OBJECTIVE
This study examined issues of heterogeneity in multiple stage development as it corresponds to qualitatively different developmental trajectories in alcohol use during adolescence.
METHOD
Using a piecewise growth mixture modeling methodology, a two-piece growth model capturing growth trajectories in adolescent alcohol use from middle school (Grades 6 through 8) to high school (Grades 9 through 12) was examined (N = 179; 54% male). It was hypothesized that (1) two stages of alcohol use development with varying trajectories would exist in these data (the first corresponding to development during middle school, followed by a second stage of continuing growth during high school) and (2) there would be multiple growth trajectories (subgroups) of alcohol use in the stage-wise development, with varying effects in initial alcohol use and growth rates of alcohol use.
RESULTS
Results indicated the tenability of the two-piece growth model of alcohol use with heterogeneity in the population comprising two distinct latent developmental trajectory classes. Class 1, with a high initial status of alcohol use at Grade 6, showed an upward increase in trajectory only during high school. Class 2, with a low initial status of alcohol use at Grade 6, showed a linear increase in middle school with a second growth spurt at high school entry and continuity in growth throughout the high school years. Analyses, incorporating time-invariate covariates, indicated varying influences of gender, early levels of deviant behavior, family structure (single vs two parent), peer encouragement and parent disapproval of alcohol use, and adolescent deviant behavior upon high school entry, on the two trajectory classes. Results also showed effects of the identified trajectories, with varying magnitudes, on later substance use in young adulthood, with Class 1 showing the strongest continuity in later substance use.
CONCLUSIONS
Findings suggest heterogenous development of alcohol use in the adolescent population, associated with varying background and covariate influences. In addition, this heterogeneity is linked to alcohol and other substance use in young adulthood.