H. Rose; S. Coulton; M. Lynskey; C. Drummond
Funders: Society for the Study of Addiction
Alcohol Research UK
Latent class analysis (LCA) was applied to indicators of beverage choice to examine whether qualitatively distinct subgroups of adolescents could be identified on the basis beverage preference. The relationship between beverage choice latent classes and outcomes of harm including, the Alcohol Use Disorders Identification Test (AUDIT) were also examined. This research aims to identify distinctive drinking patterns and harms associated in adolescent alcohol users in addition to the more traditional quantity-frequency measures.
Adolescents attending Emergency Departments between 10 and 17 years who had consumed alcohol in the past 3 months were drawn from the SIPS Junior database (n=1351). Dichotomous indicators of whether they had drunk beer, cider, alcopops, wine or spirits in the past 3 months were included in LCA, conducted for males and females separately.
LCA identified 3 latent classes for both males and females. Male latent classes demonstrated similar beverage preferences, with beer being the most prevalent drink type. Spirits were the second most prevalent drink type consumed in both male class 1 (14.0%) and class 3 (12.7%), while cider was the second most prevalent in class 2 (19.2%). Wine and alcopops were the least prevalent beverage type in all male classes. Class 1 females were the most prevalent class (56.1%) characterized by a greater prevalence of spirit consumption (26.6%) and alcopops (26.1%). Females in class 2 (17.6%) represented the smallest female class characterised by a preference for spirits (12.5%) and cider (11.3%), while class 3 (26.2%) had a greater prevalence of wine drinkers (16.2%). Multinomial regressions indicated varying levels of risk associated with beverage choice latent classes, including different rates of hazardous alcohol use indicated by the AUDIT.
There is a need to identify distinctive alcohol drinking patterns or profiles in adolescence that considers multivariate indicators in addition to the more traditional quantity-frequency measures. LCA identified subgroups of adolescents characterised by distinct patterns of beverage choice. Latent class membership was associated with varying risks for alcohol related harm.