Estimating alcohol-attributable liver disease mortality: A comparison of methods

Area of Science:

• Public Health
• Epidemiology
• Alcohol-Related Diseases

Background:

• Alcohol consumption is a significant cause of liver disease globally.
• Accurately quantifying alcohol-attributable liver disease deaths (AALDD) presents methodological challenges.

Purpose of the Study:

• To compare different methodologies for estimating AALDD in the USA.
• To identify the most accurate and comprehensive approach for AALDD estimation.

Main Methods:

• Compared a direct method with two population attributable fraction (PAF) methods.
• Utilized data on alcohol consumption prevalence, per capita sales, and relative risks from meta-analyses.
• Applied PAFs to national mortality data for alcoholic and unspecified cirrhosis.

Main Results:

• The direct method yielded higher AALDD estimates than PAF methods.
• PAF methods adjusted for per capita sales produced similar AALDD estimates across different consumption levels.
• Disaggregating non-drinkers to include former drinkers in PAF models increased AALDD estimates.

Conclusions:

• Population attributable fraction (PAF) methods are recommended for AALDD estimation.
• Adjusting for per capita alcohol sales and incorporating risks for former drinkers enhance AALDD estimate validity.
• These refined PAF approaches offer more comprehensive insights into alcohol's impact on liver disease mortality.