains poorly understood. We examined whether subjective social status (SSS), a measure of the self-perceived element of social position, was associated with mortality and its role in the associations between objective socioeconomic position (SEP) measures and mortality. We used Cox regression to model the associations between SSS, objective SEP measures and mortality in a sample of 9972 people aged C 50 years from the English Longitudinal Study of Ageing over a 10-year follow-up (2002-2013). Our findings indicate that SSS was associated with all-cause, cardiovascular, cancer and other mortality. A unit decrease in the 10-point continuous SSS measure increased by 24 and 8% the mortality risk of people aged 50-64 and C 65 years, respectively, after adjustment for age, sex and marital status. The respective estimates for cardiovascular mortality were 36 and 11%. Adjustment for all covariates fully explained the association between SSS and cancer mortality, and partially the remaining associations. In people aged 50-64 years, SSS mediated to a varying extent the associations between objective SEP measures and all-cause mortality. In people aged C 65 years, SSS mediated to a lesser extent these associations, and to some extent was associated with mortality independent of objective SEP measures. Nevertheless, in both age groups, wealth partially explained the association between SSS and mortality. In conclusion, SSS is a strong predictor of mortality at older ages, but its role in socioeconomic inequalities in mortality appears to be complex.
Ageing; Inequalities; Mortality; Social status; Socioeconomic position
People?s position in the social hierarchy is strongly linked
to health in a graded way; the higher the position the better
the health. The resulting socioeconomic inequalities in
health, the social gradient in health, have been widely
]. The burden associated with socioeconomic
inequalities is immense as each year millions of deaths and
years of potential life lost across the world are attributed to
the unequal distribution of social and economic resources
and its individual, community and societal implications
]. Research has focused on explaining socioeconomic
inequalities in health and identifying causal pathways that
might constitute targets for prevention [
explanations have been put forward about what might
explain the graded association between socioeconomic
position (SEP) and risk of ill-health and death [
while empirical research has offered evidence on many
different mediating factors ranging from unhealthy
behaviours to health insurance and from control over life to
work stress [
1, 10, 19?21
Subjective social status (SSS), a concept that refers to
self-perceptions of one?s own social position, has received
less attention in epidemiological research and its role in
socioeconomic inequalities in health remains poorly
understood. This is despite its potential to add to the
current understanding of socioeconomic inequalities in health
when used in conjunction with conventional SEP measures.
SSS is a measure of SEP as it is perceived by the
individuals themselves; one?s personal translation of objective
SEP. Thus, it is a measure of SEP as experienced and
internalised by individuals and for that reason it is expected
to be closely related to health and a series of personal
attributes including behaviours, attitudes, values and
worldviews. Further, SSS captures personal individualised
aspects of one?s social identity and socioeconomic position
] such as lifetime achievement and recognition by
others, prestige and a successful family life that conventional
SEP measures do not [
]. For that reason its use in
epidemiological research can broaden our ability to
understand socioeconomic inequality beyond conventional SEP
measures. In addition, unlike commonly used measures
that tap into specific SEP dimensions, SSS is a summary
measure of SEP that is easy to measure and thus appealing
to survey designers.
Previous research has used SSS to predict various health
], but paradoxically SSS has only rarely
been used to predict mortality . At the moment it
remains unclear how strongly SSS is associated with
mortality and what is its role in the associations between
objective SEP measures and mortality. We aimed to cover
this gap in the literature by examining whether and how
SSS might be associated with mortality at older ages. To
provide a fuller picture of this association we examined
both all-cause and cause-specific mortality. Because
evidence suggests that SSS might partially mediate the
associations between objective SEP measures and different
health outcomes, we also explored whether SSS mediated
the associations between paternal occupational class when
respondents were 14 years old, education, occupational
class, income, wealth and mortality. The broad age range
of our sample, that is C 50 years, allowed for an
exploration of age differences in the association between SSS
and mortality that can substantially add to the limited
literature on socioeconomic inequalities in health in old age
The English Longitudinal Study of Ageing (ELSA) is a
prospective observational study of community-dwellers
aged C 50 years that was designed to be nationally
representative. At baseline, in 2002?2003, the ELSA sample
comprised 11,391 individuals who previously had
participated in the Health Survey for England. The Health Survey
for England is a national health examination survey, which
each year recruits a different nationally representative
sample using a stratified probability design. ELSA has been
approved by the National Research Ethics Service and
informed consent has been obtained by the participants.
More details about ELSA can be found at:
http://www.elsaproject.ac.uk/. Our analytical sample included 9972 ELSA
participants after the exclusion of 362 participants with
proxy or partial interviews, 464 participants without valid
mortality data (most of whom did not consent to link their
interview data with the mortality records), 335 participants
who did not respond to the SSS question and were assumed
to be missing not at random and 258 participants with
missing values in covariates (excluding BMI).
We used mortality data from the Office for National
Statistics that spanned a period of ten years, from the date
of the baseline interview in 2002?2003 to February 2013.
Deaths were classified according to International
Classification of Diseases (ICD) 10th Edition. Deaths with ICD10
codes C00 to C97 were classified as cancer deaths and
those with ICD10 codes I00 to I99 as cardiovascular
deaths. All remaining deaths were classified as other.
Subjective Social Status
We measured baseline SSS, one?s perceptions of own
social position, using a drawing of a ladder with 10 rungs
]. Participants were asked to place themselves on one of
the ten rungs after they were primed to think of the ladder
as a representation of society with the use of the following
vignette: ?Think of this ladder as representing where people
stand in our society. At the top of the ladder are the people
who are the best off?those who have the most money,
most education and best jobs. At the bottom are the people
who are the worst off?who have the least money, least
education, and the worst jobs or no jobs. The higher up you
are on this ladder, the closer you are to the people at the
very top and the lower you are, the closer you are to the
people at the very bottom. Please mark a cross on the rung
on the ladder where you would place yourself?.
Respondents who had put their mark in between two rungs were
assigned to the higher of these rungs.
We used the reversed ladder score as a continuous
measure with a value range from 1 to 10 with higher values
denoting lower SSS. The distribution of the non-reversed
ladder score by age along with descriptive statistics are
presented in the online Appendix (Figures S1 and S2 and
We measured the following indicators of objective SEP:
paternal/main carer?s occupational class when respondents
aged 14 years, education, occupational class, income and
wealth. Paternal/main carer?s occupational class at age
14 years was measured using a 4-category variable
(managerial and professional occupations including running
own business, intermediate occupations mostly services
workers, routine occupations such as plant workers
including a small number of unemployed and disabled,
other including those in the armed forces). Education was
measured using a 3-category educational attainment
variable (A-level or higher, O-level/secondary education, no
educational qualifications). Occupational class was
measured using the National Statistics Socio-economic
Classification (managerial and professional occupations,
intermediate occupations, semi-routine and routine
occupations, other including those who never worked). Tertiles
of equivalised weekly household income and total net
nonpension household wealth were used to measure income
and wealth, respectively. Age, sex, marital status,
unhealthy behaviours (smoking and physical activity including
participation in sports, leisure activities and household
chores), obesity (BMI categories), and elevated depressive
symptoms (defined using the cut point of C 4 symptoms on
the 8-item CES-D, which corresponds to the cut point of
C 16 on the full 20-item CES-D [
] that has been widely
used to identify possible cases of depression) were also
measured as covariates. All these covariates were measured
at baseline in 2002?2003, except for BMI, which was
measured at ELSA wave 2, in 2004?2005. BMI was also
the only adjustment variable for which we imputed missing
values (n = 945). We did that to avoid the unnecessary
exclusion of a large number of participants from the
We examined differences in SSS by the baseline
characteristics of the sample. We estimated Cox proportional
hazard regression models of the associations between SSS
and all-cause and cause-specific mortality. We checked and
confirmed that the proportional hazards assumption was
met using the Schoenfeld residuals test and log?log plots of
survival on a categorical SSS variable. Time-to-event (in
months) was calculated as the time that elapsed from the
date of the baseline interview in 2002?2003 to the date of
death or censoring (for consenting participants not known
to be dead by the end of the study, the censoring date was
February 2013). We first estimated the unadjusted models,
which we adjusted for age, sex, and marital status, then for
smoking, physical activity and BMI, and finally for
elevated depressive symptoms. We also estimated a series of
models of the association between SSS and all-cause
mortality that were initially adjusted for age, sex and
marital status and then for each of the objective SEP
measures. We assumed that SSS, a measure of
self-perceived SEP, is a good candidate mediator of the association
between objective SEP and mortality. We examined this
assumption using a common mediation approach that
concentrated on the change in the association of interest
after adjustment for the mediating variable. We estimated
models for each objective SEP measure and its association
with all-cause mortality, which were initially adjusted for
age, sex and marital status, and then for SSS. Because
interaction analyses indicated that the association between
SSS and mortality varied by age but not sex, we stratified
all analyses by age using the cut point of 65 years as
described elsewhere [
]. Despite excluding from analyses
participants who avoided to respond to the SSS question
(N = 335), there was still a number of participants
(N = 660) with missing SSS values, who did not complete
the whole pen-and-paper questionnaire that contained the
SSS question, not just the SSS question. We assumed that
SSS values for these participants were missing at random.
Based on this assumption and to minimise non-response
bias, we imputed missing SSS values for these participants
using chained equations in STATA 14. The imputation
model included the covariates included in the analysis
model, a retirement status variable, which was a strong
predictor of SSS in our data, the mortality variable and the
Nelson?Aalen estimate of the cumulative hazard to the
For comparison reasons and to ascertain that the
imputed data are comparable to the observed data, we performed
additional analyses where we estimated models using only
the observed data (Table S2 in the online supplement).
In both age groups, male, married, non-smokers, non-obese
and physically active participants as well as those who did
not report elevated depressive symptoms on average scored
higher on the SSS scale (Table 1). As expected, there were
strong positive associations between SSS and measures of
objective SEP. The wealth differences in SSS score were
the greatest observed in our data. In the younger age group,
there was difference of 2 points between participants in the
highest and lowest wealth tertile, while in the older age
group this difference was 1.7 points.
We observed 402 and 1861 deaths in the younger and
older age groups, respectively (Table 2). In the younger
age group, all-cause mortality risk increased by 24% per
unit increase in the SSS score after adjustment for age, sex,
and marital status, while in the older age group, this
increase was smaller at 8%. SSS appeared to be associated
more strongly with CVD-related and other mortality than
with cancer-related mortality. As in all-cause mortality,
these associations were stronger in the younger age group
compared with older age group. Adjustments for unhealthy
CI confidence interval, SSS subjective social status
aUnless stated otherwise
bTo facilitate understanding, SSS has not been reversed in this table. Higher values denote higher SSS
cP values were calculated using the observed (non-imputed) data and the analysis of variance test
dThe observed (non-imputed) BMI data were used. The ??Missing?? category was not used in the calculation of the P value
eThe ??Other?? category was not used in the calculation of the P value
behaviours, BMI and elevated depressive symptoms fully
explained the association between SSS and other (in
participants aged C 65 years) and cancer mortality and
partially the associations between SSS and all-cause, CVD and
other mortality (in those aged 50?64 years).
In both age groups, the association between SSS and
allcause mortality was little affected by adjustment for most
objective SEP measures, except for the adjustment for
wealth, which explained a considerable part of it (Table 3).
The associations between measures of objective SEP and
all-cause mortality were partially explained, to a varying
extent, after adjustment for SSS (Table 4). In the younger
age group, SSS explained a large part of the associations
between education and adult occupational class and
allcause mortality, and a smaller part of the associations
between childhood occupational class, income and wealth
and all-cause mortality. In the older age group, SSS
explained a smaller part of these associations.
In a national sample of people aged C 50 years, we found
subjective social status, one?s perceptions of their own
social status, to be inversely associated with all-cause and
cause-specific mortality. These associations were stronger
in participants aged 50?64 years compared with those
aged C 65 years and were explained to a varying extent by
unhealthy behaviours, obesity and elevated depressive
symptoms. SSS partially mediated the associations
between objective SEP measures such as education and
occupational class and mortality, especially in participants
aged 50?64 years. SSS appears to explain a unique part of
mortality that no single objective SEP measure could
explain. Nevertheless, in both age groups, wealth partially
explained the association between SSS and mortality; a
strong indication that the association between SSS and
mortality can partially be attributed to SSS reflecting one?s
wealth and being a product of assets ownership and
Despite the importance of SSS to better understand
socioeconomic inequalities in health and an expanding
literature on its associations with morbidity [
very little research has focused on the association between
SSS and mortality. We are aware of only one
individuallevel study on the association between SSS and mortality
. Their findings partially concur with ours; they
examined separately men and women aged 40?65 years
and found SSS to predict mortality over 3.5 years of
follow-up in men, but not in women. Other studies have
explored the associations between self-perceptions of
specific dimensions of SEP such as self-perceived income
and wealth [
], relative deprivation [
occupational prestige [
], and perceptions about own work
] and all-cause mortality. Notwithstanding
methodological differences, our findings concur with those
of most previous studies [
Our study has strengths and limitations that need to be
acknowledged. The use of data from a survey that is
designed to be nationally representative is a strength and
makes our findings more generalizable to
communitydwellers aged C 50 years. The novelty of our findings
should also be stressed. Our study is the first to examine the
association between SSS and mortality in people aged
C 65 years and the first to examine the association between
SSS and cause-specific mortality. It is also the first
systematic attempt to explore the interrelationships between
SSS and commonly used objective SEP measures in
relation to mortality. Finally, the comprehensive assessment of
SEP and the 10-year long follow-up make our study a
thorough investigation of the association between SSS,
SEP and mortality. A weakness of our study is our inability
CI confidence interval, HR hazard ratio
Model 1 represents the unadjusted association
Model 2 is adjusted for age, sex, and marital status
Model 3 is adjusted for age, sex, marital status, smoking, physical activity, and BMI
Model 4 is adjusted for age, sex, marital status, smoking, physical activity, BMI and elevated depressive
Hazard ratios denote hazard change per unit decrease in SSS
to fully control for non-response bias. We were able to
impute missing at random SSS values and link almost all
participant data with mortality records, but our sample
remained to some extent selected as at baseline it included
community-dwellers who have survived at least to age
C 50 years. Further, the baseline household response rate
was very good at 70%, but nevertheless left some room for
non-response bias. Another weakness of our study is its
purely exploratory character. Our work neither proposed
nor tested any theoretical model of the associations
between objective SEP measures, SSS, and mortality.
However, it generated basic evidence about these
associations, which can then be used to build a well-defined
testable model of socioeconomic inequalities in mortality.
The mediation analysis presented in Table 4 is based on the
conceptual argument that SSS is most likely a product of
objective SEP and thus a good candidate mediator of the
associations between each one of the objective SEP
measures and all-cause mortality. Our approach was simple and
based on a three-variable system with a single mediator,
CI confidence interval, HR hazard ratio
Model 1 is adjusted for age, sex, and marital status
Model 2 is adjusted for age, sex, marital status and education
Model 3 is adjusted for age, sex, marital status and occupational class
Model 4 is adjusted for age, sex, marital status and paternal/carer?s occupational class when respondent was
14 years old
Model 5 is adjusted for age, sex, marital status and equivalised weekly household income tertiles
Model 6 is adjusted for age, sex, marital status and total net non-pension household wealth tertiles
Hazard ratios denote hazard change per unit decrease in SSS
which is expected to be associated with both the predictor
and the outcome and explain to a varying extent the effect
of the predictor on the outcome [
]. This approach neither
allows a simultaneous examination of direct and indirect
effects nor fully accounts for confounding [
Our findings indicate that SEP has a substantive
subjective dimension that is strongly related to all-cause
mortality in three different ways. First, SSS mediates to a
varying degree the associations between objective SEP
measures and mortality. Second, SSS to some extent
appears to be an independent predictor of mortality,
possibly as a measure of facets of social position not captured
by objective SEP measures. Third, SSS is partially
associated with mortality as a product of wealth and material
In people aged 50?64 years, SSS explained to a
considerable extent the associations between objective SEP
measures and mortality. On the basis that objective SEP is
expected to shape people?s perceptions of their standing on
the societal hierarchy and influence their social identity,
our findings likely suggest that self-perceptions of own
social status as captured by SSS is an important channel
through which objective SEP exerts a considerable part of
its effect on mortality. In people aged 50?64 years, SSS
appears to be explaining to a greater extent the associations
between education and adult social class and mortality. We
can only speculate that this might happen because social
comparisons among working age people are typically made
on the basis of education and adult occupational class and
thus these two SEP measures might be more important for
the formation of perceptions of own social status, that is
SSS, than other SEP measures in this age group. Further,
education and childhood and adult social classes are in a
sense historic SEP markers and thus expected to exert their
impact on mortality mostly indirectly via more
contemporary SEP measures such as SSS, income and wealth.
In people aged C 65 years, SSS continues to be a
significant predictor of mortality. Nevertheless, the
importance of SSS as a mediator of the associations between SEP
measures and mortality is somewhat decreased. This
change in the role of SSS in socioeconomic inequalities in
mortality likely can be attributed to its dynamic and
agedependant character. Past the age of 65 years, where most
people are retirees and no longer financially active, SSS
might be less about education and adult occupational class
and more about more dimensions of social position that are
perhaps more meaningful in this age group such as lifetime
achievement, successfulness in family life, prestige and
recognition within one?s local community. These more
individualised dimensions of SEP can also be important for
survival in old age because of their connection with the
provision of key resources such as emotional support, care
and practical help.
The observed age differences in the association between
SSS and mortality are expected. It is known that the effect
of most risk factors on mortality decreases with age
Age group: C 65 years
No of deaths
No of participants
Model 1 HR (95% CI)
Model 2 HR (95% CI)
Predictor: Equivalised weekly household income tertiles
No of deaths 115
No of participants 2453
Model 1 HR (95% CI) 1.00 (reference)
Model 2 HR (95% CI) 1.00 (reference)
Predictor: Total net non-pension household wealth tertiles
No of deaths 92
No of participants 1981
Model 1 HR (95% CI) 1.00 (reference)
Model 2 HR (95% CI) 1.00 (reference)
CI confidence interval, HR hazard ratio
Model 1 is adjusted for age, sex, and marital status
Model 2 is adjusted for age, sex, marital status and subjective social status
Hazard ratios denote the differences in the hazards between the reference category and other categories of the predictor variable
aFor clarity reasons, the HR for the small ??Other/Never worked?? category are not presented
bFor clarity reasons, the HR for the small ??Other?? category are not presented
partially as a result of survivor bias. Nevertheless, the
public health importance of SSS inequalities in people
aged C 65 years should not be underestimated. Most
deaths occur past the age of 65 years and that means that
even small differences in the relative risk of mortality
according to SSS in this age group correspond to great
differences in the number of deaths.
Regarding specific causes of death, in accordance with
previous evidence suggesting a inverse association between
objective SEP measures and CVD [
], we found that SSS is
strongly associated with CVD-related mortality in our
participants. The strength and persistence of this
association underline the importance of the subjective dimension
of SEP for cardiovascular mortality. The same applies to
the association between SSS and other mortality in
participants aged 50?64 years, which is indicative of a strong
association between the subjective aspects of SEP and
death from respiratory and other causes including suicide
and accidents. The association between SSS and
cancerrelated mortality was strong, especially among participants
50?64 years, but fully explained after adjustment for
unhealthy behaviours and obesity.
Conclusions and public health implications
In summary, our study provides substantial evidence for an
inverse association between SSS and mortality. SSS
appears to partially mediate the associations between
objective SEP measures such as education and
occupational class and mortality?especially in people aged
50?64 years. To some extent SSS appears to be associated
with mortality independent of objective SEP measures
likely because it captures facets of socioeconomic position
that no objective SEP measure does. Nevertheless, our
findings suggest that SSS is partially associated with
mortality as a product of wealth.
The implications of our work for public health are
considerable. Our findings contribute to a better
understanding of socioeconomic inequalities in health and
expand the knowledge basis for prevention strategies
aiming to reduce socioeconomic inequalities in health. It is
important to know that feelings of disadvantage and low
social status may lead to increased mortality on the top of
the pernicious effect of material disadvantage. This
knowledge can be used to fine-tune prevention strategies so
that they include empowerment as an additional target next
to the main ones of alleviation of material disadvantage
and reduction of socioeconomic inequalities in health. Our
findings also highlight the existence of important
socioeconomic inequalities in people aged C 65, which need to
be targeted by prevention strategies, and point out the need
to take into account age differences when designing
prevention strategies to tackle socioeconomic inequalities in
health in adult population.
Acknowledgements The English Longitudinal Study of Ageing is
supported by the National Institute on Aging (Grants 2RO1AG7644
and 2RO1AG017644-01A1) and a consortium of the United Kingdom
government departments (Department for Education and Skills;
Department for Environment, Food, and Rural Affairs; Department of
Health; Department of Trade and Industry; Department for Work and
Pensions; the HM Treasury Inland Revenue; the Office of the Deputy
Prime Minister; and the Office for National Statistics) coordinated by
the Economic and Social Research Council (ESRC). The National
Institute on Aging and the consortium of the United Kingdom
government departments had no role in the design and conduct of this
study; collection, management, analysis, and interpretation of the
data; and preparation, review or approval of the manuscript.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
Ethical approval All procedures performed in studies involving
human participants were in accordance with the ethical standards of
the institutional and/or national research committee and with the 1964
Helsinki declaration and its later amendments or comparable ethical
Informed consent Informed consent was obtained from all individual
participants included in the study.
Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://creative
commons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source, provide a
link to the Creative Commons license, and indicate if changes were
1. Demakakos P , Biddulph JP , Bobak M , Marmot MG . Wealth and mortality at older ages: a prospective cohort study . J Epidemiol Community Health . 2016 ; 70 ( 4 ): 346 - 53 . https://doi.org/10.1136/ jech-2015-206173.
2. Marmot MG , Shipley MJ . Do socioeconomic differences in mortality persist after retirement? 25 year follow up of civil servants from the first Whitehall study . Br Med J. 1996 ; 313 ( 7066 ): 1177 - 80 . https://doi.org/10.1136/bmj.313.7066. 1177.
3. Huisman M , Kunst AE , 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 ( 9458 ): 493 - 500 . https://doi.org/10.1016/S0140- 6736 ( 05 ) 17867 - 2 .
4. Mackenbach JP , Stirbu I , Roskam A -JR, et al. Socioeconomic inequalities in health in 22 European countries . N Engl J Med . 2008 ; 358 ( 23 ): 2468 - 81 . https://doi.org/10.1056/NEJMsa0707519.
5. Minkler M , Fuller-Thomson E , Guralnik JM . Gradient of disability across the socioeconomic spectrum in the United States . N Engl J Med . 2006 ; 355 ( 7 ): 695 - 703 . https://doi.org/10.1056/ NEJMsa044316.
6. Kaplan GA , Keil JE . Socioeconomic factors and cardiovascular disease: a review of the literature . Circulation . 1993 ; 88 ( 4 Pt 1 ): 1973 - 98 .
7. Dalstra JAA , Kunst AE , Borrell C , et al. Socioeconomic differences in the prevalence of common chronic diseases: an overview of eight European countries . Int J Epidemiol . 2005 ; 34 ( 2 ): 316 - 26 . https://doi.org/10.1093/ije/dyh386.
8. Ellis L , Coleman MP , Rachet B . How many deaths would be avoidable if socioeconomic inequalities in cancer survival in England were eliminated? A national population-based study, 1996 - 2006 . Eur J Cancer . 2012 ; 48 ( 2 ): 270 - 8 . https://doi.org/10. 1016/j.ejca. 2011 . 10 .008.
9. Mackenbach JP , Meerding WJ , Kunst AE . Economic costs of health inequalities in the European Union . J Epidemiol Community Health . 2011 ; 65 ( 5 ): 412 - 9 . https://doi.org/10.1136/jech. 2010 . 112680 .
10. Marmot M , Allen J , Goldblatt P , Boyce T . Fair society, healthy lives. The Marmot Review . Strategic review of health inequalities in England post- 2010 . 2010 . http://eprints.ucl.ac.uk/111743/. Accessed 26 Nov 2014 .
11. Marmot MG , Shipley MJ , Rose G. Inequalities in death-specific explanations of a general pattern? Lancet . 1984 ; 323 ( 8384 ): 1003 - 6 . https://doi.org/10.1016/S0140- 6736 ( 84 ) 92337 - 7 .
12. Marmot MG . Understanding social inequalities in health . Perspect Biol Med . 2003 ; 46 ( 3 Suppl) : S9 - 23 .
13. Adler NE , Boyce T , Chesney MA , et al. Socioeconomic status and health. The challenge of the gradient . Am Psychol . 1994 ; 49 ( 1 ): 15 - 24 .
14. Link BG , Phelan J . Social conditions as fundamental causes of disease . J Health Soc Behav . 1995 ;Spec No: 80 - 94 . http://www. ncbi.nlm.nih.gov/pubmed/7560851. Accessed 7 June 2017 .
15. Mackenbach JP . The persistence of health inequalities in modern welfare states: The explanation of a paradox . Soc Sci Med . 2012 ; 75 ( 4 ): 761 - 9 . https://doi.org/10.1016/j.socscimed. 2012 . 02 . 031.
16. Siegrist J , Marmot M. Health inequalities and the psychosocial environment-two scientific challenges . Soc Sci Med . 2004 ; 58 ( 8 ): 1463 - 73 . https://doi.org/10.1016/S0277- 9536 ( 03 ) 00349 - 6 .
17. Gottfredson LS . Intelligence: is it the epidemiologists' elusive ''fundamental cause'' of social class inequalities in health? J Pers Soc Psychol . 2004 ; 86 ( 1 ): 174 - 99 . https://doi.org/10.1037/ 0022 - 3514 . 86 .1.174.
18. Elovainio M , Ferrie JE , Singh-Manoux A , et al. Socioeconomic differences in cardiometabolic factors: social causation or healthrelated selection? Evidence from the Whitehall II Cohort Study, 1991 - 2004 . Am J Epidemiol. 2011 ; 174 ( 7 ): 779 - 89 . https://doi. org/10.1093/aje/kwr149.
19. van Oort FVA , van Lenthe FJ , Mackenbach JP . Material, psychosocial, and behavioural factors in the explanation of educational inequalities in mortality in The Netherlands . J Epidemiol Community Health . 2005 ; 59 ( 3 ): 214 - 20 . https://doi.org/10.1136/ jech. 2003 . 016493 .
20. Whitehead M , Pennington A , Orton L , et al. How could differences in ''control over destiny'' lead to socio-economic inequalities in health? A synthesis of theories and pathways in the living environment . Health Place . 2016 ; 39 : 51 - 61 . https://doi.org/ 10.1016/j.healthplace. 2016 . 02 .002.
21. Matthews KA , Gallo LC , Taylor SE . Are psychosocial factors mediators of socioeconomic status and health connections ? Ann N Y Acad Sci . 2010 ; 1186 ( 1 ): 146 - 73 . https://doi.org/10.1111/j. 1749- 6632 . 2009 . 05332 .x.
22. Eidlin B . Class formation and class identity: birth, death, and possibilities for renewal . Sociol Compass . 2014 ; 8 ( 8 ): 1045 - 62 . https://doi.org/10.1111/soc4.12197/full.
23. Demakakos P , Nazroo J , Breeze E , Marmot M. Socioeconomic status and health: the role of subjective social status . Soc Sci Med . 2008 ; 67 ( 2 ): 330 - 40 . https://doi.org/10.1016/j.socscimed. 2008 . 03 .038.
24. Tang KL , Rashid R , Godley J , Ghali WA . Association between subjective social status and cardiovascular disease and cardiovascular risk factors: a systematic review and meta-analysis . BMJ Open . 2016 ; 6 ( 3 ):e010137. https://doi.org/10.1136/bmjopen2015- 010137 .
25. Euteneuer F. Subjective social status and health . Curr Opin Psychiatry . 2014 ; 27 ( 5 ): 337 - 43 . https://doi.org/10.1097/YCO. 0000000000000083.
26. Quon EC , McGrath JJ . Subjective socioeconomic status and adolescent health: a meta-analysis . Health Psychol . 2014 ; 33 ( 5 ): 433 - 47 . https://doi.org/10.1037/a0033716.
27. Kopp MS , Skrabski A , La?szlo? KD, Janszky I. Gender patterns of socioeconomic differences in premature mortality: follow-up of the Hungarian Epidemiological Panel . Int J Behav Med . 2011 ; 18 ( 1 ): 22 - 34 . https://doi.org/10.1007/s12529-010-9126-5.
28. Huisman M , Read S , Towriss CA , Deeg DJH , Grundy E. Socioeconomic inequalities in mortality rates in old age in the World Health Organization Europe Region . Epidemiol Rev . 2013 ; 35 ( 1 ): 84 - 97 . https://doi.org/10.1093/epirev/mxs010.
29. Adler NE , Epel ES , Castellazzo G , Ickovics JR . Relationship of subjective and objective social status with psychological and physiological functioning: preliminary data in healthy white women . Heal Psychol . 2000 ; 19 ( 6 ): 586 - 92 .
30. Demakakos P , Pierce MB , Hardy R . Depressive symptoms and risk of type 2 diabetes in a national sample of middle-aged and older adults: the English longitudinal study of aging . Diabetes Care . 2010 ; 33 ( 4 ): 792 - 7 . https://doi.org/10.2337/dc09- 1663 .
31. White IR , Royston P . Imputing missing covariate values for the Cox model . Stat Med . 2009 ; 28 ( 15 ): 1982 - 98 . https://doi.org/10. 1002/sim.3618.
32. Pra?g P, Mills MC , Wittek R . Subjective socioeconomic status and health in cross-national comparison . Soc Sci Med . 2016 ; 149 : 84 - 92 . https://doi.org/10.1016/j.socscimed. 2015 . 11 . 044.
33. Muennig P , Kuebler M , Kim J , Todorovic D , Rosen Z. Gender differences in material, psychological, and social domains of the income gradient in mortality: implications for policy . PLoS ONE . 2013 ; 8 ( 3 ):e59191. https://doi.org/10.1371/journal.pone. 0059191 .
34. Bessudnov A , McKee M , Stuckler D. Inequalities in male mortality by occupational class, perceived status and education in Russia, 1994 - 2006 . Eur J Public Health . 2012 ; 22 ( 3 ): 332 - 7 . https://doi.org/10.1093/eurpub/ckr130.
35. Kondo N , Saito M , Hikichi H , et al. Relative deprivation in income and mortality by leading causes among older Japanese men and women: AGES cohort study . J Epidemiol Community Health . 2015 ; 69 ( 7 ): 680 - 5 . https://doi.org/10.1136/jech-2014- 205103.
36. A? berg-Yngwe M , Kondo N , Ha?gg S , Kawachi I. Relative deprivation and mortality-a longitudinal study in a Swedish population of 4,7 million, 1990 - 2006 . BMC Public Health. 2012 ; 12 : 664 . https://doi.org/10.1186/ 1471 -2458-12-664.
37. Christ SL , Fleming LE , Lee DJ , Muntaner C , Muennig PA , Caban-Martinez AJ . The effects of a psychosocial dimension of socioeconomic position on survival: occupational prestige and mortality among US working adults . Sociol Health Illn . 2012 ; 34 ( 7 ): 1103 - 17 . https://doi.org/10.1111/j.1467- 9566 . 2012 . 01456 .x.
38. Shippee TP , Rinaldo L , Ferraro KF . Mortality risk among Black and White working women: the role of perceived work trajectories . J Aging Health . 2012 ; 24 ( 1 ): 141 - 67 . https://doi.org/10. 1177/0898264311422743.
39. Preacher KJ . Advances in mediation analysis: a survey and synthesis of new developments . Annu Rev Psychol . 2015 ; 66 ( 1 ): 825 - 52 . https://doi.org/10.1146/annurev-psych010814- 015258 .
40. Richiardi L , Bellocco R , Zugna D . Mediation analysis in epidemiology: methods, interpretation and bias . Int J Epidemiol . 2013 ; 42 ( 5 ): 1511 - 9 . https://doi.org/10.1093/ije/dyt127.