Journals of Gerontology: Social Sciences
cite as: J Gerontol B Psychol Sci Soc Sci, 2020, Vol. 75, No. 9, 1937–1950
doi:10.1093/geronb/gbz068
Advance Access publication June 21, 2019
Original Article
Death by a Thousand Cuts: Stress Exposure and Black–
White Disparities in Physiological Functioning in Late Life
Courtney Boen, MPH, PhD*
Department of Sociology, Population Studies Center, and Population Aging Research Center, University of Pennsylvania,
Philadelphia.
*Address correspondence to: Courtney Boen, MPH, PhD, Department of Sociology, Population Studies Center, and Population Aging Research
Center, University of Pennsylvania, 232 McNeil Building, 3718 Locus Walk, Philadelphia, PA 19104-6299. E-mail: cboen@sas.upenn.edu
Received: September 24, 2018; Editorial Decision Date: May 20, 2019
Decision Editor: Deborah Carr, PhD
Abstract
Objectives: This paper investigates Black–White differences in stress—including diverse measures of chronic, acute,
discrimination-related, and cumulative stress exposure—and examines whether race differences in these stress measures
mediate Black–White disparities in C-reactive protein (CRP) and metabolic dysregulation in later life.
Methods: Using data from the Health and Retirement Study (HRS) (2004–2012), this study uses stepwise ordinary least
squares (OLS) regression models to examine the prospective associations between multiple stressors—including traumatic
and stressful life events, financial strain, chronic stress, everyday and major life discrimination, and measures of cumulative
stress burden—and CRP and metabolic dysregulation. Mediation analyses assessed the contribution of stress exposure to
Black–White disparities in the outcomes.
Results: Blacks experienced more stress than Whites across domains of stress, and stress exposure was strongly associated
with CRP and metabolic dysregulation. Race differences in financial strain, everyday and major life discrimination, and cumulative stress burden mediated Black–White gaps in the outcomes, with measures of cumulative stress burden mediating
the greatest proportion of the racial disparities.
Discussion: The “thousand cuts” that Blacks experience from their cumulative stress exposure across domains of social
life throughout the life course accelerate their physiological deterioration relative to Whites and play a critical role in racial
health disparities at older ages.
Keywords: Life course, Physiological functioning, Racial health disparities, Stress
Research documents stark Black–White health disparities
from mid through late life, whereby Blacks in the United
States experience earlier onset of disease, greater severity
of illness, and poorer survival rates than Whites (Williams,
Mohammed, Leavell, & Collins, 2010). Because of their
positioning in both the social class and racial hierarchies,
Blacks in the United States are exposed to greater levels
of material deprivation (Boen, 2016) and report higher
levels of related psychosocial stress (Turner & Avison,
2003) than Whites. Given documented links between psychosocial stress exposure and health (Cohen et al., 2012;
Thoits, 2010), scholars hypothesize Blacks’ cumulative exposure to stressors across the life course may play an essential role in racial health inequality (Ferraro & Shippee,
2009; Geronimus et al., 2010; Goosby, Straley, & Cheadle,
2017; Strenthal, Slopen, & Williams, 2011; Turner, 2013;
Williams, Yu, Jackson, & Anderson, 1997).
Despite a growing body of research in this area, critical
gaps in our understanding of the role of stress exposure in
the production of racial health inequality at older ages remain. First, most studies assess the associations between
single domains of stress—such as discrimination—and
© The Author(s) 2019. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved.
For permissions, please e-mail: journals.permissions@oup.com.
1937
1938
health, leaving questions about whether and how race
differences in the accumulation of stress exposure across
domains of social life are implicated in Black–White health
gaps at older ages. Few studies incorporate measures of
cumulative life course stress exposure—including diverse
indicators of acute, chronic, discrimination-related, and total
stress burden—which may underestimate the role of stress
in health inequality at older ages. Second, while studies have
examined how stress relates to single indicators of disease or
self-reported health status, fewer assess how stress exposure
impacts multiple markers of physiological functioning, which
restricts understanding of the social and biological processes
contributing to disparities in disease emergence and progression in later life. Further, because Blacks are less likely than
Whites to receive a diagnosis (Williams & Jackson, 2005),
the use of disease outcomes in studies of racial health inequality may be particularly worrisome, as it risks underestimation of both the magnitude of Black–White health gaps
and the role of social exposures in racial health disparities.
Finally, few studies of stress and health use longitudinal data
or formal mediation techniques to assess the role of stress
exposure in population health gaps, which raises concerns
about reverse causality and limits understanding of the relative contributions of stressors in racial health disparities.
Using nationally representative, longitudinal data from
the Health and Retirement Study (HRS), the current study
expands understanding of the role of stress in racial health
inequality by examining how exposure to multiple forms
of stress—including domain-specific measures of acute,
chronic, and discrimination-related stress as well as composite indicators of cumulative stress burden—contribute
to racial disparities in inflammation and metabolic risk
from mid through late life. The study begins by assessing
race gaps in physiological functioning and stress exposure.
Then, using a combination of prospective OLS regression
models and mediation analyses, the study examines the relative contributions of diverse indicators of stress exposure
to Black–White disparities in physiological functioning,
paying particular attention to the role of cumulative life
course stress exposure in the production of racial health
inequality at older ages. By expanding the conceptualization and operationalization of “racialized social stress” to
include a host of racially patterned psychosocial exposures
and assessing the relative contributions of a variety of
domain-specific and composite measures of stress to Black–
White disparities across multiple physiological systems,
this study provides new evidence of how race differences
in the accumulation of stress across domains of social life
throughout the life span contribute to stark Black–White
disparities in physiological functioning in late life.
Black–White Disparities in Health and
Physiological Function
Blacks in the United States experience higher rates of morbidity and mortality from a host of conditions, including
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
cardiovascular diseases (National Center for Health
Statistics, 2016). Still, less is known about the physiological processes and social exposures that produce racial disparities in death and disease. Increasingly, scholars
have integrated markers of biological risk and physiological dysregulation in studies of population health inequality in order to elucidate the biophysiological processes
undergirding social disparities in disease. Indeed, research
documents stark racial disparities in biological risk at older
ages, including Black–White disparities in inflammatory
and metabolic risk (Mitchell & Aneshensel, 2017; Mitchell,
Ailshire, & Crimmins, 2019), which may contribute to racial disparities across a host of disease outcomes.
Given that biological risk profiles in later life have
been linked to the accumulation of psychological, social,
and environmental exposures (Ferraro & Shippee, 2009),
scholars hypothesize that racial differences in physiological
functioning at older ages may reflect race differences in cumulative exposure to racialized risks and opportunities—
including psychosocial exposures—across the life course
(Geronimus et al., 2010). In particular, research increasingly highlights the critical role of immune function and energy metabolism in shaping health and disease risk (Finch,
2010). Both chronic inflammation—a marker of sustained
immune system activation—and metabolic disorders are
identified as key pathogenic pathways affecting mortality
risk at older ages, particularly from cardiovascular diseases
(Yang & Kozloski, 2011), a leading cause of death among
older adults. As such, understanding how social exposures
shape individual health risk and pattern population-level
racial health disparities through inflammatory and metabolic processes can inform prevention efforts aimed at
identifying the psychosocial and physiological predecessors
of multiple forms of disease, including cardiovascular
diseases, from mid through late life (Mitchell & Aneshensel,
2017; Mitchell et al., 2019).
The Role of Stress in Racial Health Inequality
A large body of literature seeks to understand the social
determinants of racial health disparities, with a number
of studies indicating that stress may be an important—but
largely underestimated—mechanistic pathway underlying
Black–White gaps in health (Ferraro & Shippee, 2009;
Geronimus et al., 2010; Goosby et al., 2017; Strenthal
et al., 2011; Turner, 2013; Williams et al., 1997). More than
40 decades of research document the substantial effects of
stress on health (Cohen et al., 2012; Thoits, 2010). Studies
on the health consequences of stress typically draw on
the stress process model (Pearlin, Menaghan, Lieberman,
& Mullan, 1981) to infer the pathways through which
stress exposure “gets under the skin” to affect disease risk.
In the face of stress, the hypothalamic-pituitary-adrenal
axis and sympathetic nervous system respond by secreting
hormones to upregulate functioning across physiological systems, including inflammatory and cardiometabolic
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
functioning (McEwen, 2007). While this upregulation in
response to immediate threats and infections is necessary
to protect health in the short term, repeated activation of
the body’s stress response systems results in physiological
“exhaustion,” where the bodily mechanisms used to defend
against stressors become inefficient and ineffective (Seyle,
1974). Repeated exposure to stress diminishes the ability
of physiological systems to downregulate (Cohen et al.,
2012), such that they become hypervigilant, operating
under “threat levels” even in the absence of an immediate
threat. In this way, long term and repeated stress exposure
can promote physiological dysregulation and increased disease and mortality risk from a host of causes (Cohen et al.,
2012; McEwen, 2007).
Research documents that stress exposure is not randomly distributed in the population. Socially disadvantaged
groups are exposed to more negative life events and greater
levels of chronic strain than advantaged groups (Pearlin,
1999) from birth through late life, in ways that shape
the life course patterning of health inequality (Ferraro &
Shippee, 2009). With regard to racial health inequities,
racism patterns exposure to a variety of stressors, both
discrimination-related and more generalized (Pearlin,
1999), in ways that burden people of color with greater
lifetime stress burden than Whites. The United States is a
racialized social system (Bonilla-Silva, 1997), with racism
operating across domains of social life and at the institutional, interpersonal, and internalized levels to shape exposure to risks and opportunities to contribute to the racial
patterning of health (Jones, 2000). While most studies on
the role of racism in shaping stress exposure and health
risk focus on perceived discrimination (Goosby et al.,
2017), racism patterns exposure to a host of stressors—
both discrimination-related and more generalized—in ways
that are linked to United States’s racialized social system.
For example, research documents striking Black–White
disparities in exposure to the deaths of family members
and friends (Umberson et al., 2017) and financial instability and insecurity (Boen, 2016)—disparities in stress exposure that are not captured by studies of perceived racial
discrimination. As such, examinations of the ways in which
racialized social stress may contribute to population health
gaps must consider how living in a racialized social system
patterns exposure to a host of psychosocial stressors,
both discrimination-related and more generalized, across
domains of social life.
Gaps in the Literature
Taken together, work on the physiological consequences
of stress, on the one hand, and the racial patterning of
stress exposure, on the other, suggests that Blacks’ greater
cumulative exposure to racialized social stress across the
life course may contribute to their disproportionate physiological deterioration relative to Whites and play a critical role in producing Black–White disparities across a host
1939
of disease outcomes (Brown & Hargrove, 2018; Ferraro
& Shippee, 2009; Geronimus et al., 2010; Goosby et al.,
2017; Strenthal et al., 2011). Despite a growing body of
literature in this area, three critical gaps in the literature
remain. First, most studies measure stress exposure use one
or two indicators of stress, which suggests that the role of
stress in Black–White disparities in health and aging has
been underestimated (Turner, 2013). As discussed, while
a growing body of literature on the role of stress in racial health disparities considers the health consequences of
racial discrimination (Goosby et al., 2017), both discrimination and more generalized indicators of social stress
are patterned by race and associated with health. As such,
failure to include multiple dimensions of stress exposure in
studies of racial health inequality risks underestimating the
contribution of stress to racial disparities in health. Further,
while evidence suggests that cumulative exposure to social
stressors play a role in Black–White health gaps (Brown &
Hargrove, 2018; Strenthal et al., 2011), few studies assess
how the accumulation of stress across domains throughout
the life span contribute to racial health gaps. Insights from
cumulative inequality theory (Ferraro & Shippee, 2009)
suggest that stressors accumulate across the life course and
across domains of social life to produce population health
inequality at older age. Consistent with this notion, research by Strenthal and colleagues (2011) and Brown and
Hargrove (2018) used count indices of high stress exposure to show that individuals with high exposure across
domains of stress had increased health risk. Still, few studies
incorporate these cumulative stress measures to more holistically estimate the role of stress in racial health inequality
and document the unique and joint contributions of diverse
measures of stress exposure to racial disparities in health.
Second, most studies of the role of stress exposure in
population health inequality use markers of disease (e.g.,
Strenthal et al., 2011) or self-reported health status (e.g.,
Brown and Hargrove, 2018) as outcomes, which restricts
understanding of how stress exposure impacts a range of
biological processes to ultimately impact health and disease risk and contribute to health inequality across a range
of outcomes. For one, studies that utilize measures of disease or diagnosis as outcomes risk misclassification error,
whereby individuals who do not yet have the disease or
have not yet been diagnosed with the disease are classified
as “well” (Aneshensel, Rutter, & Lachenbruch, 1991). The
misclassification of individuals with high levels of physiological dysregulation as “healthy” could result in an underestimation of the role of stress in shaping individual health and
mortality risk, which is a particular concern for studies of racial health gaps given documented disparities in health care
access and diagnosis that make Blacks less likely to receive a
diagnosis than Whites with similar health profiles (Williams
& Jackson, 2005). Further, because studies on stress exposure
and health usually rely on indicators of disease, diagnosis, or
self-reported health status as outcomes, less is known about
how stress exposure affects multiple physiological systems
1940
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
to predispose individuals to host of poor outcomes (Turner,
2013). Studies that consider the effects of stress on multiple,
predisease markers of physiological well-being would be
consistent with the nonspecificity hypothesis that guides the
stress process model, whereby stress exposure elicits a range
of physiological changes that, over time, serve to increase
disease risk from a host of causes (Seyle, 1974). Research
by Geronimus et al. (2010) suggests that Blacks may experience accelerated physiological deterioration relative to
Whites—in a process called “weathering”—due largely to
racial differences in social exposures across the life course.
Still, more research using multiple predisease markers of
physiological well-being is needed, as such studies can elucidate the processes through which cumulative exposure to
racialized social stress over the life course “gets under the
skin” to accelerate biological aging processes and ultimately
shape population patterns of racial disparities in disease risk
in old age (Ferraro & Shippee, 2009).
Finally, few studies of the relationship between stress
and health use longitudinal data or formally test the contribution of stress exposure to population health gaps using
mediation analyses and nationally representative data. The
use of longitudinal data can reduce concerns about reverse
causality and improve causal inference, while the use of mediation analyses with nationally representative samples can
provide new knowledge of the relative contributions of various stressors to population-level racial health disparities,
which can inform policy and intervention efforts.
Data and Methods
Aims of the Present Study
Measures
While research documents race differences in health and
physiological risk and provides strong evidence of racial
disparities in stress exposure, these two bodies of literature
have not been fully integrated to assess whether cumulative
stress exposure is a key mechanistic pathway producing
race difference in biological aging and physiological
functioning in later life. This study uses nationally representative, longitudinal data to assess relative contributions
of various measures of stress exposure to Black–White
disparities in inflammatory and metabolic risk in mid to
late life by addressing four overarching research questions:
1. What is the patterning of Black–White disparities in
stress exposure and physiological well-being in late life?
2. How do diverse measures of stress exposure relate to
markers of inflammatory and metabolic risk at older
ages?
3. Do racial differences in stress exposure account for
Black–White disparities in physiological well-being in
mid to late life?
4. Do composite measures of cumulative stress exposure,
quantified by summing high risk cut off points or by
using factor analysis, account for more of the racial
gaps in physiological functioning than the domainspecific stressors?
Data and Samples
Data for this study come from five waves of the HRS, a
nationally representative, longitudinal study of adults aged
50 years and older in the United States. The HRS collects
information about the well-being of older adults, primarily through the use of in-home interviews. More information about the design of HRS can be found in Heeringa
and Connor (1995). The HRS collected blood-based
biomarkers on a random half of the sample in 2006, and
the other half of the sample provided biomarker data in
2008. These sample respondents were then reinterviewed
in 2010 and 2012, respectively, when they again provided
biomarker samples. Measures of stress exposure were collected in a leave behind questionnaire conducted every
other year starting in 2004, though respondents did not answer questions about stress exposure at every one of these
waves. This study uses data from five waves of the HRS
that include data on physiological functioning and stress
exposure: 2004, 2006, 2008, 2010, and 2012.
The analytic samples include Black and White
respondents aged 50 years and older with valid sampling
weights. Supplementary analyses revealed that the greatest
source of missing data among eligible respondents was
stress exposure. I use multiple imputation by chained equations (MICE) procedures to adjust for missing stress exposure data (10 multiply imputed data sets).
Outcomes
This study includes two outcomes from the 2010/2012
waves of the HRS that represent critical markers of health
that are strongly predictive of disease risk in later life with
documented links to stress exposure. C-reactive protein
(CRP), a marker of inflammatory response and immune
function, is an acute-phase protein produced by the liver.
Elevated levels of circulating CRP indicate systemic inflammation (Finch, 2010). Studies document a relationship between inflammation and health risk, including prospective
associations of CRP with higher rates of coronary heart
disease and mortality risk (Harris et al., 1999). Because of
a skewed distribution, I include CRP as a log-transformed
measure.
Consistent with previous studies (Yang, Gerken,
Schorpp, Boen, & Harris, 2017), I construct a composite
measure of metabolic dysregulation, which indicates
overall level of metabolic burden using clinical markers.
For each individual metabolic marker, I construct a dummy
measure where “1” indicates high risk, with cut points for
high risk defined by clinical practice (blood pressure: systolic blood pressure > 140 mmHG or diastolic blood pressure > 90 mmHg; Hba1c ≥ 5.7%; waist circumference ≥
102 cm for males or 88 cm for females; total cholesterol ≥
240 mg/dL; high-density lipoprotein cholesterol ≤ 40 mg/
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
dL; and body mass index ≥ 30 kg/m2). I then summed the
scores from each of the markers to construct the index of
overall metabolic dysregulation, which ranges from 0 (low)
to 6 (high).
1941
education (1 = <high school, 2 = high school, 3 = some
college, 4 = bachelor’s degree or higher), total household
wealth (continuous), total household income (continuous),
and marital status (1 = married, 2 = partnered, 3 = separated
or divorced, 4 = widowed, and 5 = never married).
Key Explanatory Variables
Key explanatory variables for this study include a number
of measures indicating stress exposure. I utilize psychosocial stress exposure data from the 2004, 2006, and 2008
waves of the HRS, though respondents did not answer
questions about stress exposure at every wave.
First, I constructed standardized count indices of the
following domain-specific stressors, corresponding to
stress exposure in different domains of social life: lifetime
traumas (Krause, Shaw, & Cairney, 2004), recent stressful
life events (Turner, 2013), financial strain (Campbell,
Converse, & Rodgers, 1976), ongoing chronic strains
(Troxel, Matthews, Bromberger, & Sutton-Tyrrell, 2003),
everyday discrimination (Williams et al., 1997), and major
lifetime discrimination (Williams et al., 1997). These widely
used measures include both recent and lifetime stressors
and discrimination-related and generalized stressors.
For detailed information on the construction of the individual stress measures, see Supplementary Appendix A.
For individuals with valid stress exposure data in multiple
waves, stress exposure was included as average exposure
across the waves. Supplementary analyses utilizing the
most recent wave of stress exposure for each respondent
produced substantively similar results.
Next, I created two composite measures of stress exposure. Consistent with previous studies (Strenthal et al.,
2011), I created a measure of high risk stress burden,
which indicates the number of stressors for which the respondent is in the highest quartile (0 = low cumulative
stress, 1 = high risk on one stressor; 2 = high risk on two
stressors; 3 = high risk on three stressors; 4 = high risk
on four or more stressors). Next, I used factor analysis to
create a composite measure of cumulative stress burden
using the individual domain-specific stressors. Results
from the factor analyses revealed a one factor structure
for total stress burden (Eigen value = 2.273) with strong
item-rest correlations for all of the individual stress measures (all above 0.50), indicating strong internal consistency. While similar to the high risk stress burden measure,
the cumulative stress burden measure does not use cutoffs
and places no restrictions on reporting high levels of stress
but instead reflects total stress exposure across domains
of stress.
Other Measures
The racial disparity in the outcomes is measured by
a dummy variable, where “1” indicates Black. Other
covariates include age (continuous), gender (1 = female),
Analytic Strategy
First, I use descriptive statistics to assess Black–White
disparities in the measures of health, stress exposure, and
other characteristics, using t tests (two-tailed) and chisquare tests to formally assess race differences. Next, I use
multivariate OLS regression analyses to model the prospective associations between the stressors and the measures of CRP and metabolic dysregulation. At the time of
analysis, the HRS had released only two complete waves
of biomarker data, collected at relatively short intervals,
which prevented modeling racial disparities in trajectories
of CRP or metabolic dysregulation. As such, for both
outcomes, I exploit the temporal sequencing of the data
by modeling the outcomes in 2010–2012 as a function of
stress exposure and other covariates in 2004–2008, based
on when respondents had valid data. Supplementary
analyses with the metabolic dysregulation outcome indicated that the results were robust to alternative modeling strategies, including Poisson and negative binomial
regression.
The models for all outcomes proceed in a stepwise
fashion. Model 1 adjusts for age, gender, and race and
provides evidence of the unadjusted race gaps in the
outcomes; Model 2 builds on Model 1 by also including
the socioeconomic measures; and Models 3–8 include
each of the domain-specific stressors individually. Model 9
includes the composite measure of high risk stress burden,
and Model 10 includes the measure of cumulative stress
burden. In this way, results from Models 3 to 10 indicate
the associations of the stressors with the outcomes and
the contribution of the stress measures to Black–White
health disparities net of race differences in socioeconomic
status (SES), which is a conservative approach given both
the strong racial patterning of SES and the documented
associations between SES and stress (Lantz, House, Mero,
& Williams, 2005).
In reporting the multivariate results, I pay particular attention to whether inclusion of the stressor measures in the
models reduces the racial disparity in the markers of health.
To formally test whether the stress measures help to “explain” the race gaps in CRP and metabolic dysregulation,
I compute direct and indirect effects for each imputed
dataset, combine the estimates using Rubin’s rules for combining estimates from multiple imputed datasets (Rubin,
2004), and finally examine the equality of coefficients
across models (Preacher & Hayes, 2008). All descriptive
statistics and multivariate model estimates are weighted to
adjust for survey design effects and nonresponse.
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
1942
Results
Descriptive Statistics
Table 1 presents descriptive statistics by race, and results
indicate that Blacks have greater physiological risk than
Whites, including higher levels of CRP (p < .001) and metabolic dysregulation (p < .001). Blacks also report more psychosocial stress exposure than Whites across virtually all
domains of stress, with the exception of lifetime traumas.
Table 1 also reveals racial disparities in the composite
stress measures, with Blacks reporting more high risk stress
burden (p < .001) and cumulative stress burden (p < .001)
than Whites. Further, compared to Whites, Blacks also experience more socioeconomic disadvantage in mid to late
life, with Blacks having lower levels of education, wealth,
and income (p < .001) than Whites.
Multivariate Models
CRP
Results in Table 2 document the prospective associations
between stress exposure and log CRP (N = 7,280). Table 2
also shows the results of the mediation analyses that indicate the proportion of the race gap in log CRP “explained”
Table 1. Descriptive Statistics by Race (HRS 2004–2012) (N = 7,280)
Outcomes
Log C-reactive protein
Metabolic dysregulation
Domain-specific stressors
Lifetime traumas
Stressful events
Financial strain
Ongoing chronic strains
Everyday discrimination
Major discrimination
Composite stressors
High risk stress burden
High on 0 stressors
High on 1 stressor
High on 2 stressors
High on 3 stressors
High on 4 or more stressors
Cumulative stress burden
Sociodemographic characteristics
Age
Gender (1 = female)
Socioeconomic factors
Education
<High school
High school
Some college
BA+
Total household wealth
Total household income
Marital status
Married
Partnered
Separated or divorced
Widowed
Never married
Full sample
Whites
Blacks
(N = 7,280)
(N = 6,276)
(N = 1,004)
Mean/Prop.
Mean/Prop.
Mean/Prop.
p-value
1.173
2.464
1.150
2.410
1.385
2.988
<.001
<.001
0.160
0.058
2.456
1.521
1.682
0.091
0.160
0.057
2.417
1.514
1.664
0.086
0.165
0.072
2.821
1.591
1.849
0.142
.420
.015
<.001
.003
<.001
<.001
0.301
0.272
0.189
0.12
0.118
0.08
0.313
0.277
0.186
0.115
0.109
0.047
0.193
0.221
0.223
0.160
0.203
0.390
<.001
63.924
0.541
64.119
0.533
62.117
0.615
<.001
<.001
0.012
0.355
0.253
0.277
567,240
75,482
0.097
0.358
0.255
0.290
612,905
79,241
0.283
0.327
0.232
0.157
144,040
40,647
<.001
0.643
0.037
0.137
0.147
0.036
0.672
0.036
0.120
0.141
0.031
0.379
0.051
0.294
0.197
0.079
<.001
<.001
<.001
<.001
Note: Weighted descriptive statistics. Sample sized based on CRP analytic sample, with exception of metabolic dysregulation outcome (N = 6,542). p-value of t test
(two-tailed test) or chi-square test indicating race difference in mean/proportion.
HRS = Health and Retirement Study.
Gender (1 = female)
Proportion race gap mediated by
stressor(s)a
Sociodemographic characteristics
Age
Cumulative stress burden
High on 4 or more stressors
High on 3 stressors
High on 2 stressors
Composite stressors
High risk stress burden (high on 0
stressors is ref.)
High on 1 stressor
Major discrimination
Everyday discrimination
Ongoing chronic strains
Financial strain
Stressful life events
Domain-specific stressors
Lifetime traumas
Racial disparity
Race (1 = Black)
(SE)
(SE)
−0.001
(0.001)
0.105***
(0.021)
-
−0.003*
(0.001)
0.072***
(0.022)
-
0.142***
(0.036)
Coeff.
Coeff.
0.225***
(0.034)
Model 2
Model 1
−0.002*
(0.001)
0.079***
(0.022)
NS
0.328***
(0.081)
0.145***
(0.035)
(SE)
Coeff.
Model 3
Table 2. Stress Exposure and Log C-Reactive Protein (N = 7,280)
−0.002†
(0.001)
0.073***
(0.022)
NS
0.107
(0.096)
0.142***
(0.036)
(SE)
Coeff.
Model 4
−0.002
(0.001)
0.072***
(0.022)
0.033*
0.033*
(0.013)
0.137***
(0.036)
(SE)
Coeff.
Model 5
−0.002
(0.001)
0.067**
(0.022)
NS
0.097***
(0.026)
0.142***
(0.036)
(SE)
Coeff.
Model 6
−0.002
(0.001)
0.077***
(0.022)
0.023*
0.029†
(0.015)
0.138***
(0.036)
(SE)
Coeff.
Model 7
−0.002†
(0.001)
0.079***
(0.022)
0.072**
0.198**
(0.074)
0.133***
(0.036)
(SE)
Coeff.
Model 8
−0.001
(0.001)
0.077***
(0.022)
0.066***
0.007
(0.030)
0.066†
(0.036)
0.078
(0.050)
0.134**
(0.042)
0.133***
(0.036)
(SE)
Coeff.
Model 9
−0.001
(0.001)
0.079***
(0.022)
0.058***
(0.013)
0.067***
0.134***
(0.036)
(SE)
Coeff.
Model 10
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
1943
Coeff.
(SE)
Coeff.
(SE)
−0.038
(0.051)
0.094**
(0.034)
0.025
(0.031)
−0.026
(0.061)
0.093**
(0.034)
−0.050†
(0.027)
−0.178***
(0.027)
−0.000***
(0.000)
−0.000*
(0.000)
Model 2
Model 1
−0.045
(0.051)
0.083*
(0.034)
0.015
(0.031)
−0.016
(0.061)
0.085*
(0.033)
−0.055*
(0.027)
−0.175***
(0.027)
−0.000***
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 3
−0.040
(0.051)
0.090**
(0.034)
0.023
(0.031)
−0.028
(0.061)
0.093**
(0.034)
−0.052†
(0.027)
−0.179***
(0.027)
−0.000***
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 4
−0.042
(0.051)
0.085*
(0.035)
0.019
(0.031)
−0.027
(0.061)
0.091**
(0.034)
−0.049†
(0.027)
−0.175***
(0.027)
−0.000**
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 5
−0.046
(0.051)
0.082*
(0.035)
0.016
(0.031)
−0.023
(0.061)
0.092**
(0.034)
−0.051†
(0.027)
−0.177***
(0.027)
−0.000**
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 6
−0.041
(0.051)
0.091**
(0.034)
0.023
(0.031)
−0.024
(0.061)
0.092**
(0.034)
−0.050†
(0.027)
−0.177***
(0.027)
−0.000***
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 7
−0.043
(0.051)
0.086*
(0.034)
0.023
(0.031)
−0.032
(0.061)
0.094**
(0.034)
−0.056*
(0.027)
−0.183***
(0.027)
−0.000***
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 8
−0.053
(0.051)
0.077*
(0.035)
0.015
(0.031)
−0.024
(0.061)
0.090**
(0.034)
−0.055*
(0.027)
−0.178***
(0.027)
−0.000**
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 9
Note: Results based on OLS regression models. CRP was assessed in 2010–2012; all other covariates were measured in 2004–2008. Model estimates are weighted. HS = High school.
a
Proportion of racial disparity in CRP observed in Model 2 “explained away” by the stress measure(s) included in the model. p-value of mediation test. “NS” indicates no statistically significant mediation.
***p < .001, **p < .01, *p < .05, †p < .1.
Never married
Widowed
Separated or divorced
Marital status (Married is
reference)
Partnered
Total household income
Total household wealth
BA+
Some college
Socioeconomic factors
Education (HS is reference)
<HS
Table 2. Continued
−0.050
(0.051)
0.073*
(0.035)
0.012
(0.031)
−0.026
(0.061)
0.090**
(0.033)
−0.056*
(0.027)
−0.179***
(0.027)
−0.000**
(0.000)
−0.000*
(0.000)
(SE)
Coeff.
Model 10
1944
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
by the stress measures included in the respective models.
As indicated by the coefficient for race, the Black–White
disparity in inflammation is largest in Model 1. Including
the measures of SES in Model 2 reduces the racial disparity
over Model 1. Models 3–8 introduce each of the domainspecific stressors in a stepwise fashion. In several cases, the
Black–White disparity in log CRP is further attenuated by
the inclusion of the stress exposure measures in the models,
net of SES.
Model 3 indicates that lifetime traumas are positively
associated with inflammation (0.328, p < .001), such that
individuals who reported more traumas over the course of
their lives had higher CRP. However, results of mediation
analyses reveal that differences in lifetime traumas did not
account for a significant portion of the Black–White disparity in CRP. Model 4 reveals no significant association
between stressful life events and log CRP. Results from
Models 5 and 6 show that financial strain (0.033, p = .012)
and ongoing chronic strain (0.097, p < .001) are positively associated with inflammation. Mediation analyses
show that including the measure of financial strain in
Model 5 reduces the race gap in log CRP by approximately
3.4% (p = .015), net of SES. This suggests that, even after
adjusting for racial differences in socioeconomic resources,
Black–White differences in financial stress account for a
portion of the racial disparity in physiological inflammation. Results in Models 7 and 8 of Table 2 indicate that
everyday (0.029, p = .063) and major life discrimination
(0.198, p = .007) are positively associated with inflammation. Further, results from mediation analyses indicate that
including the measures of everyday and major life discrimination in the models reduces the race gap in inflammation
by 2.3 (p = .043) and 7.2% (p = .001), respectively.
Models 9–10 include the composite stress measures. The
measure of high risk stress burden is positively associated
with CRP, indicating that, compared to individuals who report lower levels of stress, individuals reporting high stress
across domains of stress are at increased inflammatory risk.
Mediation analyses reveal that Black–White differences in
high risk stress burden account for approximately 6.6% of
the race gap in CRP (p < .001). The measure of cumulative
stress burden is also associated with increased inflammation. Including the measure of cumulative stress burden in
Model 10 results in a 6.7% reduction of the race gap in
CRP (p < .001).
Metabolic dysregulation
Table 3 displays results of the metabolic dysregulation
models (N = 6,452). The racial disparity in metabolic risk
is greatest in Model 1 and is attenuated after adjusting
for SES in Model 2. Results from Models 3–8 reveal that,
with the exception of major life discrimination, all domainspecific stressors are positively associated with metabolic
dysregulation. Further, results from mediation analyses
indicate that racial differences in financial strain and everyday discrimination are drivers of the Black–White gap
1945
in metabolic risk net of racial differences in SES. Lifetime
traumas, stressful life events, and ongoing chronic strains—
while prospectively associated with metabolic risk—do not
account for significant portions of the Black–White gap in
metabolic risk in late life.
Results from Models 9 indicate that high risk stress
burden is positively associated with metabolic dysregulation,
particularly for those who reported high levels of exposure
on four or more of the stressors, and that race differences
in high risk stress burden partially “explain” the Black–
White gap in metabolic risk. Finally, in Model 10, the
composite measure of cumulative stress burden is strongly
associated with metabolic dysregulation (0.090, p < .001)
and accounts for approximately 3% of the Black–White
metabolic gap (p = .004), net of SES. Across the models in
Table 3, the race gap in metabolic risk is smallest in Models
7 and 10, which include the measures of everyday discrimination and cumulative stress burden, respectively.
Discussion
The divergence of Black–White health gaps through mid
and late life has led scholars to hypothesize that Blacks may
experience accelerated aging and physiological deterioration relative to Whites due to racial differences in social
exposures, including stress (Geronimus et al., 2010). Still,
while research documents that Blacks experience greater
levels of stress exposure than Whites (Turner & Avison,
2003) and that repeated exposure to stress can increase
disease risk (Cohen et al., 2012; Thoits, 2010), these two
bodies of literature have not been fully integrated to adequately assess the role of cumulative stress exposure in
Black–White health gaps in later life. Using nationally representative, longitudinal data and diverse measures of physiological functioning and stress exposure, this study offers
three contributions to understanding of the links between
racism, stress, and health inequality from mid to late life.
First, results provide strong evidence of the racial
patterning of physiological well-being and stress exposure
at older ages. Consistent with previous research (Geronimus
et al., 2010; Mitchell & Aneshensel, 2017; Mitchell et al.,
2019), results from both the descriptive and multivariate
analyses indicated that older age Blacks have higher levels
of systemic inflammation and metabolic dysregulation
than Whites. These results are consistent with the notion of
“weathering” (Geronimus et al., 2010), whereby older aged
Blacks in the United States experience accelerated physiological dysregulation compared to Whites in ways that
relate to population patterns of health and disease risk in
later life. Also, consistent with previous research (Brown &
Hargrove, 2018; Strenthal et al., 2011), results indicated
that Blacks experience more cumulative stress burden than
Whites, including more stressful events, financial strain,
ongoing chronic strain, and everyday and major life discrimination. In this way, findings show that racism, as a
system of oppression and domination, patterns exposure
Proportion race gap mediated
by stressor(s)a
Cumulative stress burden
High on 4 or more
stressors
High on 3 stressors
High on 2 stressors
Composite stressors
High risk stress burden
(high on 0 stressors is ref.)
High on 1 stressor
Major discrimination
Everyday discrimination
Ongoing chronic strains
Financial strain
Stressful life events
Domain-specific stressors
Lifetime traumas
Racial disparity
Race (1 = Black)
(SE)
(SE)
-
-
0.493***
(0.060)
Coeff.
Coeff.
0.566***
(0.058)
Model 2
Model 1
NS
0.285*
(0.144)
0.496***
(0.060)
(SE)
Coeff.
Model 3
Table 3. Stress Exposure and Metabolic Dysregulation (N = 6,452)
NS
0.359†
(0.185)
0.492***
(0.060)
(SE)
Coeff.
Model 4
0.016†
0.048†
(0.026)
0.486***
(0.060)
(SE)
Coeff.
Model 5
NS
0.113*
(0.051)
0.493***
(0.060)
(SE)
Coeff.
Model 6
0.024**
0.101***
(0.029)
0.479***
(0.060)
(SE)
Coeff.
Model 7
NS
0.199
(0.144)
0.484***
(0.061)
(SE)
Coeff.
Model 8
0.017†
(0.081)
0.078
(0.055)
0.029
(0.062)
0.139
(0.093)
0.172*
0.483***
(0.060)
(SE)
Coeff.
Model 9
0.090***
(0.025)
0.029**
0.480***
(0.060)
(SE)
Coeff.
Model 10
1946
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
−0.135
(0.112)
−0.109†
(0.061)
0.010
(0.053)
−0.069
(0.120)
−0.141
(0.112)
−0.117†
(0.061)
0.001
(0.053)
−0.060
(0.120)
−0.137
(0.112)
−0.122*
(0.061)
0.004
(0.053)
−0.078
(0.121)
0.106†
(0.060)
−0.047
(0.049)
−0.236***
(0.052)
−0.000***
(0.000)
0.000
(0.000)
−0.005*
(0.002)
0.050
(0.040)
(SE)
Coeff.
Model 4
−0.140
(0.113)
−0.119†
(0.061)
0.001
(0.053)
−0.070
(0.120)
0.101†
(0.059)
−0.041
(0.049)
−0.227***
(0.052)
−0.000**
(0.000)
0.000
(0.000)
−0.005*
(0.002)
0.045
(0.040)
(SE)
Coeff.
Model 5
−0.145
(0.113)
−0.121*
(0.061)
0.001
(0.053)
−0.066
(0.120)
0.103†
(0.059)
−0.044
(0.049)
−0.232***
(0.051)
−0.000***
(0.000)
0.000
(0.000)
−0.005*
(0.002)
0.040
(0.040)
(SE)
Coeff.
Model 6
−0.146
(0.113)
−0.117†
(0.061)
0.007
(0.053)
−0.065
(0.120)
0.102†
(0.059)
−0.042
(0.049)
−0.231***
(0.051)
−0.000***
(0.000)
0.000
(0.000)
−0.004†
(0.002)
0.060
(0.040)
(SE)
Coeff.
Model 7
−0.140
(0.112)
−0.116†
(0.061)
0.007
(0.053)
−0.076
(0.121)
0.106†
(0.060)
−0.048
(0.049)
−0.237***
(0.052)
−0.000***
(0.000)
0.000
(0.000)
−0.006*
(0.002)
0.053
(0.040)
(SE)
Coeff.
Model 8
−0.151
(0.112)
−0.127*
(0.061)
−0.001
(0.053)
−0.071
(0.121)
0.102†
(0.059)
−0.049
(0.049)
−0.234***
(0.051)
−0.000***
(0.000)
0.000
(0.000)
−0.004*
(0.002)
0.052
(0.040)
(SE)
Coeff.
Model 9
−0.153
(0.112)
−0.138*
(0.061)
−0.009
(0.053)
−0.071
(0.121)
0.101†
(0.059)
−0.051
(0.049)
−0.235***
(0.051)
−0.000**
(0.000)
0.000
(0.000)
−0.003
(0.002)
0.055
(0.040)
(SE)
Coeff.
Model 10
Note: Results based on OLS regression models. Metabolic dysregulation was assessed in 2010–2012; all other covariates were measured in 2004–2008. Model estimates are weighted. HS = High school.
a
Proportion of racial disparity in metabolic dysregulation observed in Model 2 “explained away” by the stress measure(s) included in the model. p-value of mediation test. “NS” indicates no statistically significant mediation.
***p < .001, **p < .01, *p < .05, †p < .1.
Never married
Widowed
Separated or divorced
Marital status (married is
reference)
Partnered
Total household income
Total household wealth
BA+
Some college
0.098
(0.060)
−0.046
(0.049)
−0.229***
(0.052)
−0.000***
(0.000)
0.000
(0.000)
(SE)
0.104†
(0.060)
−0.042
(0.049)
−0.231***
(0.052)
−0.000***
(0.000)
0.000
(0.000)
(SE)
(SE)
Coeff.
−0.006**
(0.002)
0.051
(0.040)
Coeff.
Coeff.
Model 3
−0.006**
(0.002)
0.046
(0.040)
Model 2
Model 1
Sociodemographic characteristics
Age
−0.004†
(0.002)
Gender (1 = female)
0.072†
(0.038)
Socioeconomic factors
Education (HS is reference)
<HS
Table 3. Continued
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
1947
1948
to a variety of acute and chronic stressors and strains—
both discrimination-related and more generalized—across
domains of social life in ways that produce stark racial
disparities in cumulative life course stress exposure.
Second, results in Tables 2 and 3 offer robust, consistent
evidence of the prospective associations between stress exposure and physiological well-being, where stress exposure
is consistently linked to greater health risk in mid to late
life. While previous research has relied largely on cross-sectional data, this study used lagged stress exposure measures
to offer a more rigorous examination of the role of stress
in shaping trajectories of health. Further, while extant research on stress and health has been largely restricted by
the use of single health or disease outcomes, findings from
this study indicate that stress exposure contributes to population health inequality through numerous physiological
processes. Both systemic inflammation and metabolic function have been increasingly identified as critical pathogenic
pathways shaping disease and mortality risk across the life
course (Finch, 2010; Yang & Kozloski, 2011), and findings
from this study consistently indicated that these physiological processes may be particularly critical pathways through
which social stressors “get under the skin” to produce population health disparities from mid through late life.
Third, this study is among the first to use formal
mediation analyses to document the unique and joint
contributions of various stress exposure measures to racial
gaps in physiological functioning and provides new evidence of the role of cumulative stress exposure in producing
Black–White gaps in physiological risk in late life. Results
from the mediation analyses indicated that Black–White
disparities in exposure to financial strain, everyday discrimination, major life discrimination, and the composite
stress measures, in particular, helped to “explain away”
significant portions of the racial gaps in the outcomes.
Importantly, racial differences in the composite indicator
of cumulative stress burden accounted for the greatest proportion of the race gaps in metabolic risk and also played a
prominent role in contributing to Black–White gaps in inflammation. Previous research on the links between stress
and health generally examines one stressor at a time, despite evidence that stressors often co-occur, meaning that
studies that utilize single stressors risk overestimating the
association between particular stressors and health (Green
et al., 2010) and underestimating the role of cumulative
stress burden in the production of health inequality. By
contrast, composite indicators of stress exposure may
better reflect the co-occurrence of stressors and, consistent
with cumulative inequality theory (Ferraro & Shippee,
2009), may more effectively capture how the accumulation of stressors over time and across domains of social
life collectively contribute to racial disparities in health
among older adults than any single stressor measure.
Results from this study indeed showed that Black–White
differences in cumulative stress burden as indicated by
the factor score mediated the greatest proportion of
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
racial disparities in both outcomes. Importantly, the proportion of the Black–White gap mediated by cumulative
stress exposure was more than twice as high for CRP
than metabolic dysregulation, suggesting that inflammatory processes may play a particularly prominent role in
linking racial disparities in cumulative stress to Black–
White gaps in health and disease risk. While stress exposure played a prominent role in producing racial gaps
in CRP, the proportion of the race gap in metabolic risk
mediated by the stressors was more modest, indicating that
other factors—such as health behaviors—may play a more
central role in shaping metabolic risk. Taken together, the
results presented here show that the physiological toll of
the “thousand small cuts” resulting from the accumulation
of stress exposure across domains of social life throughout
the life course—including experiencing multiple traumatic
and stressful events as well as repeated exposure to more
chronic stressors and strains—can accelerate the physiological functioning of older age Blacks relative to Whites. It
is worth noting that the contribution of stress exposure to
Black–White health gaps persisted net of racial differences
in socioeconomic factors, which is a relatively conservative test for mediation, given that both stress exposure is
patterned by SES (Turner & Avison, 2003) and that SES is
strongly patterned by race (Boen, 2016).
This study is not without limitations, which should
be addressed in future research. First, while the current
study exploited the temporal sequencing of the data by
assessing how stress exposure at baseline was associated
with inflammatory and metabolic risk at follow-up, future
work utilizing multiple waves of biomarkers will improve
causal inference and further elucidate how stress exposure
shapes trajectories of health inequality as individuals age.
In supplementary analysis, I ran lagged dependent variable models, regressing the outcomes in 2010/2012 on
the outcomes measured at baseline, while also adjusting
for the covariates. I found minimal evidence that baseline
stress exposure contributed to race differences in physiological changes over the period, though four years may be
too short of a time period for stress exposure to account
for Black–White differences in change in the biomarkers.
Still, more research utilizing longitudinal biomarker data
is needed. Second, while the present study integrates multiple measures of life course stress exposure, there are
other stressors and traumas not included here that may
play a role in Black–White health gaps, including vicarious
stressors related to the experiences of family and friends
(Williams & Mohammed, 2009) and neighborhood level
stressors (Clarke et al., 2014). The proportion of the race
gaps mediated by the stressors included in the present study
is modest, so more research on how the accumulation of
other stressors may also contribute to Black–White health
gaps is needed. Finally, future research should consider the
psychological, emotional, and behavioral mechanisms that
underlie the links between stress exposure and physiological functioning.
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
Taken together, these results indicate that studies of racial
health inequality should consider the role of Black–White
differences in cumulative stress burden—rather than race
differences in single measures of domain-specific stressors—in
order to more accurately capture the role of stress in racial
health inequality. Failure to measure race differences in the
accumulation of stress exposure across domains of social life
in the study of Black–White health disparities can result in
overestimating the role of any singular stressor in racial health
gaps and underestimating the total contribution of stress to racial health disparities. Importantly, the findings presented here
also suggest that studies of Black–White health disparities that
do not attempt to comprehensively measure the accumulation
of recent and lifetime traumas and events, chronic stressors
and strains, and discrimination-related stress—or even fail to
account for stress exposure at all—may risk overestimating
the race residual, or the “unexplained” racial health gap,
which is not without implications. In particular, when studies
are left with significant, unaccounted for racial disparities in
health, authors speculate about the potential explanations for
these “unexplained” gaps. As such, there has been a resurgence of biological and genetic explanations for Black–White
health differences (see Roberts, 2013 for a review), which is
concerning in the face of overwhelming evidence that social
explanations for racial health gaps have been underestimated.
This study, then, serves as both essential evidence of the key
role of stress exposure in racial health inequality and as further
proof that more research on the social origins—including the
psychosocial determinants—of racial health gaps is needed.
By linking Black–White gaps in systemic inflammation and
metabolic dysregulation to racial disparities cumulative exposure to social stressors, findings from this study indicate that
Black–White disparities in physiological functioning observed
from mid through late life in part reflect the accumulation of
racialized social stressors that occur across domains of life and
across the entire life span. The “thousand cuts” that Blacks experience across the life course as a result of living in a racially
stratified society play a role in producing their accelerated physiological deterioration in mid to late life relative to Whites.
The racial patterning of stress exposure extends beyond explicit incidences of racial discrimination to include a variety of
stressors and strains across domains of social life. As such, reducing racial health disparities at older ages will require policy
and intervention efforts targeting both the material and psychosocial factors underlying population health inequality across the
life course.
Supplementary Material
Supplementary data is available at The Journals of
Gerontology, Series B: Psychological Sciences and Social
Sciences online.
Funding
This publication was made possible by funding from the
Population Research Infrastructure Program of the National
Institutes of Health’s (NIH)’s Eunice Kennedy Shriver
1949
National Institute of Child Health and Human Development
awarded to the Population Studies Center at the University of
Pennsylvania, NIH grant number: R24 HD044964.
Acknowledgments
I would like to thank Y. Claire Yang, Karolyn Tyson,
Robert Hummer, Kathleen Mullan Harris, and Anthony
Perez for their feedback on earlier drafts of the paper. I also
want to thank the anonymous reviewers for their insightful
comments and suggestions.
Author contributions
C. Boen planned the study, conducted the data analysis,
and wrote the paper.
Conflict of Interest
None reported.
References
Aneshensel, C. S., Rutter, C. M., & Lachenbruch, P. A. (1991). Social
structure, stress, and mental health: Competing conceptual and
analytic models. American Sociological Review, 56, 166–178.
doi:10.2307/2095777
Boen, C. (2016). The role of socioeconomic factors in black-white
health inequities across the life course: Point-in-time measures,
long-term exposures, and differential health returns. Social Science
& Medicine, 170, 63–76. doi:10.1016/j.socscimed.2016.10.008
Bonilla-Silva, E. (1997). Rethinking racism: Toward a structural
interpretation. American Sociological Review, 62, 465–480.
doi:10.2307/2657316
Brown, T. H., & Hargrove, T. W. (2018). Psychosocial mechanisms
underlying older black men’s health. The Journals of Gerontology,
Series B: Psychological Sciences and Social Sciences, 73, 188–
197. doi:10.1093/geronb/gbx091
Campbell, A., Converse, P. E., & Rodgers, W. (1976). The quality of
American life: Perceptions, evaluations, and satisfactions. New
York, NY: Russell Sage Foundation.
Clarke, P., Morenoff, J., Debbink, M., Golberstein, E.,
Elliott, M. R., & Lantz, P. M. (2014). Cumulative exposure
to neighborhood context consequences for health transitions
over the adult life course. Research on Aging, 36, 115–142.
doi:10.1177/0164027512470702
Cohen, S., Janicki-Deverts, D., Doyle, W. J., Miller, G. E., Frank, E.,
Rabin, B. S., & Turner, R. B. (2012). Chronic stress, glucocorticoid
receptor resistance, inflammation, and disease risk. Proceedings of
the National Academy of Sciences of the United States of America,
109, 5995–5999. doi:10.1073/pnas.1118355109
Ferraro, K. F., & Shippee, T. P. (2009). Aging and cumulative inequality: How does inequality get under the skin? The
Gerontologist, 49, 333–343. doi:10.1093/geront/gnp034
Finch, C. E. (2010). The biology of human longevity: Inflammation,
nutrition, and aging in the evolution of lifespans. Burlington,
MA: Academic Press.
Geronimus, A. T., Hicken, M. T., Pearson, J. A., Seashols, S. J.,
Brown, K. L., & Cruz, T. D.. (2010). Do US black women
1950
experience stress-related accelerated biological aging? Human
Nature, 21, 19–38. doi:10.1007/s12110-010-9078-0
Goosby, B. J., Straley, E., & Cheadle, J. E. (2017). Discrimination,
sleep, and stress reactivity: Pathways to African American-white
cardiometabolic risk inequities. Population Research and Policy
Review, 36, 699–716. doi:10.1007/s11113-017-9439-z
Green, J. G., McLaughlin, K. A., Berglund, P. A., Gruber, M. J.,
Sampson, N. A., Zaslavsky, A. M., & Kessler, R. C. (2010).
Childhood adversities and adult psychiatric disorders in the
national comorbidity survey replication I: Associations with first
onset of DSM-IV disorders. Archives of General Psychiatry, 67,
113–123. doi:10.1001/archgenpsychiatry.2009.186
Harris, T. B., Ferrucci, L., Tracy, R. P., Corti, M. C., Wacholder, S.,
Ettinger, W. H., Heimovitz, H., Cohen, H. J., & Wallace, R.
(1999). Associations of elevated interleukin-6 and
C-reactive protein levels with mortality in the elderly. The
American Journal of Medicine, 106, 506–512. doi:10.1016/
S0002-9343(99)00066-2
Heeringa, S. G., & J. H. Connor. (1995). Technical description of
the health and retirement survey sample design. Ann Arbor, MI:
University of Michigan.
Jones, C. P. (2000). Levels of racism: A theoretic framework and a
gardener’s tale. American Journal of Public Health, 90, 1212–
1215. doi:10.2105/ajph.90.8.1212
Krause, N., Shaw, B. A., & Cairney, J. (2004). A descriptive epidemiology of lifetime trauma and the physical health
status of older adults. Psychology and Aging, 19, 637–648.
doi:10.1037/0882-7974.19.4.637
Lantz, P. M., House, J. S., Mero, R. P., & Williams, D. R. (2005). Stress,
life events, and socioeconomic disparities in health: Results from
the Americans’ Changing Lives Study. Journal of Health and Social
Behavior, 46, 274–288. doi:10.1177/002214650504600305
McEwen, B. S. (2007). Physiology and neurobiology of stress and
adaptation: Central role of the brain. Physiological reviews, 87,
873–904. doi:10.1152/physrev.00041.2006
Mitchell, U. A., & Aneshensel, C. S. (2017). Social inequalities in
inflammation: Age variations in older persons. Journal of Aging
and Health, 29, 769–787. doi:10.1177/0898264316645546
Mitchell, U. A., Ailshire, J. A., & Crimmins, E. M. (2019). Change
in cardiometabolic risk among blacks, whites, and hispanics:
Findings from the health and retirement study. The Journals of
Gerontology, Series A: Biological Sciences and Medical Sciences,
74, 240–246. doi:10.1093/gerona/gly026
National Center for Health Statistics (US). (2016). Health, United
States, 2015: With Special Feature on Racial and Ethnic Health
Disparities. Hyattsville, MD: U.S. Department of Health and
Human Services, Centers for Disease Control and Prevention,
National Center for Health Statistics.
Pearlin, L. I. (1999). The stress process revisited. In Handbook of the
sociology of mental health. “The stress process revisited” (pp.
395–415). New York, NY: Springer.
Pearlin, L. I., Menaghan, E. G., Lieberman, M. A., & Mullan, J. T.
(1981). The stress process. Journal of Health and Social
Behavior, 22, 337–356. doi: 10.2307/2136676
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling
strategies for assessing and comparing indirect effects in multiple
mediator models. Behavior Research Methods, 40, 879–891.
doi:10.3758/BRM.40.3.879
Journals of Gerontology: SOCIAL SCIENCES, 2020, Vol. 75, No. 9
Roberts, D. (2013). Fatal invention: How science, politics, and big
business re-create race in the twenty-first century. New York,
NY: The New Press.
Rubin, D. B. (2004). Multiple imputation for nonresponse in surveys.
New York: John Wiley and Sons.
Seyle, H. (1974). Stress without distress. Philadelphia, PA: Lippincott
Williams & Wilkins.
Sternthal, M. J., Slopen, N., & Williams, D. R. (2011). Racial
disparities in health: how much does stress really matter?. Du
Bois Review: Social Science Research on Race, 8, 95–113.
doi:10.1017/S1742058X11000087
Thoits, P. A. (2010). Stress and health: Major findings and policy
implications. Journal of Health and Social Behavior, 51 Suppl,
S41–S53. doi:10.1177/0022146510383499
Troxel, W. M., Matthews, K. A., Bromberger, J. T., & SuttonTyrrell, K. (2003). Chronic stress burden, discrimination,
and subclinical carotid artery disease in African American
and Caucasian women. Health Psychology, 22, 300–309.
doi:10.1037/0278-6133.22.3.300
Turner, R. J. (2013). Understanding health disparities the relevance of
the stress process model. Society and Mental Health, 3, 170–186.
doi:10.1177/2156869313488121
Turner, R. J., & Avison, W. R. (2003). Status variations in stress exposure: Implications for the interpretation of research on race,
socioeconomic status, and gender. Journal of Health and Social
Behavior, 44, 488–505. doi:10.2307/1519795
Umberson, D., Olson, J. S., Crosnoe, R., Liu, H., Pudrovska, T.,
& Donnelly, R.. (2017). Death of family members as an
overlooked source of racial disadvantage in the United
States. Proceedings of the National Academy of Sciences of
the United States of America, 114, 915–920. doi:10.1073/
pnas.1605599114
Williams, D. R., Yu, Y., Jackson, J. S., & Anderson, N. B. (1997).
Racial differences in physical and mental health: Socioeconomic status, stress and discrimination. Journal of Health
Psychology, 2, 335–351. doi:10.1177/135910539700200305
Williams, D. R., & Jackson, P. B. (2005). Social sources of racial
disparities in health. Health affairs (Project Hope), 24, 325–334.
doi:10.1377/hlthaff.24.2.325
Williams, D. R., & Mohammed, S. A. (2009). Discrimination and
racial disparities in health: Evidence and needed research.
Journal of Behavioral Medicine, 32, 20–47. doi:10.1007/
s10865-008-9185-0
Williams, D. R., Mohammed, S. A., Leavell, J., & Collins, C.
(2010). Race, socioeconomic status, and health: Complexities,
ongoing challenges, and research opportunities. Annals
of the New York Academy of Sciences, 1186, 69–101.
doi:10.1111/j.1749-6632.2009.05339.x
Yang, Y., & Kozloski, M. (2011). Sex differences in age trajectories
of physiological dysregulation: Inflammation, metabolic syndrome, and allostatic load. The Journals of Gerontology, Series
A: Biological Sciences and Medical Sciences, 66, 493–500.
doi:10.1093/gerona/glr003
Yang, Y. C., Gerken, K., Schorpp, K., Boen, C., & Harris, K. M.
(2017). Early-life socioeconomic status and adult physiological
functioning: A life course examination of biosocial mechanisms.
Biodemography and Social Biology, 63, 87–103. doi:10.1080/1
9485565.2017.1279536