Uneven reductions in high school students' alcohol use from 2007 to 2012 by age, sex, and socioeconomic strata.

BACKGROUND: Many Western countries have reported declines in adolescent alcohol use. This study examined changes in adolescent alcohol use in New Zealand between 2007 and 2012 and explored variations across sociodemographic strata.

METHODS: Data from 2 nationally representative, cross-sectional high school surveys conducted in 2007 (n = 7709) and 2012 (n = 7266) were examined. Changes in the prevalence of drinking in the past 4 weeks were examined among the total sample, as well as the frequency of drinking in the past 4 weeks and typical drinking-occasion quantity among drinkers. Only students residing in urban areas were included. Variation in changes was investigated across 4 demographic groups characterized by age (<16 years, ≥16 years) and sex. Interactions with household- and neighborhood-level socioeconomic position (SEP) identified any differential changes between socioeconomic strata.

RESULTS: From 2007 to 2012, significantly fewer students consumed alcohol in the past 4 weeks. Interaction analyses demonstrated that, among young females (<16 years), declines were significantly greater among those of high household SEP when compared with those of low household SEP. Among drinkers, reductions in the frequency of drinking were found among all demographic groups and SEP strata. Interaction analyses revealed that only young males (<16 years) showed significantly reduced typical drinking-occasion quantities. Among young females, significant interactions revealed a shift towards increasing typical drinking-occasion quantities among those of low household and neighborhood SEP, whereas their more advantaged counterparts showed no significant change over time.

CONCLUSIONS: Fewer drinking occasions characterized the major declines in adolescent drinking between 2007 and 2012. Whereas young males showed reductions in the typical quantity consumed, young females of low household and neighborhood SEP progressed towards higher typical quantities. To address the uneven distribution of alcohol-related harm and improve the targeting of harm reduction initiatives, it remains imperative to examine changes in both the overall shift and shape of the distribution curve.