Hostname: page-component-788cddb947-xdx58 Total loading time: 0 Render date: 2024-10-09T18:49:17.946Z Has data issue: false hasContentIssue false

Associations of perceived stress, depressive symptoms, and caregiving with inflammation: a longitudinal study

Published online by Cambridge University Press:  11 May 2022

Joanne Elayoubi*
Affiliation:
School of Aging Studies, University of South Florida, Tampa, FL, USA
William E. Haley
Affiliation:
School of Aging Studies, University of South Florida, Tampa, FL, USA
David L. Roth
Affiliation:
Center on Aging and Health, Division of Geriatric Medicine and Gerontology, Johns Hopkins University, Baltimore, MD, USA
Mary Cushman
Affiliation:
Department of Medicine & Pathology, University of Vermont, Burlington, VT, USA
Orla C. Sheehan
Affiliation:
Center on Aging and Health, Division of Geriatric Medicine and Gerontology, Johns Hopkins University, Baltimore, MD, USA
Virginia J. Howard
Affiliation:
Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA
Melissa deCardi Hladek
Affiliation:
School of Nursing, Johns Hopkins University, Baltimore, MD, USA
Gizem Hueluer
Affiliation:
School of Aging Studies, University of South Florida, Tampa, FL, USA
*
Correspondence should be addressed to: Joanne Elayoubi, School of Aging Studies, University of South Florida, MHC 1305, Tampa, FL 33620, USA. Phone: (813) 974-2414. Email: elayoubij@usf.edu.
Rights & Permissions [Opens in a new window]

Abstract

Objectives:

Higher inflammation has been linked to poor physical and mental health outcomes, and mortality, but few studies have rigorously examined whether changes in perceived stress and depressive symptoms are associated with increased inflammation within family caregivers and non-caregivers in a longitudinal design.

Design:

Longitudinal Study.

Setting:

REasons for Geographic And Racial Differences in Stroke cohort study.

Participants:

Participants included 239 individuals who were not caregivers at baseline but transitioned to providing substantial and sustained caregiving over time. They were initially matched to 241 non-caregiver comparisons on age, sex, race, education, marital status, self-rated health, and history of cardiovascular disease. Blood was drawn at baseline and approximately 9.3 years at follow-up for both groups.

Measurements:

Perceived Stress Scale, Center for Epidemiological Studies-Depression, inflammatory biomarkers, including high-sensitivity C-reactive protein, D dimer, tumor necrosis factor alpha receptor 1, interleukin (IL)-2, IL-6, and IL-10 taken at baseline and follow-up.

Results:

Although at follow-up, caregivers showed significantly greater worsening in perceived stress and depressive symptoms compared to non-caregivers, there were few significant associations between depressive symptoms or perceived stress on inflammation for either group. Inflammation, however, was associated with multiple demographic and health variables, including age, race, obesity, and use of medications for hypertension and diabetes for caregivers and non-caregivers.

Conclusions:

These findings illustrate the complexity of studying the associations between stress, depressive symptoms, and inflammation in older adults, where these associations may depend on demographic, disease, and medication effects. Future studies should examine whether resilience factors may prevent increased inflammation in older caregivers.

Type
Original Research Article
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - ND
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is unaltered and is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use or in order to create a derivative work.
Copyright
© International Psychogeriatric Association 2022

Introduction

It has been widely believed that there are close linkages between stress, depression, and inflammation. Several prior studies have found associations of higher inflammation levels with higher stress (Hänsel et al., Reference Hänsel, Hong, Cámara and Von Kaenel2010; Liu et al., Reference Liu, Wang and Jiang2017), poor physical and mental health outcomes, and mortality (Wirtz and Von Känel, Reference Wirtz and Von Känel2017; Glaser and Kiecolt-Glaser, Reference Glaser and Kiecolt-Glaser2005; Prior et al., Reference Prior2016). Biomarkers of inflammation have been used as objective health measures for studying mechanisms that link chronic stress to physical and mental health (Hänsel et al., Reference Hänsel, Hong, Cámara and Von Kaenel2010; Piazza et al., Reference Piazza, Almeida, Dmitrieva and Klein2010). These biomarkers include cytokines, interleukin-6 (IL-6), C-reactive protein (CRP), and tumor necrosis factor alpha receptor 1 (TNFR1).

Research findings on the association between stress, depression (or depressive symptoms), and inflammation have been mixed. Several prospective population-based studies examining inflammatory biomarkers as predictors of inflammation found higher levels of IL-6 and CRP significantly associated with depressive symptoms (Gimeno et al., Reference Gimeno2009; Valkanova et al., Reference Valkanova, Ebmeier and Allan2013; Zalli et al., Reference Zalli, Jovanova, Hoogendijk, Tiemeier and Carvalho2016). A prospective cohort study found depressive symptoms to be associated with changes in inflammatory biomarker levels over time (Stewart et al., Reference Stewart, Rand, Muldoon and Kamarck2009), while other studies had found inconsistent associations between inflammatory pathways and depression. For example, results from a cross-sectional, population-based study found higher IL-6 levels and not CRP to be associated with future major depressive episodes in older adults (Bremmer et al., Reference Bremmer2008). In addition, several meta-analyses did not find consistent associations between inflammation and depression including one that examined TNF alpha biomarker (Haapakoski et al., Reference Haapakoski, Mathieu, Ebmeier, Alenius and Kivimäki2015) and another that examined several inflammatory biomarkers (IL-2, IL-8, and IL-10) (Dowlati et al., Reference Dowlati2010). Moreover, these meta-analyses mostly used cross-sectional studies and studies that did not examine the association of inflammation on changes in depressive symptoms over time. A meta-analysis of methodologically rigorous studies reported no association between CRP and depression (Horn et al., Reference Horn, Long, Nelson, Allen, Fisher and Byrne2018).

Family caregiving is a chronically stressful life experience that leads to high levels of perceived stress and depressive symptoms and has been linked in some studies with increased inflammation (Allen et al., Reference Allen2017; Epstein-Lubow et al., Reference Epstein-Lubow, Gaudiano, Hinckley, Salloway and Miller2010; Pinquart and Sörensen, Reference Pinquart and Sörensen2003; Talley and Crews, Reference Talley and Crews2007). Family caregiving is a public health concern where caregivers are often exposed to high levels of stress usually for long periods of time (Talley and Crews, Reference Talley and Crews2007). A better understanding of the possible associations between perceived stress, depressive symptoms, and inflammation may benefit family caregivers’ mental and physical health. However, a recent systematic review and meta-analysis of caregiving and inflammation found a small but statistically significant difference (0.16 standard deviation units) across all inflammatory biomarkers between caregivers and non-caregiver controls (Roth et al., Reference Roth, Sheehan, Haley, Jenny, Cushman and Walston2019), with caregivers showing greater inflammation than controls. Few studies have examined whether caregiving is associated with increased inflammation over time compared to non-caregiving controls, and those that did found little evidence linking caregiving with increased inflammation (Allen et al., Reference Allen2017; Potier et al., Reference Potier, Degryse and De Saint-Hubert2018).

To date, relatively few studies have examined whether longitudinal changes in perceived stress and depressive symptoms are associated with changes in inflammatory biomarkers over time in family caregiving. The ability to examine these effects in a relatively low-stress general sample compared to a group that has developed substantial increases in stress over time could be valuable in understanding longitudinal relationships between perceived stress, depressive symptoms, and inflammation. The Caregiving Transitions Study (CTS), a longitudinal population-based study, has unique data to address these relationships over time (Roth et al., Reference Roth2020a). This study has the advantage of including participants who transitioned into extensive and sustained family caregiving over time, had marked increases in perceived stress and depressive symptoms over time, and matched with a case–control group (non-caregivers) who had little change in perceived stress or depressive symptoms over time (Haley et al., Reference Haley2020). In a recent report from this study, Roth and colleagues (Roth et al., Reference Roth2020b) studied changes in six inflammatory biomarkers over time and found that those who had transitioned to extensive and sustained caregiving showed increased inflammation on only one biomarker, TNFR1 with a small effect size (d = 0.14).

In the present analyses, we examined whether individual differences in perceived stress or depressive symptoms were associated with changes in inflammatory biomarkers over time, and whether these associations differed among caregivers (who showed marked increases in perceived stress and depressive symptoms over time) and a non-caregiver comparison group (who showed stability over time on these variables). We hypothesized that 1) baseline measures of perceived stress or depressive symptoms would be associated with higher inflammation at follow-up, 2) perceived stress or depressive symptoms measured at follow-up would be associated with increased inflammation, and 3) greater changes in perceived stress or depressive symptoms (from baseline to follow-up) would be associated with greater increases in inflammation over time for transitioned caregivers only. Because inflammatory biomarkers can be affected by a variety of factors including sociodemographic variables, chronic conditions such as obesity and diabetes (Ouakinin et al., Reference Ouakinin, Barreira and Gois2018; Siddiqui et al., Reference Siddiqui, Desai, Sharma, Aslam, Sinha and Madhu2019), and medications such as antidepressants and statins (Hussain and Ballantyne, Reference Hussain and Ballantyne2021; Vogelzangs et al., Reference Vogelzangs2012), we conducted both unadjusted and adjusted analyses and considered these factors as potential confounders.

Methods

Data source

REGARDS and Caregiving Transitions Study procedures and participants

The present study uses data from a national longitudinal cohort study, the REasons for Geographic and Racial Differences in Stroke (REGARDS) study and an ancillary study of REGARDS, the CTS. In 2003–2007, the REGARDS study enrolled 30,239 African American and White adults 45 or more years of age living in the United States. Residents of the “stroke belt” region of the United States (southern states with higher mortality from stroke) and African Americans were oversampled by design. Participants who provided a verbal informed consent were given a baseline computer-assisted telephone interview (CATI) that assessed demographics, medical history, and a wide variety of potential risk factors for stroke. Exclusion criteria included race other than non-Hispanic White or African American, history of chemotherapy/radiation for cancer in past 2 years and living in or transitioning to a nursing home (NH). Eligible participants had a subsequent in-home assessment where biological specimens (including blood draws), physical measurements, and medication use were taken and recorded. Semiannual follow-up CATIs were conducted to monitor changes in health, including possible strokes and cognitive functioning. A second comprehensive telephone interview and in-home visit were conducted 2013–2016, approximately 9–10 years after initial enrollment, and extensive follow-up information was gathered including blood draws. Additional information on the design, sampling, enrollment, and follow-up procedures in the REGARDS study has been documented elsewhere (Howard et al., Reference Howard2017).

Participants at the baseline CATI were asked “Are you currently providing care on an ongoing basis to a family member with a chronic illness or disability?” This could include any form of help like watching, transporting, bathing, or dressing the care recipient. Participants who replied “yes” to this question were categorized as caregivers and provided more information as to whether they cared for a spouse, resided with the care recipient, experienced any mental or emotional strain from the care provided, and have been included in several papers addressing the mental and physical health consequences of caregiving (Roth et al., Reference Roth, Perkins, Wadley, Temple and Haley2009; Roth et al., Reference Roth, Haley, Hovater, Perkins, Wadley and Judd2013; Roth et al., Reference Roth, Brown, Rhodes and Haley2018). Participants who replied “no” were categorized as non-caregivers and were considered for inclusion in the current analyses. After REGARDS 2nd telephone and in-home assessment (approximately 11.8 years after the baseline CATI), an updated caregiving status information was collected through a Caregiving Screening CATI module administered as part of the standard REGARDS semiannual follow-up CATI (Roth et al., Reference Roth2020a). The total number of participants who answered “yes” to being a caregiver at the Caregiver Screening CATI and had previously answered “no” to being a caregiver at the baseline CATI was 1229. These participants reporting being a caregiver at the 2nd CATI were asked questions on a more detailed CTS Enrollment Interview to determine their exposure to caregiving and whether they met eligibility criteria including when they initiated the caregiving role (month and year), whether caregiving was provided on a continuous basis since that time, their personal relationship to the care recipient, whether the care recipient currently resided or ever resided in a NH or assisted living facility (ALF), and the number of hours per week of caregiving provided to the care recipient. To be eligible to participate as an incident caregiver in the CTS, the transition into the family caregiving role had to occur at least 6 months after the first in-home assessment and at least 3 months before the 2nd in-home assessment of REGARDS to assure that the second blood draw was during a period of caregiving (since blood was drawn at both REGARDS in-home assessments).

Exclusion criteria for incident caregivers in the CTS included the caregiver not having usable blood samples at either of the REGARDS in-home assessments, the care recipient living in a NH/ALF or other residential care setting, caregiving duties less than 5 hours per week, or the caregiver residing more than 50 miles from the care recipient. Of the 1229 potential incident caregivers, 251 incident caregivers met eligibility criteria and accepted participation in the CTS (Haley et al., Reference Haley2020; Roth et al., Reference Roth2020a).

The 10,254 participants who were not caregivers at either assessment were considered potential non-caregivers. Once an incident caregiver was enrolled, a pool of non-caregivers was identified that individually matched the caregiver on seven demographic and health history factors including, age, sex, race, education, marital status, self-rated health at baseline, and self-reported history of serious cardiovascular disease (Haley et al., Reference Haley2020; Roth et al., Reference Roth2020a). These potential participants were randomly called until one comparison was determined to be eligible and agreed to participate. Additional matching restrictions for spouse caregivers included non-caregiving comparison participants being married, and matching restrictions for caregivers of a parent included non-caregivers having at least one living parent. These matched non-caregivers then completed the CTS Enrollment Interview to confirm that they had not been family caregivers at any point during their participation in the REGARDS study, and to complete assessments of their psychological well-being and health. Caregivers and non-caregivers with missing biomarker data from blood drawn during REGARDS 1st and 2nd in-home assessments were excluded from this study. Participants who had at least one of the six biomarker measures obtained from both REGARDS in-home assessments were included in the present study for a total number of 239 caregivers and 241 non-caregiver comparison participants. Due to missing biomarker data, caregivers and non-caregivers were not all individually matched; however, analyses showed they did not differ significantly on any of these variables (Roth et al., Reference Roth2020b). The protocols of both the REGARDS parent study and the CTS ancillary study were approved by the Institutional Review Boards of the participating institutions, and informed consent was provided by all participants.

Measurements

Perceived stress and depressive symptoms

Perceived stress and depressive symptoms were assessed during the REGARDS baseline CATI with a four-item version of the Cohen Perceived Stress Scale (PSS) (Cohen et al., Reference Cohen, Kamarck and Mermelstein1983) and a four-item version of the Center for Epidemiological Studies-Depression (CES-D) scale (Melchior et al., Reference Melchior, Huba, Brown and Reback1993), respectively. Higher scores indicated more perceived stress or depressive symptoms. At the CTS Enrollment Interview, caregivers and non-caregivers repeated the identical four-item version of the PSS but completed a 10-item CES-D measure (Andresen et al., Reference Andresen, Malmgren, Carter and Patrick1994). We used this longer version at follow-up to provide greater variability on the measure. For data analysis, we transformed the 4-item version of the CES-D to estimate a 10-item score, using regression analyses as reported in prior publications (Haley et al., Reference Haley2020; Melchior et al., Reference Melchior, Huba, Brown and Reback1993), with the two versions on different variants of the CES-D being highly correlated (r = 0.82). As noted below, we also used a cutoff score of 10 or more on the 10-item CES-D to classify participants as having clinically significant depressive symptoms, the more conservative of two cut points proposed (Andresen et al., Reference Andresen, Malmgren, Carter and Patrick1994), and as described by Haley et al. (Reference Haley2020), during sensitivity analyses.

Biomarker assays

Participants had blood samples obtained in the morning and collected by trained phlebotomists. Participants were reminded to fast 10–12 hours overnight, and visits to participant homes were conducted in the mornings to permit fasting status. Information about the processing of biological samples can be found in previous publications (Gillett et al., Reference Gillett, Boyle, Zakai, Mcclure, Jenny and Cushman2014; Howard et al., Reference Howard2005; Roth et al., Reference Roth2020a; Roth et al., Reference Roth2020b).

The six circulating blood biomarkers of inflammation assayed in this study included high-sensitivity CRP, D dimer, TNFR1, IL-2, IL-6, and IL-10. These biomarkers were selected based on review of the previous literature on measures used most in studies of inflammation, immunity, and caregiving and the availability of valid measures from stored frozen blood samples. Additional information on the measurement can be found in a previous publication based on the same biomarker data from the REGARDS study (Roth et al., Reference Roth2020b). The time interval between REGARDS 1st and 2nd in-home assessments ranged from 7.6 to 12.4 years and averaged 9.3 years for participants included in the present analyses.

Covariates

We used covariate information collected at the time of REGARDS 2nd in-home assessment, except for age which was taken at the time of the Caregiving Transitions enrollment. Other covariates included sex (women = 1, men = 0), race (Black = 1 and White = 0), education (college graduate or above = 1, and some college, high school graduate, and less than school graduate = 0), marital status (married/cohabitating = 1 and single/never married/divorced/widowed = 0). Nicotine use was indicated by a binary variable indicating whether a participant was smoking (coded as 1) or never smoked/smoked in the past (coded as 0). Using the Centers for Disease Control and Prevention (CDC, 2021) BMI obesity cutoff point of 30, obesity was indicated by a binary variable (BMI ≥ 30 = 1 and BMI < 30 = 0). Similarly, the use of statins and antidepressant medications were indicated by two binary variables (“yes” = 1; “no” = 0). Taking hypertension and diabetes medications were also indicated by binary variables (“yes” = 1 and “no” = 0).

Statistical Analyses

A log (base 2) transformation was made for each of the six biomarkers at each assessment because biomarker levels were highly skewed. We considered these variables using a strategy previously described (Jenny et al., Reference Jenny2012) in which the dependent variable is change over time (Δ) and defined as the difference between REGARDS 2nd in-home assessment (T 2) and REGARDS 1st in-home assessment (T 1 ) on the log (base 2) scores: Δ = Log2 (Y 2) − Log2 (Y 1), where Y 2 and Y 1 represent the raw values of the specific biomarker of interest at T 2 and T 1 , respectively.

Following Roth and colleagues (2020b) who used the same data set, after the log (base 2) transformations possible outliers for each biomarker were identified using the Tukey interquartile range (IQR) method (Tukey, Reference Tukey1977). The IQR, which is the difference between the 75th (Q3) and 25th (Q1) percentiles, was calculated, and all values greater than 3*IQR above Q3 were considered extreme outliers and recoded as missing. There were no values that were less than 3*IQR below Q1. CRP, TNFR1, and D dimer did not have extreme outliers; however, for biomarkers IL-2, IL-6, and IL-10, values above 0.95, 8.23, and 2.22 pg/mL, respectively, were extreme outliers and recoded as missing. For IL-2 and IL-6, 0.7% of the values were identified as extreme outliers, and for IL-10, 1.5% of the values were designated as being extreme outliers. Overall, less than 0.5% values were extreme outliers and coded missing.

Independent t-tests and chi-square tests were performed to examine group differences between caregivers and non-caregivers at T 2. We also assessed differences between caregivers and non-caregivers for perceived stress and depressive symptoms measured at baseline and changes in perceived stress and depressive symptoms from baseline to T 2. Pearson’s correlation statistical analyses were performed to examine the correlations between all covariates and T 2 inflammatory biomarkers.

Separate regression models were conducted to examine change in inflammatory levels for each of the biomarkers at T 2 as a dependent outcome, controlling for that biomarker at T 1 (centered at sample means) and evaluated whether there were interaction effects between caregiving and perceived stress or caregiving and depressive symptoms. First, we examined baseline perceived stress or baseline depressive symptoms as predictors for change in inflammatory levels for each of the biomarkers at T 2 for both caregiving and non-caregiving groups. Next, we assessed whether perceived stress or depressive symptoms at T 2 predicted change in inflammatory levels for each biomarker at T 2 in both groups. We then examined change in perceived stress or change in depressive symptoms on change in inflammatory levels at T 2. Statistical significance was evaluated at p < 0.05. These analyses were conducted first without covariate adjustment, and then with adjustment for the covariates. Analyses were performed with SAS software Version 9.4. We used data from all caregivers and non-caregivers who had available biomarker data after the recoding of outliers as missing.

Sensitivity analyses

To examine the role of clinically significant levels of depressive symptoms, we created a new binary variable that dichotomized depressive symptoms (0 = absent, 1 = present) based on whether participants had a score of 10 or greater on the CES-D, a clinically validated cutoff (Andresen et al., Reference Andresen, Malmgren, Carter and Patrick1994). We assessed changes in clinically significant depressive symptoms over time from baseline and T 2 as clinically significant depressive symptoms at follow-up minus clinically significant depressive symptoms at baseline with scores of 0, −1, and 1 indicating stability, improvement, and increase over time, respectively.

Results

Sample characteristics

Table 1 shows descriptive statistics for all study variables for both caregiver and non-caregiver groups at REGARDS 2nd in-home assessment (T 2 ). Mean age was 72 years (SD = 8) and ranged from 55 to 93 years; 65% were women; and 35.2% were Black. Due to the matching procedures, caregivers and non-caregivers did not differ statistically on sociodemographic variables including age, sex, race, education, and marital status. Groups did not differ in baseline perceived stress and depressive symptoms, but caregivers had significantly higher depressive symptoms (M = 6.96, SD = 6.29) compared to non-caregivers (M = 3.29, SD = 4.43) p < 0.001 at T 2 . Caregivers also had significantly greater changes in depressive symptoms over time (M = 3.47, SD = 6.48) compared to non-caregivers (M = −0.23, SD = 4.69, p < 0.001). Similarly, caregivers had significantly higher perceived stress at T 2 (M = 4.55, SD = 3.01) compared to non-caregivers (M = 2.53, SD = 2.56) p < 0.001. Caregivers also showed significantly greater increases in perceived stress over time, (M = 1.69, SD = 3.27) compared to non-caregivers (M = −0.52, SD = 2.81) p < 0.001.

Table 1. Descriptive characteristics of study variables

SD = standard deviation; PSS = Perceived Stress Scale; CES-D = Center for Epidemiological Studies Depression Scale.

a Measured in years at Caregiving Transitions enrollment.

b Education = 1 for having a college degree or above and some college, high school graduate, and less than school graduate = 0.

c Covariates measured from REasons for Geographic and Racial Differences in Stroke (REGARDS) 2nd in-home assessment that include marital status = 1 for married/cohabitating and single/never married/divorced/widowed = 0, nicotine use = 1 for smoking and never smoked/smoked in the past = 0, and obesity using BMI > or equal to 30 = 1 and BMI < 30 = 0. All medication use were binary-coded “yes” = 1 and “no” = 0.

d Cohen’s 4-item measure of perceived stress from REGARDS baseline computer-assisted telephone interview (CATI).

e Cohen’s 4-item measure of perceived stress at Caregiving Transitions enrollment.

f Depressive symptoms from REGARDS baseline CATI.

g Depressive symptoms measured at Caregiving Transitions enrollment.

p-Value for t-tests or chi-square assessed differences between caregivers and non-caregivers.

Bolded values indicate significant results at p < 0.05.

With respect to medication use and chronic diseases at T 2 , caregivers reported being more likely to take diabetes medications (23.85%) compared to non-caregivers (17.84%) p = 0.011. Caregivers also reported higher likelihood of antidepressant medication use (18.83%) compared to non-caregivers (10.79%) p = 0.013, and caregivers were more likely to be obese (43.51%) compared to non-caregivers (34.44%) p = 0.042.

Perceived stress at baseline, follow-up, and changes over time

Results from unadjusted to adjusted models did not change; therefore, only adjusted models are reported. We first examined baseline perceived stress, then perceived stress at T 2 , followed by change in perceived stress with inflammatory biomarker levels at T 2 controlling for the same biomarker at T 1 for both caregivers and non-caregivers. Table 2 shows the association of T 1 biomarker levels with T 2 biomarker levels for all six biomarkers, indicating that individual differences in biomarker levels were relatively stable over time. The results for our first hypothesis examining baseline perceived stress and inflammatory biomarker levels at T 2 controlling for the same biomarker at T 1 for both caregivers and non-caregivers show no significant associations. There were no significant effects for our second hypothesis examining perceived stress at T 2 with inflammatory levels at T 2 . However, in support of our third hypothesis, there were significant effects for changes in perceived stress over time, and the interaction of caregiver status over time, with IL-10 at T 2. Examination of the interaction effect for caregiving and perceived stress revealed that an increase in perceived stress over time was significantly associated with an increase in IL-10 at T 2 for caregivers only (b = 0.04, SE = 0.02, p = 0.02).

Table 2. Estimates from regression analyses of predictors on inflammatory biomarker levels at follow-up

N = 480. T 1 Biomarkers are inflammatory biomarkers assessed from REasons for Geographic and Racial Differences in Stroke (REGARDS) 1st in-home assessment. Models presented are adjusted for covariates including age measured at caregiving transition enrollment; sex, race, education, and marital status, nicotine use, obesity, hypertension and diabetes medication, antidepressants, and statins measured at REGARDS 2nd in-home assessment.

SE = standard error; b = estimated coefficients; PSS = Perceived Stress Scale; CES-D = Center for Epidemiological Studies Depression Scale.

a Cohen’s 4-item measure of perceived stress from REGARDS baseline computer-assisted telephone interview (CATI).

b Cohen’s 4-item measure of perceived stress at Caregiving Transitions enrollment.

c Depressive symptoms from REGARDS baseline CATI.

d Depressive symptoms measured at Caregiving Transitions enrollment.

Bolded values indicate significant results at p < 0.05.

Depressive symptoms at baseline, follow-up, and changes over time

We next examined the role of depressive symptoms at baseline, depressive symptoms at T 2 , and change in depressive symptoms with inflammatory biomarker levels at T 2 controlling for the same biomarker at T 1 . An initial analysis was conducted without covariate adjustment (data not shown). Table 2 shows the association of T 1 biomarker levels with T 2 biomarker levels for all six inflammatory biomarkers. Baseline depressive symptoms were associated with higher levels of D dimer (b = 0.04, SE = 0.02, p = 0.03), but there were no significant findings to support our hypotheses that baseline, T 2 , or changes in depressive symptoms were associated with inflammation for the sample as a whole or for caregivers compared to non-caregivers.

Our sensitivity analyses replacing T 1 and T 2 depressive symptoms with the binary variable for clinically significant depressive symptoms in the same regression analyses led to similar findings. Baseline clinically significant depressive symptoms were associated with higher levels of D dimer (b = 0.73, SE = 0.27, p = 0.01), but there were no other significant associations found for baseline or T 2 clinically significant depressive symptoms. Similarly, there were no significant associations when examining change in clinically significant depressive symptoms over time on inflammation.

Covariates and their associations with Inflammatory biomarkers

Table 3 refers to the correlations between potential confounders and T 2 biomarkers for the whole sample. Age was associated with higher D dimer, IL-6, IL-2, and TNFR1 and lower CRP. Black race was associated with higher inflammatory biomarkers levels for most biomarkers except IL-2 and TNFR1. Obesity and taking hypertension and diabetes medications were associated with higher levels of IL-6 and TNFR1 inflammatory biomarkers.

Table 3. Correlations between covariates and inflammatory biomarkers measured at follow-up

N = 480.

a Measured in years at Caregiving Transitions enrollment.

b Education = 1 for having a college degree or above and some college, high school graduate, and less than school graduate = 0.

c Covariates measured from REasons for Geographic and Racial Differences in Stroke (REGARDS) 2nd in-home assessment include marital status = 1 for married/cohabitating and single/never married/divorced/widowed = 0, nicotine use = 1 for smoking and never smoked/smoked in the past = 0, and obesity using BMI > or equal to 30 = 1 and BMI < 30 = 0. All medication use were binary-coded “yes” = 1 and “no” = 0.

Bolded values indicate significant results at p < 0.05.

Discussion

The results from this study add to the growing literature on the impact of stress and depressive symptoms with inflammatory processes associated with long-term family caregiving. Our previous results from the CTS showed few significant differences in changes in inflammatory biomarker levels between caregivers and non-caregivers (Roth et al., Reference Roth2020b). Concerning our first hypothesis, we did find that baseline depressive symptoms were significantly associated with D dimer; however, baseline depressive symptoms were not associated with other biomarkers, and baseline perceived stress was not associated with higher inflammation. Contrary to our second hypothesis, perceived stress or depressive symptoms at T 2 did not predict increased inflammation. In support of our third hypothesis, a greater change in perceived stress was associated with an increase in inflammation for IL-10 only in individuals who transitioned into caregiving, while non-caregivers had a significant decrease in inflammation for the same biomarker. Of note, however, these three significant effects found among 72 analyses for the association of depressive symptoms, perceived stress, caregiving, and inflammatory biomarkers represent only 4% of these analyses, below a chance level with the 0.05 significance level used.

The CTS allowed us to examine indicators of stress in inflammation in both non-caregivers and in family caregivers by using biomarker data before and after the onset of caregiving. Our findings showed that although caregivers had much higher increases in perceived stress and depressive symptoms after the onset of caregiving compared to an individually matched comparison group of non-caregivers, perceived stress and depressive symptoms on all occasions (baseline, follow-up, and change over time) were inconsistently related to increased inflammatory biomarker levels, either for non-caregivers or for incident caregivers.

It is noteworthy that IL-10 is an anti-inflammatory cytokine (Saxton, Reference Saxton2021) and animal studies in rodents showed reductions in IL-10 with psychological distress (Voorhees et al., Reference Voorhees2013). These increases in IL-10 biomarker with heightened levels of perceived stress may imply caregivers had an increase in anti-inflammatory processes that contributed to a reduction in overall systematic inflammation.

When examining the associations of covariates on markers of inflammation at follow-up, we found several covariates to be positively associated with inflammation. With respect to sociodemographic variables, age and race were the strongest indicators significantly associated with increased inflammation. Prior reviews showed inflammation as a significant contributor to age-related diseases (Franceschi and Campisi, Reference Franceschi and Campisi2014; Singh and Newman, Reference Singh and Newman2011), and a prior longitudinal study examining the effects of race and ethnicity on inflammation reported Blacks having higher baseline CRP and greater changes in CRP over time compared to Whites (Zahodne et al., Reference Zahodne, Kraal, Zaheed, Farris and Sol2019). Surprisingly, a decrease in CRP over time was associated with older age, and a decrease in IL-2 and TNFR1 was found among Blacks.

With respect to covariates for chronic conditions and medications, obesity, and taking hypertension and diabetes medications were all significantly associated with increases in IL-6 and TNFR1. Obesity and taking diabetes medications were associated with an increase in CRP, and obesity and taking hypertension medications were significantly associated with an increase in D dimer. Statin use was significantly associated with decreases in CRP; however, antidepressants were significantly associated with increases in CRP and TNFR1. These findings are consistent with prior literature linking obesity and diabetes (Fried et al., Reference Fried, Von Stockert, Haslbeck, Lamers, Schoevers and Penninx2020; Mclaughlin et al., Reference Mclaughlin2021; Ouakinin et al., Reference Ouakinin, Barreira and Gois2018; Siddiqui et al., Reference Siddiqui, Desai, Sharma, Aslam, Sinha and Madhu2019) and medications such as antidepressants and statins (Hussain and Ballantyne, Reference Hussain and Ballantyne2021; Vogelzangs et al., Reference Vogelzangs2012) with inflammation. Depressive symptoms may be associated with inflammation by a common pathophysiology with obesity and metabolic conditions (Fried et al., Reference Fried, Von Stockert, Haslbeck, Lamers, Schoevers and Penninx2020; Lamers et al., Reference Lamers, Milaneschi, De Jonge, Giltay and Penninx2018). Moreover, the strong associations between diet, obesity, and inflammation (Aleksandrova et al., Reference Aleksandrova, Koelman and Rodrigues2021; Koelman et al., Reference Koelman, Egea Rodrigues and Aleksandrova2022) warrants the need for future studies to examine the possible lifestyle mechanisms (such as diet and exercise) associated with changes in inflammation and incident disease. Although it was not assessed in this study, social support has been shown to reduce inflammation (Uchino et al., Reference Uchino, Trettevik, Kent de Grey, Cronan, Hogan and Baucom2018); future research studies need to investigate the impact of social resources on inflammation in family caregiving. The fact that inflammation was generally associated with these demographic, disease, and medication variables in a manner consistent with the literature also suggests that measurement problems in these biomarkers are not a likely explanation for the lack of association with caregiving, perceived stress, or depressive symptoms.

This study had several strengths. Caregivers in the CTS study had to have a certain threshold of caregiving duties to meet eligibility, and the caregiving and non-caregiving samples were carefully matched on many sociodemographic and health variables at baseline. These are major methodological strengths that distinguish the CTS from most other studies of the health effects of family caregiving. There were some design components in our study that may have made it difficult to detect associations between caregiving, perceived stress, depressive symptoms, and inflammation. Participants averaged over 70 years of age at the follow-up assessment and had high prevalence of obesity and use of diabetes and hypertension medications and statins. In addition, the caregivers were significantly more likely than non-caregivers to report obesity and diabetes medications and were nearly twice as likely to be using antidepressants. Although we adjusted for these factors, it is possible they obscured associations that might occur in younger individuals without these characteristics. More frequent repeated measures of blood biomarker assays are needed in future studies to examine the time frame of any possible associations between perceived stress, depressive symptoms, onset of caregiving, and inflammation. For example, it is possible that onset of caregiving has acute effects on inflammation, but that adaptation occurs over time. Despite these limitations, this study uses a unique set of data that allowed us to examine indicators of well-being and inflammation before and after the transition of the caregiving role while accounting for multiple covariates.

Conclusions

While family caregiving is generally viewed as a chronically stressful experience that impacts caregiver well-being, it can also be an experience that brings positivity, meaning, and purpose to a caregiver. These positive attributes in caregiving may be serving a stress-buffering effect (Roth et al., Reference Roth, Fredman and Haley2015; Roth et al., Reference Roth, Brown, Rhodes and Haley2018) on inflammation and may explain why caregivers with significantly heightened perceived stress and depressive symptoms after transitioning into the caregiving role did not exhibit increased inflammation compared to non-caregivers even after controlling for an extensive number of covariates. Moreover, a study using REGARDS reported perceived stress and depressive symptoms as independent predictors of mortality for non-caregivers but not in caregivers (Roth et al., Reference Roth, Brown, Rhodes and Haley2018). This finding in addition to our present study which found no associations between perceived stress, depressive symptoms, and inflammation among caregivers suggests that caregiving may be a unique source of stress associated with psychological and health benefits that mitigates the negative effects of stress on inflammation through increased resiliency. Although the narrative of stress, depressive symptoms, and caregiving increasing inflammation is widely believed to be accurate, this study which used a large sample size of caregivers and demographically matched comparisons group supports the need to establish a more balanced view of family caregiving to better understand and identify caregivers who are more vulnerable to adverse health effects due to the stress of caregiving.

Conflict of interest

None.

Source of funding

This work was supported by a cooperative agreement (U01 NS041588) cofounded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), NIH, Department of Health and Human Services. The Caregiving Transitions Study was further supported by an investigator-initiated grant (RF1 AG050609) from the NIA. Additional support was provided by the Johns Hopkins University Claude D. Pepper Older Americans Independence Center funded by the NIA (P30 AG021334). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or NIA. Representatives of the NINDS were involved in the review of the manuscript but were not directly involved in the collection, management, analysis, or interpretation of the data.

Description of authors’ role

J.E., W.E.H., G.H., D.L.R. V.J.H., and M.C. designed research; J.E. and G.H. performed research; M.C. oversaw biomarker assays; G.H. and J.E. analyzed data; W.E.H., G.H., D.L.R., V.J.H., M.C., O.C.S., and M.H. contributed revisions to the paper; and J.E. wrote the paper.

Acknowledgements

We thank the investigators, staff, and participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at https://www.uab.edu/soph/regardsstudy/.

References

Aleksandrova, K., Koelman, L. and Rodrigues, C. E. (2021). Dietary patterns and biomarkers of oxidative stress and inflammation: a systematic review of observational and intervention studies. Redox Biology, 42, 101869. DOI 10.1016/j.redox.2021.101869.CrossRefGoogle ScholarPubMed
Allen, A. P. et al. (2017). A systematic review of the psychobiological burden of informal caregiving for patients with dementia: focus on cognitive and biological markers of chronic stress. Neuroscience & Biobehavioral Reviews, 73, 123164. DOI 10.1016/j.neubiorev.2016.12.006.CrossRefGoogle ScholarPubMed
Andresen, E. M., Malmgren, J. A., Carter, W. B. and Patrick, D. L. (1994). Screening for depression in well older adults: evaluation of a short form of the CES-D. American Journal of Preventive Medicine, 10, 7784. DOI 10.1016/S0749-3797(18)30622-6.CrossRefGoogle Scholar
Bremmer, M. et al. (2008). Inflammatory markers in late-life depression: results from a population-based study. Journal of Affective Disorders, 106, 249255. DOI 10.1016/j.jad.2007.07.002.CrossRefGoogle ScholarPubMed
Centers for Disease Control and Prevention . (2021). Defining Adult Weight & Obesity. Available at: https://www.cdc.gov/obesity/adult/defining.html; accessed 6 October 2021.Google Scholar
Cohen, S., Kamarck, T. and Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385396. DOI 10.2307/2136404.CrossRefGoogle ScholarPubMed
Dowlati, Y. et al. (2010). A meta-analysis of cytokines in major depression. Biological Psychiatry, 67, 446457. DOI 10.1016/j.biopsych.2009.09.033.CrossRefGoogle ScholarPubMed
Epstein-Lubow, G., Gaudiano, B. A., Hinckley, M., Salloway, S. and Miller, I. W. (2010). Evidence for the validity of the American Medical Association’s caregiver self-assessment questionnaire as a screening measure for depression. Journal of the American Geriatrics Society, 58, 387388. DOI 10.1111/j.1532-5415.2009.02701.x.CrossRefGoogle ScholarPubMed
Franceschi, C. and Campisi, J. (2014). Chronic inflammation (inflammaging) and its potential contribution to age-associated diseases. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences, 69, S4S9. DOI 10.1093/gerona/glu057.CrossRefGoogle ScholarPubMed
Fried, E. I., Von Stockert, S., Haslbeck, J., Lamers, F., Schoevers, R. and Penninx, B. (2020). Using network analysis to examine links between individual depressive symptoms, inflammatory markers, and covariates. Psychological Medicine, 50, 26822690. DOI 10.1017/S0033291719002770.CrossRefGoogle ScholarPubMed
Gillett, S. R., Boyle, R. H., Zakai, N. A., Mcclure, L. A., Jenny, N. S. and Cushman, M. (2014). Validating laboratory results in a national observational cohort study without field centers: the Reasons for Geographic and Racial Differences in Stroke cohort. Clinical Biochemistry, 47, 243246. DOI 10.1016/j.clinbiochem.2014.08.003.CrossRefGoogle Scholar
Gimeno, D. et al. (2009). Associations of C-reactive protein and interleukin-6 with cognitive symptoms of depression: 12-year follow-up of the Whitehall II study. Psychological Medicine, 39, 413423. DOI 10.1017/S0033291708003723.CrossRefGoogle ScholarPubMed
Glaser, R. and Kiecolt-Glaser, J. K. (2005). Stress-induced immune dysfunction: implications for health. Nature Reviews Immunology, 5, 243251. DOI 10.1038/nri1571.CrossRefGoogle ScholarPubMed
Haapakoski, R., Mathieu, J., Ebmeier, K. P., Alenius, H. and Kivimäki, M. (2015). Cumulative meta-analysis of interleukins 6 and 1β, tumour necrosis factor α and c-reactive protein in patients with major depressive disorder. Brain, Behavior, and Immunity, 49, 206215. DOI 10.1016/j.bbi.2015.06.001.CrossRefGoogle ScholarPubMed
Haley, W. E. et al. (2020). Effects of transitions to family caregiving on well-being: a longitudinal population-based study. Journal of the American Geriatrics Society, 68, 28392846. DOI 10.1111/jgs.16778.CrossRefGoogle ScholarPubMed
Hänsel, A., Hong, S., Cámara, R. J. and Von Kaenel, R. (2010). Inflammation as a psychophysiological biomarker in chronic psychosocial stress. Neuroscience & Biobehavioral Reviews, 35, 115121. DOI 10.1016/j.neubiorev.2009.12.012.CrossRefGoogle ScholarPubMed
Horn, S. R., Long, M. M., Nelson, B. W., Allen, N. B., Fisher, P. A. and Byrne, M. L. (2018). Replication and reproducibility issues in the relationship between c-reactive protein and depression: a systematic review and focused meta-analysis. Brain, Behavior, and Immunity, 73, 85114. DOI 10.1016/j.bbi.2018.06.016.CrossRefGoogle ScholarPubMed
Howard, G. et al. (2017). Racial differences in the incidence of cardiovascular risk factors in older black and white adults. Journal of the American Geriatrics Society, 65, 8390. DOI 10.1111/jgs.14472.CrossRefGoogle ScholarPubMed
Howard, V. J. et al. (2005). The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology, 25, 135143. DOI 10.1159/000086678.CrossRefGoogle ScholarPubMed
Hussain, A. and Ballantyne, C. M. (2021). New approaches for the prevention and treatment of cardiovascular disease: focus on lipoproteins and inflammation. Annual Review of Medicine, 72, 431446. DOI 10.1146/annurev-med-100119-013612.CrossRefGoogle ScholarPubMed
Jenny, N. S. et al. (2012). Long-term assessment of inflammation and healthy aging in late life: the Cardiovascular Health Study All Stars. Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences, 67, 970976. DOI 10.1093/gerona/glr261.CrossRefGoogle Scholar
Koelman, L., Egea Rodrigues, C. and Aleksandrova, K. (2022). Effects of dietary patterns on biomarkers of inflammation and immune responses: a systematic review and meta-analysis of randomized controlled trials. Advances in Nutrition, 13, 101115. DOI 10.1093/advances/nmab086.CrossRefGoogle Scholar
Lamers, F., Milaneschi, Y., De Jonge, P., Giltay, E. and Penninx, B. (2018). Metabolic and inflammatory markers: associations with individual depressive symptoms. Psychological Medicine, 48, 11021110. DOI 10.1017/S0033291717002483.CrossRefGoogle ScholarPubMed
Liu, Y.-Z., Wang, Y.-X. and Jiang, C.-L. (2017). Inflammation: the common pathway of stress-related diseases. Frontiers in Human Neuroscience, 11, 316. DOI 10.3389/fnhum.2017.00316.CrossRefGoogle ScholarPubMed
Mclaughlin, A. P. et al. (2021). The influence of comorbid depression and overweight status on peripheral inflammation and cortisol levels. Psychological Medicine, 18. DOI 10.1017/S0033291721000088.Google ScholarPubMed
Melchior, L. A., Huba, G., Brown, V. B. and Reback, C. J. (1993). A short depression index for women. Educational and Psychological Measurement, 53, 11171125. DOI 10.1177/0013164493053004024.CrossRefGoogle Scholar
Ouakinin, S. R., Barreira, D. P. and Gois, C. J. (2018). Depression and obesity: integrating the role of stress, neuroendocrine dysfunction and inflammatory pathways. Frontiers in Endocrinology, 9, 431. DOI 10.3389/fendo.2018.00431.CrossRefGoogle ScholarPubMed
Piazza, J. R., Almeida, D. M., Dmitrieva, N. O. and Klein, L. C. (2010). Frontiers in the use of biomarkers of health in research on stress and aging. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 65, 513525.CrossRefGoogle ScholarPubMed
Pinquart, M. and Sörensen, S. (2003). Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis. Psychology and Aging, 18, 250267. DOI 10.1093/geronb/gbq049.CrossRefGoogle ScholarPubMed
Potier, F., Degryse, J.-M. and De Saint-Hubert, M. (2018). Impact of caregiving for older people and pro-inflammatory biomarkers among caregivers: a systematic review. Aging Clinical and Experimental Research, 30, 119132. DOI 10.1007/s40520-017-0765-0.CrossRefGoogle ScholarPubMed
Prior, A. et al. (2016). The association between perceived stress and mortality among people with multimorbidity: a prospective population-based cohort study. American Journal of Epidemiology, 184, 199210. DOI 10.1093/aje/kwv324.CrossRefGoogle ScholarPubMed
Roth, D. L. et al. (2020a). Transitions to family caregiving: enrolling incident caregivers and matched non-caregiving controls from a population-based study. Aging clinical and Experimental Research, 32, 18291838. DOI 10.1007/s40520-019-01370-9.CrossRefGoogle ScholarPubMed
Roth, D. L. et al. (2020b). The transition to family caregiving and its effect on biomarkers of inflammation. Proceedings of the National Academy of Sciences, 117, 1625816263. DOI 10.1073/pnas.2000792117.CrossRefGoogle ScholarPubMed
Roth, D. L., Brown, S. L., Rhodes, J. D. and Haley, W. E. (2018). Reduced mortality rates among caregivers: does family caregiving provide a stress-buffering effect? Psychology and Aging, 33, 619629. DOI 10.1037/pag0000224.CrossRefGoogle ScholarPubMed
Roth, D. L., Fredman, L. and Haley, W. E. (2015). Informal caregiving and its impact on health: a reappraisal from population-based studies. The Gerontologist, 55, 309319. DOI 10.1093/geront/gnu177.CrossRefGoogle Scholar
Roth, D. L., Haley, W. E., Hovater, M., Perkins, M., Wadley, V. G. and Judd, S. (2013). Family caregiving and all-cause mortality: findings from a population-based propensity-matched analysis. American Journal of Epidemiology, 178, 15711578. DOI 10.1093/aje/kwt225.CrossRefGoogle ScholarPubMed
Roth, D. L., Perkins, M., Wadley, V. G., Temple, E. M. and Haley, W. E. (2009). Family caregiving and emotional strain: associations with quality of life in a large national sample of middle-aged and older adults. Quality of Life Research, 18, 679688. DOI 10.1007/s11136-009-9482-2.CrossRefGoogle Scholar
Roth, D. L., Sheehan, O. C., Haley, W. E., Jenny, N. S., Cushman, M. and Walston, J. D. (2019). Is family caregiving associated with inflammation or compromised immunity? A meta-analysis. The Gerontologist, 59, e521e534. DOI 10.1093/geront/gnz015.Google ScholarPubMed
Saxton, R. A. et al. (2021). Structure-based decoupling of the pro-and anti-inflammatory functions of interleukin-10. Science, 371, 3815. DOI 10.1126/science.abc8433.CrossRefGoogle ScholarPubMed
Siddiqui, A., Desai, N. G., Sharma, S. B., Aslam, M., Sinha, U. K. and Madhu, S. V. (2019). Association of oxidative stress and inflammatory markers with chronic stress in patients with newly diagnosed type 2 diabetes. Diabetes/metabolism Research and Reviews, 35, e3147. DOI 10.1002/dmrr.3147.CrossRefGoogle ScholarPubMed
Singh, T. and Newman, A. B. (2011). Inflammatory markers in population studies of aging. Ageing Research Reviews, 10, 319329. DOI 10.1016/j.arr.2010.11.002.CrossRefGoogle ScholarPubMed
Stewart, J. C., Rand, K. L., Muldoon, M. F. and Kamarck, T. W. (2009). A prospective evaluation of the directionality of the depression-inflammation relationship. Brain, Behavior, and Immunity, 23, 936944. DOI 10.1016/j.bbi.2009.04.011.CrossRefGoogle ScholarPubMed
Talley, R. C. and Crews, J. E. (2007). Framing the public health of caregiving. American Journal of Public Health, 97, 224228. DOI 10.2105/AJPH.2004.059337.CrossRefGoogle ScholarPubMed
Tukey, J. (1977). Exploratory Data Analysis (Vol. 2, pp. 131–160). Reading, PA: Addison-Wesley.Google Scholar
Uchino, B. N., Trettevik, R., Kent de Grey, R. G., Cronan, S., Hogan, J. and Baucom, B. R. (2018). Social support, social integration, and inflammatory cytokines: a meta-analysis. Health Psychology, 37, 462471. DOI 10.1037/hea0000594.CrossRefGoogle ScholarPubMed
Valkanova, V., Ebmeier, K. P. and Allan, C. L. (2013). CRP, IL-6 and depression: a systematic review and meta-analysis of longitudinal studies. Journal of Affective Disorders, 150, 736744. DOI 10.1016/j.jad.2013.06.004.CrossRefGoogle ScholarPubMed
Vogelzangs, N. et al. (2012). Association of depressive disorders, depression characteristics and antidepressant medication with inflammation. Translational Psychiatry, 2, e79e79. DOI 10.1038/tp.2012.8.CrossRefGoogle ScholarPubMed
Voorhees, J. L. et al. (2013). Prolonged restraint stress increases IL-6, reduces IL-10, and causes persistent depressive-like behavior that is reversed by recombinant IL-10. PloS One, 8, e58488. DOI 10.1371/journal.pone.0058488.CrossRefGoogle ScholarPubMed
Wirtz, P. H. and Von Känel, R. (2017). Psychological stress, inflammation, and coronary heart disease. Current Cardiology Reports, 19, 110. DOI 10.1007/s11886-017-0919x.CrossRefGoogle ScholarPubMed
Zahodne, L. B., Kraal, A. Z., Zaheed, A., Farris, P. and Sol, K. (2019). Longitudinal effects of race, ethnicity, and psychosocial disadvantage on systemic inflammation. SSM-Population Health, 7, 100391. DOI 10.1016/j.ssmph.2019.100391.CrossRefGoogle ScholarPubMed
Zalli, A., Jovanova, O., Hoogendijk, W., Tiemeier, H. and Carvalho, L. (2016). Low-grade inflammation predicts persistence of depressive symptoms. Psychopharmacology, 233, 16691678. DOI 10.1007/s00213-015-391.CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Descriptive characteristics of study variables

Figure 1

Table 2. Estimates from regression analyses of predictors on inflammatory biomarker levels at follow-up

Figure 2

Table 3. Correlations between covariates and inflammatory biomarkers measured at follow-up