Age, period and cohort effects on alcohol-related risky behaviours in Australia from 2001 to 2016.

Abstract

The aim of this study is to examine age, period and birth cohort trends in the prevalence of any alcohol-related risky behaviour and to compare these trends between men and women. We used an age-period-cohort analysis of repeated cross-sectional survey data from the Australian National Drug Strategy Household Survey from 2001 to 2016. Participants were 121 281 people aged 14-80 years who reported consuming alcohol in the past 12 months. Any risky behaviour undertaken while under the influence of alcohol in the past 12 months (e.g. operating a motor vehicle) was measured: male or female. Controlling for age and cohort, cubic spline models showed that any alcohol-related risky behaviour declined with time among participants who consumed alcohol [2016 versus 2007 rate ratio (RR) = 0.80, 95% confidence interval (CI) = 0.76-0.84]. Risky behaviour peaked in the 1954 birth cohort (1954 versus 1971 RR = 1.42, 95% CI = 1.30-1.55) and then steadily declined with more recent birth cohorts (2002 versus 1971 RR = 0.32, 95% CI = 0.27-0.39). Risky behaviour peaked at age 21 years, followed by steady decline and stabilization at approximately age 70 years. Males were overall twice as likely as females to report alcohol-related risky behaviour (RR = 2.10, 95% CI = 1.84-2.39), but this effect was smaller in cohorts born after 1980 [1980 prevalence rate ratios (PRR) = 2.09, 95% CI = 1.81-2.43; 2002 PRR = 1.31, 95% CI = 1.03-1.68]. Alcohol-related risky behaviour in Australia has declined generally since 2001, with rates for recent cohorts having the sharpest decline. Risky behaviour remains most prevalent in young adults, and the male-female gap in risky behaviour is closing for more recent birth cohorts. These trends are consistent with alcohol consumption trends observed in Australia and world-wide.