Biases inherent in all-cause mortality studies: implications for shaping the 2025–2030 dietary guidelines for Americans on alcohol consumption

In a commentary, Kevin Shields, Timothy Niami and colleagues argue that the net impact of alcohol on health at low doses depends on the underlying risks of diseases and injuries causally linked to alcohol and thus varies across countries. As such, informing the public about the health risks of alcohol consumption requires country-specific information. The Review of Evidence on Alcohol and Health by the National Academies of Sciences, Engineering, and Medicine (NASEM) aims to inform the 2025–2030 Dietary Guidelines for Americans. The report compares the impact of consuming ≤1 drink/day (≤14 g of ethanol) for women and ≤2 drinks/day (≤28 g of ethanol) for men on all-cause mortality, relative to never consuming alcohol. However, the ability of this analysis to draw conclusions is limited by several factors.
The commentary argues that NASEM all-cause mortality analysis provides little insight how the health of Americans is affected by alcohol consumption at or near the 2020–2025 Dietary Guidelines upper threshold of 2 drinks/day for men and 1 drink/day for women. Moreover, due to the inherent limitations of all-cause mortality analysis, such analyses should not be used to guide alcohol and health recommendations. Instead, cause-specific approaches that model the health impact of alcohol consumption based on specific causes of death should be prioritised.
Source: Shield K, Keyes K, Martinez P, Milam AJ, Rehm J, Naimi TS. Biases inherent in all-cause mortality studies: implications for shaping the 2025-2030 dietary guidelines for Americans on alcohol consumption. Lancet Reg Health Am. 2025 Apr 4;46:101083.