Effect of soy protein on testosterone levels.

To the Editors: Goodin et al. (1) reported a 19% decrease in mean serum testosterone levels among 12 men over a 4-week period in response to the daily consumption of 56 g isolated soy protein. These data are inconsistent with the results of a meta-analysis recently completed by the authors of this letter. This inconsistency, as judged by Fig. 1 in the paper by Goodin et al. (1), may stem largely from the change that occurred in just one subject (SP #2). In this regard, we would ask Goodin et al. (1) to consider reanalyzing the results (or providing the data so that we may do so) excluding this subject, who clearly appears to be an outlier. His baseline testosterone level was more than twice that of any other subject and f50% greater than the normal reference range as stated in the paper. Additionally, this subject experienced an f40% decrease in testosterone within just 4 weeks and a further 30% or so decrease during the 2-week post-soy consumption period, whereas the mean value for all other subjects increased during this 2-week period. Parenthetically, we are interested in knowing whether Goodin et al. (1) are able to offer a possible explanation for such an elevated baseline level. In addition, for the record, we would ask that Goodin et al. (1) list the correct values for testosterone and luteinizing hormone levels, as the data listed in the text do not match the data in the figures. As well, it would be helpful if Goodin et al. (1) could provide additional details about the study so we have a better understanding of the experimental design. Specifically, the isoflavone content of the soy product used would be of interest, as would details about the assays used.