A minor coding error occurred in the analysis programs that generated results included in the article cited above. In Table 1, the proportions of postmenopausal hormone use should be 61.3% (rather than 37.0%) for case subjects and 63.3% (rather than 35.4%) for control subjects in the Nurses’ Health Study (NHS). In Supplementary Table 2, the proportions of postmenopausal hormone use should be 70.2%, 64.0%, 55.9%, and 62.7% (rather than 29.8%, 33.3%, 43.8%, and 34.8%) in the 1st, 2nd, 3rd, and 4th quartiles of lignan metabolites in the NHS, respectively.
The online version reflects these changes.