RESUMO
BACKGROUND: We sought to apply modified labeling theory in a cross-sectional study of alcohol use disorder (AUD) to investigate the mechanisms through which perceived alcohol stigma (PAS) may lead to the persistence of AUD and risk of psychiatric disorder. METHODS: We conducted structural equation modeling (SEM) including moderated mediation analyses of two waves (W1 and W2) of data from the National Epidemiologic Survey on Alcohol and Related Conditions. We analyzed validated measures of PAS, perceived social support, social network involvement, and psychiatric disorders among (n=3608) adults with two or more DSM-5 AUD symptoms in the first two of the three years between the W1 and W2 survey. Cross-sectional analyses were conducted owing to the assessment of PAS only at W2. RESULTS: Per mediation analyses, lower levels of perceived social support explained the association of PAS with past-year AUD and past-year internalizing psychiatric disorder at W2. The size of the mediated relationship was significantly larger for those classified as labeled (i.e., alcoholic) per their prior alcohol treatment or perceived need (n=938) as compared to unlabeled (n=2634), confirming a hypothesis of moderated mediation. Unexpectedly, mediation was also present for unlabeled individuals. CONCLUSIONS: Lower levels of social support may be an important intermediate outcome of alcohol stigma. Longitudinal data are needed to establish the temporal precedence of PAS and its hypothesized intermediate and distal outcomes. Research is needed to evaluate direct measures of labeling that could replace proxy measures (e.g., prior treatment status) commonly employed in studies of the stigma of psychiatric disorders.