Energy adjustment methods applied to alcohol analyses

When alcohol consumption is related to outcome, associations between alcohol type and health outcomes may occur simply because of the ethanol in the beverage type. When one analyzes the consequences of consumption of beer, wine, and spirits, the total alcohol intake must therefore be taken into account. However, owing to the linear dependency between total alcohol intake and the alcohol content of each beverage type, the effects cannot be separated from each other or from the effect of ethanol. In nutritional epidemiology, similar problems regarding intake of macronutrients and total energy intake have been addressed, and four methods have been proposed to solve the problem: energy partition, standard, density, and residual. The aim of this study was to evaluate the usefulness of the energy adjustment methods in alcohol analyses by using coronary heart disease as an example. Data obtained from the Copenhagen City Heart Study were used. The standard and energy partition methods yielded similar results for continuous, and almost similar results for categorical, alcohol variables. The results from the density method differed, but nevertheless were concordant with these. Beer and wine drinkers, in comparison with findings for nondrinkers, had lower risk of coronary heart disease. Except for the case of men drinking beer, the effect seemed to be associated with drinking one drink per week. The standard method derives influence of substituting alcohol types at constant total alcohol intake and complements the estimates of adding consumption of a particular alcohol type to the total intake. For most diseases, the effect of ethanol predominates over that of substances in the beverage type, which makes the density method less relevant in alcohol analyses. (C) 2003 Elsevier Inc. All rights reserved.