The Effects of Social Network Characteristics on Suicidal Thoughts Among Refugees in the Southwest U.S

Background and Purpose: Refugees face complex and severe mental health challenges, often intensified by trauma and violence across the migration journey. In particular, suicide has become a critical public health concern in this population, with suicidal ideation reported as high as 70.6%. However, the factors contributing to suicide risk among refugees remain poorly understood. In the general population, social support networks play a key role in promoting mental wellness and reducing suicide risk. Psychosocial interventions that enhance support networks have proven effective, yet most research relies on aggregated measures of social support, overlooking the structural and functional dimensions of social networks. To address this gap, this study used social network data to examine how network characteristics are associated with suicidal thoughts among refugees resettled in Arizona.

Methods: Egocentric network data were collected (July-September 2022) from 150 refugees resettled in Phoenix, Arizona. A community-based participatory research approach was used, with community leaders from each group serving as active partners. Surveys were self-administered at ethnic-based community organizations or in participants’ homes, in both English and native languages. Social networks were assessed through egocentric network analysis, capturing variables such as network size, relationship types, quality, residence location of members, interaction frequency, and relational homophily. Suicidal thoughts were measured using the Wish to be Dead Scale. Multivariate logistic regression was used to examine associations between social network variables and suicidal thoughts. Due to the modest sample size, network variables were grouped by type and analyzed in three separate models, each controlling for the same covariates, including sociodemographic characteristics, social support, trauma, and discrimination.

Results: Among respondents, 26.0% reported experiencing suicidal thoughts. The majority were male (53.0%) with an average age of 41.4 years, and had lived in the U.S. for an average of 8.45 years. Larger network size was significantly associated with higher odds of suicidal thoughts (OR=2.61, p<.001). In contrast, having a child (OR=0.01, p<.05) or other family members (OR=0.01, p=.05) in one’s network was linked to significantly lower odds. Refugees with more emotionally supportive network members were also less likely to report suicidal thoughts (OR=0.15, p<.001). Notably, general perceived social support was not associated with suicidal thoughts in any model.

Conclusions and Implications: To our knowledge, this is the first study to use egocentric network data to examine social networks and suicidal thoughts among refugees in Arizona. Findings highlight the importance of analyzing detailed network characteristics rather than relying solely on general social support measures, which were not associated with suicidal thoughts. Notably, larger network size was linked to higher odds of suicidal thoughts, suggesting that simply expanding one’s network may not be protective and could even be harmful if it includes superficial or stressful ties. These findings underscore the need for psychosocial interventions that prioritize fostering meaningful, emotionally supportive relationships over increasing network size. The protective effect of emotional support reinforces that the quality—not just the quantity—of social connections is key in refugee suicide prevention efforts.