Flaherty, Brian PMagis-Weinberg, LucíaGomez Juarez, Nancy2025-05-122025-05-122025GomezJuarez_washington_0250O_27922.pdfhttps://hdl.handle.net/1773/53040Thesis (Master's)--University of Washington, 2025Background: Aging with HIV may pose challenges in African settings, where resources are limited and older women often rely on social networks for assistance. We compared the social networks of middle-aged and older Kenyan women with and without HIV recruited from two populations: the Mombasa Cohort, a long-term open cohort study of women who engage in sex work, and the general population in Mombasa. We examined associations between HIV status and contact frequency and relationship quality.Methods: In this cross-sectional study, we used an egocentric network approach to examine social network composition of "egos" (participants) and their "alters" (individuals with whom participants communicated in the past 6 months). Computer-assisted interviews collected ego and alter characteristics. Two multilevel models for clustered data were fitted for each outcome: unadjusted models included only the ego's HIV status as the main predictor, while adjusted models included HIV status and potential confounding variables. An exploratory analysis evaluated recruitment source as the main predictor. Results: After adjusting for sociodemographic factors, women with HIV (WWH) reported slightly lower levels of emotional support from their networks compared to women without HIV. Stratified analyses by recruitment source and adjusting for confounders showed no differences by HIV status among the Mombasa Cohort participants. In contrast, WWH from the general population reported lower levels of emotional, financial, and affectionate support compared to their HIV-negative counterparts. In exploratory analysis with recruitment source as the primary predictor, Mombasa Cohort women reported lower levels of emotional support, and had higher odds of having non-daily contact with their alters, feeling less close to their alters, receiving low affectionate support, and experiencing low positive social interactions compared to general population women. In contrast, Mombasa Cohort women reported higher levels of financial support compared to general population women. Conclusions: Our findings contribute to the broader understanding of the relationship between HIV, social networks, and social support, highlighting the importance of considering a history of sex work when developing support interventions for older women living with HIV in Kenya and similar settings.application/pdfen-USnoneHIVKenyaolder womensocial frailtysocial networkssocial supportQuantitative psychologyPublic healthPsychologyThesis