particularly well suited to data imputation methods, which are most accurate when large amounts of auxiliary data are available for the prediction model, as is the case in ALSPAC. However, continued participation in ALSPAC is known to be correlated with child sex and ethnicity (females and whites are more likely to continue participation), as well as household income, with low-income individuals being less likely to have sustained participation (Boyd et al., 2013). Finally, we note that rates of alcohol use are higher among adolescents in the United Kingdom as compared to the United States (Hibell et al., 2011; Johnston et al., 2011). By the age 15.5 assessment, virtually all of the ALSPAC participants (99.5%) report that they have had a full drink of alcohol, with 77% reporting at least 1 episode of binge drinking and 69% reporting that they have at least 1 time thought that they should cut back on their drinking. However, a recent report of substance use among 15- to 16-year-old individuals from 36 European countries found that alcohol use in the United Kingdom was only slightly above average compared with other European countries (Hibell et al., 2011).
