Full Title: Effects of Soy on Health Outcomes
August 2005
View or download Summary/Report
Structured Abstract
Objectives: Systematic review to describe the range of soy products and outcomes that have been studied, to summarize the effects of soy consumption to prevent a wide variety of medical conditions in healthy adults, and to summarize adverse events related to soy consumption.
Data Sources: Investigators searched MEDLINE®, EMBASE®, and the Commonwealth Agricultural Bureau (CAB) databases. Additional studies were identified in bibliographies of selected reviews and by technical experts.
Review Methods: English-language, prospective studies of soy products in adults of at least 4 weeks' duration were included. Researchers reviewed outcomes related to cardiovascular health, menopausal symptoms and reproductive health, endocrine function, tumor-related biomarkers, bone health, neurocognitive health, kidney function, and glucose metabolism. Soy supplements (including soy milk) were used in about three quarters of all the trials analyzed in this report, with soy foods used in the remainder of the trials. Most used soy protein with isoflavones, one-third used isoflavones alone, and a few used soy protein without isoflavones. Textured soy protein and soy flour were the most common soy foods investigated.
Results: Meta-analysis indicates that consumption of soy products appears to exert a small benefit on low density lipoprotein (LDL) and no significant effect on high density lipoprotein (HDL). Across studies, there is the possible suggestion that higher doses of soy protein are associated with greater LDL reduction among those with elevated baseline LDL, but not with HDL or triglycerides. Dose of isoflavones was not associated with effect for any lipid. Higher baseline LDL or triglycerides may also be associated with net effect for these 2 lipids; the effect of baseline HDL is unclear. In individual studies, the effect of dose and baseline was generally inconsistent. Meta-analysis of blood pressure (BP) found no effect of soy consumption. No association was found between baseline BP, soy protein or isoflavone dose, and effect on BP. No significant effect of soy products was found for several markers of inflammation, vascular function, or lipid oxidation.
Although the effect of soy products on menopausal symptoms are inconsistent across studies, the evidence of a benefit was stronger from the randomized trials of soy isoflavone supplements, but not of other soy products among post-menopausal women. This effect was not seen in the few studies among peri-menopausal women or those treated for breast cancer. Soy products do not appear to affect menstrual cycle length or estradiol level in pre-menopausal women, thyroid stimulating hormone, bone markers, or glucose metabolism.
Small numbers of studies or inconsistency among studies precluded drawing conclusions regarding the effect of soy protein on other endocrine markers and other evaluated outcomes. For all outcomes, no soy protein or isoflavone dose-effect response or soy product type difference in effect was apparent across studies.
Conclusions: Soy products appear to exert a small benefit on LDL and triglycerides; these effects may be of small clinical effect in individuals, although possibly large enough to have a population-wide effect. The inconsistent association between soy protein dose and effect, and the lack of association between soy isoflavone dose and effect, limit possible determination of an appropriate amount of soy product needed for lipid reduction. Soy products may reduce menopausal symptoms in post-menopausal women. The current literature does not support other effects of soy products. However, other than menopausal- and menstrual-related symptoms, no clinical outcomes were evaluated. The evidence from human studies does not suggest any worrisome adverse events beyond mild gastrointestinal intolerance. Conclusions were often limited due to small numbers of studies or heterogeneity across studies.
Given the large amount of heterogeneity and inadequate reporting, particularly related to soy protein and isoflavone dose, many questions remain as to whether specific soy products in adequate doses may be of benefit in specific populations. Further, well-conducted studies are needed to clarify the effect of soy dose on lipid parameters and to determine whether soy components other than protein or isoflavones may be responsible for the lipid effects seen.
Download Report
Effects of Soy on Health Outcomes
Evidence-based Practice Center: Tufts-New England Medical Center
Topic Nominators: National Center for Complementary and Alternative Medicine and the Office of Dietary Supplements, National Institute of Health (NIH)
Current as of August 2005
Internet Citation:
Effects of Soy on Health Outcomes, Structured Abstract. August 2005. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/clinic/tp/soytp.htm
540 Gaither Road Rockville, MD 20850