AIMS: To compare beverage-specific per capita consumption and total alcohol consumption's associations with cirrhosis mortality rates in multiple countries. DESIGN: Pooled cross-sectional time-series analysis. SETTING: Australia, Canada, New Zealand, the United Kingdom and the United States during the years 1953-1993. MEASUREMENTS: National level data on per capita total alcohol, beer, wine and spirits consumption and standardized all-cause cirrhosis mortality rates. FINDINGS: Significant associations with cirrhosis mortality are found for both total ethanol and spirits. Spirits consumption is found to make up the majority of the effect of alcoholic beverage consumption on cirrhosis mortality and the model including only spirits is found to fit the data at least as well as the model including only total ethanol consumption. The lag relationship between all alcohol types and cirrhosis is found to be short with only present and 1 year's lagged consumption having significant associations. CONCLUSIONS: Spirits consumption rather than beer or wine is associated with cirrhosis mortality in this group of primarily beer-drinking countries. This finding offers important clues to understanding the drinking behaviors associated with cirrhosis mortality on the individual level.
Kerr et al. (Wed,) studied this question.