Alcohol 35 (2005) 3–12 Theoretical article A physiologically based model for ethanol and acetaldehyde metabolism in human beings David M. Umulis 1 , Nihat M. Gu ¨rmen, Prashant Singh, H. Scott Fogler * University of Michigan, Department of Chemical Engineering, 2300 Hayward Street, Ann Arbor, MI 48109-2136, USA Received 19 August 2004; received in revised form 2 November 2004; accepted 7 November 2004 Abstract Pharmacokinetic models for ethanol metabolism have contributed to the understanding of ethanol clearance in human beings. However, these models fail to account for ethanol’s toxic metabolite, acetaldehyde. Acetaldehyde accumulation leads to signs and symptoms, such as cardiac arrhythmias, nausea, anxiety, and facial flushing. Nevertheless, it is difficult to determine the levels of acetaldehyde in the blood or other tissues because of artifactual formation and other technical issues. Therefore, we have constructed a promising physiologically based pharmacokinetic (PBPK) model, which is an excellent match for existing ethanol and acetaldehyde concentration–time data. The model consists of five compartments that exchange material: stomach, gastrointestinal tract, liver, central fluid, and muscle. All compartments except the liver are modeled as stirred reactors. The liver is modeled as a tubular flow reactor. We derived average enzymatic rate laws for alcohol dehydrogenase (ADH) and acetaldehyde dehydrogenase (ALDH), determined kinetic parameters from the literature, and found best-fit parameters by minimizing the squared error between our profiles and the experimental data. The model’s transient output correlates strongly with the experimentally observed results for healthy individuals and for those with reduced ALDH activity caused by a genetic deficiency of the primary acetaldehyde-metabolizing enzyme ALDH2. Furthermore, the model shows that the reverse reaction of acetaldehyde back into ethanol is essential and keeps acetaldehyde levels approximately 10-fold lower than if the reaction were irreversible.  2005 Elsevier Inc. All rights reserved. Keywords: Alcohol metabolism; Acetaldehyde dehydrogenase (ALDH); ALDH deficiency; Physiologically based pharmacokinetic (PBPK) model; Alcohol dehydrogenase (ADH); Michaelis–Menten kinetics 1. Introduction Pharmacokinetic models for in vivo ethanol elimination have evolved significantly during the past 70 years, from the inception of a pseudo zero-order elimination process (Widmark, 1932) to the current physiologically based models such as those developed by Derr (1993), Levitt (2002), and Norberg (2001). Although the models continually improve in their ability to predict time trajectories for ethanol concentration, they fail to account for the production and interaction of ethanol’s major metabolite, acetaldehyde. Acetaldehyde is highly toxic, with a 50% lethal dose (LD 50 ) concentration approximately 10 times lower than that for ethanol in rats (Brien & Loomis, 1983). Acetaldehyde expo- sure leads to a number of well-known signs and symptoms, * Corresponding author. Tel.: +1-734-763-1361; fax: +1-734-763-0459. E-mail address: sfogler@umich.edu (H.S. Fogler). 1 Present address: University of Minnesota, Department of Chemical Engineering and Materials Science, 421 Washington Avenue SE, Minneapolis, MN 55455, USA. Accepting Editor: T.R. Jerrells 0741-8329/05/$ – see front matter  2005 Elsevier Inc. All rights reserved. doi: 10.1016/j.alcohol.2004.11.004 such as cardiac arrhythmias, nausea, anxiety, and facial flushing (Condouris & Havelin, 1987; Peng et al., 1999; Yamamoto et al., 2000). In this article, we present a physiologically based model with reversible enzyme kinetics that accurately predicts si- multaneously the concentrations of both ethanol and acetal- dehyde in the blood as a function of time. 2. Methods 2.1. Rate law derivation The rate law for ethanol metabolism is based on the alcohol dehydrogenase (ADH) reaction pathway because it is the largest contributor to ethanol oxidation. The first assumption is that the concentration of the oxi- dized form of nicotinamide adenine dinucleotide (NAD + ) reaches its rate-limiting state shortly after ingestion and remains constant. Ethanol elimination is approximately zeroth order, supporting the suggestion that the reaction is limited by the amount of enzyme, co-substrate, or both. The enzymatic reaction, accounting for the NAD + co-substrate, is