Epidemiology-based risk assessment using the benchmark dose/margin of exposure approach: the example of ethanol and liver cirrhosis

Abstract

A novel approach to derive a threshold dose with respect to alcohol-related harm, the benchmark dose (BMD) methodology, is introduced to provide a basis for evidence-based drinking guidelines. This study is the first to calculate a BMD for alcohol exposure using epidemiological cohort data. With this BMD we will be able to calculate the margin of exposure (MOE) for alcohol consumption, which can be used for comparative risk assessment and applied to setting public health policy. Benchmark dose-response modelling of epidemiological data gathered during a recent systematic review and meta-analysis of alcohol consumption as a risk factor for liver cirrhosis morbidity and mortality. For a benchmark response (BMR) of 1.5%, the resulting BMD values were 30.9 g/day for males and 29.7 g/day for females; the corresponding lower one-sided confidence values were 25.7 and 27.2 g/day, respectively. The intake scenario for the Canadian population resulted in an MOE of 1.23. Intake scenarios for individuals as based on the Canadian drinking guidelines led to MOE values between 0.96 and 1.91. Using an uncertainty factor of 10, the acceptable daily intake for alcohol would be 2.6 g/day. The BMD approach was feasible in developing evidence-based guidelines for low-risk drinking. As our calculated MOEs result around unity (i.e. 1) for moderate drinking, it is evident that the current guidelines correspond very well to low risk on the dose-response curve. The BMD methodology therefore validates current guidelines. The results again highlight the health risk associated with alcohol consumption.