Our results should be interpreted within the context of the following limitations. First, COGA is a high-risk sample with most participants from extended families enriched for AUDs. Findings may not be generalizable to other populations or samples ascertained with different risk profiles. Second, we used lifetime diagnoses of AUD and psychiatric comorbidities in view of incomplete age of onset information for these conditions within the COGA sample. As such, our analyses do not provide insights regarding the direction of the effects. Future work incorporating timing of diagnoses would be valuable. Third, our analyses included only participants’ AUD and psychiatric comorbidities information. Partner effects may play a role as well. For example, there is evidence that a spouse’s AUD is associated with their partner’s as well as their own marital adjustment (Cranford et al., 2011). Spousal discordance in their alcohol use patterns is linked with greater marital dissatisfaction (Homish & Leonard, 2007) and greater likelihood of divorce (Ostermann et al., 2005). Thus, incorporating dyadic data is an important next step in future research. Fourth, we focused on legal marriage, and first marriage/divorce for those with multiple marriages. Future research can consider marriage-like cohabiting relationships as well as remarriages.
