TO THE EDITOR: We would like to thank Dr. Hirsch for the interesting suggestion that alcoholics' preference for stronger sucrose solutions may result from chemosensory adjustment of the olfactory system to excessive alcohol intake. Although we did not have the opportunity to review this issue in our article, we agree that it is an important consideration.

The main finding of our pilot study was that a majority of alcoholic subjects have a preference for stronger sucrose concentrations, significantly more than what we found in nonalcoholic comparison subjects. We analyzed several mechanisms that may underlie this phenomenon, including the possibility that it may be a "consequence of heavy drinking that has altered taste sensitivity." Although this factor, including possible alterations in smell and taste, needs to be further evaluated, we cited evidence in support of the hypothesis that sweet liking in alcoholism may represent an underlying neurobiological trait. First, animal experiments have indicated an association between preference for concentrated sweets and alcohol use in rats known to have a genetically determined propensity to prefer or reject alcohol before exposure to it (1). Second, 35% of our alcoholic subjects showed a preference for lower sucrose concentration, but there was no difference in the pattern or duration of alcohol intake between "sweet-liking" and "sweet-disliking" alcoholics. Finally, 16% of our nonalcoholic subjects had a preference for the most concentrated sucrose solution despite their minimal alcohol consumption. This last observation is in agreement with the results from work in healthy subjects, which indicated that sweet liking or sweet disliking are psychophysical traits that are relatively stable over time (2–4).

Nevertheless, we agree with Dr. Hirsch that further studies are needed to fully explain the mechanism of our observed association between alcoholism and a preference for stronger sweet solutions and that investigations of olfactory and taste sensitivity are important. One strategy that we are currently using to address the potentially confounding effects of heavy drinking is to study subjects at high risk or low risk for alcoholism.