In addition, we did not gather information on the reason for divorce, which might be perceived as a positive life event (e.g., in the case of an abusive parent or parent with a substance use disorder). Our measure of perceived stress may not have preceded the first drink for earlier onsetters; when possible, it is important to assess proximal factors that are closely spaced in time to a given event. Although we selected youth for whom divorce did not precede drinking onset, it is still possible that the negative effects of the divorce process were already present in the years leading up to the divorce (Arkes, 2013).
