Abstract
African Americans are at heightened risk for coronary heart disease (CHD), with biologic pathways poorly understood. We examined the role of allostatic load (AL) in the association of depressive symptoms with incident CHD among 2,670 African American men and women in the prospective Jackson Heart Study. Depressive symptoms were quantified using the Center for Epidemiologic Studies Depression Scale (CES-D). Incident CHD was ascertained by self-report, death certificate survey, and adjudicated medical record surveillance. Baseline AL was quantified using biologic parameters of metabolic, cardiovascular, immune, and neuroendocrine subsystems and as a combined meta-factor. Sequential models adjusted for demographic, socioeconomic, and behavioral covariates, stratified to examine differences by sex. Greater depressive symptomatology was associated with greater metabolic, cardiovascular, and immune AL (p-values≤0.036) and AL meta-factor z-scores (p=0.007), with findings driven by observations among females. Each 1-point increase in baseline depressive symptomatology, and 1-SD increase in metabolic AL, neuroendocrine AL, and AL meta-factor z-scores was associated with 33%, 88%, 39%, and 130% increases in CHD risk, respectively (p-values <0.001). Neuroendocrine AL and AL meta-factor scores predicted incident CHD among males but not females in stratified analyses. Metabolic AL partially mediated the association of depressive symptoms with incident CHD (5.79% mediation, p=0.044), a finding present among females (p=0.016) but not males (p=0.840). Among African American adults, we present novel findings of an association between depressive symptomatology and incident CHD, partially mediated by metabolic AL. These findings appear to be unique to females, an important consideration in the design of targeted interventions for CHD prevention.
Keywords: Depression, Allostatic load, metabolic, women, coronary heart disease, African American
1. Introduction
Cardiovascular disease (CVD) remains the leading cause of mortality in the U.S. (Kochanek, Murphy, Xu, & Arias, 2017), with African Americans at heightened risk for incident disease and CVD-related death compared to whites (Cunningham et al., 2017; Writing Group Members et al., 2016). Moreover, even when controlling for a number of relevant covariates CVD presents at earlier ages, post-diagnosis survival is shortened, and risk for sudden death is heightened in African American versus white populations (Thomas, K. L., Honeycutt, Shaw, & Peterson, 2010; Thorpe et al., 2016; Zhao et al., 2019). For example, in mediation analyses of 15,069 participants, Zhao et al. (2019) found that 34.7% of the excess risk of sudden cardiac death in African Americans compared to whites remained unaccounted for after adjusting for individual socioeconomic, behavioral, clinical, and biological parameters (e.g., income, education, smoking, diabetes mellitus, body mass index, blood pressure). Racial and ethnic minorities with coronary artery disease are also less likely to receive effective pharmacotherapy (Tran et al., 2017). Though, racial disparities in CVD events have also been noted across treatment tiers among those with similar cardiovascular risk (e.g., greater probability of coronary heart disease [CHD] among hypertensve African Americans versus hypertensive whites receiving one, two, three, and four anti-hypertensive medications (Tajeu, Mennemeyer, Menachemi, Weech-Maldonado, & Kilgore, 2017). Together, such data suggest that traditional and bias-associated risk do not fully account for racial disparities in CVD outcomes. As such, while a substantial literature outlines the presence and persistence of racial disparities in cardiovascular health, the identification of non-traditional risk factors that may contribute to CVD among African Americans remains an important area of inquiry.
In this regard, depressive symptoms have been increasingly recognized as a non-traditional risk factor that may contribute to the development and progression of poor cardiovascular health (Eurelings et al., 2018; Wandell et al., 2018; Wu, Sun, Wang, Li, & Ma, 2018). For example, depressive symptoms have been shown to precede incident CVD during follow-up periods of 5 – 15 years (Eurelings et al., 2018; Wandell et al., 2018). Recent work suggests that depressive symptoms may be a particularly salient predictor of CHD among African Americans compared to non-Hispanic whites (O’Brien et al., 2015). Despite such findings and those suggesting that African Americans may bear a disproportionate burden of depressive symptomatology (Mukherjee, Trepka, Pierre-Victor, Bahelah, & Avent, 2016), including as a function of racial discriminatory exposures (Paradies et al., 2015), studies explicating the link between depressive symptoms and CHD among African Americans have received relatively little attention.
Moreover, depressive symptoms have been linked to dysregulation across a number of biological parameters, namely obesity (Ambrosio et al., 2018; Srinivas, Rajendran, Anand, & Chockalingam, 2018), high blood pressure (Srinivas et al., 2018), and inflammation (Ambrosio et al., 2018), including the new onset or progression of biological dysregulation in prospective investigations. In fact, when such parameters are held constant in aggregate, the magnitude of racial disparities in CVD-related mortality is significantly attenuated (Duru, Harawa, Kermah, & Norris, 2012), suggesting dysregulation of key biological parameters may help explain the association between depressive symptoms and CHD. However, the mediational role of biological subsystems of potential importance in the association between depressive symptoms and CHD among African Americans remains understudied. Similarly, the prevalence of depression is known to differ among men and women and the structure of psychological stress-responsive biological dysregulation appears to differ by sex (Buckwalter et al., 2016; Labaka, Goni-Balentziaga, Lebena, & Perez-Tejada, 2018). For example accumulating evidence suggests that women are more likely to exhibit an inflammatory phenotype than men in the context of major depressive disorder (Labaka et al., 2018). The identification of sex differences in the pathways linking depressive symptoms to incident CHD could provide new opportunity to design tailored preventive interventions.
To address these knowledge deficits and identify novel targets in the prevention of CHD among African Americans, we draw upon the allostatic load (AL) model, which posits that the repeated application of adaptive responses to psychological stress or distress leads to cumulative and quantifiable wear and tear on the body (McEwen, 1998). AL can be estimated as a composite meta-factor (an aggregate measure of underlying subsystems) and partitioned into subsystems (i.e., metabolic [waist circumference, triglyceride/high density lipoprotein ratio, low density lipoprotein, hemoglobin A1c], cardiovascular [heart rate, systolic blood pressure, diastolic blood pressure), immune [high sensitivity C-reactive protein], neuroendocrine [cortisol, aldosterone] (e.g., Seeman, T. et al., 2008; Seeman, T. E., Singer Rowe, Horwitz, & McEwen, 1997)), providing a framework from which to examine the relative importance of biological dysregulation across and within subsystems in the association or depressive symptoms with incident CHD. Specifically, we examined associations among depressive symptoms and AL indices assessed at baseline and incident CHD over a period of approximately 10 years among the adults participating in the prospective Jackson Heart Study (JHS). Given the extant literature, we hypothesized that greater depressive symptomatology at baseline would predict heightened risk for incident CHD and that this association would be partially mediated by the AL meta-factor. We also hypothesized that subscale analyses would reveal metabolic and cardiovascular AL subsystem contributions among the full sample with the potential for unique immune AL subsystem contributions among women in the development of depressive symptom-associated CHD.
2. Methods
2.1. Study design
The JHS is a prospective study of risk factors for the development and progression of CVD in a cohort of 5,306 African American adults, aged 21–94 years at baseline from the tri-county area (Hinds, Madison, Rankin counties) of metropolitan Jackson, Mississippi. Details about the study design have been described elsewhere (Taylor et al., 2005). Briefly, JHS participants were enrolled and followed longitudinally at three examinations occurring from 2000 – 2004 (i.e., baseline), 2005 – 2008 (i.e., examination 2), and 2009 – 2013 (i.e., examination 3) and through annual phone calls. In the current analyses, participants were excluded if they had baseline CHD (n=400), if CES-D score was missing (n=1713), if baseline or follow-up CHD data was missing (n=117), or if covariates were missing (n=406). After exclusions, the analytic cohort included 2,670 adults (mean age 53.4 ± 12.5 years; 66% female). Demographic and clinical summaries for excluded versus included JHS participants are presented in Supplemental Table 1. Excluded JHS participants were older, more likely to be male, had less education, were less likely to report a management or professional occupation, were more likely to smoke, were less likely to engage in physical activity, were more likely to be taking medications, and exhibited less favorable biological parameters (e.g., greater waist circumference, heart rate, systolic blood pressure, triglycerides, triglyceride/high density lipoprotein ratio, hemoglobin A1c, serum cortisol) than included JHS participants (p values ≤ 0.05). Both groups had 4% incident CHD among those with available data. The JHS was approved by the institutional review boards of the University of Mississippi Medical Center, Jackson State University, and Tougaloo College. All participants provided informed consent.
2.2. Depressive Symptoms
Depressive symptoms over the last week were quantified at baseline using the Center for Epidemiologic Studies Depression Scale (CES-D), a 20-item self-report measure (Radloff, 1977). Participants were asked to indicate whether each item (e.g., “I had trouble keeping my mind on what I was doing”) was experienced rarely or none of the time (score=0), some or a little of the time (score=1), occasionally or a moderate amount of time (score=2), or most or all of the time (score=3) during the past week (some items reverse coded). CES-D scores range from 0–60, with higher scores indicative of greater depressive symptomatology and scores ≥16 indicative of clinically significant depressive symptomatology (Thomas, J. L., Jones, Scarinci, Mehan, & Brantley, 2001). The scale has been extensively used and validated, including among African American community-dwelling populations (Atkins, 2014). Internal reliability was high in this sample (α=0.82).
2.3. Coronary Heart Disease
Incident CHD served as the outcome of interest. Methods for ascertaining cardiovascular events in the JHS cohort have been described previously (Keku et al., 2005). Briefly, CVD events were ascertained through a combination of active and passive surveillance. Annual follow-up included interviews with participants and next of kin to ascertain health events, such as cardiac events, hospitalizations, or death, and through questionnaires completed by physicians and medical examiners or coroners and reviewed by the medical record abstraction unit to generate diagnosis information. These diagnoses were reviewed and adjudicated by trained medical personnel. Cardiovascular illness hospitalizations were identified and adjudicated as described previously (Keku et al., 2005). Hospitalization data were obtained from the hospital discharge index from all catchment area hospitals and annual follow-up data and data from non-catchment area hospitals were obtained after patient consent. Death certificates from state vital statistics offices were surveyed for potential CVD events. The self-reported data from annual follow-up were reconciled with the hospital discharge index data. The primary diagnoses based on international Classification of Diseases, Ninth Revision, Clinical Modification (ICD 9-CM) codes were reviewed and adjudicated by trained medical personnel. For the current analyses, we assessed the CHD occurrence between 2000 and 2011.
2.4. Allostatic Load
Baseline AL was quantified according to biological parameters in one of four AL subsystems: 1) metabolic (i.e., waist circumference, triglyceride/high density lipoprotein [HDL] ratio, low density lipoprotein [LDL], hemoglobin A1c [HbA1c]), 2) cardiovascular (i.e., heart rate [HR], systolic blood pressure [SBP], diastolic blood pressure [DBP]), 3) immune (i.e., high sensitivity C-reactive protein [hs-CRP]), and 4) neuroendocrine (i.e., serum cortisol, serum aldosterone). The continuous scores for each biological parameter were transformed into standardized z-scores. The standardized z-scores for each biological parameter were then averaged within each respective AL subsystem to create a subsystem score. The AL subsystem scores were then averaged to create a composite AL meta-factor.
AL subsystem parameters were chosen to reflect the secondary markers (i.e., metabolic, cardiovascular, immune) and primary mediators (i.e., neuroendocrine) of AL commonly assessed in the AL literature (e.g., Seeman, T. et al., 2008; Seeman, T E. et al., 1997) and available for use in the JHS dataset. We generated z-scores, which are indicative of individual differences from the sample mean for each parameter. Calculation of z-scores allows for examination of incremental effects across diverse parameters and appears to show slightly improved predictive power for health outcomes than alternative AL indices (Seplaki, Goldman, Glei, & Weinstein, 2005). Employing an approach used previously (e.g., Friedman, Karlamangla, Gruenewald, Koretz, & Seeman, 2015), we averaged AL subsystem scores to allow for the creation of an AL meta-factor that is not unduly influenced by the number of biological parameters in each AL subsystem.
Certified technicians and nurses measured participants’ waist circumference (average of two measurements around the umbilicus) and resting seated blood pressure, the average of two measurements at 5-minute intervals using an appropriately sized cuff with standard Hawksley random-zero instruments.
Fasting blood samples were drawn at baseline in the supine position and processed using a standardized protocol. Plasma and serum were prepared from samples by sedimentation in a refrigerated centrifuge within two hours of blood collection, stored at −70°C, and sent to central laboratories (University of Minnesota) (Carpenter et al., 2004; Taylor et al., 2005). Fasting serum LDL (mg/dL [Friedewald equation]), HDL, and triglycerides were assayed using standard techniques (Carpenter et al., 2004). HbA1c concentrations were measured as previously described (Joseph et al., 2016). Using a Hitachi 911 analyzer (Roche Diagnostics, Indianapolis, IN), hs-CRP was measured by the immunoturbidimetric CRP-Latex assay (Kamiya Biomedical Company, Seattle, WA) (Effoe, Correa, Chen, Lacy, & Bertoni, 2015). Measurement was done in duplicate, and any duplicates that were not within a 3 assay SD from one another were rerun. The interassay coefficient of variation on control samples was 4.5% (at an hs-CRP level of 0.45 mg/L) and 4.4% (at an hs-CRP level of 1.56 mg/L). Serum aldosterone was measured by radioimmunoassay (Siemens) and the intra-assay coefficients of variation were 8.7% and 6.2% for low and high concentrations. Morning serum cortisol was measured by chemiluminescent immunoassay performed on an immunoassay system (ADVIA Centaur; Siemens). Intra-assay coefficients of variation, were 9.1% and 7.7% for high and low cortisol concentrations, respectively.
2.5. Covariates
Baseline information was obtained during clinic visits or at home using standardized questionnaires including: demographics (age, sex), occupation (management/professional versus not), level of education (≥ Bachelor’s degree versus < Bachelor’s degree), current smoking status, and physical activity assessed using an interviewer-administered physical activity questionnaire, modified from the Baeke physical activity survey (Baecke, Burema, & Frijters, 1982). This instrument was identical to the one used during the Kaiser Physical Activity survey, which showed good validity and reliability in a multiethnic sample (Ainsworth, Sternfeld, Richardson & Jackson, 2000). Physical activity was categorized according to the American Heart Association 2020 cardiovascular health guidelines as poor, intermediate, or ideal health, as described previously (Joseph et al., 2016).
2.6. Statistical analyses
Baseline characteristics of participants are presented by CES-D less than versus greater than or equal to 16 (at risk for clinical depression), using two-sample t-tests for normally distributed continuous variables (mean, standard deviation), Wilcoxon two-sample nonparametric tests for non-normally distributed continuous variables (median, interquartile range), or Chi-square tests for categorical variables as appropriate.
Multivariable regression models examined associations between baseline depressive symptoms and AL at baseline. Cox proportional hazard models were built to examine associations among baseline depressive symptoms and incident CHD and baseline AL and incident CHD, run in separate models. Hazard ratios (HR, 95% confidence interval-CI) were calculated per 1-point increment for depressive symptomology and per 1-standard deviation (SD) increment of continuously-measured AL. The mediational role of AL in the association between depressive symptoms and incident CHD was also assessed through formal mediation analyses (R package ‘mediation’) (Imai, Keele, & Tingley, 2010; Tingley, Yamamoto, Hirose, Keele, & Imai, 2014). The average indirect effects of the exposure (i.e., depressive symptoms) on the outcome (i.e., CHD) through the mediating variables (i.e., AL subsystem and meta-factor z-scores) and proportions mediated were quantified.
For all analyses, sequential modeling was performed to adjust for potentially confounding variables as follows: Model 1: unadjusted; Model 2: Model 1 + demographic variables (age, sex); Model 3: Model 2 + socioeconomic variables (education, occupation); and Model 4: Model 3 + behavioral variables (smoking status, physical activity). For all models, interaction effects by sex were examined and the sample was stratified to examine differences by sex. For models assessing AL, subscale analyses were performed, with metabolic, cardiovascular, immune, and neuroendocrine subsystem scores serving as independent variables. The proportional hazards assumption was assessed using Schoenfeld residuals and no significant violations were noted. Statistical significance was defined as two-sided alpha=0.05 for the primary outcome, the main mediation analysis. The association of baseline depressive symptoms and AL and the association of baseline depressive symptoms with incident CHD, are estimated separately to establish the foundation for the main mediation analysis. The sex-stratified and AL subscale findings were exploratory in nature and p-values were reported at a nominal level. Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC).
3. Results
3.1. Participant Characteristics
Among 2,670 adults mean age 53.4 ± 12.5 years, 66% female, during a median follow-up period of 9.9 years (IQR 9, 10.7), 106 CHD events occurred. Table 1 shows the profile of participants stratified by clinically non-significant (CES-D<16) versus significant (CES-D≥16) depressive symptomatology. Among the full cohort, the prevalence of clinically significant depressive symptoms was 20.9%. Women were more likely to exhibit significant depressive symptoms than men (23.7% vs. 15.6%; p<0.001). Compared to those with a CES-D score <16, participants with a CES-D score ≥16 were also younger and more likely to report current smoking, less than a high school education, non-professional occupations, and poor physical activity (p-values<0.050). In regards to AL biological parameters, participants with CES-D scores ≥16 had a greater waist circumference, hs-CRP, and immune, and meta-factor AL z-scores and lower LDL than participants with CES-D scores <16 (p-values<0.050). CHD incidence was 6% vs. 4% among participants with CES-D ≥16 vs. <16 (p=0.031).
Table 1:
Baseline Characteristics* | <16(n=2112) | ≥16(n=558) | Total (n=2670) | P-value |
---|---|---|---|---|
Age (years) | 53.8 (12.4) | 51.8 (13.0) | 53.4 (12.5) | 0.001 |
Sex (female) | 1335 (63%) | 414 (74%) | 1749 (66%) | <.001 |
High school or greater education (yes) | 1857 (88%) | 452 (81%) | 2309 (86%) | <.001 |
Management/Professional occupation (yes) | 948 (45%) | 153 (27%) | 1101 (41%) | <.001 |
Smoking status (yes) | 191 (9%) | 94 (17%) | 285 (11%) | <.001 |
Physical activity | <.001 | |||
Poor | 900 (43%) | 286 (51%) | 1186 (44%) | |
Intermediate | 722 (34%) | 191 (34%) | 913 (34%) | |
Ideal | 490 (23%) | 81 (15%) | 571 (21%) | |
Waist circumference (cm) | 99.7 (16.1) | 101.4 (16.9) | 100.1 (16.3) | 0.033 |
Triglycerides (mg/dL) | 100.5 (52.7) | 104.6 (61.4) | 101.3 (54.7) | 0.113 |
High-density lipoprotein (mg/dL) | 52.2 (14.6) | 51.9 (13.9) | 52.1 (14.4) | 0.661 |
Low-density lipoprotein (mg/dL) | 127.7 (35.8) | 121.3 (35.4) | 126.4 (35.8) | <.001 |
Triglyceride/HDL ratio† | 1.7 (1,2.7) | 1.8 (1,2.8) | 1.7 (1,2.7) | 0.605 |
Hemoglobin A1c (%) | 5.8 (1.1) | 5.9 (1.2) | 5.8 (1.1) | 0.401 |
Heart rate | 63.8 (10) | 64.7 (10.5) | 64 (10.2) | 0.062 |
Systolic blood pressure (mmHg) | 125.9 (15.7) | 126.1 (16.3) | 125.9 (15.8) | 0.739 |
Diastolic blood pressure (mmHg) | 76 (8.5) | 76.3 (8.5) | 76 (8.5) | 0.446 |
High-sensitivity C-reactive protein (mg/L)† | 0.25 (0.1,0.5) | 0.33 (0.1,0.7 | ) 0.26 (0.1,0.6) | <.001 |
Cortisol (ug/dL) | 9.6 (4) | 9.4 (4) | 9.6 (4) | 0.309 |
Plasma aldosterone (ng/dL)† | 4.5 (3, 7.3) | 4.1 (2, 7.4) | 4.4 (3, 7.3) | 0.062 |
Allostatic load metabolic Z-score | −0.02(0.6) | −0.02 (0.6) | −0.02 (0.6) | 0.976 |
Allostatic load cardiovascular Z-score | −0.04 (0.7) | 0.01 (0.7) | −0.03 (0.7) | 0.147 |
Allostatic load immune Z-score | −0.06 (1) | 0.14 (1) | −0.01 (1) | <.001 |
Allostatic load neuroendocrine Z-score | −0.007 (0.8) | −0.07 (0.8) | −0.02 (0.8 | 0.110 |
Allostatic load meta-factor Z-score | −0.03 (0.5) | 0.01 (0.5) | −0.02 (0.5) | 0.038 |
Incident coronary heart disease (yes) | 75 (4%) | 31 (6%) | 106 (4%) | 0.031 |
Note: p values of <0.05 considered statistically significant
Mean (SD) or percentages are listed, p-values calculated using chi-square (categorical variables) and two sample t-test (parametric continuous variables)
Median (IQR), p-values calculated using Wilcoxon two sample nonparametric test (nonparametric continuous variables)
3.2. Depressive Symptoms and Allostatic Load
The sequentially adjusted associations among depressive symptoms and AL indices at baseline are presented in Table 2. Among the full cohort, the fully adjusted models revealed that 1-point greater CES-D score was associated with a 0.003 unit positive difference in metabolic AL z-score (p=0.036), a 0.004 unit positive difference in cardiovascular AL z-score (p=0.031), and a 0.007 unit positive difference in immune AL z-score (p=0.009). The association of CES-D with the AL meta-factor z-score, which serves as a composite index of all AL subsystems, was also statistically significant (B=0.003, p=0.007). Depressive symptomatology was not associated with neuroendocrine subsystem dysregulation (B=−0.0007, p=0.737).
Table 2.
AL Z-Scores | Model | Overall (n = 2,670) | Women (n = 1,749) | Men (n = 921) | |||
---|---|---|---|---|---|---|---|
Beta (95% CI) | p | Beta (95% CI) | p | Beta (95% CI) | p | ||
Metabolic | 1 | 0.002 (−0.001,0.005) | 0.210 | 0.005 (0.001,0.008) | 0.009 | −0.001 (−0.007, 0.004) | 0.702 |
2 | 0.004 (0.001, 0.007) | 0.005 | 0.006 (0.003, 0.009) | 0.001 | −0.0006 (−0.006, 0.005) | 0.840 | |
3 | 0.004 (0.0006, 0.006) | 0.018 | 0.005 (0.002, 0.009) | 0.002 | −0.001 (−0.007, 0.004) | 0.681 | |
4 | 0.003 (0.0002, 0.006) | 0.036 | 0.005 (0.001,0.008) | 0.005 | −0.002 (−0.007, 0.004) | 0.590 | |
Cardiovascular | 1 | 0.004 (0.0007, 0.007) | 0.018 | 0.006 (0.002, 0.010) | 0.003 | 0.00008 (−0.006, 0.006) | 0.980 |
2 | 0.005 (0.002, 0.008) | 0.003 | 0.007 (0.003, 0.010) | 0.001 | 0.0004 (−0.006, 0.007) | 0.897 | |
3 | 0.005 (0.001, 0.008) | 0.007 | 0.006 (0.002, 0.010) | 0.001 | 0.00003 (−0.006, 0.006) | 0.992 | |
4 | 0.004 (0.0003, 0.007) | 0.031 | 0.005 (0.002, 0.009) | 0.005 | −0.001 (−0.007, 0.005) | 0.705 | |
Immune | 1 | 0.012 (0.007, 0.017) | <.001 | 0.010 (0.004, 0.015) | 0.001 | 0.006 (−0.003, 0.016) | 0.161 |
2 | 0.009 (0.005, 0.014) | <.001 | 0.011 (0.005, 0.016) | <.001 | 0.007 (−0.002, 0.016) | 0.138 | |
3 | 0.008 (0.003, 0.013) | 0.002 | 0.009 (0.003, 0.015) | 0.002 | 0.005 (−0.004, 0.014) | 0.275 | |
4 | 0.007 (0.002, 0.011) | 0.009 | 0.008 (0.002, 0.013) | 0.008 | 0.003 (−0.006, 0.012) | 0.474 | |
Neuroendocrine | 1 | −0.003 (−0.006, 0.001) | 0.190 | −0.0002 (−0.005, 0.004) | 0.933 | −0.001 (−0.008, 0.006) | 0.756 |
2 | 0.0002 (−0.003, 0.004) | 0.909 | 0.0006 (−0.004, 0.005) | 0.783 | −0.0008 (−0.007, 0.006) | 0.820 | |
3 | −0.0001 (−0.004, 0.004) | 0.940 | 0.0003 (−0.004, 0.005) | 0.905 | −0.001 (−0.008, 0.006) | 0.731 | |
4 | −0.0007 (−0.004, 0.003) | 0.737 | −0.0001 (−0.005, 0.004) | 0.951 | −0.002 (−0.009, 0.005) | 0.577 | |
Meta-factor | 1 | 0.004 (0.002, 0.006) | <.001 | 0.005 (0.002, 0.008) | <.001 | 0.001 (−0.003, 0.005) | 0.618 |
2 | 0.005 (0.002, 0.007) | <.001 | 0.006 (0.003, 0.009) | <.001 | 0.001 (−0.003, 0.006) | 0.496 | |
3 | 0.004 (0.002, 0.006) | 0.001 | 0.005 (0.003, 0.008) | <.001 | 0.0007 (−0.004, 0.005) | 0.757 | |
4 | 0.003 (0.0008, 0.005) | 0.007 | 0.004 (0.002, 0.007) | 0.001 | −0.0004 (−0.005, 0.004) | 0.869 |
Note: p values of <0.05 considered statistically significant multivariable regression models examined associations among baseline depressive symptoms and AL at baseline; Model 1 unadjusted; Model 2 includes model 1 plus age, sex; Model 3 includes Model 2 plus education, occupation; Model 4 includes Mode 3 + smoking, physical activity; AL = allostatic load
A statistically significant interaction effect by sex was also noted in examining the association between CES-D scores and AL metabolic subsystem z-scores (p=0.048), with CES-D scores significantly associated with metabolic AL among women (p=0.005) but not men (p=0.590) in stratified analyses. Interaction effects by sex approached but did not reach statistical significance in examining associations among depressive symptoms, cardiovascular AL, and the AL meta-factor (p=0.068 and p=0.054, respectively), with results stratified by sex shown in Table 2. For these AL parameters, significant associations with depressive symptoms were witnessed among women but not men.
3.3. Depressive Symptoms, Allostatic Load, and Coronary Heart Disease
The sequentially adjusted associations of depressive symptoms and AL with CHD are presented in Table 3. Among the full cohort, in the fully adjusted models, each 1-point increase in CES-D score was associated with a 3.3% higher risk of CHD (p=0.007). Each 1-unit increase in metabolic AL subsystem z-score was associated with an 88% higher risk of CHD (p<0.001). Each 1-unit increase in neuroendocrine AL subsystem z-score was associated with a 39% higher risk of CHD (p=0.008). When AL was examined as a meta-factor, each 1-unit increase in AL z-score was associated with a 130% higher risk of CHD (p<0.001).
Table 3:
Predictor | Model 1 | Model 2 | Model 3 | Model 4 | |
---|---|---|---|---|---|
HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
Overall | CES-D | 1.025 (1.00, 1.05)* | 1.036 (1.01, 1.06)* | 1.034 (1.01, 1.06)* | 1.033 (1.01, 1.06)* |
(n = 2,670) | Metabolic AL z-score | 2.01 (1.53, 2.64)* | 1.90 (1.41,2.55)* | 1.88 (1.40, 2.53)* | 1.88 (1.39, 2.53)* |
Cardiovascular AL z-score | 1.26 (0.95, 1.67) | 1.23 (0.93, 1.63) | 1.21 (0.92, 1.61) | 1.20 (0.90, 1.60) | |
Immune AL z-score | 1.14 ( 0.94, 1.38) | 1.17 (0.95, 1.43) | 1.16 (0.95, 1.43) | 1.16 (0.94, 1.42) | |
Neuroendocrine AL z-score | 1.47 (1.17, 1.83)* | 1.40 (1.09, 1.78)* | 1.40 (1.10, 1.79)* | 1.39 (1.09, 1.78)* | |
AL meta-factor z-score | 2.45 (1.64, 3.68)* | 2.35 (1.53, 3.63)* | 2.32 (1.51,3.58)* | 2.30 (1.49, 3.57)* | |
Women | CES-D | 1.033 (1.01, 1.06)* | 1.047 (1.02, 1.08)* | 1.044 (1.02, 1.07)* | 1.044 (1.02, 1.07)* |
(n = 1,749) | Metabolic AL z-score | 1.95 (1.38, 2.77)* | 1.81 (1.23, 2.67)* | 1.79 (1.22, 2.65)* | 1.78 (1.21,2.63)* |
Cardiovascular AL z-score | 1.27 (0.87, 1.85) | 1.24 (0.74, 1.82) | 1.21 (0.83, 1.78) | 1.20 (0.82, 1.76) | |
Immune AL z-score | 1.06 (0.83, 1.36) | 1.07 (0.83, 1.37) | 1.05 (0.82, 1.36) | 1.05 (0.81, 1.36) | |
Neuroendocrine AL z-score | 1.32 (0.99, 1.76) | 1.19 (0.87, 1.63) | 1.18 (0.86, 1.62) | 1.18 (0.86, 1.62) | |
AL meta-factor z-score | 2.03 (1.20, 3.44)* | 1.84 (1.05, 3.23)* | 1.78 (1.01,3.15)* | 1.77 (1.00, 3.14) | |
Men | CES-D | 1.017 (0.98, 1.06) | 1.017 (0.98, 1.06) | 1.014 (0.97, 1.06) | 1.014 (0.97, 1.06) |
(n = 921) | Metabolic AL z-score | 2.04 (1.30, 3.20)* | 2.03 (1.28, 3.24)* | 2.01 (1.27, 3.19)* | 2.02 (1.27, 3.21)* |
Cardiovascular AL z-score | 1.22 (0.80, 1.86) | 1.22 (0.81, 1.84) | 1.22 (0.81, 1.84) | 1.20 (0.79, 1.82) | |
Immune AL z-score | 1.48 (1.08, 2.05)* | 1.37 (0.98, 1.91) | 1.38 (0.99, 1.92) | 1.37 (0.97, 1.92) | |
Neuroendocrine AL z-score | 1.74 (1.17, 2.57)* | 1.86 (1.25, 2.76)* | 1.88 (1.26, 2.79)* | 1.86 (1.25, 2.76)* | |
AL meta-factor z-score | 3.31 (1.75, 6.28)* | 3.36 (1.72, 6.55)* | 3.35 (1.73, 6.51)* | 3.33 (1.71,6.49)* | |
Note: p values of <0.05 considered statistically significant; Models run separately for each listed predictor; Model 1 unadjusted; Model 2 includes model 1 plus age, sex; Model 3 includes Model 2 plus education, occupation; Model 4 includes Model 3 + smoking, physical activity; AL = allostatic load
statistically significant association
Associations among depressive symptoms, AL, and CHD did not significantly differ by sex in examining interaction effects in the fully adjusted models (p≥0.082). However, given the sex differences noted in the associations among depressive symptoms and AL, we also provide these results stratified by sex (Table 3). Point estimates for the association of depressive symptoms with CHD were similar among males and females and was significant among females (HR 1.044, 95%CI 1.015, 1.073). Stratifying by sex also revealed higher HRs for males versus females for the metabolic AL subsystem z-score though, as noted, these differences were not statistically significant (Males: HR 2.02, 95%CI 1.27, 3.21; Females: HR 1.78, 95%CI 1.21, 2.63). The neuroendocrine AL subsystem (Males: HR 1.86, 95%CI 1.25, 2.76; Females: HR 1.18, 95%CI 0.86, 1.62) and AL meta-factor (Males: HR 3.33, 95%CI 1.71, 6.49; Females: HR 1.77, 95%CI 1.00, 3.14) associations failed to reach statistical significance in females but were predictive of incident CHD in males.
3.4. Allostatic Load in the Association of Depressive Symptoms with Coronary Heart Disease
The fully adjusted model for testing mediation by AL in the association of depressive symptoms with incident CHD is presented in Table 4. In examining the AL meta-factor z-score among the full cohort, there was a significant mediation effect, with percent mediation estimated at 7.55% (p=0.016). In examining the AL subsystem scores, metabolic AL was found to mediate 5.79% of the association between depressive symptoms and incident CHD (p=0.044). No additional indirect effects were supported according to the alternative AL subsystems.
Table 4:
Overall (n = 2,670) | Women (n = 1,749) | Men (n = 921) | ||||
---|---|---|---|---|---|---|
% Mediated | p | % Mediated | p | % Mediated | p | |
Metabolic AL z-score | 5.79 | 0.044 | 6.18 | 0.016 | −2.16 | 0.840 |
Cardiovascular AL z-score | 1.78 | 0.254 | 1.54 | 0.500 | −0.19 | 0.920 |
Immune AL z-score | 2.29 | 0.256 | 0.17 | 0.934 | 1.66 | 0.810 |
Neuroendocrine AL z-score | −0.57 | 0.744 | −0.03 | 0.952 | −2.95 | 0.780 |
AL Meta-factor z-score | 7.55 | 0.016 | 4.94 | 0.116 | −0.20 | 0.990 |
Note: p values of <0.05 considered statistically significant; Models adjusted for age, sex, education, occupation, smoking and physical activity; AL = allostatic load
Next, given the aforementioned results, tests of mediation were repeated stratifying by sex. Metabolic AL significantly mediated the association between depressive symptoms and incident CHD among females (6.18% mediated, p=0.016). In contrast, there was no support for a mediated effect of metabolic AL in the association between CES-D score and CHD risk among males (−2.16% mediated, p=0.840). The AL meta-factor also failed to reach statistical significance as a mediator among either group alone (p values≥0.116). Consistent with our previous findings, the direct effect of depressive symptoms on incident CHD could also be detected among females (p=0.010) but not males (p=0.420) in the metabolic AL mediational model.
4. Discussion
This study provides evidence that, among a diverse cohort of African American adults, each 1-point increase in baseline depressive symptoms is associated with a 3.3% greater risk of incident CHD and is partially mediated by greater overall AL and metabolic AL. This report expands important literature demonstrating (Brown, Stewart, Stump, & Callahan, 2011; O’Brien et al., 2015; Sims et al., 2015) but also refuting (Moise et al., 2016) an association between depressive symptoms and CHD-related outcomes in studies with significant African American representation. Extending prior work, we present findings produced from formal mediation analyses suggesting a key biologic pathway (i.e., metabolic) by which depressive symptomatology may be linked to CHD in African Americans. Findings from tests of mediation are of particular importance considering that there is biologic plausibility for a number of correlates of depressive symptomatology to play a role in cardiovascular risk. A more complete understanding of CHD risk pathways serves as a critical foundation toward progress in targeted CHD prevention.
The metabolic AL subsystem score found to mediate the association between depressive symptoms and incident CHD was composed of abdominal adiposity (i.e., waist circumference), cholesterol (i.e., triglyceride/HDL ratio. LDL), and blood glucose (i.e., HbA1c) markers. In the context of depressive symptoms, we observed less favorable waist circumference, in particular. The observed associations are in line with prior literature, particularly among African Americans (Beydoun et al., 2016; Grossniklaus et al., 2012), with some data suggesting that this relationship is not due to increased depression-associated dietary energy density (Grossniklaus e. al., 2012). It is also important to consider that, with increased abdominal adiposity, use of fat stores as a source of energy has been increasingly appreciated as resulting in increased hepatic triglyceride production within very LDL particles and the accumulation of small HDL particles (which fail to offer the expected protection) and small LDL particles (which result in lower measured LDL levels) (Bosomworth, 2013). The triad of high triglyceride levels, low HDL levels, and increased LDL particle number (termed “atherogenic dyslipidemia”) is thought to be particularly detrimental to the health of the coronary arteries (Bosomworth, 2013), with the findings from this report suggesting that this may be a particularly important pathway in the development of depressive symptom-associated CHD.
We also noted associations among depressive symptoms and immune AL (i.e., hs-CRP). This finding is in line with prior psychoneuroimmunologic studies linking depressive symptoms and markers of inflammation, with inflammation increasingly implicated in chronic disease (reviewed by Kiecolt-Glaser, Derry, & Fagundes, 2015). In fact, some evidence suggests the presence of an inflammatory subtype of depression more likely to demonstrate resistance to traditional pharmacotherapies (Beijers, Wardenaar, van Loo, & Schoevers, 2019; Jeng et al., 2018), prompting pursuit of adjuvant or alternative treatment approaches among these individuals (reviewed by lonescu & Papakostas, 2017).
In the current study, hs-CRP did not predict CHD in the assessed sample, which is interesting considering that hs-CRP has been considered for inclusion in predictive models of CVD (US Preventive Services Task Force et al., 2018). Though, the usefulness of hs-CRP as a reflection of chronic, low-grade inflammation has been debased and the U.S. Preventive Services Task Force recently determined that the current evidence is insufficient to warrant recommendation of hs-CRP testing, calling for high quality prospective studies among diverse populations (US Preventive Services Task Force et al., 2018). While the current study suggests minimal predictive value of a single assessment of hs-CRP for CHD among a large African American cohort, more work is needed to determine the potential for incremental predictive value when hs-CRP is added to traditional risk scoring systems.
We also showed that overall, depressive symptoms were not associated with neuroendocrine AL (i.e., cortisol and aldosterone) but neuroendocrine AL predicted increased risk for CHD. In fact, participants with clinically significant depressive symptoms showed marginally lower aldosterone levels versus those without depressive symptoms. This finding is unexpected, as multiple studies (e.g., Segeda, Izakova, Hlavacova, Bednarova, & Jezova, 2017) have implicated higher aldosterone burden in the context of depressive symptoms, with increased activity of the renin-angiotensin-aldosterone system (RAAS) implicated in hypertension and CVD (Joseph et al., 2017). This discrepancy may be related to methodological differences (e.g., population, sampling) but may also reflect the mounting of a compensatory response among those combatting depressive symptoms. For example, De Vos et al. (2018) reported that, among African Americans (n=68) but not non-Hispanic whites (n=127), chronic depression predicts elevations in DBP and concomitant reductions in renin over a 3-year follow-up period. The authors posited that renin suppression reflected an unsuccessful defense against volume-loading hypertension, as aldosterone levels and estimated glomerular filtration rate remained unchanged. The role of the RAAS in CHD in the context of depression warrants further exploration.
Importantly, this report reveals differences by sex in the role of AL in the association of depressive symptoms with CHD. In fact, among males (n=921), baseline depressive symptoms failed to predict incident CHD, AL as a meta-factor, or any AL subsystem score. Though, in males, metabolic and neuroendocrine AL and the AL meta-factor were associated with CHD. In females (n=1,749), baseline depressive symptoms predicted CHD risk and greater AL (metabolic, cardiovascular, immune, and meta-factor). Greater metabolic AL, in particular, predicted CHD, with support for the variable as a mediator. This suggests that findings related to the metabolic pathway in depressive symptom-associated CHD were driven by females and highlight the benefit of AL subsystem analyses. Moreover, the menopausal transition and greater severity of menopausal symptoms (including depressed mood) have been linked to dysregulation across several metabolic parameters (Cengiz, Kaya, Suzen Caypinar, & Alay, 2019; Gurka, Vishnu, Santen, & DeBoer, 2016), which may to help to explain the findings of the current study and highlight the importance of continued work in this area.
Patways linking neuroendocrine AL and CHD may also show differences by sex, with future studies focusing on more nuanced depictions of neuroendocrine function having potential to shed light on this possibility. Specifically, while a single morning serum cortisol value holds strong potential for clinical translation, correlates well with alternative indices (e.g., salivary cortisol (Restituto et al., 2008)), and predicts concurrent metabolic dysregulation (e.g., higher fasting plasma glucose, HbA1c (Ortiz et al., 2019)), in-depth assessments of diurnal patterns or total output over a period of time may provide additional insight.
Similarly, there remains a need for explication of pathways, mechanisms, and biology underlying sex differences in CVD as well as clinical translation for diagnosis and treatment, evolving due to increased inclusivity of females in cardiovascular research demonstrating sex differences in aging, cardiac function, coronary artery anatomy, and blood pressure regulation (Taqueti, 2018). Here, we add to the body of literature identifying sex-specific cardiovascular outcomes in response to depressive symptoms and AL among African Americans. Consistent with our hypothesis, AL contributions to CHD risk differed by sex, which mirrors work from others who have identified higher AL among African American women associated with metabolic atherogenic dyslipidemia (Chyu & Upchurch, 2018).
Strengths of this report include the assessment of a large, socioeconomically diverse, contemporary, African American community-based cohort with rigorously ascertained physiologic and laboratory measures and over a decade of follow-up of adjudicated CHD outcomes. The availability of CES-D data also provided the opportunity to assess associations among continuously estimated depressive symptoms, AL parameters, and incident CHD, which is not possible when relying solely on the presence or absence of a clinical diagnosis of depression. This being said, the CES-D is well-validated but relies on self-report. Data on clinical diagnosis or treatment of depression at baseline exam were not available, prohibiting their examination. Some evidence does suggest that antidepressants alter risk for CVD (Coupland et al., 2016; Hamer, Batty, Seldenrijk, & Kivimaki, 2011). Individuals with self-reported CHD at baseline were excluded from analyses. Therefore, the current report provides estimates of associations among depressive symptoms, AL indices, and incident CHD diagnosis. Though, the potential for unrecognized pathophysiologic features of CHD among analyzed individuals must also be recognized (Vigili de Kreutzenberg et al., 2017).
Despite these and other strengths, there are some additional potential limitations. First, the participants in the JHS are from one geographic area in the southeastern U.S. and may not be representative of all African Americans. A number of JHS participants were also excluded from analyses due to baseline CHD (n=400) or data missingness for CES-D scores (n=1713), CHD data (n=117), or covariates (n=406), producing an analytical sample that was generally of higher socioeconomic status and healthier than the excluded sample. The generalizability of this report and potential for selection bias must be considered with this in mind. Second, CES-D and AL were analyzed according to baseline measures only. As such, while a growing literature suggests that depressive symptoms are capable of contributing to the progression or even establishment of dysregulation in the biological systems under investigation (Ambrosio et al., 2018; Niles & O’Donovan, 2019; Srinivas et al., 2018), the cross-sectional nature of this data must be fully acknowledged and highlights the importance of future longitudinal investigations of detailed AL indices and incident CVD over lengthy follow-up periods. Third, exploratory analyses of AL subscale and sex-stratified statistical associations were interpreted without correction for multiple comparisons and require verification in future studies. Therefore, some caution is warranted in the interpretation of results.
In conclusion, this report provides data indicating that depressive symptom-associated CHD risk is partially mediated by greater metabolic AL burden among African American adults, with stratified analyses suggesting that this mediational pathway is unique to females. We also identified concurrent associations among depressive symptoms and AL (metabolic, cardiovascular, immune, meta-factor) among females only and predictive value of neuroendocrine AL for incident CHD among males only. These findings suggest that application of the AL framework may be particularly fruitful in elucidation of biologic pathways linking psychological parameters to CHD. In particular, determination of the clinical significance of the statistically significant associations noted in the present analyses will be an important area of future inquiry. Similarly, future studies examining risk factors and mechanisms underlying CHD should carefully consider time-varying depressive symptoms and potential differences by sex, which is critical to the design of interventions for the targeted prevention and treatment of CHD among African American men and women.
Supplementary Material
Highlights.
African Americans are at heightened risk for coronary heart disease.
Biologic pathways are poorly understood.
In this study, depressive symptoms were associated with allostatic load in females.
Depressive symptoms and allostatic load predicted incident coronary heart disease.
Metabolic allostatic load was a partial mediator in the pathway among females.
Acknowledgements
The Jackson Heart Study (JHS) is supported and conducted in collaboration with Jackson State University (HHSN268201800013I), Tougaloo College (HHSN268201800014I), the Mississippi State Department of Health (HHSN268201800015I/HHSN26800001) and the University of Mississippi Medical Center (HHSN268201800010I, HHSN268201800011I and HHSN268201800012I) contracts from the National Heart, Lung, and Blood Institute (NHLBI) and the National Institute for Minority Health and Health Disparities (NIMHD). Preparation of this manuscript was supported by the National Institute of Nursing Research (K23NR017902, SLG) and the National Institute of Diabetes and Digestive and Kidney Diseases (K23DK117041, JJJ) of the National Institutes of Health The authors also wish to thank the staff and participants of the JHS. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services.
Footnotes
Conflicts of Interest
We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.
Declaration of interest: None.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Ainsworth BE, Sternfeld B, Richardson MT, & Jackson K (2000). Evaluation of the kaiser physical activity survey in women. Medicine and Science in Sports and Exercise, 32(7), 1327–1338. [DOI] [PubMed] [Google Scholar]
- Ambrosio G, Kaufmann FN, Manosso L, Platt N, Ghisleni G, Rodrigues ALS, … Kaster MP (2018). Depression and peripheral inflammatory profile of patients with obesity. Psychoneuroendocrinology, 91, 132–141. doi:S0306-4530(17)31279-9 [pii] [DOI] [PubMed] [Google Scholar]
- Atkins R (2014). Validation of the center for epidemiologic studies depression scale in black single mothers. Journal of Nursing Measurement, 22(3), 511–524 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Baecke JA, Burema J, & Frijters JE (1982). A short questionnaire for the measurement of habitual physical activity in epidemiological studies The American Journal of Clinical Nutrition, 36(5), 936–942. [DOI] [PubMed] [Google Scholar]
- Beijers L, Wardenaar KJ, van Loo HM, & Schoevers RA (2019). Data-driven biological subtypes of depression: Systematic review of biological approaches to depression subtyping. Molecular Psychiatry, doi 10.1038/s41380-019-0385-5 [doi] [DOI] [PubMed] [Google Scholar]
- Beydoun MA, Fanelli-Kuczmarski MT, Shaked D, Dore GA, Beydoun HA, Rostant OS, … Zonderman AB (2016). Alternative pathway analyses indicate bidirectional relations between depressive symptoms, diet quality, and central adiposity in a sample of urban US adults. The Journal of Nutrition, 146(6), 1241–1249. doi: 10.3945/jn.115.229054 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bosomworth NJ (2013). Approach to identifying and managing atherogenic dyslipidemia: A metabolic consequence of obesity and diabetes. Canadian Family Physician Medecin De Famille Canadien, 59(11), 1169–1180. doi:59/11/1169 [pii] [PMC free article] [PubMed] [Google Scholar]
- Brown JM, Stewart JC, Stump TE, & Callahan CM (2011). Risk of coronary heart disease events over 15 years among older adults with depressive symptoms. The American Journal of Geriatric Psychiatry : Official Journal of the American Association for Geriatric Psychiatry, 19(8), 721–729. doi: 10.1097/JGP.0b013e3181faee19 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Buckwalter JG, Castellani B, McEwen B, Karlamangla AS, Rizzo AA, John B, … Seeman T (2016). Allostatic load as a complex clinical construct: A case-based computational modeling approach. Complexity, 21(Suppl 1), 291–306. doi: 10.1002/cplx.21743 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carpenter MA, Crow R, Steffes M, Rock W, Heilbraun J, Evans G, … Sarpong D (2004). Laboratory, reading center, and coordinating center data management methods in the jackson heart study. The American Journal of the Medical Sciences, 328(3), 131–144. doi : S0002-9629(15)33984-7 [pii] [DOI] [PubMed] [Google Scholar]
- Cengiz H, Kaya C, Suzen Caypinar S, & Alay I (2019). The relationship between menopausal symptoms and metabolic syndrome in postmenopausal women. Journal of Obstetrics and Gynaecology : The Journal of the Institute of Obstetrics and Gynaecology, 39(4), 529–533. doi: 10.1080/01443615.2018.1534812 [doi] [DOI] [PubMed] [Google Scholar]
- Chyu L, & Upchurch DM (2018). A longitudinal analysis of allostatic load among a multi-ethnic sample of midlife women: Findings from the study of women’s health across the nation. Women’s Health Issues : Official Publication of the Jacobs Institute of Women’s Health, 28(3) 258–266. doi:S1049-3867(17)30150-0 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Coupland C, Hill T, Morriss R, Moore M, Arthur A, & Hippisley-Cox J (2016). Antidepressant use and risk of cardiovascular outcomes in people aged 20 to 64: Cohort study using primary care database. BMJ (Clinical Research Ed), 352, i1350. doi:10.1136/bmj.i1350. doi: 10.1136/bmj.i1350 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cunningham TJ, Croft JB, Liu Y, Lu H, Eke PI, & Giles WH (2017). Vital signs: Racial disparities in age-specific mortality among blacks or african americans - united states, 1999–2015. MMWR.Morbidity and Mortality Weekly Report, 66(17), 444–456. doi: 10.15585/mmwr.mm6617e1 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- De Vos AC, Malan L, Seedat YK, Cockeran M, & Malan NT (2018). Chronic depression symptoms desensitize renin activity to protect against volume-loading hypertension in blacks: The SABPA study. Physiology & Behavior, 194, 474–480. doi:S0031-9384(18)30418-9 [pii] [DOI] [PubMed] [Google Scholar]
- Duru OK, Harawa NT, Kermah D, & Norris KC (2012). Allostatic load burden and racial disparities in mortality. Journal of the National Medical Association, 104(1–2) 89–95. doi:S0027-9684(15)30120-6 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Effoe VS, Correa A, Chen H, Lacy ME, & Bertoni AG (2015). High-sensitivity C-reactive protein is associated with incident type 2 diabetes among african americans: The jackson heart study. Diabetes Care, 38(9), 1694–1700. doi: 10.2337/dc15-0221 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Eurelings LS, van Dalen JW, Ter Riet G, Moll van Charante EP, Richard E, van Gool WA, & ICARA Study Group. (2018). Apathy and depressive symptoms in older people and incident myocardial infarction, stroke, and mortality: A systematic review and meta-analysis of individual participant data. Clinical Epidemiology, 10, 363–379. doi: 10.2147/CLEP.S150915 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Friedman EM, Karlamangla A S, Gruenewald TL, Koretz B, & Seeman TE (2015). Early life adversity and adult biological risk profiles. Psychosomatic Medicine, 77(2), 176–185. doi: 10.1097/PSY.0000000000000147 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grossniklaus DA, Dunbar SB, Gary R, Tohill BC, Frediani JK, & Higgins MK (2012) Dietary energy density: A mediator of depressive symptoms and abdominal obesity or independent predictor of abdominal obesity? European Journal of Cardiovascular Nursing : Journal of the Working Group on Cardiovascular Nursing of the European Society of Cardiology, 11(4), 423–431. doi: 10.1016/j.ejcnurse.2011.03.008 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gurka MJ, Vishnu A, Santen RJ, & DeBoer MD (2016). Progression of metabolic syndrome severity during the menopausal transition. Journal of the American Heart Association, 5(8), 10.1161/JAHA.116.003609. doi: 10.1161/JAHA.116.003609 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hamer M, Batty GD, Seldenrijk A, & Kivimaki M (2011). Antidepressant medication use and future risk of cardiovascular disease: The Scottish health survey. European Heart Journal, 32(4), 437–442. doi: 10.1093/eurheartj/ehq438 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Imai K, Keele L, & Tingley D (2010). A general approach to causal mediation analysis. Psychological Methods, 15(4), 309–334. doi: 10.1037/a0020761 [doi] [DOI] [PubMed] [Google Scholar]
- lonescu DF, & Papakostas GI (2017). Experimental medication treatment approaches for depression. Translational Psychiatry, 7(3), e1068. doi: 10.1038/tp.2017.33 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jeng JS, Li CT, Chen MH, Lin WC, Bai YM, Tsai SJ, … Sung, Y J. (2018). Repeated low-grade infections predict antidepressant-resistant depression: A nationwide population-based cohort study. The Journal of Clinical Psychiatry, 79(1), 10.4088/JCP.17m11540. doi: 10.4088/JCP.17m11540 [doi] [DOI] [PubMed] [Google Scholar]
- Joseph JJ, Echouffo-Tcheugui JB, Kalyani RR, Yeh HC, Bertoni AG, Effoe VS, … Golden SH (2016). Aldosterone, renin, and diabetes mellitus in african americans: The jackson heart study. The Journal of Clinical Endocrinology and Metabolism, 101(4), 1770–1778. doi: 10.1210/jc.2016-1002 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Joseph JJ, Echouffo-Tcheugui JB, Kalyani RR, Yeh HC, Bertoni AG, Effoe VS, … Golden SH (2017). Aldosterone, renin, cardiovascular events, and all-cause mortality among african americans: The jackson heart study. JACC.Heart Failure, 5(9), 642–651. doi:S2213-1779(17)30385-2 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Keku E, Rosamond W, Taylor HA Jr,, Garrison R, Wyatt SB, Richard M, … Sarpong D (2005). Cardiovascular disease event classification in the jackson heart study: Methods and procedures. Ethnicity & Disease, 15(4 Suppl 6), S6–62–70. [PubMed] [Google Scholar]
- Kiecolt-Glaser JK, Derry HM, & Fagundes CP (2015). Inflammation: Depression fans the flames and feasts on the heat. The American Journal of Psychiatry, 172(11), 1075–1091. doi: 10.1176/appi.ajp.2015.15020152 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kochanek KD, Murphy S, Xu J, & Arias E (2017). Mortality in the united states, 2016. NCHS Data Brief, (293)(293), 1–8. [PubMed] [Google Scholar]
- Labaka A, Goni-Balentziaga O, Lebena A, & Perez-Tejada J (2018). Biological sex differences in depression: A systematic review. Biological Research for Nursing, 20(4) 383–392. doi: 10.1177/1099800418776082 [doi] [DOI] [PubMed] [Google Scholar]
- McEwen BS (1998). Stress, adaptation, and disease. allostasis and allostatic load. Annals of the New York Academy of Sciences, 840, 33–44. [DOI] [PubMed] [Google Scholar]
- Moise N, Khodneva Y, Richman J, Shimbo D, Kronish I, & Safford MM (2016). Elucidating the association between depressive symptoms, coronary heart disease, and stroke in black and white adults: The REasons for geographic and racial differences in stroke (REGARDS) study. Journal of the American Heart Association, 5(8), 10.1161/JAHA.116.003767. doi: 10.1161/JAHA116.003767 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mukherjee S, Trepka MJ, Pierre-Victor D, Bahelah R, & Avent T (2016). Racial/ethnic disparities in antenatal depression in the united states: A systematic review. Maternal and Child Health Journal, 20(9), 1780–1797. doi: 10.1007/s10995-016-1989-x [doi] [DOI] [PubMed] [Google Scholar]
- Niles AN, & O’Donovan A (2019). Comparing anxiety and depression to obesity and smoking as predictors of major medical illnesses and somatic symptoms. Health Psychology: Offcial Journal of the Division of Health Psychology, American Psychological Association, 38(2), 172–181. doi: 10.1037/hea0000707 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- O’Brien EC, Greiner MA, Sims M, Hardy NC, Wang W, Shahar E, … Curtis LH (2015). Depressive symptoms and risk of cardiovascular events in blacks: Findings from the jackson heart study. Circulation.Cardiovascular Quality and Outcomes, 8(6), 552–559. doi: 10.1161/CIRCOUTCOMES.115.001800 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ortiz R, Kluwe B, Odei JB, Echouffo Tcheugui JB, Sims M, Kalyani RR, … Joseph JJ (2019). The association of morning serum cortisol with glucose metabolism and diabetes: The jackson heart study. Psychoneuroendocrinology, 103, 25–32. doi:S0306-4530(18)30715-7 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paradies Y, Ben J, Denson N, Elias A, Priest N, Pieterse A, … Gee, G. (2015). Racism as a determinant of health: A systematic review and meta-analysis. PloS One, 10(9), e0138511. doi: 10.1371/journal.pone.0138511 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401 [Google Scholar]
- Restituto P, Galofre JC, Gil MJ, Mugueta C, Santos S, Monreal JI, & Varo N (2008). Advantage of salivary cortisol measurements in the diagnosis of glucocorticoid related disorders. Clinical Biochemistry, 41(9), 688–692. doi:101016/j.clinbiochem.2008.01.015 [doi] [DOI] [PubMed] [Google Scholar]
- Seeman T, Merkin SS, Crimmins E, Koretz B, Charette S, & Karlamangla A (2008). Education, income and ethnic differences in cumulative biological risk profiles in a national sample of US adults: NHANES III (1988–1994). Social Science & Medicine (1982), 66(1), 72–87. doi:S0277-9536(07)00483-2 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Seeman TE, Singer BH, Rowe JW, Horwitz RI, & McEwen BS (1997). Price of adaptation–allostatc load and its health consequences. MacArthur studies of successful aging. Archives of Internal Medicine, 157(19), 2259–2268. [PubMed] [Google Scholar]
- Segeda V, Izakova L, Hlavacova N, Bednarova A, & Jezova D (2017). Aldosterone concentrations in saliva reflect the duration and severity of depressive episode in a sex dependent manner. Journal of Psychiatric Research, 91, 164–168. doi:S0022-3956(17)30193-0 [pii] [DOI] [PubMed] [Google Scholar]
- Seplaki CL, Goldman N, Glei D, & Weinstein M (2005). A comparative analysis of measurement approaches for physiological dysregulation in an older population. Experimental Gerontology, 40(5), 438–449. doi:S0531-5565(05)00039-2 [pii] [DOI] [PubMed] [Google Scholar]
- Sims M, Redmond N, Khodneva Y, Durant RW, Halanych J, & Safford MM (2015). Depressive symptoms are associated with incident coronary heart disease or revascularization among blacks but not among whites in the reasons for geographical and racial differences in stroke study. Annals of Epidemiology, 25(6), 426–432. doi: 10.1016/j.annepidem.2015.03.014 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Srinivas S, Rajendran S, Anand K, & Chockalingam A (2018). Self-reported depressive symptoms in adolescence increase the risk for obesity and high BP in adulthood. International Journal of Cardiology, 269, 339–342. doi:S0167-5273(18)33005-5 [pii] [DOI] [PubMed] [Google Scholar]
- Tajeu GS, Mennemeyer S, Menachemi N, Weech-Maldonado R & Kilgore M (2017). Cost-effectiveness of antihypertensive medication: Exploring race and sex differences using data from the REasons for geographic and racial differences in stroke study. Medical Care, 55(6), 552–560. doi: 10.1097/MLR.0000000000000719 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Taqueti VR (2018). Sex differences in the coronary system. Advances in Experimental Medicine and Biology, 1065, 257–278. doi: 10.1007/978-3-319-77932-4_17 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Taylor HA, Jr,, Wilson JG, Jones DV, Sarpong DF, Srinivasan A, Garrison RJ, … Wyatt SB (2005). Toward resolution of cardiovascular health disparities in african americans: Design and methods of the jackson heart study. Ethnicity & Disease, 15(4 Suppl 6), S6–4–17 [PubMed] [Google Scholar]
- Thomas JL, Jones GN, Scarinci IC, Mehan DJ, & Brantley PJ (2001). The utility of the CES-D as a depression screening measure among low-income women attending primary care clinics. the center for epidemiologic studies-depression. International Journal of Psychiatry in Medicine, 31(1), 25–40. [DOI] [PubMed] [Google Scholar]
- Thomas KL, Honeycutt E, Shaw LK, & Peterson ED (2010). Racial differences in long-term survival among patients with coronary artery disease. American Heart Journal, 160(4), 744–751. doi: 10.1016/j.ahj.2010.06.014 [doi] [DOI] [PubMed] [Google Scholar]
- Thorpe RJ Jr, Fesahazion RG, Parker L, Wilder T, Rooks RN, Bowie JV, … LaVeist TA (2016). Accelerated health declines among african americans in the USA. Journal of Urban Health : Bulletin of the New York Academy of Medicine, 93(5), 808–819. doi: 10.1007/s11524-016-0075-4 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tingley D, Yamamoto T, Hirose K, Keele L, & Imai K (2014). Mediation: R package for causal mediation analysis. Journal of Statistical Software, 59(5) [Google Scholar]
- Tran HV, Waring ME, McManus DD, Erskine N, Do VTH, Kiefe C, & Goldberg RJ (2017). Underuse of effective cardiac medications among women, middle-aged adults, and racial/ethnic minorities with coronary artery disease (from the national health and nutrition examination survey 2005 to 2014). The American Journal of Cardiology, 120(8), 1223–1229. doi:S0002-9149(17)31168-2 [pii] [DOI] [PubMed] [Google Scholar]
- US Preventive Services Task Force, Curry SJ, Krist AH, Owens DK, Barry MJ, Caughey AB, … Wong JB (2018). Risk assessment for cardiovascular disease with nontraditional risk factors: US preventive services task force recommendation statement. Jama, 320(3), 272–280. doi: 10.1001/jama.2018.8359 [doi] [DOI] [PubMed] [Google Scholar]
- Vigili de Kreutzenberg S, Solini A, Vitolo E, Boi A, Bacci S, Cocozza S, … Baroni MG (2017). Silent cororary heart disease in patients with type 2 diabetes: Application of a screening approach in a follow-up study. Journal of Diabetes and its Complications, 31(6), 952–957 doi:S1056-8727(17)30111-3 [pii] [DOI] [PubMed] [Google Scholar]
- Wandell P, Carlsson AC., Holzmann MJ, Arnlov J, Sundquist J, & Sundquist K (2018). The association between relevant co-morbidities and prevalent as well as incident heart failure in patients with atrial fibrillation. Journal of Cardiology, 72(1), 26–32. doi:S0914-5087(18)30003-0 [pii] [DOI] [PMC free article] [PubMed] [Google Scholar]
- Writing Group Members, Mozaffarian D., Benjamin EJ, Go AS, Arnett DK, Blaha MJ, … Stroke Statistics Subcommittee. (2016). Executive summary: Heart disease and stroke statistics–2016 update: A report from the american heart association. Circulation, 133(4), 447–454. doi: 10.1161/CIR.0000000000000366 [doi] [DOI] [PubMed] [Google Scholar]
- Wu Y, Sun D, Wang B, Li Y, & Ma Y (2018). The relationship of depressive symptoms and functional and structural markers of subclinical atherosclerosis: A systematic review and meta-analysis. European Journal of Preventive Cardiology, 25(7), 706–716. doi: 10.1177/2047487318764158 [doi] [DOI] [PubMed] [Google Scholar]
- Zhao D, Post WS, Blasco-Colmenares E, Cheng A, Zhang Y, Deo R, … Gullar E. (2019). Racial differences in sudden cardiac death. Circulation, 139(14), 1688–1697. doi: 10.1161/CIRCULATIONAHA.118.036553 [doi] [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.