Some early studies of the onset age–alcohol use disorder association (Grant and Dawson, 1997; Prescott and Kendler, 1999) employed retrospective reporting of onset age and symptomatology, raising the possibility that some of the association was because of response biases. Recall of onset ages is subject to “telescoping bias,” whereby older individuals recollect events as occurring closer in time (i.e., at later onset) than younger individuals (Johnson and Schultz, 2005). If the recall of alcohol-related problems is also subject to an interval-dependent reporting bias, it is possible that the early-onset problem-drinking association is spuriously inflated (i.e., by younger individuals reporting both more symptoms and earlier onset ages). However, this concern is reduced by the replication of the association between early onset and problematic use in several studies using prospective longitudinal designs (e.g., Buchmann et al., 2009; Pedersen and Skrondal, 1998; Pitkänen et al., 2005).
