Abstract
Although social relationships impact health and social dynamics play a key role in shaping HIV vulnerability, results from intentional efforts to build or strengthen social relationships have not been synthesized to understand if and how such interventions work to prevent HIV in low- and middle-income countries (LMICs). We conducted a systematic review of HIV prevention interventions implemented in LMICs, published between 2010 and 2022, that used pre/post or multi-arm methods to evaluate interventions that focused on building/strengthening social relationships to impact HIV-related outcomes. We searched PubMed, CINAHL, EMBASE, Sociological Abstracts, and PsycInfo on August 16, 2022, complemented by hand-searching and secondary reference searching. We used a standardized form for data abstraction and assessed risk of bias using the Evidence Project tool. Results were synthesized narratively, and studies were classified in an emergent typology based on the function of social tie building within the intervention. Fifty-one articles presenting results from 28 studies met the inclusion criteria and were included. Within these studies, we identified five types of social tie interventions, including community-wide social mobilization ("collaboration", n = 3), formation of collectives to address both upstream and downstream health-related factors ("collectivization", n = 13), forming or strengthening groups to enhance peer support and build skills ("clubs", n = 4), expanding personal networks among individuals ("companionship", n = 2), and strengthening ties between heterogeneous groups/non-peers ("connections", n = 2). Four studies addressed two or more types of social ties strengthening and were classified as "cross-cutting." Across these categories, most studies found that interventions were associated with some positive health-related changes, such as reduced HIV incidence, increased condom use, and increased health service utilization. However, some interventions fell short of their stated goals, especially those striving to impact upstream social and structural factors. Overall, results suggest that social ties can be intentionally altered to effect change; however, disparate contexts and implementation dynamics likely contributed to variation seen across outcomes and impact. Inconsistent measurement of social ties and use of theory made it challenging to determine whether interventions were explicitly trying to alter ties, and if so, to what extent tie building/strengthening impacted intervention effectiveness. To continue advancing our understanding of social tie interventions, more efforts are needed to operationalize theory, measure social tie constructs, describe intervention context and implementation outcomes, and apply innovative study designs.