Layers of stigma : structural, institutional, relational, and psychological vulnerability factors associated with internalized stigma among people receiving injectable opioid agonist treatment

Opioid use disorder (OUD) is among the most heavily stigmatized health conditions in North America. Among people receiving injectable opioid agonist treatment (iOAT), an evidence-based option for severe OUD, internalized stigma is associated with reduced treatment engagement and poorer health. Less is known about how structural vulnerabilities, healthcare experiences, provider relationships, and psychological vulnerability relate to stigma when examined together. This study examined discrimination-related stigma among current and former iOAT clients in British Columbia, guided by the Health Stigma and Discrimination Framework and Modified Labeling Theory. Cross-sectional baseline data were drawn from the PORTIA cohort (N = 139, recruited across seven Lower Mainland clinic sites). Discrimination-related stigma was measured using the OSTQOL-38 Discrimination subscale (range 0-4). Variables were entered sequentially across six theory-based blocks, from sociodemographic and structural factors to psychological vulnerability, using hierarchical proportional odds ordinal regression with multiple imputation pooled across 20 datasets. The sample had a mean age of 51.75 years (SD 10.94); 24.5% were women or gender-diverse and 21.6% Indigenous. The mean Discrimination score was 1.95 (SD 1.24). Financial insecurity, harm reduction service access, negative healthcare experiences, and infrequent shared decision-making were initially associated with higher scores. These associations no longer met the conventional threshold for significance after post-traumatic stress disorder (PTSD) symptom severity was added, although the modest sample and wide confidence intervals limited precision. PTSD severity was the only factor meeting the conventional threshold for significance in the final model (OR = 1.06 per PCL-5 unit, 95% CI 1.04 to 1.09), about 79% higher odds per 10-unit increase, and remained similar after social connectedness was added. In this exploratory analysis, PTSD symptom severity was the strongest concurrent correlate and the only factor meeting the conventional threshold for significance in the fully adjusted model. This does not indicate that the earlier associations are unimportant; the modest sample limited the precision of their adjusted estimates. Given the cross-sectional design, the findings do not establish causal relationships. They highlight the potential relevance of trauma-informed care, relationally responsive practice, and attention to the structural conditions shaping clients' lives in iOAT services.