Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Jul 18.
Published in final edited form as: J Natl Med Assoc. 2017 Aug 2;110(4):371–377. doi: 10.1016/j.jnma.2017.07.006

Positive Social Interaction Offsets Impact of Low Socioeconomic Status on Stress

Eva N Woodward a,1, Jennifer L Walsh a,b,c,d, Theresa E Senn e, Michael P Carey b,c,d
PMCID: PMC6639010  NIHMSID: NIHMS1009677  PMID: 30126563

Abstract

Background

Stress is associated with unhealthy behaviors and premature morbidity and mortality, especially among those of low socioeconomic status (SES). Clarifying the roles of stress-related risk and protective factors can guide interventions designed to reduce stress and improve health among socioeconomically disadvantaged populations.

Purpose

(1) Replicate prior research showing that lower SES is associated with higher stress in a predominantly racial minority, socioeconomically disadvantaged sample, and (2) test the hypothesis that different types of social support (a protective factor) mitigate the deleterious effects of SES on self-reported perceived stress.

Methods

Low-income patients (N = 508, 54% male, 68% Black, Mage = 28) from a publicly-funded clinic completed self-report measures as part of a larger trial. Structural equation modeling tested the hypothesized associations among SES, social support, and stress.

Results

Individuals of lower SES, β = −0.27 (0.08), p < .01, and lower overall social support, β = −0.47 (0.05), p < .001, reported higher stress. Social support moderated associations between SES and stress, with participants with lower SES benefitting the most from social support. Positive social interaction was the strongest moderator, β = 0.20 (0.08), p = .01.

Conclusions

The associations among SES, stress, and social support corroborate prior research. One specific type of social support (viz., positive social interaction) was particularly important for decreasing stress among socioeconomically disadvantaged persons. Future research to confirm the associations among these factors is needed to inform theory and intervention design.

Keywords: social support, socioeconomic status, stress, resilience, protective factor

1.0. Introduction

Individuals from socioeconomically disadvantaged backgrounds, on average, demonstrate worse health behaviors,1 health outcomes,2 and die earlier 3 than their wealthier, more educated, and employed peers. One proposed mechanism to explain these health disparities is stress.3 Greater stress contributes to several unhealthy behaviors for those with low socioeconomic status (SES),4 with low SES being indirectly related to unhealthy behaviors through perceived stress.5 Thus, low SES contributes to high stress, and both low SES itself and high stress put individuals at higher risk for unhealthy behaviors. Identifying modifiable protective factors may inform stress management interventions for low SES individuals.

Social support may be a protective factor to offset stress, possibly contributing to decreases in unhealthy behaviors among socioeconomically disadvantaged individuals.6–8 One limitation of current research is that social support has not been studied more granularly to understand how different types may help to offset stress for socioeconomically disadvantaged individuals.9 Types of social support include affectionate, emotional/informational, tangible, and positive social interaction.10 See definitions and example survey questions assessing the four social support types in Table 1. Our sample provided an opportunity to study different types of social support among a sample of mostly low SES and racial minority individuals, for whom stress may be particularly harmful to health.11

Table 1.

Definitions and example survey items of social support types.

Social support type Definition Example item (MOS-SS)
Positive Social Interaction Uplifting encounters with others, perhaps over shared interests “Someone to get together with for relaxation”)
Affectionate Physical touch and feelings of love “Someone to love and make you feel wanted”
Tangible Providing logistic assistance, such as help with a task “Someone to help with daily chores if you were sick”
Emotional/lnformational Advice giving, problem solving, and emotional validation in a crisis “Someone to give you good advice about a crisis”
Open in a new tab

Source: Sherboume and Stewart (1991, Medical Outcomes Study-Social Support (MOS-SS) survey.

To replicate prior research in a novel sample (predominantly racial minority and urban dwelling), we examined whether lower SES would be associated with higher stress. To extend prior research, we tested the novel hypothesis that social support would mitigate effects of SES on stress. We explored different types of social support to increase specificity of findings.

2.0. Method

2.1. Participants

In this secondary analysis, participants were patients attending a publicly funded sexually transmitted disease (STD) clinic and in a randomized controlled trial (RCT) to evaluate a sexual risk reduction intervention12. Inclusion criteria for the RCT were age 16 or older and sexual risk behavior past three months. Exclusion criteria were severe mental impairment, current inpatient substance use treatment, HIV infection, and planning to move out of area in the next year. Of 2,766 patients approached, 2,677 (97%) agreed to be screened, 1,322 (49%) were eligible, and 1,010 (76%) consented and completed baseline surveys. Participants were randomly assigned to complete a general health (n = 508) or a sexual health (n = 502) survey. The current sample consisted of general health survey participants. We chose the general health survey participants because the questions about the main variables (social support, stress) were assessed in that survey and not in the sexual health survey.

2.2. Procedures

A research assistant met with patients in a private room and obtained verbal consent for screening. Those who were eligible and interested provided written, informed consent. Participants completed an audio computer-assisted self-interview, allowing low-literacy individual to participate, in a private room. Participants viewed an intervention video as part of the RCT (unrelated to stress) and were reimbursed $30. All procedures were approved by participating institutional review boards.

2.3. Measures

2.3.1. Demographic information

We obtained information on participant demographics; a race dummy variable was created: white, black, and other.

2.3.2. Socioeconomic status (SES)

SES was a latent factor indicated by these categorical variables: annual family income, highest grade completed in school (education), and current employment status.

2.3.3. Perceived stress

Three items from the Perceived Stress Scale (PSS)7 assessed stress in the last month: “How often have you felt difficulties were piling up so high that you could not overcome them?”; “How often have you felt that you were unable to control the important things in your life?”; and “How often have you felt that things were going your way?”. Participants rated each item 0 (never) to 4 (very often); higher scores indicate higher stress levels (including recoding item 3). The PSS has been reliable and valid with urban populations;13,14 internal consistency was adequate in this study (Cronbach’s α = 0.68). Items served as indicators of a latent stress construct.

2.3.4. Social support

The 19-item Medical Outcomes Study-Social Support survey10 assessed perceived support. Participants were asked, “How often is each of the following types of support available to you if you need it?” with responses from 1 (none of the time) to 5 (all of the time). The scale included four types of support: emotional / informational (“Someone to give you good advice about a crisis”), tangible (“Someone to help with daily chores if you were sick”), affectionate (“Someone to love and make you feel wanted”), and positive social interaction (“Someone to get together with for relaxation”). We summed each type and types served as indicators for a latent social support factor. Following recommendations, each type score was transformed on a 0–100 scale10; higher values indicated higher support. This measure is reliable in racial minority samples (α = 0.93)15 and also in our sample (α = .97).

2.4. Data Management and Analysis

Data were analyzed using structural equation modeling (SEM) in Mplus.16 We controlled for age, race, and sex in all models.17 Model fit was assessed using the comparative fit index (CFI), the Tucker-Lewis index (TLI), and the root-mean-square error of approximation (RMSEA). Indicators of acceptable fit are CFI > .90, TLI > .90, and RMSEA < .06.18,19 To test moderation, we created latent interaction terms using the Mplus function XWITH. Herein, we report standardized coefficients.

Responses were missing for 0.39% to 3.54% of participants per variable. The variable with the most missing responses was income; all other variables were missing less than 0.40% of cases. We used multiple imputation to replace missing values, a widely-accepted method for dealing with missing data.20

3.0. Results

3.1. Descriptive Statistics

Participants were 508 individuals (54% male, 68% Black, Mage = 28). Participants reported feeling moderately stressed. Their reported levels of social support were similar to other samples.10 See Table 2 for sample characteristics.

Table 2.

Characteristics of 508 individuals from a publicly-funded STD clinic in upstate New York.

Characteristic N (%)
Sex (assigned at birth)
 Male 272 (54)
 Female 236 (47)
Age M = 27.9, SD = 9.0, Range = 16–63
Marital Status
 Married 25 (5)
 Divorced 27 (5)
 Separated 30 (6)
 Single (never married) 421 (83)
 Widowed
Annual Family Income
 < $15,000 258 (51)
 $15,000 to $30,000 155 (31)
 $30,000 to $45,000 50 (10)
 > $45,000 27 (5)
Employment Status
 Unemployed 253 (50)
 Part-time 100 (20)
 Full-time 155 (31)
Education Level
 < 12th grade 136 (27)
 GED 91 (18)
 High school diploma 96 (19)
 Some college (23)
 College degree or more 69 (14)
Race
 African American 345 (68)
 White 92 (18)
 Other 92 (18)
Perceived Stress M 5.4, SD 2.5, Range = 0–12
Social Support (Overall) M = 67.7, SD = 24.7, Range = 0–100
 Affectionate M = 72.6, SD = 28.1, Range = 0–100
 Emotional/Informational M = 65.6, SD = 27.3, Range = 0–100
 Tangible M = 64.8, SD = 30, Range 0–100
 Positive Social Interaction M = 72.4 SD = 26.6, Range 0–100
Open in a new tab

3.2. Measurement Model

We tested the fit of measurement models including three latent constructs (SES, social support, and stress) and correlations among all constructs. To identify latent constructs, variance was fixed at 1.21 With this method, the sample mean of each latent construct is 0 and each 1-unit change in a latent construct can be interpreted as a change of 1 standard deviation (SD). We allowed correlations between (a) residuals of the highly-correlated social support types affection and positive social interaction22 and between (b) stress items that were not reverse scored. The measurement model fit the data well, χ2 (30, N = 508) = 34.54, CFI = 0.99, TLI = 0.99, RMSEA = 0.02. All factor loadings were positive and significant, ps < .001.

3.3. Structural Model and Direct Associations

We tested a structural model including a latent social support construct, which included all four social support types. Overall fit indices were unavailable because latent interaction terms were included in the model.23 However, the structural model in Figure 1 fit well prior to the addition of interaction terms, χ2 (57, N = 508) = 82.88, CFI = 0.97, TLI = 0.95, RMSEA = 0.03, and accounted for 32% of the variance in stress.

Figure 1.

Figure 1.

Open in a new tab

Results of structural equation model testlng associations between SES, social support and stress. *p < 0.05. ** < 0.000. Age, race, and sex were covariates. SES, stress, and social support are latent factors. For the Perceived Stress Scale (PSS), the second item had a low factor loading in this sample and was dropped.

Prior to the addition of interaction terms, individuals with lower SES, β = −0.27 (0.08), p < .01, and lower overall social support, β = −0.47 (0.05), p < .001, reported higher stress. In separate models, higher stress was also associated with lower social support of each type: affectionate, β = −0.37 (0.05), p < .001, tangible, β = −0.40 (0.05), p < .001, positive social interaction, β = −0.33 (0.05), p < .001, and emotional/informational, β = −0.41 (0.05), p < .001. Individuals with higher SES also reported higher overall social support, β = 0.12 (0.06), p < .05. SES was not directly related to affectionate support, β = 0.02 (0.06), p = .67, but was positively related to tangible support, β = 0.11 (0.05), p < .05, positive social interaction, β = 0.11 (0.06) p < .05, and emotional/informational support, β = 0.15 (0.05) p < .01.

3.4. Social Support as a Hypothesized Protective Factor, Moderating Associations between SES and Stress

After adding interaction terms, overall social support moderated the association between SES and stress, β = 0.19 (0.09), p < .05. Although there was a strong, negative SES-stress association for those one SD below the mean in social support, β = −0.55 (0.15), p < .001, this association disappeared for those one SD above the mean, β = −0.13 (0.13), p = .32. This pattern of results, whereby social support reduced stress particularly for those with low SES, can be seen in Figure 2a.

Flgure 2.

Flgure 2.

Open in a new tab

Soclal support moderates the association between socloeconomlc status and stress. (A) Overall social support as a latent construct moderates the association between SES and stress, ß = 0.19 (0.09), p < 0.05. (B) Positive social interaction as a moderator of the association between SES and stress, ß = 0.20 (0.08), p < 0.01. The figure depicts the association between SES and stress for those with below average SD below the mean) and above overage (I SD above the mean) levels of social support and positive social interaction. as well as 95% confidence intervals, Age. sex, and race included as covariates.

We also investigated social support types separately as moderators of the association between SES and stress. These models showed that three types of social support offset the negative impact of SES on stress, including positive social interaction, β = 0.19 (0.07), p < .01, affectionate support, β = 0.17 (0.09), p = .05, and tangible support, β = 0.16 (0.08), p < .05. Emotional/informational social support was non-significant, β = 0.12 (0.10), p = .26. Next, we included all significant interactions in the same model, systematically pruning the model for parsimony. We found affectionate support to have the weakest effect on the SES-stress association, β = −0.06 (0.15), p = .69, followed by tangible support, β = 0.06 (0.07), p = .47. Following the removal of these non-significant interaction terms, positive social interaction remained a significant moderator of the SES-stress association, β = 0.20 (0.08), p = .01, indicating that, for those with below average positive social interaction, lower SES was associated with higher stress. Although there was a strong, negative SES-stress association for those one SD below the mean in positive social interaction, β = −0.57 (0.15), p < .001, this association disappeared for those one SD above the mean, β = −0.12 (0.12), p = .32. Thus, as seen in Figure 2b, at lower levels of SES, those with higher levels of positive social interaction reported less stress than those with lower levels of positive social interaction. When considering all types simultaneously, positive social interaction showed evidence of playing the most important role in offsetting the impact of SES on stress.

4.0. Discussion

We utilized data from a sample largely comprised of racial minority, socioeconomically disadvantaged individuals at a public clinic to examine one protective factor (viz., social support) to better understand associations between SES and stress. Lower SES was associated with higher stress, beyond age, sex, and race, consistent with prior literature.24 People with higher social support reported lower stress, consistent with the stress-buffering, or protective, effects of social support.7 Consistent with hypotheses and one previous study, overall social support mitigated the negative impact of low SES on stress.25

A novel contribution of this study involved the investigation of a more nuanced typology of social support to increase the specificity of our findings. Emotional/informational support had the strongest direct associations with stress in our socioeconomically disadvantaged sample, meaning that higher levels of advice giving, problem solving, and emotional validation were directly related to lower stress. However, positive social interaction emerged as the most important type of social support in offsetting the impact of SES on stress, particularly for those who were very socioeconomically disadvantaged. Positive social interaction may be especially helpful in stress reduction for these individuals because it allows them to experience positive affect in a way that tangible support (e.g., childcare) would not and in a way distinct from the general, nonspecific positive affect that affectionate support offers. It may also reduce stress for the very socioeconomically disadvantaged through social comparison with others, prompting cognitive reappraisal.26

Our results suggest a process of resilience, in which positive social interaction may be a protective factor that offsets typically expected higher levels of stress among socioeconomically disadvantaged and/or racial minority individuals.27 Resilience is a not a trait; rather, it is a dynamic process that can be represented by adaptation or posttraumatic growth among a population from whom there is evidence of higher likelihood of a negative trajectory due to adversity.28 Adaptation or posttraumatic growth can be positive traits, developmental milestones, behaviors, or health outcomes.29 Specifically, these results support the protective model of resilience, which posits that, despite adversity (low SES), a negative outcome (stress) can be mitigated by a protective factor (social support).30,31 Resilience research has begun to influence and improve health interventions for disadvantaged populations, with more calls for resilience research on individuals who are at high risk for health problems.27,30,32 Because SES can be difficult to change, resilience research can help researchers learn from these individuals who may have protective factors to cope with adversity as well as—or better than—higher SES peers. Other research might simulate our methods by investigating hypothesized protective factors as moderators of traditionally negative health outcome pathways for populations with adversity.

Because stress is linked to unhealthy behaviors and poor health outcomes, stress management is vital to disease prevention for socioeconomically disadvantaged individuals. Stress management interventions with socioeconomically disadvantaged individuals might seek to increase positive social interaction, specifically, to allay distress, improve health behaviors,33 and delay premature morbidity and mortality. Continued investigation might explore whether positive social interaction can be generalized as a protective factor against stress for other populations. By identifying positive social interaction as particularly important for low-SES individuals, our analyses showcase the value of investigating types of social support. By doing so, we can better identify specific targets for intervention. Positive social interaction should be viewed as only one of many approaches to preventing unhealthy behaviors, morbidity, and early mortality among socioeconomically disadvantaged populations.27,34,35

4.1. Strengths and Limitations

Our sample included socioeconomically disadvantaged, primarily Black and African-American adults—an important sample in which to study stress due to higher stress levels that originate from socioeconomic hardship and racial discrimination.17,36 However, participants in our study came from one STD clinic, possibly limiting generalizability. A further sample limitation is that we did not have specific information about ethnicity, national origin, or immigrant status for participants; thus, results might be different for those different subgroups. There may be something about the population in this sample that enhances responses to positive social interaction and other populations might respond better to other social support types. Our study benefitted from the use of an advanced statistical approach (viz., SEM), which allowed us to account for measurement error and inter-correlations among variables. However, the data were cross-sectional, limiting our ability to assess causal relationships (i.e., whether changes in social support were associated with changes in stress). A longitudinal study assessing relations among SES, stress, and types of social support would clarify temporal associations. Although we were able to investigate types of social support, improving on prior research, limitations of secondary data prevented us from considering other protective factors against stress (e.g., personality traits, coping styles). It would be helpful for future research to also specify the therapeutic agent of positive social interaction, possibly including relaxation, distraction, or humor.

4.2. Conclusions

Overall social support was protective against stress for low SES individuals, and one specific type—positive social interaction—played the strongest role in mitigating the negative effect of very low SES on stress. Whereas the lowest SES individuals without high support reported more stress than high SES peers, the lowest SES with high positive social interaction reported similar levels of stress as their lower-risk, higher SES peers. Facilitating positive social interaction appears to be an important and feasible stress reduction intervention target for low-SES and/or racial minority individuals.

References

  • 1.Stringhini S, Dugravot A, Shipley M, Goldberg M, Zins M, Kivimäki M, et al. Health Behaviours, Socioeconomic Status, and Mortality: Further Analyses of the British Whitehall II and the French GAZEL Prospective Cohorts. PLOS Med. 2011. February 22;8(2):e1000419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Adler NE, Boyce T, Chesney MA, Cohen S, Folkman S, Kahn RL, et al. Socioeconomic status and health. The challenge of the gradient. Am Psychol. 1994. January;49(1):15–24. [DOI] [PubMed] [Google Scholar]
  • 3.Sapolsky R Sick of Poverty. Sci Am. 2004;293:7. [DOI] [PubMed] [Google Scholar]
  • 4.Adler NE, Rehkopf DH. U.S. Disparities in Health: Descriptions, Causes, and Mechanisms. Annu Rev Public Health. 2008;29:235–52. [DOI] [PubMed] [Google Scholar]
  • 5.Senn TE, Walsh JL, Carey MP. The mediating roles of perceived stress and health behaviors in the relation between objective, subjective, and neighborhood socioeconomic status and perceived health. Ann Behav Med. 2014;48(2):215–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Webb Hooper M, Baker EA, McNutt MD. Associations between coping, affect, and social support among low-income African American smokers. Addict Behav. 2013. November;38(11):2736–40. [DOI] [PubMed] [Google Scholar]
  • 7.Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychol Bull. 1985;98(2):310–57. [PubMed] [Google Scholar]
  • 8.Rios R, Aiken LS, Zautra AJ. Neighborhood contexts and the mediating role of neighborhood social cohesion on health and psychological distress among Hispanic and non-Hispanic residents. Ann Behav Med. 2012;43(1):50–61. [DOI] [PubMed] [Google Scholar]
  • 9.Thoits PA. Mechanisms Linking Social Ties and Support to Physical and Mental Health. J Health Soc Behav. 2011. June 1;52(2):145–61. [DOI] [PubMed] [Google Scholar]
  • 10.Sherbourne CD, Stewart AL. The MOS social support survey. Soc Sci Med. 1991;32(6):705–14. [DOI] [PubMed] [Google Scholar]
  • 11.Williams DR, Yu Y, Jackson JS, Anderson NB. Racial Differences in Physical and Mental Health Socio-economic Status, Stress and Discrimination. J Health Psychol. 1997. July 1;2(3):335–51. [DOI] [PubMed] [Google Scholar]
  • 12.Carey MP, Senn TE, Walsh JL, Coury-Doniger P, Urban MA, Fortune T, et al. Evaluating a Brief, Video-Based Sexual Risk Reduction Intervention and Assessment Reactivity with STI Clinic Patients: Results from a Randomized Controlled Trial. AIDS Behav. 2015. July;19(7):1228–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24(4):385–96. [PubMed] [Google Scholar]
  • 14.Roberti JW, Harrington LN, Storch EA. Further psychometric support for the 10-item version of the perceived stress scale. J Coll Couns. 2006. Fal;9(2):135–47. [Google Scholar]
  • 15.Brittain K, Loveland-Cherry C, Northouse L, Caldwell CH, Taylor JY. Sociocultural differences and colorectal cancer screening among African American men and women. Oncol Nurs Forum. 2012;39(1):100–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Muthén LK, Muthén BO. Mplus. Stat Anal Latent Var User’s Guide. 1998;3. [Google Scholar]
  • 17.American Psychological Association Task Force on Socioeconomic Status. Report of the APA Task Force on Socioeconomic Status [Internet]. Washington D.C.: American Psychological Association; 2007. [cited 2016 Dec 21] p. 1–46. Available from: http://www.apa.org/pi/ses/resources/publications/index.aspx [Google Scholar]
  • 18.Hu L, Bentler P. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model Multidiscip J. 1999;6:1–55. [Google Scholar]
  • 19.Kline RB. Principles and Practice of Structural Equation Modeling. New York, NY: Guilford Press; 1998. [Google Scholar]
  • 20.Schafer JL. Multiple imputation: A primer. Stat Methods Med Res. 1999;8(1):3–15. [DOI] [PubMed] [Google Scholar]
  • 21.Kline RB. Principles and Practice of Structural Equation Modeling. New York, NY: The Guilford Press; 2011. [Google Scholar]
  • 22.Leung J, Pachana NA, McLaughlin D. Social support and health‐related quality of life in women with breast cancer: A longitudinal study. Psychooncology. 2014;23(9):1014–20. [DOI] [PubMed] [Google Scholar]
  • 23.Muthén LK. Mplus not reporting model fit index [Internet]. StatModel. 2014. Available from: http://www.statmodel.com/discussion/messages/11/19585.html?1408576087
  • 24.Gallo LC, Bogart LM, Vranceanu A-M, Matthews KA. Socioeconomic status, resources, psychological experiences, and emotional responses: A test of the reserve capacity model. J Pers Soc Psychol. 2005;88(2):386–99. [DOI] [PubMed] [Google Scholar]
  • 25.John-Henderson NA, Stellar JE, Mendoza-Denton R, Francis DD. Socioeconomic status and social support: Social support reduces inflammatory reactivity for individuals whose early-life socioeconomic status was low. Psychol Sci. 2015; [DOI] [PubMed] [Google Scholar]
  • 26.Lazarus RS, Folkman S. Stress, appraisal, and coping. New York: Springer; 1984. [Google Scholar]
  • 27.Chen E, Miller GE. “Shift and persist” strategies: Why low socioeconomic status isn’t always bad for health. Perspect Psychol Sci. 2012;7:23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Masten AS, Wright MO. Resilience over the lifespan: Developmental perspectives on resistance, recovery, and transformation In: Reich JW, Zautra AJ, Hall JS, Reich JW, Zautra AJ, Hall JS, editors. Handbook of adult resilience [Internet]. New York, NY, US: Guilford Press; 2010. p. 213–37. Available from: http://search.ebscohost.com/login.aspx?direct=true&db=psyh&AN=2010-10101-011&site=ehost-live [Google Scholar]
  • 29.Fergus S, Zimmerman MA. ADOLESCENT RESILIENCE: A Framework for Understanding Healthy Development in the Face of Risk. Annu Rev Public Health. 2005;26(1):399–419. [DOI] [PubMed] [Google Scholar]
  • 30.Dunkel Schetter C, Dolbier C. Resilience in the context of chronic stress and health in adults. Soc Personal Psychol Compass. 2011;5(9):634–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Masten AS, Reed M-GJ. Resilience in development In: Snyder CR, Lopez SJ, Snyder CR, Lopez SJ, editors. Handbook of positive psychology [Internet]. New York, NY, US: Oxford University Press; 2002. p. 74–88. Available from: http://search.ebscohost.com/login.aspx?direct=true&db=psyh&AN=2002-02382-006&site=ehost-live [Google Scholar]
  • 32.Herrick AL, Stall R, Goldhammer H, Egan JE, Mayer KH. Resilience as a research framework and as a cornerstone of prevention research for gay and bisexual men: Theory and evidence. AIDS Behav. 2014;18(1):1–9. [DOI] [PubMed] [Google Scholar]
  • 33.Uchino BN. Social support and health: A review of physiological processes potentially underlying links to disease outcomes. J Behav Med. 2006;29(4):377–87. [DOI] [PubMed] [Google Scholar]
  • 34.Blue S, Shove E, Carmona C, Kelly MP. Theories of practice and public health: understanding (un)healthy practices. Crit Public Health. 2016. January 1;26(1):36–50. [Google Scholar]
  • 35.Borrelli B Smoking cessation: Next steps for special populations research and innovative treatments. J Consult Clin Psychol. 2010;78(1):1–12. [DOI] [PubMed] [Google Scholar]
  • 36.Williams DR, Mohammed SA. Discrimination and racial disparities in health: evidence and needed research. J Behav Med. 2009;32(1):20–47. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES