The aim of the study was to examine for Australia whether the link between population alcohol consumption and liver disease mortality varies over time, using 71 years of data.

Overall and gender-specific rates of liver disease mortality were analysed in relation to total alcohol consumption as well as for different beverage types by using autoregressive integrated moving average (ARIMA) time series methods. Separate models were developed for the entire time period and for two sub-periods (1935-1975, 1976-2006).

A 1-l increase in adult per capita consumption of pure alcohol led to a rise of ∼10% in overall liver disease mortality rates and a 11 and 9% increase in female and male liver disease mortality, respectively. The strength of the relationship between per capita consumption and liver disease mortality diminished over time. Spirits consumption was found to be the main driving factor in liver mortality rates between 1935 and 1975, while beer consumption was found to be the most significant predictor in liver diseases in the last three decades. In a comparative perspective, the effect of per capita alcohol consumption on liver disease in Australia is similar to the USA, Southern and Eastern Europe countries, but weaker than in Canada and western European countries.

An increase in per capita alcohol consumption in Australia is likely to lead to an increase in liver disease. Changes in the most important beverage over the study period suggest substantial shifts in drinking patterns and preferences among the heaviest Australian drinkers.