cannot rule out reverse causality, and it is also possible that bi-directional relationships exist between these variables. For example, AUD persistence and/or internalizing disorder may reduce social support, lead people to perceive more stigma, and/or heighten sensitivity to the potential effects of stigma. Scant research in the broader stigma literature exists that has utilized prospective designs, which is a significant methodological limitation of this literature (Hatzenbuehler, 2009; Livingston and Boyd, 2010).
