There are a number of limitations to our study suggesting cautious interpretation of findings, including incomplete assessment of major domains of risk. For example, we do not know why parents separated, and reason for relationship dissolution may have important implications. While post-hoc tests of interactions were largely nonsignificant, if substance dependence was the primary cause of separation and removal from the home of the alcoholic or other drug-addicted parent worked to reduce offspring risk, our understanding of parental separation as a risk-factor would need reconsidering. Data on prior separations is also limited, leaving open the possibility that some parents may have separated before final dissolution, and that parents in families coded as intact previously separated and since reconciled.
