Extreme Weather, Resource Insecurities, and Mobility As Drivers of Relational-Level HIV Vulnerability Among Refugee Adolescents and Youth In a Ugandan Humanitarian Setting

Background and Purpose: Refugee settlements are disproportionately exposed to extreme weather events (EWE) such as flooding, drought, and extreme heat. EWE are linked with HIV vulnerabilities through complex pathways, including resource insecurity, displacement, and disrupted relational dynamics. Yet few studies have explored linkages between EWE-related risks and HIV vulnerabilities among refugee youth in climate-affected humanitarian settings. To address this knowledge gap, we aimed to identify distinct subgroups based on patterns of risks related to EWE exposure, resource insecurities, and recent mobility, and to assess associations between these risk profiles and relational-level HIV vulnerabilities among refugee youth in Bidi Bidi refugee settlement, Uganda.

Methods: Among a purposive sample of refugee youth aged 16-24 in Bidi Bidi refugee settlement, we conducted Latent Class Analysis (LCA) to examine EWE-related risk classes (resource insecurities [water insecurity, food insecurity, sanitation insecurity], past 12-month EWE exposure [cumulative exposure, exposure to multiple types of EWE], and mobility [past 6-month EWE-related overnight travel from residence]). We then conducted multivariable regression, adjusted for gender, employment, and length of time in Uganda, to assess associations between latent class membership and relational-level HIV vulnerabilities (sexual relationship power, past-year intimate partner violence, condom use self-efficacy, reproductive autonomy).

Results: Most participants (n=400, 50% women; mean age: 19.49, standard deviation: 2.3) reported experiencing multiple types of past-year EWE, with 61.5% (n=246) reporting 2-4 types, 30.0% (n=12) reporting 5 or more, and 8.5% (n=34) reporting one. When it came to past-year EWE frequency, more than three-quarters (77.2%, n=308) reported experiencing more than one type of EWE, at least more than once. The three-class model was chosen as the optimal model based on statistical fit and interpretability. Class profiles included: Class 1 (n=72, 18.0%; moderate risk, characterized by moderate resource insecurity, high mobility, and high EWE exposure); Class 2 (n=34, 8.5%; lower risk, characterized by water insecurity, low mobility, low EWE exposure); and Class 3 (n=294, 73.5%; higher risk, characterized by severe resource insecurity, moderate mobility, high EWE exposure). Compared to Class 2 (lower risk), Class 1 (moderate risk) and Class 3 (higher risk) reported lower sexual relationship power (Class 1: adjusted beta coefficient (aβ)= -3.63, 95% confidence interval (CI)= -6.27, -0.98; Class 3: aβ= -4.24, 95%CI= -6.80, -1.68) and condom use self-efficacy (Class 1: aβ= -2.55, 95%CI= -4.21, -0.90; Class 3: aβ= -3.09, 95%CI= -4.69, -1.48), and higher past-year intimate partner violence (Class 1: Odds ratio (OR)=4.81, 95%CI=1.00, 23.03; Class 3: OR=5.59, 95%CI=1.19, 26.36). Class 1 (moderate risk) reported lower reproductive autonomy (aβ= -0.27, 95%CI= -0.54, -0.01) than Class 2 (lower risk).

Conclusions and Implications: Refugee youth participants' differing experiences of resource insecurity, extreme weather event exposure, and weather-related mobility patterns were in turn associated with a range of relational-level HIV vulnerabilities. Humanitarian and health systems can integrate HIV prevention with climate adaptation in climate-affected humanitarian settings. As the frequency and severity of EWE continue to increase globally and disproportionately affect humanitarian settings, centering the experiences of refugee youth—particularly in low-resource humanitarian contexts such as in Uganda—can advance sexual and reproductive health and rights.