You are on page 1of 11

Stone, G. C., Cohen, F., Adler, N. E., & Associates (1979).

Health psychologyA handbook: Theories, applications,


and challenges of a psychological approach to the health
care system. San Francisco: JosseyBass.

Health Disparities Through a Psychological


Lens
Nancy E. Adler
University of California, San Francisco

There is growing concern in the United States about


avoidable, unjust differences in health associated with
sociodemographic characteristics, such as socioeconomic
status and race/ethnicity. This concern has sparked
research to identify how disparities develop and how they
can be reduced. Studies showing that disparities occur at
all levels of socioeconomic status, not simply at the very
bottom, suggest that psychosocial factors play an important
role. The author discusses both content and process issues
in psychological research on disparities.
Keywords: health disparities, socioeconomic status, stress,
behavior, subjective status

If one asks the average person why some people are


healthier than others, common explanations are family history and genetics, lack of health care, and/or lifestyle.
There has been growing awareness, however, that health is
also influenced by ones social position. Differences in
health and longevity are not just between individuals but
also between groups of people defined by social and demographic characteristics. A recent U.S. analysis showed a
gap of over 35 years between the groups with the shortest
and longest life expectancies at birth (Murray et al., 2006).
This difference represents 90% of the gap in life expectancy between the shortest- and longest-living groups across
the globe. Concern about such health disparities has galva-

Editors Note
Nancy E. Adler received the Award for Distinguished Scientific Applications of Psychology. Award winners are invited to deliver an award address at the APAs annual
convention. A version of this award address was delivered
at the 117th annual meeting, held August 6 9, 2009, in
Toronto, Ontario, Canada. Articles based on award addresses are reviewed, but they differ from unsolicited articles in that they are expressions of the winners reflections
on their work and their views of the field.
November 2009 American Psychologist

nized calls for action. For example, in the Healthy People


2010 initiative, the Department of Health and Human Services set two overarching health goals for the nation for
the first decade of the 21st century. One was to improve
the health of the U.S. population, and the second was to
eliminate disparities in health. The aims informed a large
set of specific objectives with measurable outcomes. Distressingly, a recent review of progress shows little movement toward achieving the goal of eliminating health disparities (see the Healthy People 2010 Midcourse Review,
n.d.). Of 195 objectives that had information on race/
ethnicity, 14 showed increased disparities, whereas 24
decreased. For objectives on which there were data regarding education, 3 showed decreases in disparities, but
l4 showed increases. Given this discouraging assessment, health disparities remain an issue in setting goals
for the next decade.
Interest in health disparities has been fueled by research
documenting their existence and magnitude. In turn, concern about these disparities has fostered further research on
their nature and causes, leading to geometric growth of
work in this area (Adler & Ostrove, 1999). Although earlier health research included sociodemographic variables,
such as education or race/ethnicity, these were primarily
used as controls. The focus on health disparities made sociodemographic variables a focus of study in their own
right (Adler & Stewart, in press).
My own interest in health disparities was the result of
one-trial learning at a meeting that arose out of the
MacArthur Foundation Network on Health-Promoting and
Disease-Preventing Behavior. At the time, my research was
testing how to expand rational models of decision making
to account for seemingly irrational health-risking behaviors.
I was fortunate to participate in the network, which put this
work in a broader context. One issue discussed by the network was how to conceptualize and measure an individuals overall state of health and whether there are common
pathways to multiple diseases. I was charged with convening a meeting to explore the feasibility of developing a
biobehavioral battery of measures that would predict an
individuals overall health status. Along with several psychologists, I invited a social epidemiologist, Leonard Syme
to this meeting. After asking why there were so many psychologists in attendance, Len declared that we could have a
very brief meeting because such a battery could be composed of one question. That item would ask about a persons social class. He showed us recently published data
from the Whitehall study of British civil servants. The
study assessed over 10 years the health and longevity of
civil servants across the range of occupational grades. The
key finding was that at each successive drop in occupational grade level, mortality rates increased. It was not simply that the lowest ranked workers had higher mortality
663

than the top-ranked workers, but that mortality increased at


each successive step down in occupational grade.
These findings challenged basic assumptions about the
determinants of mortality, especially in the United States.
Prevailing explanationspoverty and lack of health care
clearly could not account for the Whitehall findings. The
increased risk of mortality among those of relatively lower
rank occurred in a working population, all of whom had
access to health care and none of whom was in poverty.
Most striking was the significant difference in mortality
between high-level civil servants who were well-paid professionals and those one level above them at the very top.
Indicative of the separation across fields, few of the
psychologists were aware of these findings. We asked Len
what might explain the pattern of mortality, and he suggested possible differences in engagement for those at different levels. At this point, the tone of the meeting
changed; Len acknowledged the relevance of psychological
constructs, and the psychologists became excited about a
new challenge: how to explain the health disparities associated with position in the socioeconomic hierarchy.
Members of the network published two articles as a result of that meeting. One (Adler, Boyce, Chesney, Folkman, & Syme, 1993) addressed the medical community at
a time when there was particular interest in health care
reform and the problems of the uninsured. This article,
which was published in the Journal of the American Medical Association, showed why universal coverage, although
important, would not on its own solve the problem of
health disparities. A second article, published in the American Psychologist (Adler et al., 1994), challenged the psychological community to become engaged with explaining
the socioeconomic status (SES) health gradient. Adler et
al. (1994) presented evidence of the gradient and of their
inability to explain it, and suggested that psychosocial factors associated with socioeconomic position and with
health might account for some of the association.
In the ensuing years, more psychologists have become
involved in work related to disparities in health and in
other domains. The Task Force on Socioeconomic Status
established by the American Psychological Association
dealt with a wide range of issues linked to social class and
SES, including both mental and physical health. Task force
members came from different subfields of psychology and
contributed theoretical and empirical work on inequality,
class, and their intersection with other bases of social disadvantage, such as race/ethnicity, gender, disability, sexual
orientation, and age. The task forces report was adopted
by the American Psychological Association Council of
Representatives in 2006. The council established an ongoing committee and an Office on Socioeconomic Status
within the Public Interest Directorate to encourage and coordinate research, education, and interventions to reduce
inequality and disadvantage. This advance places work on
664

health disparities in a broader context within psychology.


However, health disparities research is also linked to other
disciplines; psychologists represent only a small portion of
researchers addressing the problem. Next, I describe the
current state of knowledge and issues in research on disparities, discuss my research and experiences as a health
disparities researcher, and consider implications for the role
of psychology in this research.
Understanding Health Disparities
Definition
Health disparities refer to the unequal distribution of disease and mortality across different groups. Specific definitions vary by which groups or social processes are included
(Carter-Pokras & Baquet, 2002). Some definitions limit the
term to differences in morbidity and mortality associated
with race and ethnicity, whereas others include differences
by characteristics such as gender, SES, and area of residence. In addition, some use the term to refer only to differences among groups in health care access or quality,
whereas others also include health status. The definitions
all converge, however, on differences that are avoidable
and unjust. The former criterion suggests that disparities
primarily result from social rather than biological factors
and could be changed by modifying policy or practice. The
latter suggests that as a matter of justice, all individuals
should be able to achieve the optimal level of health afforded them on the basis of their genetic endowment. In
recent years, the terms health inequality and health inequities have increasingly been used to emphasize the injustice
of differences in health among groups.
Magnitude
There are substantial differences in health and longevity
among groups in the United States. The data presented earlier by Murray et al. (2006) used ecological data on multiple characteristics (race/ethnicity, gender, geography) associated with life expectancy. Most studies have examined
individual sociodemographic factors in relation to specific
diseases, with particular focus on racial and ethnic differences. Large epidemiological studies and national data sets
typically include measures of race/ethnicity and education
but not other aspects of SES. In contrast, British and European data typically classify individuals in terms of occupational class.
In the United States, African Americans have the highest mortality on most causes of death, along with Native
Americans. Hispanics are at high risk of some diseases,
such as diabetes, but have relatively lower rates of cardiovascular disease. Asian Americans generally have the lowest rates of disease, although this classification masks substantial variation across groups of Asian Americans,
including some with poorer health than the general populaNovember 2009 American Psychologist

tion. Although Whites are often taken as the reference


group for comparison with ethnic minorities, they do not
have the best outcomes; non-Hispanic Whites are second
after African Americans in rates for many diseases and in
mortality (Adler, 2006).
Health disparities associated with SES are greater than
those associated with race and ethnicity, however. Differences in health status and mortality between the most and
least affluent within any given racial/ethnic group are
larger than differences between racial/ethnic groups at the
same socioeconomic level. For example, at age 25, the difference in life expectancy between White and Black men is
4.4 years. Within each group, the difference between those
with higher versus lower incomes is almost double that
difference (7.9 and 8.6 years for Whites and Black, respectively). Differences in SES account for much (though not
all) of the differences in health by race/ethnicity. Discriminatory social and economic policies that limit opportunities
for education, housing, high-paying jobs, and so forth contribute to poorer health in disadvantaged groups. In addition, the burden and direct experiences of discrimination
can add to the risk of disease (Williams, 1999).
Disparities Over the Life Course
Health disparities develop even before birth. Babies born to
mothers who experience social disadvantage are more
likely to be born prematurely or to be small for their gestational age. Low birth weight, in turn, is implicated in later
development of disease (Barker, 1995). Although U.S.
children are generally healthy, early indicators of future
disease risk can be seen in those from poorer families.
School-age children with greater accumulation of disadvantage show elevated risk on biological indicators, including
blood pressure and abdominal fat deposition, that foreshadow earlier onset of disease (Evans, 2003). The accumulation of effects associated with disadvantage leads to
increasing disparities over the life course, with the largest
differences seen in middle and late adulthood. The gap narrows after age 65, in part because of differential selection
of those who survive to age 65 and in part because of
safety nets that begin at this age.
Most health disparities researchers have interpreted associations between health and factors such as income and
education in terms of effects of the former on the latter.
However, economists and others have noted that causality
works in both directions and have shown that poor health
may affect SES. For example, childhood illness can limit
future educational achievement and opportunities for work
and income (Case, Fertig, & Paxson, 2005), and in later
life, poor health may force early retirement and reduce income (Smith, 1999). The focus on the impact of health on
SES is especially important in the developing world where
economic development may occur more rapidly among
populations who are freed from the ravages of disease. IsNovember 2009 American Psychologist

sues of causality will continue to be debated, and new analytic tools may be helpful in establishing causal direction
(Adler & Rehkopf, 2008).
The association of SES and health can best be modeled
as dynamic. Figure 1 shows this interaction but does not
indicate the relative strength of the influence in each direction or at different life stages. Although precise values remain to be determined, current research suggests the paths
from SES to health are more powerful than the reverse.
These paths are also more modifiable by social intervention. The dynamic interplay over the life course points to
the importance of interventions early in life because early
experiences can set individuals on different trajectories that
will reverberate through their entire life.
Psychological Understanding of Health Disparities
As with any complex problem, health disparities cannot be
solved by any one discipline. Increasingly, team science is
engaging researchers from disparate fields to tackle such
problems (Adler & Stewart, in press). The meeting of the
research network described earlier helped launch a successor MacArthur network to address the question of how the
SES gradient occurs. The organizing question for the network was How does SES get under the skin to affect
health? Though psychologists are overrepresented, network members encompass fields ranging from neuroscience
and medicine to economics, sociology, and social epidemiology. We identified a number of pathways linking SES to
health. Some, such as differential access and quality of
health care and issues of environmental justice regarding
differential exposure to toxins and carcinogens, were already well known and studied. Not surprisingly, perhaps,
we focused primarily on psychosocial pathways.
Health Behaviors
Health behaviors are responsible for an estimated 40% of
premature mortality in the United States (McGinnis &
Foege, 1993; McGinnis, Williams-Russo, & Knickman,
2002). The major contributors are tobacco use, lack of exercise, underconsumption of fruits and vegetables, and
overconsumption of fats and sugar. All these behaviors
show a graded association with SES, mirroring the SES
gradient in morbidity and mortality. Empirical tests of mediation have found that health behaviors account for up to one
third of the association of SES and health.
Modifying these health behaviors is a promising avenue
for reducing disparities. Health psychologists have done a
great deal of work on understanding health-risk behaviors
and developing interventions to modify them. Relatively
little of this work has been done in conjunction with health
disparity researchers or has been explicitly framed as an
effort to reduce disparities, however. One reason for this
may be the tension between behavioral research, which
takes primarily an individual focus, and disparity research,
665

Figure 1
Socioeconomic Status and Health Over the Life Course

Note. From Reaching for a Healthier Life: Facts on Socioeconomic Status and Health in the U.S. (Figure 2, p. 9) by N. E. Adler & J. Stewart et al., 2007. Copyright 2007
by the John D. and Catherine T. MacArthur Foundation Research Network on Socioeconomic Status and Health.

which focuses on the social context. The predominant


models of health behaviors, such as the health belief model
(Becker, 1974) and the transtheoretical model of change
(Prochaska, DiClemente, & Norcross, 1992) place the onus
for change on the individual. Interventions such as motivational interviewing (Hettema, Steele, & Miller, 2005) aim
to increase the individuals motivation to change and provide necessary information about how to do so. Some disparity researchers worry that these approaches risk blaming
the victim with insufficient attention to the environmental
drivers of the behaviors, a topic to which I return later.
These researchers focus primarily on the social context in
which these processes occur.
Stress Pathway
The lower people are on the social hierarchy, the greater
their exposure to both acute and chronic stress. Because
some cope with stress by engaging in behaviors that bring
relief in the short run but damage health over time (e.g.,
smoking, overeating, substance use), health behaviors are
an indirect pathway from stress to health. Stress may also
directly affect health; as with health behaviors, the bodys
responses to a threat are functional in resolving the immediate threat, but the body pays a price over time. Building
on his animal model work on stress and the hypothalamic
pituitaryadrenal axis, network member Bruce McEwen
elaborated on the concept of allostasis and allostatic load
that can result from chronic stress exposure associated with
lower SES (McEwen, 1998). The core idea is that the
stress response is immediately adaptive but is problematic
in the long term. How much of a price the body pays depends on individual and social resources that can influence
the frequency and severity of stress exposure and/or can
buffer physiological responses to it. Network member
Karen Matthews, along with Linda Gallo, developed the
concept of reserve capacity, the cumulative set of psycho666

social resources, to encompass these processes (Matthews,


Gallo, & Taylor, in press). Sense of control may be an especially important resource. Within the Whitehall study,
another network member, Michael Marmot, showed,
among other things, that an increasing sense of control as
one moved up in occupational grade accounted for much of
the health effect of occupational level (Marmot, Bosma,
Hemingway, Brunner, & Stansfeld, 1997).
Although the specific measurement of allostatic load is
still evolving (Seeman, Epel, Gruenewald, Karlamangla, &
McEwen, in press), as is that of reserve capacity, this work
provides an important conceptual understanding of how
lower social status associated with SES could result in
higher rates of disease and mortality. This work provides a
plausible psychophysiological pathway by which distal social processes could result in the patterns of morbidity and
mortality associated with SES.
Unpacking SES
Most studies use a single measure of SES to represent the
variable. However, the conceptualization and measurement
of SES is potentially problematic. The aspects of SES that
engender a stress response have not been established. The
three core components of SESincome, education, and
occupationreflect different types of resources and are not
always highly correlated with one another. Moreover, associations among the indicators are affected by other bases of
social status. Reflecting the wage gap between men and
women, for example, the correlation between income and
education is weaker for women than for men (Winkleby,
Jatulis, Frank, & Fortmann, 1992).
Measures of each component of SES are complex. Income can be assessed in terms of ones personal income or
that of the household, and the implications of a given level
of income depend on how many people are in the household and on where you live. Education is assessed by years
November 2009 American Psychologist

of schooling and/or highest degree. These measures do not


take into account the quality of the education and the implications of a degree from more versus less prestigious schools.
Occupations vary on a number of dimensions, including their
prestige, financial rewards, and qualifications; in the United
States, the coding scheme typically used is outdated.
In recent years, more attention has been paid to understanding what it is about each aspect of SES that affects
health. Questions asked have included the following: How
does more money buy better health beyond buying better
health care? Does education improve health by increasing
knowledge and problem-solving skills, by engaging individuals in different social networks and exposing them to
different norms, and by providing a credential that increases life opportunities, or is education a marker for individual traits, like conscientiousness or future orientation,
that are linked both to educational attainment and to better
health? What aspects of occupation are health promoting or
health damaging beyond provision of income and health
insurance?
At the same time that researchers identify the active
ingredients of each component of SES, they may learn
something from understanding why all of them generally
act in a similar way in relation to health. Although each
component of SES represents a different type of resource,
common to each is that they place individuals in a social
hierarchy. The possibility that social status itself could
have biological consequences is supported by several
strands of research. For example, the finding that the SES
health gradient operates within countries but operates less
strongly across them suggests that absolute income matters
less than relative income in determining health. Above a
minimal threshold, a given income will be linked to relative health status compared to others in ones country but
not compared to people in other countries. Thus, individuals in the United States with low SES have higher incomes
than do middle-income individuals in less affluent countries, but they do not necessarily have better health. Indeed,
despite our high per capita income (and spending far more
per capita on health care than any other nation), the United
States ranks near the bottom of industrialized nations on a
number of health indicators, including infant mortality and
overall life expectancy. As a graphic example, McCord and
Freeman (1990) reported that Black men in Harlem were
less likely to live to the age of 65 than were men in Bangladesh, despite their comparatively higher incomes. Marmot (2006) has argued that the experiences involving low
control, lack of autonomy, and diminished social participation, which are associated with low SES, contribute to unequal social patterning of health.
Animal Models
Suggestive findings from animal studies shed light on possible effects of social ordering. Dominance hierarchies in
November 2009 American Psychologist

animals are not equivalent to socioeconomic patterning in


humans, but findings from animal models are consistent
with the idea that relative status can affect health. Hierarchies allocate resources in a way that maximizes survival
for the group, but this may be at the cost of individual
members. Nonhuman species vary considerably in their
social organization, and effects of rank vary depending on
the species, its social organization, and the physical surroundings (Sapolsky, 2005). Under unstable conditions,
when dominance needs to be constantly re-established,
dominant animals are more vulnerable, but more often
lower ranked animals are adversely affected.
Health-damaging effects of low status result not just
from physical exposures and exclusion from resources but
from social encounters. Robert Sapolskys (2005) study of
free-living baboons revealed adverse biological consequences of subordinate position even though the animals
inhabited an area that provided ample food. All animals,
irrespective of rank, had adequate nutrition; ironically, the
favorable environment freed the baboons from the constant
hunt for food and gave them time to harass one another.
Lower ranked animals, subjected to chronic social encounters that reinforced their lower status, exhibited neuroendocrine and metabolic profiles similar to those observed in
lower SES humans. Sapolsky (2005) described this parallel, noting that it is a testimony to the power of humans,
after inventing material technology and the unequal distribution of its spoils, to corrosively subordinate its havenots (p. 652).
My colleague Tom Boyce, a behavioral pediatrician,
observed that when placed in groups, young children form
dominance hierarchies that are reminiscent of those he
studied in a colony of monkeys. In the Peers and Wellness
study, my colleagues and I are studying the psychosocial
and biological precursors and consequences of hierarchies
in children entering kindergarten. Preliminary results are
pointing to adverse consequences of lower rank. As children age, they become more aware of their position in
broader society as well as in their immediate group. These
perceptions may shape self-perceptions and engender social
emotions, such as shame, which may in turn affect health
(Dickerson, Gruenewald, & Kemeny, 2004; Kemeny,
Gruenewald, & Dickerson, 2004). My research on subjective status, described next, suggests that such self-perceptions may play an important role in health and the creation
of health disparities.
Subjective Status
Humans, unlike other species, experience multiple hierarchies and bases of social ordering. Complex human societies dont consist of a single pecking order. In addition to
SES, people judge each other by a range of physical and
social attributes. The individual components of SES may
not place a person in the same relative position on the
667

social ladder. For example, high school dropouts may


become millionaires, and poorly paid clergy are highly
respected in their communities. Class notes from the
Harvard Magazine provide a good illustration. An alumnus
reported that he was unemployed, abandoned by his girlfriend, and in a body cast following a motorcycle accident.
He wrote, however, that people were still impressed when
he told them he had graduated from Harvard. How did he
integrate his relative status across domains? Was he protected by the status afforded him by his educational credential? Did he average across life domains or did he give
greater salience to his Harvard degree? Was his self-evaluation affected by his treatment by others, and did their
ways of summing across domains determine how they
treated him?
Prompted by these different streams of evidence, I began a series of studies to determine if people can reliably
report on their relative standing in society and, if so,
whether their perception matters for their health. A small
literature on peoples self-identification in terms of social
class existed (e.g., Jackman & Jackman, 1973). These studies relied on use of labels, and the vast majority of people
identified themselves as middle class. I wanted to use
something that was less value bound and could be used
across a wide range of samples. I developed a visual scale,
similar to earlier work by Cantril (1965), that used a drawing of a ladder to assess subjective social status (SSS),
with the following instructions:
Think of this ladder as representing where people stand in the
United States. At the top of the ladder are the people who are
the best offthose who have the most money, the most education, and the most respected jobs. At the bottom are the
people who are worst offwho have the least money, least
education, and the least respected jobs or no job. 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.

When used in other countries, the instructions are translated, and either the name of the country is substituted for
United States or it reads, imagine that everyone in society
is . . . . Using a drawing of a ladder implicitly conveys the
sense that some people are higher than others, independent
of the basis for their placement on the ladder, which fits
with common terminology referring to the social ladder,
and can be used in different cultures and societies because
ladders are almost universal.
The measure (referred to hereinafter as the ladder) has
proven to be useful. The fact that it assesses a complex
variable with a single item has both disadvantages and advantages. Disadvantages are that it doesnt differentiate the
various dimensions of status or allow for full psychometric
testing. However, it has been relatively easy to include in a
variety of studies. Epidemiologists who design large population studies are notorious among psychologists for asking
668

for one or two items to assess psychological concepts that


require a dozen or more items for adequate assessment.
Because of its brevity, epidemiologists and others have
been willing to include the ladder in their surveys. As a
result, it has been used with a range of populations, including a nationally representative sample of the United States
(Operario, Adler, & Williams, 2004) and of Hungary
(Kopp, Skrabski, Rethelyi, Kawachi, & Adler, 2004), a
multiethnic sample of pregnant women (Ostrove, Adler,
Kuppermann, & Washington, 2000), British civil servants
participating in the ongoing Whitehall study (Singh-Manoux, Adler, & Marmot, 2003; Singh-Manoux, Marmot, &
Adler, 2005), impoverished women in Mexico participating
in the Oportunidades program (Fernald & Adler, 2008) and
their adolescent children (Ritterman et al., 2009), elderly
men and women in Taiwan (Hu, Adler, Goldman, Weinstein, & Seeman, 2005) and England (Wright & Steptoe,
2005), Cherokee and Anglo adolescents in Appalachia
(Brown et al., 2008), and a national survey of Asian Americans (Leu et al., 2008).
These studies have revealed strong and consistent associations between SSS and self-rated health and depressive
symptoms, which remain significant even when adjusted
for objective SES. Although this supports an independent
impact of subjective status on health, alternative explanations include a spurious association due to an underlying
third factor, such as neuroticism or negative affect or
mono-method bias, any of which could lead people to rate
their health and their social status in a similar way. These
can be refuted with several types of data. Findings from
Operario et al. (2004) address potential confounding by
negative affect. In a national phone survey, SSS was significantly related to self-rated health even when researchers
controlled for income, education, and negative affect. Furthermore, the association of global health with income and
with education drops to an even greater extent when adjusted for negative affect than does the association with
SSS. This pattern suggests that negative affect mediates
rather than confounds the association. Both lower objective
SES and lower SSS appear to affect health in part through
increasing negative affect, as well as having a direct association.
Data from the Whitehall II study are also consistent
with a causal impact of SSS on health. Ladder rankings in
Wave 5 of the study showed significant cross-sectional associations with global health, with the prevalence of angina, diabetes, and depression, and (in men only) with respiratory illness. In addition, SSS predicted change in
health status. Controlling for health status at Wave 5, SSS
measured at Wave 5 predicted health in Wave 6. Shared
variance due to underlying third variables would have been
controlled for in adjusting for the contemporaneous health
report. The residual still predicted subsequent health.
November 2009 American Psychologist

Stronger evidence that the association of SSS and selfrated health is not due to confounding emerges from studies that do not rely on self-report of health outcomes but
rely instead on biological markers. Although third factors
and reverse causality could potentially account for associations, these seem unlikely given the variety of indicators
used. A common aspect of the diverse measures is that
they reflect chronic stress exposure. In a community sample of women from New Haven, Adler, Epel, Castellazzo,
and Ickovics (2000) found that those reporting lower SSS
showed greater physiological arousal as indicated by a
faster heart rate and longer sleep latency (time to fall
asleep) even after adjustment for education, income, and
negative affect. Lower SSS was related to a significantly
greater rise in morning cortisol in a sample of older adults
in England, whereas objective SES was not (Wright &
Steptoe, 2005), and was related to greater diurnal cortisol
and abdominal fat deposition in a sample of diabetic men
(Epel & Adler, 2001). The idea that SSS affects health
through stress pathways is also supported by imaging data
showing a smaller volume of grey matter in an area of the
brain that encodes stress, the perigenual area of the anterior
cingulate cortex (Gianaros et al., 2007). The evidence
comes from the study reported by Cohen et al. (2008), in
which healthy volunteers gave ladder rankings before they
were exposed to a rhinovirus. Those who placed themselves lower on the ladder were more likely to become
infected, independent of potential confounders, including
traditional SES measures, psychological characteristics
such as self-esteem and positive and negative emotional
style, and health habits. Consistent with the finding reported earlier on sleep, the only variables that reduced the
association were sleep efficiency and sleep duration.
What might account for the independent association of
SSS with the biological indicators and with global health?
As noted earlier, reverse causation cannot be ruled out;
individuals who are ill or who have biological risk factors
may perceive themselves as having relatively lower standing in society apart from their socioeconomic resources. In
addition, although the studies provide different ways to
assess potential confounding, it is not possible to rule this
out altogether. My hypothesis is that global self-rated
health relates to SSS above and beyond its association with
the objective components of SES in part because the
former is a more sensitive and comprehensive indicator of
social position than the latter. Current measures of objective SES are relatively imprecise in assessing the actual
resources and experiences associated with each SES indicator. As discussed earlier, for example, measures of education do not account for quality of schooling or the prestige
associated with ones credential. In determining where they
stand on the ladder, people know their objective status and
may also factor in the value and implications of each component for their life. They may also consider subtler facNovember 2009 American Psychologist

tors, such as interpersonal interactions and whether these


indicate respect or devaluation.
SSS ratings may operate like global self-rated health, a
single item that asks individuals to assess their overall
health compared to that of others their age. Responses to
this item predict mortality even after adjusting for all
known objective indicators of health (Idler & Benyamini,
1997). Self-rated health may be a stronger predictor of subsequent health and survival because people know the objective indicators of their health and interpret them in line
with subtler indicators, including their vigor and sense of
well-being. It is logical that SSS would show a stronger
association with global health than with specific diseases.
Overall social disadvantage is associated with greater vulnerability to a wide range of diseases. Allostatic load resulting from the chronic stress of lower status fosters dysregulation of many systems in the body (McEwen, 1998;
Sapolsky, 2005; Seeman et al., in press), and development
of a specific disease may depend on determinants such as
genetic risk and specific psychosocial and environmental
exposures.
Individual- Versus Social-Level Determinants
My work on disparities and subjective status draws on work
from a range of fields and provides some insight on how our
field relates to them. Psychology sits at the nexus of the social
sciences on one side and the biological sciences on the other.
Health psychology, drawing from all domains of psychology,
has a similar scope and position. This enables health psychologists to bridge to biomedical sciences as well as to medical
social sciences and public health. It also exposes us to criticisms from the former that we are too soft and from the latter
that we are too hard (in terms of being reductionist). Perhaps
this makes us just right in our capacity to deal with complex
issues, such as health disparities.
The terminology used to refer to disciplines is value
laden. When I first chaired the curriculum committee for
the University of California, San Francisco, medical
school, material was viewed as either basic science or clinical. Psychosocial issues were part of the latter and seen as
peripheral. When the curriculum was redesigned, the committee argued successfully that social and behavioral sciences are also basic sciences underpinning the practice of
medicine. Our vocabulary is slowly shifting to refer to basic biological and social/behavioral science. Having now
spent many years in a biomedical environment, I am familiar with critiques that psychology is a soft science. The
phenomena psychologists are trying to explain are multidetermined and messier than are those studied at the bench.
Psychologists might reasonably argue that if biological sciences are the hard sciences, psychology is a harder science
because it aims to explain more complex phenomena (and
our social science colleagues might add that theirs are the
hardest).
669

There is encouraging evidence of an expanded awareness of psychological processes in biomedical sciences.


The problem of health disparities and the puzzle of the
SES gradient provide more impetus in this direction. Biomedical sciences alone cannot solve them, and psychologists have the opportunity to demonstrate the power of psychosocial research. To accomplish this, however, we have
to expand our own explanatory models. Solving the problem of health disparities requires not only moving beyond
biomedical sciences to include psychological determinants
of health, but reaching further to include social determinants. Disciplines such as sociology, anthropology, and
social epidemiology have been studying social factors that
foster health disparities. Reminiscent of the iconic New
Yorker Magazine drawing of a New Yorkers view of the
United States, where fine differentiations are made among
boroughs of Manhattan but west of New Jersey it all looks
the same, biomedical scientists may see little difference
between social and behavioral sciences. However, differences in intellectual roots, assumptions, methods, and levels of analysis can sometimes create tensions. Two influential articles, both of which challenge the biomedical
approach to understanding disease, exemplify some of
these differences. First, McGinnis and Foege (1993) analyzed the actual causes of death. They observed that although people may die of cancer or heart disease, the actual causes are tobacco use, sedentary lifestyle, and so
forth. They estimated annual mortality associated with the
major underlying contributors, most of which were behavioral. Second, Link and Phelan (1995) argued that the fundamental causes of disease are social conditions. They observed that the mechanisms linking SES and health change
over time, but the fundamental association of poverty and
social disadvantage with health remains. This implicitly
challenges the utility of studying more proximal determinants of disease and mortality, including behavior.
The actual causes of death resonate with health psychology and the interdisciplinary field of behavioral medicine,
whereas the fundamental causes resonate with the field of
public health. The public health perspective emphasizes
social context and the importance of intervening upstream
to modify the conditions that expose individuals to diseasecausing agents rather than focusing on modifying what the
individual does himself or herself to contribute to the exposure. John Snows response in 1854 to his discovery that
contaminated water from a communal well was responsible
for a cholera outbreak (Snow, 1855) helped define the public health approach. Rather than attempt to get all individuals to boil the water drawn from the well, Snow removed
the pump handle to prevent further exposure. The public
health community continues the focus on modifying the
vector of transmission rather than the person. Some in public health are critical of attempts to identify more proximal
determinants. In relation to disparities, they emphasize the
670

unequal and unjust distribution of material resources and


the political and economic processes that bring these about,
and they are critical of psychosocial approaches (e.g.,
Lynch, Smith, Kaplan, & House, 2000). Two types of concerns underlie this criticism. The first is that individual-level
research, particularly if examining health behaviors, fosters
the view that disadvantaged groups are responsible for their
greater health risk (Lynch, Kaplan, & Salonen, 1997). The
second is that focusing on the characteristics of individuals
deflects attention from unjust societal conditions and the
institutional processes that create them (Krieger, 2001).
It is my belief that individual processes and structural
contexts both need to be addressed to successfully address
health disparities. We can see this in attempts to address
the obesity epidemic. This epidemic began earlier and is
more widespread among those lower on the SES hierarchy.
Psychological interventions focus on motivating individuals
to change their diet and to engage in more exercise. Social
interventions aim to change the environment to facilitate
healthier behavior, focusing on resources and constraints in
the community, including cost and availability of fresh
fruits and vegetables, density of fast-food outlets, and access to recreational facilities. In a recent article, Judith
Stewart and I tried to bridge the gap between these two
perspectives by suggesting that both could be encompassed
by a social justice perspective (Adler & Stewart, 2009).
Given the strong contribution of behavior to health, we
argued that access to the resources needed to allow people
to engage in health-promoting behaviors is an issue of social justice. Making an analogy to the environmental justice concept, which argues that it is unjust for disadvantaged communities to be disproportionately exposed to
toxic substances and environmental hazards, we proposed
the concept of behavioral justice, which asserts that it is
unjust for disadvantaged communities to lack resources
needed for engaging in health-promoting behaviors. However, although access to resources is necessary, it is often
not sufficient to maintain health-promoting behaviors. In
addition to assuring adequate resources, we need to understand how to motivate behavior and how to maintain it.
We have much work to do to eliminate or even substantially reduce health disparities. My experience in the
MacArthur network convinced me of the value of interdisciplinary work and of the central role that psychology can
play in accomplishing this goal. The scope of our discipline allows us to bridge to both more molecular and more
molar fields. Although this can sometimes be challenging,
it is not only worthwhile but crucial for progress in addressing key societal problems such as health disparities.
Authors Note
Preparation of this article was partially supported by the
John D. and Catherine T. MacArthur Foundation Research
Network on Socioeconomic Status and Health. I thank JuNovember 2009 American Psychologist

dith Stewart for her help on this article, my network colleagues for their colleagueship, Elissa Epel for her collaboration, and the postdoctoral students in psychology and
medicine for their input.
Correspondence concerning this article should be addressed to Nancy E. Adler, Health Psychology Program,
3333 California Street, Suite 465, University of California,
San Francisco, CA 94118. E-mail: nancy.adler@ucsf.edu
References
Adler, N. E. (2006). Overview of health disparities. In
G. E. Thompson, F. Mitchell, & M. Williams (Eds.), Examining the health disparities research plan of the National Institutes of Health: Unfinished business (pp. 129
188). Washington, DC: National Academic Press.
Adler, N. E., Boyce, T., Chesney, M. A., Cohen, S., Folkman, S., Kahn, R. L., & Syme, S. L. (1994). Socioeconomic status and health. The challenge of the gradient.
American Psychologist, 49, 1524.
Adler, N. E., Boyce, T., Chesney, M. A., Folkman, S., &
Syme, S. L. (1993). Socioeconomic inequalities in health.
No easy solution. Journal of the American Medical Association, 269, 3140 3145.
Adler, N. E., Epel, E. S., Castellazzo, G., & Ickovics, J. R.
(2000). Relationship of subjective and objective social status with psychological and physiological functioning: Preliminary data in healthy White women. Health Psychology,
19, 586 592.
Adler, N. E., & Ostrove, J. M. (1999). Socioeconomic status and health: What we know and what we dont. In N. E.
Adler, M. Marmot, B. S. McEwen, & J. Stewart (Eds.),
Annals of the New York Academy of Sciences: Vol. 896.
Socioeconomic status and health in industrialized nations:
Social, psychological and biological pathways (pp. 315).
New York: New York Academy of Sciences.

Adler, N. E., & Stewart, J. (with Cohen, S., Cullen, M.,


Diez Roux, A., Dow, W., et al.). (2007). Reaching for a
healthier life: Facts on socioeconomic status and health
in the U.S. Retrieved September 3, 2009, from the John
D. and Catherine T. MacArthur Foundation Research
Network on Socioeconomic Status and Health website:
http://www.macses.ucsf.edu/News/Reaching%20for%20
a%20Healthier%20Life.pdf
Barker, D. J. (1995). Fetal origins of coronary heart disease. British Medical Journal, 311, 171174.
Becker, M. H. (Ed.). (1974). The health belief model and
personal health behavior. Health Education Monograph, 2,
324 473.
Brown, R. A., Adler, N. E., Worthman, C. M., Copeland,
W. E., Costello, E. J., & Angold, A. (2008). Cultural and
community determinants of subjective social status among
Cherokee and White youth. Ethnicity and Health, 13, 289
303.
Cantril, H. (1965). The patterns of human concerns. New
Brunswick, NJ: Rutgers University Press.
Carter-Pokras, O., & Baquet, C. (2002). What is health
disparity? Public Health Reports, 117, 426 434.
Case, A., Fertig, A., & Paxson, C. (2005). The lasting impact of childhood health and circumstance. Journal of
Health Economics, 24, 365389.
Cohen, S., Alper, C. M., Doyle, W. J., Adler, N. E.,
Treanor, J. J., & Turner, R. B. (2008). Objective and subjective socioeconomic status and susceptibility to the common cold. Health Psychology, 27, 268 274.
Dickerson, S. S., Gruenewald, T. L., & Kemeny, M. E.
(2004). When the social self is threatened: Shame, physiology, and health. Journal of Personality, 72, 11911216.

Adler, N. E., & Rehkopf, D. H. (2008). U.S. disparities in


health: Descriptions, causes, and mechanisms. Annual Review of Public Health, 29, 235252.

Epel, E., & Adler, N. (2001). Subjective social status and


physiological functioning: Relations with cortisol rhythm
and central fat distribution in diabetic men [Abstract]. Psychosomatic Medicine, 63, 91190.

Adler, N. E., & Stewart, J. (2009). Reducing obesity: Motivating action while not blaming the victim. Milbank Quarterly, 87, 49 70.

Evans, G. W. (2003). A multimethodological analysis of


cumulative risk and allostatic load among rural children.
Developmental Psychology, 39, 924 933.

Adler, N. E., & Stewart, J. (in press). Health disparities


across the lifespan: Meaning, methods and mechanisms. In
N. E. Adler & J. Stewart (Eds.), The biology of disadvantage. New York: New York Academy of Sciences.

Fernald, L. C. H., & Adler, N. E. (2008). Blood pressure


and socio-economic status in low-income women in Mexico: A reverse gradient? Journal of Epidemiology & Community Health, 62, e8.

November 2009 American Psychologist

671

Gianaros, P. J., Horenstein, J. A., Cohen, S., Matthews,


K. A., Brown, S. M., Flory, J. D., et al. (2007). Perigenual
anterior cingulated morphology covaries with perceived
social standing. Social Cognitive and Affective Neuroscience, 2, 161173.
Healthy People 2010 midcourse review. (n.d.). Retrieved
August 1, 2008, from www.healthypeople.gov/data/
midcourse/html/execsummary/introduction.htm
Hettema, J., Steele, J., & Miller, W. R. (2005). Motivational interviewing. Annual Review of Clinical Psychology,
1, 91111.
Hu, P., Adler, N. E., Goldman, N., Weinstein, M., & Seeman, T. E. (2005). Relationship between subjective social
status and measures of health in older Taiwanese. Journal
of the American Geriatrics Society, 53, 483 488.
Idler, E. L., & Benyamini, Y. (1997). Self-Rated health
and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38, 2137.
Jackman, M. R., & Jackman, R. W. (1973). An interpretation of the relation between objective and subjective social
status. American Sociological Review, 38, 569 582.
Kemeny, M. E., Gruenewald, T. L., & Dickerson, S. S.
(2004). Shame as the emotional response to threat to the
social self: Implications for behavior, physiology, and
health. Psychological Inquiry, 15, 153160.
Kopp, M., Skrabski, A., Rethelyi, J., Kawachi, I., & Adler,
N. E. (2004). Self-rated health, subjective social status and
middle-aged mortality in a changing society. Behavioral
Medicine, 30, 6570.
Krieger, N. (2001). Theories for social epidemiology in the
21st century: An ecosocial perspective. International Journal of Epidemiology, 30, 668 677.
Leu, J., Yen, I. H., Gansky, S. A., Walton, E., Adler,
N. E., & Takeuchi, D. T. (2008). The association between
subjective social status and mental health among Asian
immigrants: Investigating the influence of age at immigration. Social Science & Medicine, 66, 11521164.
Link, B. G., & Phelan, J. (1995). Social conditions as fundamental causes of disease. Journal of Health and Social
Behavior, 35, 80 94.
Lynch, J. W., Kaplan, G. A., & Salonen, J. T. (1997). Why
do poor people behave poorly? Variation in adult health
behaviours and psychosocial characteristics by stages of the
672

socioeconomic lifecourse. Social Science & Medicine, 44,


809 819.
Lynch, J. W., Smith, G. D., Kaplan, G. A., & House, J. S.
(2000). Income inequality and mortality: Importance to
health of individual income, psychosocial environment, or
material conditions. British Medical Journal, 320, 1200
1204.
Marmot, M. G. (2006). Status syndrome: A challenge to
medicine. Journal of the American Medical Association,
295, 1304 1307.
Marmot, M. G., Bosma, H., Hemingway, H., Brunner, E.,
& Stansfeld, S. (1997). Contribution of job control and
other risk factors to social variations in coronary heart disease incidence. Lancet, 350, 235239.
Matthews, K. M., Gallo, L., & Taylor, T. (in press). Are
psychological factors mediators of SES and health connections? A progress report and blueprint for the future. In
N. E. Adler & J. Stewart (Eds.), The biology of disadvantage. New York: New York Academy of Sciences.
McCord, C., & Freeman, H. P. (1990). Excess mortality in
Harlem. New England Journal of Medicine, 322, 173177.
McEwen, B. S. (1998). Protective and damaging effects of
stress mediators. New England Journal of Medicine, 338,
171179.
McGinnis, J. M., & Foege, W. H. (1993). Actual causes of
death in the United States. Journal of the American Medical Association, 270, 22072212.
McGinnis, J. M., Williams-Russo, P., & Knickman, J. R.
(2002). The case for more active policy attention to health
promotion. Health Affairs (Millwood), 21, 78 93.
Murray, C., Kulkarni, S., Michaud, C., Tomijima, N., Bulzacchelli, M., Iandiorio, T., et al. (2006). Eight Americas:
Investigating mortality disparities across races, counties,
and race counties in the United States. PLoS Medicine,
3(9), e260.
Operario, D., Adler, N. E., & Williams, D. R. (2004). Subjective social status: Reliability and predictive utility for
global health. Psychology and Health, 19, 237246.
Ostrove, J. M., Adler, N. E., Kuppermann, M., & Washington, A. E. (2000). Objective and subjective assessments
of socioeconomic status and their relationship to self-rated
health in an ethnically diverse sample of pregnant women.
Health Psychology, 19, 613 618.
November 2009 American Psychologist

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C.


(1992). In search of how people change: Applications to
addictive behaviors. American Psychologist, 47, 1102
1114.
Ritterman, M. L., Fernald, L. C., Ozer, E. J., Adler, N. E.,
Gutierrez, J. P., & Syme, S. L. (2009). Objective and subjective social class gradients for substance use among Mexican adolescents. Social Science & Medicine. Advance online publication.
Sapolsky, R. M. (2005, April 29). The influence of social
hierarchy on primate health. Science, 308, 648 652.
Seeman, T., Epel, E., Gruenewald, T., Karlamangla, A., &
McEwen, B. (in press). Socio-economic differentials in
peripheral biology: Cumulative allostatic load. In N. E.
Adler & J. Stewart (Eds.), The biology of disadvantage.
New York: New York Academy of Sciences.
Singh-Manoux, A., Adler, N. E., & Marmot, M. G. (2003).
Subjective social status: Its determinants and its association
with measures of ill-health in the Whitehall II study. Social
Science & Medicine, 56, 13211333.
Singh-Manoux, A., Marmot, M. G., & Adler, N. E. (2005).
Does subjective social status predict health and change in
health status better than objective status? Psychosomatic
Medicine, 67, 855 861.

November 2009 American Psychologist

Smith, J. P. (1999). Healthy bodies and thick wallets: The


dual relation between health and economic status. Journal
of Economic Perspectives, 13, 145166.
Snow, J. (1855). On the mode of communication of cholera
(2nd ed.). London: John Churchill.
Williams, D. R. (1999). Race, socioeconomic status, and
health. The added effects of racism and discrimination. In
N. E. Adler, M. Marmot, B. S. McEwen, & J. Stewart
(Eds.), Annals of the New York Academy of Sciences: Vol.
896. Socioeconomic status and health in industrialized nations: Social, psychological and biological pathways (pp.
173188). New York: New York Academy of Sciences.
Winkleby, M. A., Jatulis, D. E., Frank, E., & Fortmann,
S. P. (1992). Socioeconomic status and health: How education, income, and occupation contribute to risk factors for
cardiovascular disease. American Journal of Public Health,
82, 816 820.
Wright, C. E., & Steptoe, A. (2005). Subjective socioeconomic position, gender and cortisol responses to waking in
an elderly population. Psychoneuroendocrinology, 30, 582
590.

673