Soy Isoflavones and Bone Mineral Density in Perimenopausal and Postmenopausal Western Women: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Background: Several randomized controlled trials (RCTs) have compared the effect of phytoestrogens (PEs) vs. placebos on bone density after menopause, with inconsistent results. Methods: We performed a systematic review to assess the overall effect of PEs on bone mineral density (BMD) in menopausal Western women. We searched for all RCTs comparing PEs with placebos conducted on perimenopausal or postmenopausal Western women, published from January 1990 to February 2010. The main outcome measure was the lumbar spine (LS) BMD. Results: We identified 17 studies on soy isoflavone (IFs) bone-sparing effects. Some studies did not report a difference between treated and untreated women, whereas others supported a significant role of IFs on slowing bone loss, although these studies suffered from an internal lack of consistency, as a positive effect emerged in some bone districts but not in others. Data on LS BMD were available in 12 studies including 1433 subjects overall. The effect of PEs on BMD (mg/cm(2)) was not statistically significant (mean difference 9.86 mg/cm(2), 95% confidence interval [CI] -2.64-22.36) under a random-effects model. Excluding the genistein study, however, analyses of IF mixtures did not show a bone-sparing effect (0.73, 95% CI -2.79-4.25). No increasing effect emerged when dose and treatment duration were increased. Conclusions: Our review and meta-analysis suggest that IF mixtures are not effective in decreasing bone loss in perimenopausal and postmenopausal Western women. The role of isolated genistein and individual genetic capacity to metabolize IFs is still open to evaluation.