Trajectories of Adolescent Alcohol Use in the Year Following a Brief Alcohol Intervention

Abstract

Objective: Brief interventions have become increasingly popular for youth who engage in problematic drinking behavior. The purpose of this study was to examine the alcohol use trajectories of adolescents over a 12-month period following the receipt of a brief intervention.

Method: The current sample came from two independent studies and consisted of 206 adolescents (ages 13-19; 52% male) recruited through an emergency department or community institution (e.g., courts, schools). Timeline followback methods were used at four points over 1 year to obtain daily estimates of alcohol use behaviors, with daily data then aggregated at the monthly level to examine trajectories of total drinks consumed and maximum drinks on one occasion. Using latent growth curve analysis, we expected a general pattern of increasing use over time, with lower use during the month immediately following completion of the intervention.

Results: Models with random intercepts, random linear slopes, and fixed quadratic trends provided good fit to the data for both total drinks and maximum drinks. For each outcome, there was an immediate decrease and then a gradual increase up to the 3- and 6-month assessments, with decreases seen in the months following assessments. Older age, White race, non-Hispanic ethnicity, and greater prior substance use were associated with greater initial levels of use and growth over time.

Conclusions: Interindividual differences were observed in alcohol use trajectories over time for high-risk adolescents following an alcohol use intervention. Subsequent research may demonstrate more uniform and permanent modification of trajectories by incorporating intervention-related materials into follow-up contacts.

Figure 1.

Mean total number of drinks consumed per month across the 12 months after baseline

Figure 2.

Mean maximum number of drinks consumed in 1 day across 12 months after baseline

Figure 3.

Latent growth curve analysis of total drinks per month. Variables were log + 1 transformed. Residual error variances at (a) Months 7 and 8 and (b) Months 9 and 10 were held constant. Because the variance of the quadratic trend was fixed at 0, any potential factor covariances with this parameter were also fixed at 0. Model fit: χ²(71) = 140.68, p < .001; comparative fit index = .94; root mean square error of approximation < .07. **p < .01; ***p < .001.

Figure 4.

Latent growth curve analysis of maximum daily drinks per month. Variables were log + 1 transformed. Residual error variances at (a) Months 5 and 6 and (b) Months 7 and 8 were held constant. Because the variance of the quadratic trend was fixed at 0, any potential factor covariances with this parameter were also fixed at 0. Model fit: χ²(71) = 107.90, p = .003; comparative fit index = .95; root mean square error of approximation = .05. Max. drinks = maximum number of drinks. *p < .05; **p < .01; ***p < .001.