We would like to congratulate Levis and colleagues1 for a well-designed study and thoughtful analysis on the effect of soy isoflavone supplementation on bone loss and menopausal symptoms. However, there are some issues with this trial that warrant further analysis and explanation.
Through 2 years, subject attrition rates were 18.8% with soy isoflavone and 34.1% for placebo (P = .008). While the authors state that the baseline characteristics of the subjects who completed the trial were similar to those who withdrew, this explanation is inadequate. Were the reasons for study discontinuation available for comparison? It is unlikely that a discrepancy of this magnitude would occur by chance alone.