Unique Syndemic-Related Clusters in a Sample of Sexual Minority Men

HIV remains an urgent health concern and a significant disparity for sexual minority men (SMM). According to syndemic theory, social marginalization negatively impacts the psychosocial health profile of SMM, which increases their HIV risk. However, few studies have explored whether there are meaningful sub-groups based on unique patterns of individuals’ syndemic stressors. Identifying latent clusters could inform on which syndemic stressors to treat, which targeted treatments to employ, and for whom. We recruited 160 HIV-negative SMM via online sources targeting SMM populations. Participants completed online questionnaires on demographics, syndemic stressors, HIV-related resilience resources (behaviors, cognition/emotions, relationships, & SES ), and HIV risk. Syndemic stressors included depression, anxiety, stress, suicidality, alcohol and drug abuse, partner abuse (physical, psychological, sexual), and childhood sexual abuse. We conducted 10 separate multivariate linear regression analyses, one with each syndemic stressor as a dependent variable (DV) with all other variables as predictors, to examine the interrelations among the syndemic stressors. We performed a hierarchical cluster analysis across syndemic stressors to examine potential latent clusters. Lastly, we examined associations between cluster membership and resilience resources, as well as HIV risk, to explore potential clinical implications of these clusters. Results demonstrated different patterns of unique variance depending on the designated DV, with nine of the 10 models demonstrating significance. Cluster analysis results revealed that a 3-cluster solution best characterized our sample: alcohol abuse cluster (n = 51), sexual abuse victimization cluster (n = 35), and mental health cluster (n = 74). ANOVAs demonstrated significant mean differences of cognitions/emotions based on cluster membership, F(2,148) = 8.63, p \u3c .001, with a significant difference between Cluster 3 (mental health; M = 53.86, SD = 7.38) and Cluster 1 (alcohol abuse; M = 59.02, SD = 5.81), p \u3c .001, and a marginally significant difference between Cluster 2 (sexual abuse victimization; M = 57.18, SD = 6.76), p = .054. Cluster membership was not associated with other resilience resources nor HIV risk. Our results suggest that there exist latent subgroups that present with different syndemic-related treatment needs and can help advise treatment planning to reflect the unique needs of our sample