Dissertations, Theses, and Capstone Projects

Date of Degree

9-2025

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Psychology

Advisor

Tyrel J Starks

Committee Members

Christopher Stults

Mark Brennan-Ing

Jennifer Ford

Sarit Golub

Subject Categories

Behavior and Behavior Mechanisms

Keywords

Couples, HIV, Medication Adherence, Sexual Minority Men, Interdependence

Abstract

Sexual minority men (SMM) report higher incidence and prevalence rates of HIV infection relative to their heterosexual counterparts. The burden of HIV infection has garnered research attention focused on HIV risk reduction and HIV disease management (such as medication adherence). More recently, attention has been paid to the association of comorbid chronic health conditions and ART adherence among people living with HIV (PLWH) more broadly. Yet, limited attention has been paid to the association of comorbidities among SMM living with HIV specifically, despite the high rates of HIV among this population and various other risk factors that put SMM at an increased risk for developing chronic health conditions. Additionally, none of the previous research on antiretroviral therapy (ART) adherence and comorbidities has accounted for group differences among individuals living with HIV with and without an AIDS diagnostic history, despite findings highlighting an association between AIDS diagnostic history, ART suboptimal adherence, and comorbid illnesses among PLWH.

Previous research among partnered SMM living with HIV suggests that romantic relationships can have a meaningful impact on health outcomes, as relationship functioning has been repeatedly associated with HIV related health outcomes, including ART adherence. Much of this existing work examining relationship functioning and HIV health outcomes has applied interdependence theory as the framework for understanding of the role of romantic relationships on HIV disease management. But various constructs of interdependence theory are less regularly studied. Specifically, the role of a shared appraisal of health threats or a team-based approach to health threats (facilitated by cognitive interdependence) has been posited to have a positive influence on health outcomes. However, investigations of this construct are still quite limited among partnered SMM living with HIV. An additional research gap is that this framework has yet to be applied to the study of ART adherence among partnered SMM living with HIV and comorbidities.

The current dissertation is the first to empirically test the association between comorbidities, relationship functioning, cognitive interdependence, and ART adherence outcomes among a diverse sample of partnered SMM living with HIV. The aforementioned constructs and the role of AIDS diagnostic history were examined across two studies.

Study 1 is a secondary data analysis of self-reported data collected through an online survey used to determine eligibility for active studies recruiting SMM across the US. The sample was composed of 518 partnered SMM living with HIV. Participants reported their demographics, their partner’s demographics, relationship characteristic (including relationship length, sexual agreement and couple serostatus), comorbidities (which was dichotomized as zero/one and two or more), relationship functioning (using the Perceived Relationship Quality Components scale), cognitive interdependence (using the Inclusion of Other in Self scale) and ART medication adherence (using the visual analog scale). ART adherence was dichotomized into optimal (≥ 90%) adherence and suboptimal (< 90%) adherence. On average, participants were 43.33 years of age, 54.2% of the sample was non-white and 66.0% reported having two or more comorbidities (e.g., multimorbidities).

A path model was specified in which ART medication adherence (a dichotomous variable representing optimal and suboptimal adherence) was the final endogenous outcome. ART adherence was modeled as a logistic distribution and regressed on cognitive interdependence, relationship functioning, and comorbidities (presence or absence of multimorbidities, which is defined as two or more comorbidities). Cognitive interdependence was then regressed on relationship functioning and comorbidities.

Results revealed significant positive direct associations between relationship functioning and ART adherence (OR = 1.03; p = .001), as well as relationship functioning and cognitive interdependence (β = 0.51; p < .001). Multimorbidities were negatively associated with ART adherence (OR = 0.50; p = .002) and not significantly associated with cognitive interdependence (OR = 0.50; p = .002). There were no significant indirect pathways identified between relationship functioning and ART adherence (B = 0.002; p = .66) and multimorbidities and ART adherence (B = -0.003; p = .72) through cognitive interdependence. As such, an alternative model was tested where the positioning of relationship functioning, and cognitive interdependence was reversed. The indirect path linking cognitive interdependence and ART adherence through relationship functioning in the alternative model was significant and positive (B = 0.11; p = .001). The pathway between multimorbidities and ART adherence through relationship functioning was non-significant (B = -0.04; p = .26).

Study 2 examined data collected from 1137 partnered SMM living with HIV across the US, using the online survey used to determine eligibility for active studies employed in Study 1. Measurements of demographics, partner characteristics, relationship characteristics, relationship functioning, comorbidities, cognitive interdependence and ART adherence were identical to Study 1. The Study 2 sample was also asked about AIDS diagnostic history and the presence of comedication to treat self-reported comorbidities. On average, participants were 42.18 years of age, 54.2% of the sample was non-white, 66.9% reported multimorbidities, and 74.2% of the sample did not have a history of AIDS. Group differences by AIDS diagnostic history were examined using sequential testing of model constraints. A fully-constrained model was compared to three alternative models which consecutively freed the interceptions/thresholds, error variances, and regression coefficients across groups. The presence of significant group differences between those with and without an AIDS diagnosis is indicated when freeing constraints yields significantly better fit to the data than a more constrained model.

There was no evidence that model results differed across groups defined by the presence or absence of AIDS diagnosis. More specifically, models in which the intercepts of endogenous variables were freed (χ2 (4) = 6.08, p = .19), models in which the error variances of endogenous variables were freed (χ2 (4) = 0.14, p = 1.00), and models in which the regression coefficients for all predictors were freed (χ2 (4) = 39.68, p = 0.35) did not have a significantly better fit than a fully constrained model.

As such, the single group multivariate model tested in Study 1 was replicated amongst the Study 2 sample, with AIDS diagnostic history included as a covariate. Results revealed significant positive direct associations between relationship functioning and ART adherence (OR = 1.04, p < .001), as well as relationship functioning and cognitive interdependence (β = .51, p < .001). Multimorbidities were not significantly associated with ART adherence (OR = -.05, p = .22) or cognitive interdependence (β = -.01, p = .59). Similar to previous investigations, AIDS diagnostic history was negatively associated with ART medication adherence (OR = 0.70, p = .02). There were no significant indirect effects identified between relationship functioning and ART adherence (B = 0.000 p = .89) or multimorbidities and ART adherence (B = 0.000; p = .89) through cognitive interdependence. In the alternative model (in which relationship functioning and cognitive interdependence was reversed) the indirect pathway between cognitive interdependence and ART adherence through relationship functioning was statistically significant and positive (B = 0.12; p < .001). The pathway between multimorbidities and ART adherence through relationship functioning was significant and negative (B = -0.05; p = .03).

Taken together, these two studies showed an association between relationship related factors and ART adherence outcomes among partnered SMM living with HIV. Additionally, these findings support the association between relationship functioning and cognitive interdependence posited in interdependence theory. These findings also highlight the important role of comorbidities and romantic relationships on HIV health outcomes, which lend support for the continued investigation of these factors. Finally, examinations of both multigroup and single group multivariate models suggest that a single group analysis is a reasonable methodological approach for the examination of the associations between relationship functioning, cognitive interdependence, comorbidities and ART adherence.

Download

Share

COinS