Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: Psychosom Med. 2021 Sep 1;83(7):767–776. doi: 10.1097/PSY.0000000000000979

Perceived social support and latent herpesvirus reactivation: Testing main and stress-buffering effects in an ethnically diverse sample of adults

Yanping Jiang 1, Samuele Zilioli 1,2, Raymond P Stowe 3, Rebecca Rubinstein 4, M Kristen Peek 5, Malcolm P Cutchin 6
PMCID: PMC8419084  NIHMSID: NIHMS1722188  PMID: 34267086

Abstract

Objective:

Perceived social support is consistently associated with physical health outcomes, and one potential physiological mechanism underlying this association is immune function. In this study, we tested both the main and stress-buffering effects of perceived social support on cellular immunity measured via latent herpesvirus reactivation.

Methods:

Data were collected from a community-based sample of 1,443 ethnically diverse adults between the ages of 25 and 90 years. Participants self-reported measures of perceived social support, stressful life events, daily hassles, and perceived stress, and provided a blood sample to assess antibody titers of the herpes simplex virus-1 and Epstein-Barr virus (EBV).

Results:

In accordance with the main effect hypothesis, results indicated that perceived social support was directly associated with EBV-viral capsid antigen (EBV-VCA) antibody titers (β = −0.06, 95% CI = [−0.12, −0.01], p = .029). Perceived social support, however, did not interact with stressful life events, daily hassles, or perceived stress to influence latent herpesvirus reactivation (ps > .05). Neither race/ethnicity nor age moderated any of the interactions between perceived social support and the stress measures on latent herpesvirus reactivation (ps >.10).

Conclusion:

Overall, the current study supports the main effect hypothesis, according to which higher levels of perceived social support are associated with lower levels of latent herpesvirus reactivation.

Keywords: Perceived social support, main effect model, stress-buffering model, herpesvirus reactivation, stress

Introduction

Herpesviruses are ubiquitous pathogens that persist for life in their hosts after primary infection (1). The most common types of herpesviruses found among humans are the herpes simplex virus type 1 (HSV-1), Epstein-Barr virus (EBV), and cytomegalovirus. In the US, it was estimated that about 68% of the population ages 12 years and older were HSV-1 seropositive (2) and that almost 90% had EBV antibody titers by 19 years of age (3). The prevalence of herpesvirus infection increases with age. For example, the prevalence of HSV-1 seropositivity can reach 90% among adults ages 70 years and older (2). Notably, herpesviruses remain latent within infected cells after primary infection. Latent herpesviruses, however, may be reactivated (4). The reactivation of latent herpesviruses, though often without clinical symptoms, reflects the compromised ability of the cellular immune system to maintain virus latency (4, 5). Studies among adults have shown that latent herpesvirus reactivation, which is indexed by elevated herpesvirus antibody titers, is associated with increased levels of C-reactive protein and interleukin-6, two common biomarkers of systemic inflammation (6, 7). Elevated herpesvirus antibody titers have also been related to severe morbidity and mortality among intensive care unit patients (8, 9).

Multiple social factors, including stress exposure, have been suggested to contribute to latent herpesvirus reactivation (5). For example, early experiences of deprivation, such as institutionalization, were found to be related to elevated HSV-1 antibodies among adolescents (10). Increased antibody titers to HSV-1 and EBV were also reported among adults with low socioeconomic status (SES; 11). Fagundes and Jaremka found that attachment anxiety, arguably a measure of chronic psychological stress, was associated with increased EBV antibodies (12). These studies buttress the hypothesis that various aspects of psychosocial stress are associated with latent herpesvirus reactivation. Despite this association, what is not fully understood is whether protective psychosocial factors exist that may reduce the risk of latent herpesvirus reactivation among individuals experiencing psychosocial stress. This research question is particularly important to inform effective interventions to reduce the impact of stress on the cellular immune system.

One established factor contributing to better physical and mental health is perceived social support, which is the perceived availability of interpersonal relationships supporting a person’s certain needs (e.g., emotional support, informational support; 13). Individuals reporting more supportive interpersonal relationships have better mental health and lower morbidity and mortality rates (14-19). One key pathway by which perceived social support matters for health is that perceived social support, acting as a coping mechanism, can buffer the adverse effect of stress on health (20). A few studies have examined such stress-buffering effects of perceived social support on the immune system (21-23), which is recognized as one of the key biological intermediaries underlying the link between perceived social support and physical health (for reviews, see 24, 25). For example, in a sample of healthy adults injected with the common cold virus, perceived social support was found to protect against the increased risk of developing common cold associated with interpersonal conflicts (21). Similarly, Kiecolt-Glaser and colleagues found a stronger association between perceived social support and cellular immunity among caregivers of people with dementia than among noncaregivers (23). Perceived social support was also found to mitigate IL-6 reactivity associated with anger in a sample of older women (22).

Although stress-buffering is critical, perceived social support can promote better health regardless of an individual’s stress level (26). The idea that social support may have a direct effect on health irrespective of stress levels was originally proposed by Cohen and Wills (20) in what they named the main-effect model (vs. stress-buffering model). Consistent with this hypothesis, previous studies have shown that perceived social support dampens the activity of the immune system (25, 27-29), though null findings have also been reported (30). For example, a recent meta-analysis reported a robust relationship between perceived social support and decreased levels of pro-inflammatory cytokines (25). Although the effect of perceived social support on the immune system has been largely examined, the majority of the existing literature focuses on inflammation. Further, few studies have tested both main and stress-buffering effects models on latent herpesvirus reactivation simultaneously.

Accordingly, the primary purpose of the current study was to examine the main effect of perceived social support and its interactive effects with multiple aspects of stress (i.e., recent stressful life events, daily hassles, perceived stress) on latent EBV and HSV-1 reactivation in an ethnically diverse sample of adults. We hypothesized that higher levels of perceived social support would be related to decreased levels of antibody titers to EBV and HSV-1. We further hypothesized that there were interactive effects of perceived social support and stress on latent herpesvirus reactivation, such that the beneficial effects of perceived social support on latent herpesvirus reactivation would be stronger among people reporting higher levels of stress.

Method

Participants and Procedure

Data for this study were drawn from the Texas City Stress and Health Study, a study on the social determinants of health among Hispanics, Whites, and African Americans living in Texas City. Data were collected between July 2004 and June 2006. Participant recruitment has been described in more detail elsewhere (31). Briefly, a stratified random sample was drawn after housing units in Texas City were classified as either Hispanic (i.e., at least one adult self-identified as Hispanic in the household) or not (i.e., no adults self-identified as Hispanic in the household). Investigators then randomly selected (1) one adult ages 25+ from one in eight non-Hispanic White or non-Hispanic African American households, (2) one adult ages 25-64 from all Hispanic households with no older adults (ages 65 and above), and (3) all Hispanic adults ages 65 and above from the list. About 80% of the selected adults agreed to participate in the study, resulting in a sample size of 2,706 adults. Of the 2,706 adults, blood samples were provided by 1,457 adults (53.8%, age range of 25 to 90 years). Among the 1,457 adults, 37.3% were US-born Hispanics, 14.8% were foreign-born Hispanics, 36.7% were non-Hispanic Whites, 10.3% were non-Hispanic Blacks, and 1.0% were “other” race/ethnicity (e.g., American Indian/Alaska Native, Asian). Among the Hispanic participants in our sample, about 89% self-identified as of Mexican origin. Given that the category “other” included a very small portion of the sample, which would render between race/ethnicity comparison complicated to interpret, participants from this group were dropped from the analyses, resulting in a final sample of 1,443 adults.

All participants were interviewed in English or Spanish in their homes. The interview consisted of multiple questionnaires assessing demographic, social, psychological, behavioral, and health indicators. Blood samples were drawn from participants by a trained phlebotomist in participants’ homes or at a centrally located clinic between 9:00 am and 12:00 pm in the morning. Blood samples were centrifuged to obtain plasma and then stored at −70°C until testing. Prior to the interview, informed consent was obtained from all participants. The study protocol was approved by the Institutional Review Board at the University of Texas Medical Branch.

Measures

Herpesvirus Antibody Titers

Immunoglobulin G antibody titers to HSV-1, EBV viral capsid antigen (EBV-VCA), and EBV early antigen (EBV-EA) in plasma were determined by the indirect immune-fluorescence assay as previously described (11). Of note, proteins of the EBV-VCA complex are late antigens released during productive viral replication, whereas the incomplete reactivity is indicated by elevated EBV-EA antibodies. The assay was performed using commercially prepared substrate slides and control sera (Microgen Laboratories, La Marque, Texas, USA, and Bion Enterprises, Park Ridge, Illinois, USA). Plasma was tested in dilutions of 1:10, 1:40, 1:160, and 1:640 for EBA-EA and of 1:10, 1:40, 1:160, and 1:640, 1:2,560, and 1:10,240 for EBV-VCA and HSV-1. The endpoint titer was determined as the highest dilution of plasma still being able to demonstrate immunofluorescence-positive cells. All specimens were blind coded and batch analyzed to minimize variation between assays, and the standard error of the assay was ± one dilution. HSV-1, EBV-VCA, and EBV-EA antibody titers were analyzed as three separate continuous outcome variables. Antibody titers were log-transformed to reduce skewness.

Perceived Social Support

Perceived social support was assessed using the Medical Outcomes Study Social Support Survey (13), which consisted of 19 items to measure four domains of social support (i.e., emotional/informational, tangible, affectionate, positive social interaction). Sample items were “someone who shows you love and affection,” and “someone you can count on to listen to you when you need to talk.” Participants were asked to rate how often these kinds of support were available to them on a five-point Likert scale (0 = none of the time to 4 = most of the time). A mean score for each domain of social support was calculated by averaging the responses on the corresponding items. The four domain scores were strongly correlated with each other (rs = 0.60 – 0.81); thus, a composite index of perceived social support was created by averaging scores on the 19 items. Higher scores on this variable reflect higher levels of perceived social support. The Cronbach alpha for this scale was 0.97.

Stressful Life Events

Stressful life events were assessed using a 31-item checklist of negative life events (32). For each event (e.g., illness, marital separation, financial crisis, job change), participants were asked to report whether they experienced it during the last year (0 = no, 1 = yes). For 23 of the items, participants were also asked to report whether the events occurred to their partner/boyfriends/girlfriends. For 17 items, participants were asked to also consider whether the events occurred to their parents, and for 11 items, participants were asked to report whether any of their other relatives or friends experienced the event. Given that this study focused on stressful life events experienced by the participants, we only considered events directly experienced by the participants. A composite index for stressful life events was created by summing participants’ responses on 31 items, with a higher score reflecting more stressful life events.

Daily Hassles

Daily hassles were assessed using an adapted 28-item daily hassles scale (33). Participants were asked to rate how each concern (e.g., responsibilities, job difficulties, money/time constraints, family/friends concerns, neighborhood issues) bothered them on a daily basis on a 4-point scale (0 = not at all bothered to 3 = bothered a great deal). A composite index of daily hassles was created by averaging participants’ responses on all items, with a higher score reflecting greater daily hassles. The Cronbach alpha for this scale was 0.92.

Perceived Stress

Perceived stress was assessed using the 10-item Perceived Stress Scale (34, 35). Participants were asked to report how often they felt or thought a certain way (e.g., “be upset because of something that happened unexpectedly,” “unable to control the important things in your life”) during the last month on a 5-point Likert scale (1 = never to 5 = very often). Four positively stated items were reverse coded, and a mean score was calculated by averaging responses to all items. Higher scores reflect higher levels of perceived stress. The Cronbach alpha for this scale was 0.86.

Sociodemographic Covariates

The sociodemographic variables included sex (0 = male, 1 = female), age, race/ethnicity (dummy coded: non-Hispanic White vs. non-Hispanic Black; non-Hispanic White vs. US-born Hispanic; non-Hispanic White vs. foreign-born Hispanic), and marital status (0 = others, 1 = married). Annual household income was assessed on an 8-point scale ranging from 1 = less than 10,000 to 8 = 75,000 or more, and education was assessed by asking participants the years of schooling completed. A composite index of socioeconomic status (SES) was created by averaging z-scored income and education.

Health and Behavioral Covariates

A few health and behavioral variables identified from prior studies on latent herpesvirus reactivation were included as covariates in the analyses (6, 36). Participants were also asked to report whether they had each of 11 health conditions (0 = no, 1 = yes), including stroke/brain hemorrhage, cancer, diabetes, high blood pressure, overweight, chronic respiratory problems, arthritis, depression, anxiety, heart attack, skin problems, and gall bladder disease. A composite health conditions index was created by summing participants’ responses on this checklist. Body mass index (BMI) was calculated using participants’ height and weight assessed in the clinic or during the home visit.

Depressive symptoms were also included as a covariate, given that it has been associated with herpesvirus antibody titers (37). Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale (38) on a 4-point scale, ranging from 0 = rarely or none of the time to 3 = most or all of the time. Four positively valenced items were reversed coded. A sum score was calculated for depressive symptoms, with a high score reflecting a higher level of depressive symptoms. The Cronbach’s alpha for this scale was 0.93.

We also included smoking, physical activity, and sleep as behavioral covariates. Specifically, smoking was assessed by asking participants whether they ever smoked at least 100 cigarettes in the entire life (0 = no, 1 = yes). Physical activity was assessed by asking participants whether they engaged in any moderate or vigorous activities for at least 10 minutes during the previous seven days (0 = no, 1 = yes). Sleep was assessed using a 4-item scale, which asked how often participants experienced “trouble falling asleep,” “wake up several times per night,” “trouble staying asleep,” and “wake up after your usual amount of sleep feeling tired and worn out” during the past month (39). The response options ranged from 1 = all of the time to 5 = none of the time, and a composite index of sleep was created by averaging participants’ responses on the four items. The Cronbach alpha for the sleep scale was 0.84.

Data Analyses

The χ2 test for categorical variables and one-way analysis of variance for continuous variables were employed to test group differences by race/ethnicity. Multivariate regression analyses were used to test the main effects of perceived social support on three types of herpesvirus antibody titers, as well as the interactive effects of perceived social support and each stress measure (i.e., stressful life events, daily hassles, perceived stress) on herpesvirus antibody titers. The interaction terms were calculated after mean-centering, and simple slope analyses were employed if there were any significant interactive effects (40). Simple slopes were graphed at different values (i.e., 0.5 SD and 1 SD below the mean, the mean, and 0.5 SD and 1 SD above the mean) of the moderators using an open-source interActive tool developed by McCabe and colleagues (41).

Missing data were handled with listwise deletion (42), given that the incidence of missing data was low (i.e., 0.3%) and the assumption of missing completely at random was not rejected (Little’s test χ2(146) = 137.06, p = .69).

Results

Table 1 displays participants’ characteristics by race/ethnicity. Overall, the majority of the participants (i.e., 99.9%, N = 1,443) were seropositive for both EBV and HSV-1. Non-Hispanic White participants were older, had higher SES, and lower HSV-1 antibody titers, compared to non-Hispanic Blacks and US-born and foreign-born Hispanics. Non-Hispanic Black participants were less likely to be married, had higher BMI, and reported higher levels of daily hassles, perceived stress, and depressive symptoms than non-Hispanic Whites and US-born and foreign-born Hispanics. Compared to US-born Hispanics, foreign-born Hispanics were younger, more likely to be married, and less likely to have ever smoked. Foreign-born Hispanics reported lower SES, fewer health conditions, better sleep quality, and a lower level of stressful life events than their US-born counterparts.

Table 1.

Characteristics of study participants by race/ethnicity

Variables All Non-Hispanic White US-born Hispanic Foreign-born Hispanic Non-Hispanic Black
N 1,443 535 543 215 150
Female a 60.22 55.33 61.14 65.12 67.33
Married a 57.21 57.20 55.80 72.90c,d 40.00c,d,e
Age 51.75(15.93) 55.57(15.67) 50.88(15.96)c 44.96(14.48)c,d 50.95(15.05)c,e
Socioeconomic status −0.01(0.85) 0.32(0.72) −0.06(0.84)c −0.63(0.88)c,d −0.13(0.73)c,e
Health conditions 2.63(2.13) 3.17(2.10) 2.42(2.04)c 1.57(1.76)c,d 3.04(2.36)d,e
Body mass index 30.64(7.23) 29.94(7.42) 31.05(6.98) 29.61(5.66) 33.14(8.69)c,d,e
Depressive symptoms 9.23(11.19) 9.57(10.96) 8.90(11.39) 7.04(10.10)c 12.36(12.02)c,d,e
Smoked a 51.11 60.19 49.63c 31.63c,d 52.00e
Had physical activity a 73.33 76.55 73.25 68.84 68.67
Sleep 3.60(1.14) 3.46(1.13) 3.64(1.17) 4.00(1.09)c,d 3.40(1.03)e
Perceived social support 3.38(0.80) 3.31(0.83) 3.44(0.77)c 3.54(0.71)c 3.19(0.86)d,e
Stressful life events 1.08(1.56) 1.09(1.58) 1.15(1.62) 0.80(1.27)d 1.25(1.64)e
Daily hassles 0.60(0.51) 0.56(0.47) 0.59(0.54) 0.57(0.50) 0.78(0.54)c,d,e
Perceived stress 2.32 (0.73) 2.30(0.74) 2.32(0.74) 2.24(0.64) 2.50(0.71)c,d,e
EBV-VCAb 6.20(1.01) 6.21(1.07) 6.22(0.98) 6.05(1.00) 6.36(0.93)e
EBV-EAb 2.76(0.83) 2.77(0.83) 2.76(0.84) 2.59(0.76)c 2.94(0.87)e
Herpes simplex virus-1b 6.47(0.91) 6.28(1.03) 6.56(0.83)c 6.59(0.76)c 6.67(0.78)c
Open in a new tab

Note. EBV-VCA = Epstein-Barr virus viral capsid antigen; EBV-EA = Epstein-Barr virus early antigen.

a

Presented as percentage. Means (standard deviations) were presented.

b

Natural log of antibody titers.

c

Significantly different from non-Hispanic White at p < .05

d

Significantly different from US-born Hispanic at p < .05

e

Significantly different from foreign-born Hispanic at p < .05

Table 2 displays bivariate correlation coefficients between study variables. Perceived social support was negatively correlated with HSV-1 and EBV-VCA antibody titers (r = −0.06, p = .026; r = −0.07, p = .005; respectively), but not EBV-EA antibody titers (r = −0.03, p = .29). Perceived social support was also negatively correlated with stressful life events (r = −0.13, p < .001), daily hassles (r = −0.26, p <.001), and perceived stress (r = −0.28, p <.001). None of the three stress measures were correlated with herpesvirus antibody titers, except for three unexpected negative correlations: one between stressful life events and HSV-1 antibody titers (r = −.05, p = .044), one between daily hassles and EBV-EA antibody titers (r = −0.06, p = .033), and one between perceived stress and EBV-EA antibody titers (r = −0.06, p = .033) 1. SES was correlated with perceived social support (r = 0.11, p < .001), daily hassles (r = −0.08, p = .003), perceived stress (r = −0.12, p < .001), and antibody titers to EBV-VCA (r = −0.07, p = .007) and HSV-1 (r = −0.21, p < .001). Depressive symptoms were correlated with perceived social support (r = −0.29, p < .001), stressful life events (r = 0.31, p < .001), daily hassles (r = 0.50, p < .001), and perceived stress (r = 0.62, p < .001), but not herpesvirus antibody titers (r = 0.04, p = .13 for HSV-1; r = 0.03, p = .28 for EBV-VCA; r = 0.02, p = .49 for EBV-EA).

Table 2.

Correlation coefficients between study variables

Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Female -
2. Married −.17*** -
3. Age −.12*** −.06* -
4. SES −.09*** .17*** −.13*** -
5. Health conditions .08** −.10*** .37*** −.06* -
6. BMI .11*** −.06* −.09*** −.01 .27*** -
7. Depressive symptoms .17*** −.13*** −.09** −.13*** .38*** .11*** -
8. Smoked −.22*** −.05 .03 −.03 .11*** −.06* .13*** -
9. Had PA −.09** .07** −.13*** .18*** −.05 .05 −.03 .05* -
10. Sleep −.11*** .12*** .07** .04 −.34*** −.09** −.62*** −.15*** −.02 -
11. PSS .04 .23*** −.07** .11*** −.18*** −.05 −.29*** −.10*** −.01 .20*** -
12. SLE .02 −.20*** −.22*** −.02 .12*** .10*** .31*** .11*** .07** −.26*** −.13*** -
13. Daily hassles .18*** −.12*** −.29*** −.08** .15*** .13*** .50*** .13*** .09** −.42*** −.26*** .34*** -
14. Perceived stress .18*** −.10*** −.20*** −.12*** .21*** .11*** .62*** .08** .02 −.44*** −.28** .31*** .57*** -
15. HSV-1 .04 −.05 .13*** −.21*** .06* .05* .04 −.05 −.07* .00 −.06* −.05* .04 .01 -
16. EBV-VCA .10*** −.07* .12*** −.07** .09*** .08** .03 .02 −.03 −.01 −.07** −.04 −.00 .01 .20*** -
17. EBV-EA .08** −.04 .13*** .02 .10*** .01 .02 .02 −.04 −.02 −.03 −.00 −.06* −.06* .10*** .13*** -
Open in a new tab

Note. SES = socioeconomic status; BMI = body mass index; PA = physical activity; PSS = perceived social support; SLE = stressful life events; HSV-1 = herpes simplex virus type 1; EBV-VCA = Epstein-Barr virus viral capsid antigen; EBV-EA = Epstein-Barr virus early antigen.

*

p <.05

**

p < .01

***

p < .001.

Main and Stress-buffering Effects of Perceived Social Support

Regression analyses showed that perceived social support was associated with decreased EBV-VCA and HSV-1 antibody titers (β = −0.07, 95% confidence interval [CI] = [−0.13, −0.02], p = .005; β = −0.06, 95% CI = [−.011, −0.01], p = .026), but not with EBV-EA antibody titers (β = −0.03, 95% CI = [−0.08, 0.02], p = .29). Controlling for sociodemographic, health, and behavioral covariates, perceived social support remained significantly associated with EBV-VCA antibody titers (β = −0.06, 95% CI = [−0.12, −0.01], p = .029, see Table 3), whereas the association between perceived social support and HSV-1 antibody titers became nonsignificant (β = −0.05, 95% CI = [−0.10, 0.01], p = .092).

Table 3.

Associations between perceived social support and herpesvirus antibody titers

HSV-1
 EBV-VCA
 EBV-EA
Variables β 95% CI β 95% CI β 95% CI
Perceived social support −0.05 (−0.10, 0.01) −0.06* (−0.12, −0.01) −0.02 (−0.08, 0.03)
Female 0.02 (−0.09, 0.13) 0.23*** (0.12, 0.35) 0.19** (0.08, 0.31)
Non-Hispanic Black 0.36*** (0.18, 0.54) 0.11 (−0.08, 0.29) 0.26** (0.07, 0.44)
US-born Hispanic 0.27*** (0.15, 0.39) 0.01 (−0.11, 0.14) 0.05 (−0.08, 0.17)
Foreign-born Hispanic 0.27** (0.09, 0.46) −0.10 (−0.29, 0.09) −0.04 (−0.23, 0.14)
Married −0.01 (−0.12, 0.10) −0.01 (−0.12, 0.10) −0.01 (−0.12, 0.10)
Age 0.15*** (0.09, 0.21) 0.11*** (0.05, 0.17) 0.13*** (0.07, 0.20)
SES −0.14*** (−0.20, −0.09) −0.05 (−0.11, 0.01) 0.06 (−0.01, 0.12)
Health conditions −0.01 (−0.07, 0.06) 0.01 (−0.06, 0.07) 0.03 (−0.04, 0.09)
BMI 0.05 (−0.00, 0.11) 0.07* (0.01, 0.13) 0.00 (−0.06, 0.06)
Depressive symptoms 0.03 (−0.04, 0.10) −0.02 (−0.09, 0.05) −0.03 (−0.10, 0.04)
Smoked −0.07 (−0.18, 0.04) 0.09 (−0.03, 0.20) 0.06 (−0.05, 0.17)
Had physical activity −0.03 (−0.15, 0.09) −0.02 (−0.13, 0.10) −0.07 (−0.19, 0.05)
Sleep 0.01 (−0.05, 0.08) 0.02 (−0.05, 0.08) −0.02 (−0.09, 0.05)
R 2 0.09 0.04 0.04
Open in a new tab

Note. SES = socioeconomic status; BMI = body mass index; HSV-1 = herpes simplex virus type 1; EBV-VCA = Epstein-Barr virus viral capsid antigen; EBV-EA = Epstein-Barr virus early antigen.

*

p <.05

**

p < .01

***

p < .001.

Moderation analyses revealed no interactive effects between perceived social support and stressful life events (β = −0.02, 95% CI = [−0.07, 0.04], p = .57), daily hassles (β = 0.01, 95% CI = [−0.04, 0.07], p = .68), or perceived stress (β = −0.01, 95% CI = [−0.06, 0.05], p = .75) on EBV-VCA antibody titers. Similarly, perceived social support did not interact with stressful life events (β = 0.01, 95% CI = [−0.05, 0.06], p = .81), daily hassles (β = −0.01, 95% CI = [−0.06, 0.04], p = .71), or perceived stress (β = 0.01, 95% CI = [−0.04, 0.07], p = .67) to impact EBV-EA antibody titers. There were also no interactive effects between perceived social support and stressful life events (β = −0.01, 95% CI = [−0.07, 0.04], p = .67), daily hassles (β = 0.00, 95% CI = [−0.05, 0.06], p = .90), or perceived stress (β = 0.04, 95% CI = [−0.02, 0.09], p = .19) on HSV-1 antibody titers. As shown in Table 4, all interactive effects remained statistically nonsignificant after controlling for covariates.

Table 4.

Associations between the interaction terms of perceived social support by stress measures and herpesvirus antibody titers

HSV-1 EBV-VCA EBV-EA
Model for SLE
Variables β 95% CI β 95% CI β 95% CI
PSS −0.04 (−0.10, 0.01) −0.06* (−0.12, −0.01) −0.02 (−0.08, 0.04)
SLE −0.05 (−0.11, 0.01) −0.05 (−0.11, 0.01) 0.02 (−0.04, 0.08)
PSS × SLE −0.01 (−0.07, 0.04) −0.01 (−0.06, 0.05) −0.00 (−0.06, 0.05)
R 2 0.09 0.05 0.04
Model for daily hassles
PSS −0.04 (−0.10, 0.02) −0.07* (−0.13, −0.01) −0.03 (−0.09, 0.02)
Daily hassles 0.07* (0.00, 0.13) −0.02 (−0.08, 0.05) −0.07* (−0.14, −0.01)
PSS × daily hassles 0.02 (−0.04, 0.07) 0.03 (−0.03, 0.08) −0.01 (−0.05, 0.05)
R 2 0.09 0.04 0.04
Model for perceived stress
PSS −0.06* (−0.12, −0.01) −0.07* (−0.13, −0.01) −0.05 (−0.11, 0.01)
Perceived stress −0.01 (−0.07, 0.06) −0.04 (−0.11, 0.03) −0.11** (−0.17, −0.04)
PSS × perceived stress 0.05 (−0.00, 0.11) 0.01 (−0.04, 0.07) 0.02 (−0.03, 0.08)
R 2 0.09 0.04 0.04
Open in a new tab

Note. PSS = perceived social support; SLE = stressful life events; HSV-1 = herpes simplex virus type 1; EBV-VCA = Epstein-Barr virus viral capsid antigen; EBV-EA = Epstein-Barr virus early antigen. Covariates, including sex, race/ethnicity, marital status, age, socioeconomic status, health conditions, body mass index, depressive symptoms, smoking, physical activity, and sleep, were controlled for each moderation model.

*

p <.05

**

p < .01

***

p < .001.

Given that most of our stress measures did not show significant correlations with herpesvirus antibody titers, we performed ancillary analyses to test the stress-buffering hypothesis by including SES and depressive symptoms as alternative measures for stress. After controlling for covariates, results showed that there were no interactive effects between perceived social support and SES on antibody titers to HSV-1 (β = −0.02, 95% CI = [−0.08, 0.03], p = .45), EBV-VCA (β = −0.03, 95% CI = [−0.09, 0.02], p = .25), or EBV-EA (β = −0.03, 95% CI = [−0.09, 0.03], p = .27). However, depressive symptoms interacted with perceived social support to impact antibody titers to HSV-1 (β = 0.07, 95% CI = [0.01, 0.13], p = .014) and EBV-VCA (β = 0.06, 95% CI = [0.01, 0.12], p = .033), but not EBV-EA (β = 0.05, 95% CI = [−0.01, 0.11], p = .095). Simple slope analyses revealed negative associations between perceived social support and antibody titers to HSV-1 and EBV-VCA at 0.5 SD and 1 SD below the mean and at the mean, but not at 0.5 SD and 1 SD above the mean of depressive symptoms (see Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Simple slopes of the associations between perceived social support and antibody titers to Epstein-Barr virus viral capsid antigen (EBV-VCA) (Panel A) and antibody titers to herpes simplex virus type 1 (HSV-1) (Panel B) at different levels (i.e., 0.5 SD and 1 SD below the mean, the mean, and 0.5 SD and 1 SD above the mean) of depressive symptoms. Continuous variables were standardized. CI = confidence interval; PTCL = percentile.

In addition, as reported in Table 1, there were racial/ethnic differences in perceived social support. Further, previous reports from our team, including the current analysis, have shown age and racial/ethnic differences in latent herpesvirus reactivation (6, 11). Thus, to further explore the stress-buffering effects of perceived social support, we tested whether age and race/ethnicity moderated the interaction between perceived social support and various stress measures. After controlling for covariates, results showed that there were no three-way interactions among age, stress measures, and perceived social support on herpesvirus antibody titers (ps > .20). Similarly, the effects of the three-way interactions among race/ethnicity, stress measures, and perceived social support on herpesvirus antibody titers were not statistically significant (ps > .10).

Discussion

Enhanced immune function has been proposed as one of the biological pathways through which supportive interpersonal relationships benefit health (24). The purpose of this study was to examine both main and stress-buffering effects of perceived social support on latent herpesvirus reactivation, utilized as an index of the compromised ability of immune cells to maintain virus latency (4). With an ethnically diverse sample of adults, this study demonstrated that perceived social support was directly related to decreased herpesvirus antibody titers but did not interact with several measures of stress to influence herpesvirus antibody titers.

The main effect of perceived social support on herpesvirus antibody titers suggests that perceived social support may have a direct impact on cellular immunity. Our result is in line with recent work that highlights the need to conceptualize social support as a critical factor for health promotion in the absence of adversity (26). Feeney and Collins (2015) argue that supportive interpersonal relationships can promote health in multiple ways, not only as a resource for coping with adversity but also as a resource for enabling individuals to maximize opportunities to achieve optimal growth and development in the absence of stressors. The positive effect of perceived social support on herpesvirus antibody titers is also consistent with previous studies documenting a significant association between social support on decreased systemic inflammation (25). This finding extends the existing literature by suggesting that latent herpesvirus reactivation, as a marker of the compromised cellular immune system, may be another potential immune process linking supportive interpersonal relationships to better physical health. However, our main effects of perceived social support on cellular immunity should be interpreted with caution given their small effect sizes and the fact that, after controlling for covariates, only results concerning EBV-VCA antibody titers remained significant.

No evidence was found in favor of the stress-buffering hypothesis; that is, there were no interactive effects between perceived social support and our stress measures on herpesvirus antibody titers. Further, neither age nor race/ethnicity moderated the interaction effects of perceived social support and each stress measure on herpesvirus antibody titers. These results are somewhat inconsistent with the previous study demonstrating a stronger buffering effect of perceived social support on cellular immunity among caregivers of people with dementia than noncaregivers (23). Our results, however, are not surprising, given that mixed findings have been reported among studies of stress-buffering effects on other immune processes, including inflammation (e.g., 22, 43). The lack of stress-buffering effects in this study suggests that perceived social support may enhance cellular immune process not necessarily though enhancing individuals’ ability to cope with stress. Rather, supportive interpersonal relationships may impact cellular immune system directly or indirectly through a variety of psychological (e.g., positive affect and perceived control) and physiological (e.g., regulation of the hypothalamic-pituitary-adrenal [HPA]) mechanisms (44, 45). For example, high levels of perceived social support can provide a sense of stability and predictability in an individual’s life situation, which may promote positive affect and perceived control (20). Positive affect and perceived control, in turn, have been linked to systemic inflammation (46-48). Also, studies indicate that perceived social support could impact cortisol secretion, the end product of HPA axis activity (24), which subsequently modifies systematic inflammation and cellular immunity (49, 50).

The null results concerning the stress-buffering hypothesis could also be explained by the fact that most of our direct measures of stress were not correlated with herpesvirus antibody titers. The lack of an association between stress and herpesvirus antibody titers differs from previous studies reporting adverse effects of stress on cellular immunity (5, 10). There are two possible reasons for this unexpected finding. One possible explanation is related to the oversampling of adults ages 65 years and older (24.3% of the sample) in our study. On average, older adults tend to report decreased levels of stress (as indicated by the negative correlations between age and stress measures in our sample) and exhibit increased levels of herpesvirus antibody titers (as indicated by the positive correlations between age and antibody titers in our sample) compared to younger adults (51, 52). These patterns of associations may have attenuated the relationship between stress and herpesvirus antibody titers. Supporting this explanation, when controlling for age, significant positive associations between some of our stress measures and herpesvirus antibody titers emerged in our study. Another explanation for our findings may be due to some unexamined environmental risk factors. Participants in our study had been living close to oil refineries and petrochemical processing plants. It is possible that living in a place with significant and chronic environmental stressors might have obscured the association between stress and herpesvirus antibody titers (31). Of note, our sample had a relatively higher prevalence of seropositivity for EBV and HSV-1 than the ones documented in more nationally representative samples (e.g., 99% vs. 68% for HSV-1, 2). In line with this explanation, one previous report from the same study did not find significant associations between stress assessed by stressful life events and daily hassles and allostatic load (53), a measure of cumulative biological risk that includes excessive activation of the immune system (54).

An unexpected finding from our study concerns the negative correlations between perceived social support and antibody titers to HSV-1 and EBV-VCA among participants who reported low and average (vs. high) levels of depressive symptoms. These findings are somewhat inconsistent with the stress-buffering hypothesis, which would suggest stronger protective effects of perceived social support on health at higher levels of stress (i.e., depressive symptoms in these analyses) (20). In other words, participants in our sample with lower depressive symptoms might have gained more health benefits from perceived social support than participants with higher depressive symptoms. This finding should be interpreted with caution, given that depressive symptoms are commonly recognized as a psychological sequela of stress rather than a measure of stress itself.

Some limitations should be considered when interpreting the results of the current study. First, the nature of the cross-sectional design limits our ability to demonstrate a temporal relationship between perceived social support and latent herpesvirus reactivation. Future longitudinal studies are needed to elucidate how the relationship between perceived social support and latent herpesvirus reactivation unfolds over time. Second, this study only assessed one aspect of interpersonal relationships, namely perceived social support. Other aspects of interpersonal relationships, including social integration and received social support, may also have an impact on latent herpesvirus reactivation. For example, a recent meta-analysis indicates that social integration, as a broader measure of interpersonal relationships, compared to perceived social support, appears to have a stronger main effect on inflammation (25). Also, the lack of data to distinguish the sources (e.g., family, workers) of perceived social support constrains the opportunities to examine whether different resources of perceived social support may exert distinct effects on latent herpesvirus reactivation. Prior studies have suggested the need to distinguish the sources of social support in testing the buffering effects for different types of stressors (20). Lastly, given the complex nature of stress, our findings should be replicated in future studies using more recent stress measures designed to measure multiple dimensions of stress, such as the Stress and Adversity Inventory, which provides an assessment of the severity, frequency, timing, and duration of life stressors (55).

Nevertheless, to our knowledge, this study provided the first evidence to suggest an association between perceived social support and antibody titers to herpesviruses through main but not stress-buffering pathways in an ethnically diverse sample of adults. Overall, perceived social support seems to be directly associated with latent herpesvirus reactivation, above and beyond sociodemographic and health confounders.

Conflicts of Interest and Source of Funding:

The authors have no conflicts of interest to disclose. Data collection for the present study was supported by the National Institute of Health (P50 CA105631). The preparation of the manuscript was partly supported by the Faculty Competition for Postdoctoral Fellowship from Wayne State University to Dr. Samuele Zilioli.

Abbreviations:

BMI

Body mass index

EBV-VCA

Epstein-Barr virus viral capsid antigen

EBV-EA

Epstein-Barr virus early antigen

HSV-1

Herpes simplex virus type 1

SES

Socioeconomic status

PSS

Perceived social support

SLE

Stressful life events

CI

Confidence interval

Footnotes

1

These unexpected negative correlations between some stress measures and herpesvirus antibody titers may be driven by the oversampling of adults ages 65 years and older, who reported lower levels of stress but had higher levels of herpesvirus antibody titers than younger adults. After controlling for age, a positive correlation between daily hassles and HSV-1 antibody titers (r = 0.08, p = .002) emerged, and the negative correlations between stress and herpesvirus antibody titers reported in Table 2 became statistically nonsignificant (ps >.10).

References

  • 1.Suzich JB, Cliffe AR. Strength in diversity: Understanding the pathways to herpes simplex virus reactivation. Virology. 2018;522:81–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Schillinger JA, Xu F, Sternberg MR, Armstrong GL, Lee FK, Nahmias AJ, McQuillan GM, Louis ME, Markowitz LE. National seroprevalence and trends in herpes simplex virus type 1 in the United States, 1976–1994. Sexually transmitted diseases. 2004;31:753–60. [DOI] [PubMed] [Google Scholar]
  • 3.Balfour HH Jr, Sifakis F, Sliman JA, Knight JA, Schmeling DO, Thomas W. Age-specific prevalence of Epstein–Barr virus infection among individuals aged 6–19 years in the United States and factors affecting its acquisition. The Journal of infectious diseases. 2013;208:1286–93. [DOI] [PubMed] [Google Scholar]
  • 4.Glaser R, Kiecolt-Glaser J. Stress-associated immune modulation and its implications for reactivation of latent herpesviruses. INFECTIOUS DISEASE AND THERAPY SERIES. 1994;13:245-. [Google Scholar]
  • 5.Glaser R, Pearson GR, Jones JF, Hillhouse J, Kennedy S, Mao H, Kiecolt-Glaser JK. Stress-related activation of Epstein-Barr virus. Brain, behavior, and immunity. 1991;5:219–32. [DOI] [PubMed] [Google Scholar]
  • 6.Murdock KW, Fagundes CP, Peek MK, Vohra V, Stowe RP. The effect of self-reported health on latent herpesvirus reactivation and inflammation in an ethnically diverse sample. Psychoneuroendocrinology. 2016;72:113–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Bennett JM, Glaser R, Malarkey WB, Beversdorf DQ, Peng J, Kiecolt-Glaser JK. Inflammation and reactivation of latent herpesviruses in older adults. Brain, behavior, and immunity. 2012;26:739–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Libert N, Bigaillon C, Chargari C, Bensalah M, Muller V, Merat S, de Rudnicki S. Epstein-Barr virus reactivation in critically ill immunocompetent patients. Biomedical journal. 2015;38. [DOI] [PubMed] [Google Scholar]
  • 9.Ong DS, Bonten MJ, Spitoni C, Verduyn Lunel FM, Frencken JF, Horn J, Schultz MJ, van der Poll T, Klein Klouwenberg PM, Cremer OL. Epidemiology of multiple herpes viremia in previously immunocompetent patients with septic shock. Clinical Infectious Diseases. 2017;64:1204–10. [DOI] [PubMed] [Google Scholar]
  • 10.Shirtcliff EA, Coe CL, Pollak SD. Early childhood stress is associated with elevated antibody levels to herpes simplex virus type 1. Proceedings of the National Academy of Sciences. 2009;106:2963–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Stowe RP, Peek MK, Perez NA, Yetman DL, Cutchin MP, Goodwin JS. Herpesvirus reactivation and socioeconomic position: a community-based study. Journal of Epidemiology & Community Health. 2010;64:666–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fagundes CP, Jaremka LM, Glaser R, Alfano CM, Povoski SP, Lipari AM, Agnese DM, Yee LD, Carson WE III, Farrar WB. Attachment anxiety is related to Epstein–Barr virus latency. Brain, behavior, and immunity. 2014;41:232–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sherbourne CD, Stewart AL. The MOS social support survey. Social science & medicine. 1991;32:705–14. [DOI] [PubMed] [Google Scholar]
  • 14.Blazer DG. Social support and mortality in an elderly community population. American journal of epidemiology. 1982;115:684–94. [DOI] [PubMed] [Google Scholar]
  • 15.George LK, Blazer DG, Hughes DC, Fowler N. Social support and the outcome of major depression. The British Journal of Psychiatry. 1989;154:478–85. [DOI] [PubMed] [Google Scholar]
  • 16.Barth J, Schneider S, Von Känel R. Lack of social support in the etiology and the prognosis of coronary heart disease: a systematic review and meta-analysis. Psychosomatic medicine. 2010;72:229–38. [DOI] [PubMed] [Google Scholar]
  • 17.Tomaka J, Thompson S, Palacios R. The relation of social isolation, loneliness, and social support to disease outcomes among the elderly. Journal of aging and health. 2006;18:359–84. [DOI] [PubMed] [Google Scholar]
  • 18.Chu PS, Saucier DA, Hafner E. Meta-analysis of the relationships between social support and well-being in children and adolescents. Journal of Social and Clinical Psychology. 2010;29:624–45. [Google Scholar]
  • 19.Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS medicine. 2010;7:e1000316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychological bulletin. 1985;98:310. [PubMed] [Google Scholar]
  • 21.Cohen S, Janicki-Deverts D, Turner RB, Doyle WJ. Does hugging provide stress-buffering social support? A study of susceptibility to upper respiratory infection and illness. Psychological science. 2015;26:135–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Puterman E, Epel ES, O’Donovan A, Prather AA, Aschbacher K, Dhabhar FS. Anger is associated with increased IL-6 stress reactivity in women, but only among those low in social support. International journal of behavioral medicine. 2014;21:936–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kiecolt-Glaser JK, Dura JR, Speicher CE, Trask OJ, Glaser R. Spousal caregivers of dementia victims: longitudinal changes in immunity and health. Psychosomatic medicine. 1991;53:345–62. [DOI] [PubMed] [Google Scholar]
  • 24.Uchino BN. Social support and health: a review of physiological processes potentially underlying links to disease outcomes. Journal of behavioral medicine. 2006;29:377–87. [DOI] [PubMed] [Google Scholar]
  • 25.Uchino BN, Trettevik R, Kent de Grey RG, Cronan S, Hogan J, Baucom BR. Social support, social integration, and inflammatory cytokines: A meta-analysis. Health Psychology. 2018;37:462. [DOI] [PubMed] [Google Scholar]
  • 26.Feeney BC, Collins NL. A new look at social support: A theoretical perspective on thriving through relationships. Personality and Social Psychology Review. 2015;19:113–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Baron RS, Cutrona CE, Hicklin D, Russell DW, Lubaroff DM. Social support and immune function among spouses of cancer patients. Journal of Personality and Social Psychology. 1990;59:344. [DOI] [PubMed] [Google Scholar]
  • 28.Miyazaki T, Ishikawa T, Iimori H, Miki A, Wenner M, Fukunishi I, Kawamura N. Relationship between perceived social support and immune function. Stress and Health: Journal of the International Society for the Investigation of Stress. 2003;19:3–7. [Google Scholar]
  • 29.Muscatell KA, Eisenberger NI, Dutcher JM, Cole SW, Bower JE. Links between inflammation, amygdala reactivity, and social support in breast cancer survivors. Brain, behavior, and immunity. 2016;53:34–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.McDade TW, Hawkley LC, Cacioppo JT. Psychosocial and behavioral predictors of inflammation in middle-aged and older adults: the Chicago health, aging, and social relations study. Psychosomatic medicine. 2006;68:376–81. [DOI] [PubMed] [Google Scholar]
  • 31.Peek MK, Cutchin MP, Freeman D, Stowe RP, Goodwin JS. Environmental hazards and stress: evidence from the Texas City Stress and Health Study. Journal of Epidemiology & Community Health. 2009;63:792–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Turner RJ, Wheaton B, Lloyd DA. The epidemiology of social stress. American sociological review. 1995:104–25. [Google Scholar]
  • 33.Kanner AD, Coyne JC, Schaefer C, Lazarus RS. Comparison of two modes of stress measurement: Daily hassles and uplifts versus major life events. Journal of Behavioral Medicine. 1981. [DOI] [PubMed] [Google Scholar]
  • 34.Cohen S, Williamson GM. Perceived stress in a probability sample of the United States. In: Spacapan S, Oskamp S, editors. Newbury, CA: Sage; 1988. p. 31–67. [Google Scholar]
  • 35.Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. Journal of health and social behavior. 1983:385–96. [PubMed] [Google Scholar]
  • 36.Rector JL, Dowd JB, Loerbroks A, Burns VE, Moss PA, Jarczok MN, Stalder T, Hoffman K, Fischer JE, Bosch JA. Consistent associations between measures of psychological stress and CMV antibody levels in a large occupational sample. Brain, behavior, and immunity. 2014;38:133–41. [DOI] [PubMed] [Google Scholar]
  • 37.Miller GE, Freedland KE, Duntley S, Carney RM. Relation of depressive symptoms to C-reactive protein and pathogen burden (cytomegalovirus, herpes simplex virus, Epstein-Barr virus) in patients with earlier acute coronary syndromes. The American journal of cardiology. 2005;95:317–21. [DOI] [PubMed] [Google Scholar]
  • 38.Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Applied psychological measurement. 1977;1:385–401. [Google Scholar]
  • 39.Jenkins CD, Stanton B-A, Niemcryk SJ, Rose RM. A scale for the estimation of sleep problems in clinical research. Journal of clinical epidemiology. 1988;41:313–21. [DOI] [PubMed] [Google Scholar]
  • 40.Preacher KJ, Curran PJ, Bauer DJ. Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of educational and behavioral statistics. 2006;31:437–48. [Google Scholar]
  • 41.McCabe CJ, Kim DS, King KM. Improving present practices in the visual display of interactions. Advances in Methods and Practices in Psychological Science. 2018;1:147–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Scheffer J. Dealing with missing data. 2002. [Google Scholar]
  • 43.Mezuk B, Roux AVD, Seeman T. Evaluating the buffering vs. direct effects hypotheses of emotional social support on inflammatory markers: The Multi-Ethnic Study of Atherosclerosis. Brain, behavior, and immunity. 2010;24:1294–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Uchino BN, Bowen K, Carlisle M, Birmingham W. Psychological pathways linking social support to health outcomes: A visit with the “ghosts” of research past, present, and future. Social science & medicine. 2012;74:949–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Kiecolt-Glaser JK, Gouin J-P, Hantsoo L. Close Relationships, Inflammation, and Health. Neuroscience and biobehavioral reviews. 2010;35:33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Stellar JE, John-Henderson N, Anderson CL, Gordon AM, McNeil GD, Keltner D. Positive affect and markers of inflammation: Discrete positive emotions predict lower levels of inflammatory cytokines. Emotion. 2015;15:129. [DOI] [PubMed] [Google Scholar]
  • 47.Hostinar CE, Ross KM, Chen E, Miller GE. Modeling the association between lifecourse socioeconomic disadvantage and systemic inflammation in healthy adults: The role of self-control. Health Psychology. 2015;34:580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Pressman SD, Cohen S. Does positive affect influence health? Psychological bulletin. 2005;131:925. [DOI] [PubMed] [Google Scholar]
  • 49.Sephton SE, Dhabhar FS, Keuroghlian AS, Giese-Davis J, McEwen BS, Ionan AC, Spiegel D. Depression, cortisol, and suppressed cell-mediated immunity in metastatic breast cancer. Brain, behavior, and immunity. 2009;23:1148–55. [DOI] [PubMed] [Google Scholar]
  • 50.DeSantis A, DiezRoux A, Hajat A, Aiello A, Golden SH, Jenny N, Seeman T, Shea S. Associations of salivary cortisol levels with inflammatory markers: the Multi-Ethnic Study of Atherosclerosis. Psychoneuroendocrinology. 2012;37:1009–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Bergdahl J, Bergdahl M. Perceived stress in adults: prevalence and association of depression, anxiety and medication in a Swedish population. Stress and Health: Journal of the International Society for the Investigation of Stress. 2002;18:235–41. [Google Scholar]
  • 52.Stowe RP, Kozlova EV, Yetman DL, Walling DM, Goodwin JS, Glaser R. Chronic herpesvirus reactivation occurs in aging. Experimental gerontology. 2007;42:563–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Buschmann RN, Prochaska JD, Cutchin MP, Peek MK. Stress and health behaviors as potential mediators of the relationship between neighborhood quality and allostatic load. Annals of epidemiology. 2018;28:356–61. [DOI] [PubMed] [Google Scholar]
  • 54.McEwen BS. Stress, adaptation, and disease: Allostasis and allostatic load. Annals of the New York academy of sciences. 1998;840:33–44. [DOI] [PubMed] [Google Scholar]
  • 55.Slavich G, Epel E. The Stress and Adversity Inventory (STRAIN): An automated system for assessing cumulative stress exposure. Los Angeles: University of California, Los Angeles. 2010. [Google Scholar]

RESOURCES