Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: Dr Hollman and colleagues raise concerns about the lignan values used for our analyses. The lignan metabolites, enterolactone and enterodiol, are formed from the precursors matairesinol and secoisolariciresinol. Since the metabolites are not found in plant foods but are actually derived from enzymatic conversion of the lignan precursors, we opted to not only sum the lignans (ie, to estimate “total lignans”) but to analyze the precursors from the metabolites separately. There are limited published food nutrient data on lignan metabolites because these values are obtained through a metabolic in vitro laboratory assay. In our study, the median value of intake for the metabolites was only approximately 350 μg/d, compared with the median value of intake for precursors of about 5.4 mg/d (ie, approximately 5400 μg/d). Thus, including lignan metabolites in our assessment of “total lignans” likely had a negligible effect. Moreover, if we overestimated intake, this should not have differentially occurred for the cases and controls. This type of misclassification would tend to move the odds ratio toward 1, so that if it had occurred, our measured odds ratios would have actually underestimated the effect size.
Schabath MB, Spitz MR. Phytoestrogens and Risk of Lung Cancer—Reply. JAMA. 2006;295(7):755-756. doi:10.1001/jama.295.7.755-b