After the initial assessment, the participants are followed up at approximately biennial intervals. To avoid potential effects associated with population stratification, we limited our analyses here to the European American subset of the sample. Individuals who answered “No” to the screening question about lifetime drinking at all assessments were excluded from the analysis, as were individuals who had missing information on drunkenness at all assessments. We used data from ages 14 to 25 due to sparseness of data at earlier/later timepoints. This yielded a final N of 1070 individuals (49% male). Because data collection remains on-going, 545 individuals had 3 assessments (51%) at the time analyses were conducted, 378 had two assessments (35%), and 147 had one assessment. Individuals contribute to the mean and trajectory parameters that encompass the age period for which their data is available; this allows all individuals to contribute to the analyses.
