To determine the relative strengths of association of parental AUDs and parental separation, we used post-hoc Wald tests to compare the largest hazard associated with parental separation (i.e., the younger age categories) with (1) the hazard associated with one parent with an AUD and (2) the hazard associated with 2 AUD-affected parents for the youngest age category, for all outcomes. Parental separation had a stronger association with first use of alcohol than did having one parent with an AUD (HRs=2.16 versus 1.28, Wald χ2(1)=10.7, p<.01) but did not differ from the hazard associated with 2 affected parents (HRs=2.16 versus 1.59, χ2(1)=3.1, p=08). This pattern was also observed for tobacco and cannabis, where parental separation had a significantly stronger association with initiation risk than did having one AUD-affected parent, and had risk equivalent to that of having 2 AUD-affected parents (tobacco: HRs=1.89 versus 1.22, χ(1)=13.0, p<.001 and HRs= 1.89 and 1.81, χ(1)=.08; cannabis: HRs=2.04 versus 1.33, χ(1)=16.02, p<.,001 and HRs=2.04 versus 2.73, χ(1)=2.0, p=.16). For sexual debut, parental separation was associated with greater hazard than was having one or two AUD-affected parents (HRs=2.34 versus 1.19, χ(1)=44.6, p<.0001 and versus 1.43, χ(1)=19.3, p<.0001).
