Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Psychology (Arts and Sciences)

Date of Defense


First Committee Member

Michael Antoni

Second Committee Member

Brian Doss

Third Committee Member

Gail Ironson

Fourth Committee Member

Sara Czaja

Fifth Committee Member

Armando Mendez


Chronic fatigue syndrome (CFS) is a debilitating illness that is characterized by heterogeneous, systemic symptoms that impact patients’ and their caregiving partners’ quality of life. Extant literature has found hypothalamic-pituitary-adrenal (HPA) axis and neuroimmune biomarkers associated with the disorder, but the impact of relationship satisfaction and patients’ communication satisfaction about symptoms on these CFS-relevant biological markers and on CFS patients’ CFS symptom severity has not been examined. Like others suffering from incapacitating chronic illnesses, CFS patients are often homebound, on disability, and/or face unemployment; however, people suffering from CFS report significantly less social support and more stigma from society, as compared to other patient populations coping with the challenges of chronic and acute illnesses. CFS patients, at times, also report a lack of understanding from their surrounding support network. Because of the unique challenges associated with this commonly misunderstood illness with no definitively known cause or cure, interventions have been developed to synergistically ameliorate the mental and physical toll this disorder causes. Specifically, cognitive behavioral stress management (CBSM) has been developed to concurrently improve coping skills, stress, mood, physical symptoms, HPA and neuroimmune functioning in many patient populations, including CFS patients. To address the needs of CFS patients, and other patient populations for whom in-person therapy appointments are not feasible, this intervention has been adapted to dissemination via telephone and videophone/tablet. The present dissertation study involved an analysis of the effects of relationship satisfaction (Dyadic Adjustment Scale, DAS), depression (Center for Epidemiologic Studies-Depression, CES-D, and patient symptom disclosure satisfaction (PSDS) on fatigue severity (Fatigue Symptom Index, FSI), overall CFS symptom severity (CDC CFS Symptom Scale), salivary evening cortisol, and serum pro-inflammatory cytokines in 150 patients with CFS who were enrolled, with their caregiving partner, into a randomized controlled trial testing the efficacy of a 10-week remotely-delivered group CBSM program versus an attention-matched health promotion (HP) control program. Depression and PSDS were examined as indirect variables of the hypothesized effects of relationship satisfaction on baseline CFS-related variables (Aim 1) using structural equation modeling (SEM) in Mplus. For Aim 2, changes in depression and PSDS were also studied longitudinally as mediators of the effects of intervention assignment on CFS-related outcomes (5 Month Follow Up- T2). Specifically, the intervention effects on the 5-month change in the mediators (T2-T1) and on the T2 outcomes were examined using repeated measures analysis of covariance (ANCOVA) and path analysis using structural equation modeling (SEM). Parallel multiple mediation modeling was used to test the indirect effects of the two mediators simultaneously. An examination of cross-sectional, direct and indirect effect analyses for Aim 1 showed that relationship satisfaction (DAS Total) did not have a significant direct effect on the outcomes of interest (CFS symptom severity, fatigue severity, evening cortisol, and IL-6, and TNF-α). While relationship satisfaction did not exert a significant direct effect on any of the outcomes at baseline, greater relationship satisfaction consistently predicted greater PSDS and less depression in all models. For baseline analyses of CFS symptoms, both depression and PSDS significantly related to greater fatigue severity (p’s < 0.05), while only depression was associated with greater CFS symptom severity (p < 0.01). Analyses of baseline inflammatory markers revealed that greater PSDS significantly related to greater TNF-α, even when relationship satisfaction was included in the model (p=0.02). IL-6 or evening cortisol was not significantly related to depression or PSDS. The second part of the dissertation (Aim 2) compared the effectiveness of the interventions CBSM vs HP on the outcomes of interest measured at the 5 month follow-up using repeated measures analysis of variance (RANOVA). To examine the mechanism of action of the interventions, SEM was used to estimate the direct effect of group assignment and indirect effects of change in PSDS and depression (T2-T1) on the 5-month outcomes. Relationship satisfaction was not included in the longitudinal models as a covariate or predictor because it did not relate to CFS-related outcomes variables at baseline in Aim 1. There were no time by treatment intervention effects on any of the outcomes (CFS symptom severity, fatigue severity, evening cortisol, and IL-6, and TNF-α), nor on depression or PSDS at 5-months. Depression severity scores decreased and PSDS scores increased over time in both treatment arms, as expected, though this effect was not significant. Though the participants were randomly assigned to each intervention, the HP group started the trial with significantly more depressive symptoms, worse relationship quality, and more severe CFS and fatigue symptoms at baseline (all p’s < 0.05), which may explain the lack of significant group effects. Using SEM, a between-group analysis of intervention effects showed that group assignment had a direct effect on fatigue severity, such that those assigned to the HP group experienced greater fatigue severity at 5 months (T2) (p= 0.02). Indirect effect analysis showed that greater magnitude of change in depression severity scores (becoming more depressed between T1 and T2 in both treatment arms) experienced greater fatigue severity at 5M. When baseline fatigue severity was added as a covariate, the previously significant group effects were no longer significant (p= 0.65), but the effect of change in depression remained significant (p<0.01). Like fatigue severity outcomes, the direct effects and parallel mediation modeling showed that group assignment had a direct effect on CFS symptom severity. Those assigned to the HP group experienced greater CFS symptom severity at 5 months (T2). Indirect effect analysis showed that greater magnitude of change in PSDS severity scores (becoming more satisfied with communication between T1 and T2 within both treatment arms) predicted decreased fatigue severity at 5M. Importantly, these effects became non-significant when baseline CFS symptom severity was added to the model as a covariate (both p’s >0.05). Group assignment did not exert any effects on evening cortisol, IL-6, or TNF-α, nor were there any significant indirect effects on the biological outcomes by change in depression or PSDS. In conclusion, both sets of analyses (ANCOVA and SEM) showed no intervention effects on change in the hypothesized mediators and outcomes at 5 months. There were several study limitations and strengths. Importantly, this study did not look at the final follow-up time point at 9-months post-baseline, where there may have been intervention effects. By chance, the participants randomly assigned to the attention-matched health-promotion control condition (HP) began the intervention with significantly worse depressive and CFS symptoms, and worse relationship quality. Additionally, the effects of the HP could have been too strong (therapeutically) as it was delivered on an individual dyad basis rather than a group format used for the CBSM dyads. This could account for the lack of significant differences in effectiveness between treatment arms. One strength of the study was that it provided evidence that depression and PSDS may contribute to better relationship satisfaction in couples dealing with CFS. Second there was evidence that depression may also contribute to less fatigue. Further research should consider intervention design modifications, and delve deeper into mechanisms of change, especially considering dyad-relevant variables (e.g., effects of relationship satisfaction, perceived valence of partner comments) and stress management processes (e.g. perceived skill in using CBSM techniques), that may inform interventions aimed at improving mental and physical well-being of CFS patients and their caregiving partners, among other patient populations coping with chronic illnesses.


chronic fatigue syndrome; fatigue; behavioral medicine; stress management; relationship satisfaction; depression