Purpose Syndemic theory posits that social and structural inequities enable health conditions to cluster and interact, worsening outcomes. Despite widespread use in HIV research, most syndemic studies rely on additive indices without rigorously testing factor structure or interconnections. This study compares exploratory factor analysis (EFA) and network analysis to examine the structure of syndemic conditions — including depression, anxiety, internalized homophobia, sexual orientation stigma, racial discrimination, polydrug use, childhood sexual abuse, and physical abuse — influencing PrEP uptake among Latino men who have sex with men (LMSM) in Miami-Dade County, Florida.Methods Secondary data from 130 LMSM recruited through community organizations were analyzed. Continuous syndemic indicators (sexual orientation stigma, internalized homonegativity, discrimination stress) were retained in their native scale and modeled using a mixed graphical model (MGM) framework (Haslbeck & Waldorp, 2018), which natively accommodates mixed variable types without requiring dichotomization. Binary and count variables were modeled accordingly. The original median-split EBICglasso approach was retained as a sensitivity analysis. EFA with oblique rotation was conducted to assess latent structure. Bootstrap stability of node centrality was quantified via case-drop CS-coefficients (Epskamp et al., 2018).Results Both methods revealed convergent syndemic patterns. The primary MGM analysis identified three retained edges in the current syndemic model: depression–anxiety (edge weight = 4.53), depression–sexual abuse (1.63), and physical abuse–sexual abuse (0.64). In the expanded model incorporating discrimination, a fourth edge emerged between discrimination and internalized homonegativity (0.20), consistent with minority stress theory. Depression and anxiety were the most strongly connected nodes across both models, and findings were robust in the median-split sensitivity analysis. EFA grouped depression, anxiety, sexual orientation stigma, and discrimination on a shared latent dimension. The expanded model demonstrated improved sampling adequacy (KMO = 0.80) and model fit (lower BIC).Conclusions Depression and anxiety function as central syndemic indicators among LMSM. Interventions targeting these mental health conditions may reduce broader syndemic burden and improve PrEP uptake in this population.