College students' drinking patterns: trajectories of AUDIT scores during the first four years at university.

Aims: Changes in AUDIT score trajectories were examined in a student population during their first 4 years at a university, including high-risk consumers and a subsample of low-risk consumers.

Method: 359 students were selected for the present study, comprising all high-risk consumers (the 27% with highest scores, i.e. 11 for males and 7 for females) and a randomized sample of low-risk consumers (n = 177 and 182, respectively). The Alcohol Use Disorder Identification Test (AUDIT) was used as screening instrument. Trajectory analyses were made using a semiparametric group-based model.

Results: In the low-AUDIT group, five distinct trajectories were identified: three stable non-risky consumption groups (83%) and two increasing groups (17%; from non-risky to risky). In the high-AUDIT group, three groups were identified: two stable high groups (58%) and one decreasing group (from risky to non-risky consumption; 41%). In the integrated model, stable risky consumption comprised 16% of the total sample, decreasing consumption 11%, increasing consumption comprised 13% and stable non-risky consumption 60% of the sample. Gender influenced the trajectories.

Conclusion: The pattern of changes in risk consumption is similar to that found in corresponding US studies.