Commentary on Rossow et al. (2014) and Norström & Svensson (2014): we want to believe-or do we have to?

Declines in alcohol consumption in Australia: some challenges to the theory of collectivity

BACKGROUND AND AIMS

There is significant debate about whether or not changes in per-capita alcohol consumption occur collectively across the entire distribution of drinking. This study used data from a decade of declining drinking in Australia to test the collectivity of drinking trends.

DESIGN

Repeated cross-sectional surveys (2010, 2013, 2016, 2019), analysed with quantile regression techniques assessing trends in drinking for 20 quantile groups.

SETTING

Australia.

PARTICIPANTS

A general population sample (total n = 85 891; males = 39 182, females = 46 709) aged 14 years and over.

MEASUREMENTS

Past-year volume of alcohol consumed was measured using standard graduated frequency survey questions. Models were stratified by sex and age group.

FINDINGS

Throughout the whole population, alcohol consumption had declined in all percentile groups, with the largest proportional declines evident for light and moderate drinkers [e.g. drinkers in the 25th percentile declined by 32.7%; 95% confidence interval (CI) = -41.6, -22.3% per wave]. Broadly collective declines were also found for younger men and women with significant declines in every percentile group, but older groups showed some evidence of polarization. For example, women aged 45-64 years significantly increased their consumption (2.9% per wave, 95% CI = 0.3-5.5%), while consumption for those in the 25th percentile fell significantly (-16.7%, 95% CI = -27.6, -4.2%).

CONCLUSIONS

The declines in Australian drinking since 2010 have included important deviations from the collectivity predicted by Skog's influential theory of collectivity of drinking, with markedly different patterns evident among different demographic groups.