Perceived alcohol stigma was not assessed in the W1 NESARC survey, and thus we are unable to determine if stigma was associated with participation in the W2 survey. Because stigma may be greater among non-respondents than among respondents, differential response to W2 based on perceived alcohol stigma would be likely to bias findings of this study conservatively (i.e., result in underestimations of stigma in the study). These cross-sectional data lacked temporal ordering for perceived alcohol stigma and psychiatric disorders. Scant longitudinal data on perceived stigma significantly limits the field’s knowledge about causal relationships and mediating processes (Livingston and Boyd 2010). Although we used DSM-5 AUD for our index condition, DSM-IV diagnoses were used for the comorbid psychiatric conditions. Our identification of differences in perceived alcohol stigma across psychiatric comorbidity subtypes is important in its own right, but future studies may wish to explore whether the outcomes of perceiving more stigma are worse across comorbidity profiles. Last, past-year psychiatric diagnoses were assessed, whereas it may be useful for future studies to examine the relationship between perceived stigma and psychopathology at the time of the interview.
