Figure 1. Chemical structures of genistein, diadzein, glycitein and estradiol
 |
|
|
The Agency for Healthcare Research and Quality (AHRQ), through its Evidence-based Practice Centers (EPCs), sponsors the development of evidence reports and technology assessments to assist public- and private-sector organizations in their efforts to improve the quality of health care in the United States. This report was requested and funded by the National Center for Complementary and Alternative Medicine (NCCAM) and the Office of Dietary Supplements, National Institutes of Health. The reports and assessments provide organizations with comprehensive, science-based information on common, costly medical conditions and new health care technologies. The EPCs systematically review the relevant scientific literature on topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to developing their reports and assessments.
To bring the broadest range of experts into the development of evidence reports and health technology assessments, AHRQ encourages the EPCs to form partnerships and enter into collaborations with other medical and research organizations. The EPCs work with these partner organizations to ensure that the evidence reports and technology assessments they produce will become building blocks for health care quality improvement projects throughout the Nation. The reports undergo peer review prior to their release.
AHRQ expects that the EPC evidence reports and technology assessments will inform individual health plans, providers, and purchasers as well as the health care system as a whole by providing important information to help improve health care quality.
We welcome comments on this evidence report. They may be sent by mail to the Task Order Officer named below at: Agency for Healthcare Research and Quality, 540 Gaither Road, Rockville, MD 20850, or by e-mail to epc@ahrq.gov.
Carolyn M. Clancy, M.D.
Director
Agency for Healthcare Research and Quality
Stephen E. Straus, M.D.
Director, National Center for Complementary and Alternative Medicine
National Institutes of Health
Paul M. Coates, Ph.D.
Director, Office of Dietary Supplements
National Institutes of Health
Jean Slutsky, P.A., M.S.P.H.
Director, Center for Outcomes and Evidence
Agency for Healthcare Research and Quality
Kenneth S. Fink, M.D., M.G.A., M.P.H.
Director, EPC Program
Agency for Healthcare Research and Quality
Margaret Coopey, R.N., M.G.A., M.P.S.
EPC Program Task Order Officer
Agency for Healthcare Research and Quality
We would like to acknowledge with appreciation the following members of the Technical Expert Panel for their advice and consultation to the Evidence-based Practice Center during preparation of this report.
J. Mark Cline, DVM, PhD, DACVP
Associate Professor of Pathology/Comparative Medicine
Comparative Medicine Clinical Research Center
Wake Forest University School of Medicine
Winston-Salem, NC
Alice H. Lichtenstein, DSc
Stanley N. Gershoff Professor of Nutrition Science and Policy
Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA
Cardiovascular Nutrition Laboratory, Jean Mayer USDA Human Nutrition Research Center on Aging, Boston, MA
Christopher Gardner, PhD
Assistant Professor of Medicine
Stanford Prevention Research Center
Stanford University School of Medicine
Stanford, CA
Mark Messina, PhD
Adjunct Associate Professor
Department of Nutrition
Loma Linda University
President, Nutrition Matters, Inc,
Port Townsend, WA
Jay R. Kaplan, PhD
Professor of Comparative Medicine and Anthropology
Wake Forest University School of Medicine
Medical Center Boulevard
Winston-Salem, NC
Heather Miller, PhD, MFS
Senior Advisor for Women's Health
National Center for Complementary and Alternative Medicine
National Institutes of Health
Bethesda, MD
Donna Kritz-Silverstein, PhD
Professor, Family and Preventive Medicine
University of California School of Medicine
University of California, San Diego
La Jolla, CA
Richard Nahin, PhD, MPH
Senior Advisor for Scientific Coordination
National Center for Complementary and Alternative Medicine
National Institutes of Health
Bethesda, MD
Anne L. Thurn, PhD
Director, Evidence-based Review Program
Office of Dietary Supplements
National Institutes of Health
Bethesda, MD
Context. Soy products, including both protein and isoflavones, have been touted for a number of clinical benefits related to a putative estrogen-like effect. However, potential risks of chronic soy consumption are also of concern.
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. We searched MEDLINE®, EMBASE, and the Commonwealth Agricultural Bureau (CAB) databases. Additional studies were identified in bibliographies of selected reviews and by technical experts.
Study Selection. English-language, prospective studies of soy products in adults, of at least 4 weeks' duration were included. We 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. Eligibility criteria were adjusted for several outcomes.
Data Extraction. Selected studies were extracted for study design, demographics, amount of soy product, soy protein, and isoflavones, control, outcomes. Based on these data, studies were graded for quality and applicability.
Data Synthesis. We screened almost 4,800 abstracts and retrieved 599 full text articles, of which 178 were eligible for review. 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. Among studies with soy protein, the range of soy protein consumed daily was 14 to 154 g, with a median of 36 g per day (equivalent to over a pound of tofu daily). Among studies with soy isoflavones, the range of isoflavones consumed daily was 10 to 185 mg, with a median of 80 mg per day. These ranges were the same for all lipid profile studies. There is a large degree of heterogeneity among the studies in terms of soy products evaluated, soy protein and isoflavone doses, study durations, background diet, controls used, and study design. No study evaluated clinical cardiovascular events. Meta-analysis indicates that consumption of soy products appears to exert a small benefit on low density lipoprotein (LDL) - the summary net change was -5 (95% confidence interval [CI] -8 to -3) mg/dL - and on triglycerides - net change -8 (95% CI -11, -5) mg/dL. No significant effect was seen on high density lipoprotein (HDL) - net change +0.6 (95% CI -0.5, +1.8). 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. The net effect on systolic BP was -1 (95% CI -3, +1) mm Hg, and on diastolic BP -1, (-2, +0) mm Hg. 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. The few studies that directly compared soy doses (generally isoflavone doses) for the most part also found no difference in effect based on dose. In general, soy products were well-tolerated, although study withdrawal due to aversion was more common in soy treatment arms than control arms. The most common adverse events reported were gastrointestinal or menstrual complaints although they were few in number.
Conclusions. A wide variety of soy products and formulations have been investigated for a large number of conditions. However, a large proportion of the studies suffer from poor reporting or study design, limiting 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.
Authors: Balk E, Chung M, Chew P, Ip S, Raman G, Kupelnick B, Tatsioni A, Sun Y, Wolk B, DeVine D, Lau J
The aims of this evidence report are to summarize the current evidence on the health effects of soy and its isoflavones on the following: cardiovascular diseases, menopausal symptoms, endocrine function, cancer, bone health, reproductive health, kidney diseases, cognitive function, and glucose metabolism. In addition, safety issues and drug interactions of using soy and its isoflavones, as reported in the literature, are summarized. This report also summarizes the formulations of soy products and/or soy food used in clinical trials. The report was requested and funded by the National Center for Complementary and Alternative Medicine (NCCAM) and the Office of Dietary Supplements at the National Institutes of Health (NIH) and was conducted through the Evidence-based Practice Center (EPC) program at the Agency for Healthcare Research and Quality (AHRQ).
There is increasing interest in soy and health since the U.S. Food and Drug Administration approved a health claim in October 1999 for use on food labels stating that a daily diet containing 25 grams of soy protein, also low in saturated fat and cholesterol, may reduce the risk of heart disease. This claim was based on the beneficial results in reducing plasma low-density lipoprotein (LDL) levels from dozens of human controlled clinical trials.1 The health claim, however, covers only soy protein, since research results surrounding soy isoflavones were controversial.2 This report summarizes the current evidence on the health effects of soy and its isoflavones.
Five general questions are addressed in this report:
In the clinical trial literature, what formulations of soy were used? At what dose? For what purpose(s) (e.g., trial endpoints)?
Does current clinical trial evidence indicate that whole soy products and individual constituents of soy have an effect on:
Cardiovascular events, risk factors, and measures;
Menopausal symptoms;
Endocrine function;
Cancer and tumor-related biomarkers;
Osteoporosis and osteoporosis risk factors;
Reproductive health;
Kidney function; and
Other outcomes, based on results of Key Question 1, above?
What is the scientific evidence of a dose-response effect of different forms of soy and individual constituents of soy for the conditions specified in Key Question 1?
What are the frequency and type(s) of adverse events associated with consumption of soy that are reported in the scientific literature (both trials and epidemiology)?
What is the scientific evidence of a dose-response effect of whole soy products and individual soy constituents on their safety?
This report encompasses several health conditions and many outcomes of interest. Therefore, specific inclusion criteria were needed for each of the health conditions and sometimes for different outcomes of the same health condition. The common inclusion criteria for studies analyzed in this report consist of: human subjects 13 years and older; prospective studies including randomized controlled trials, cohorts, crossover and non-randomized comparison studies; at least five subjects in the soy arm; any health condition; quantification of the amount of soy; and reported outcomes of interest. In general, the minimum duration for all serum marker, urine marker, and vascular outcome studies was 4 weeks (exceptions are noted below, under “Specific Inclusion Criteria for Health Conditions Examined”).
For assessments of adverse events, we also included prospective observation studies and case-control studies, with no limitations on study size or duration, or quantification of soy product.
In addition to the health conditions of interest listed under Key Question 3, the Technical Expert Panel (TEP) convened by the EPC suggested the category of neurocognitive outcomes. NCCAM was also interested in knowing about research that might have been done in other health conditions. Therefore, our literature search was conducted to broadly include soy studies for any health conditions. We screened all citations to identify health conditions not on the list agreed upon with the TEP. During our review process, we included the additional category of endocrine function.
We accepted studies that used soy supplements and foods that quantified the amount of soy ingredients or products. We categorized various soy products and soy food into the following groups:
Refined soy products
Isolated soy protein with isoflavones
Isolated soy protein without isoflavones
Textured soy protein
Soy-derived isoflavone
Genistein/genistin
Daidzein/daidzin
Glycitein/glycitin
Soy/soya food products (ingested amount must be quantified)
Whole soy beans (edamame)
Soy flour
Soy drink (soy milk)
Tofu (bean curd)
Miso
Other processed soy bean products (tempeh, natto, okara, etc.)
For the purpose of this report, all study arms with a soy product of any type were considered to be soy interventions. Only study arms with a non-soy intervention were categorized as controls.
In addition to the common inclusion criteria listed above, with input from TEP members we established the following additional criteria and specific outcomes for each of the specific health conditions.
Cardiovascular Outcomes: These included total cholesterol, LDL, high density lipoprotein (HDL), triglycerides, lipoprotein(a) [Lp(a)], blood pressure (BP), C-reactive protein (CRP), homocysteine, endothelial function, systemic arterial compliance, and oxidized LDL. We also sought studies of clinical cardiovascular outcomes (e.g., death, myocardial infarction, angina) but found none. The list of outcomes was determined in consultation with the TEP, based on expert opinion of the likelihood of an effect on the outcomes, clinical importance, and estimates of the numbers of studies likely to be available.
Because of the relatively large number of available studies reporting on lipids, triglycerides, and blood pressure, it was decided with the TEP to limit inclusion of these studies to randomized controlled trials with a minimum of 10 subjects consuming a soy product. For all cardiovascular outcomes, we required a minimum duration of 4 weeks.
Menopausal Symptoms: Studies evaluated peri-menopausal women, post-menopausal women, or women on breast cancer therapies with menopausal symptoms. A minimum duration of 4 weeks was required for studies of menopausal symptoms.
Endocrine Function: We included in our analyses the following endocrine markers: testosterone, follicle stimulating hormone (FSH), total estradiol and thyroid stimulating hormone (TSH). In addition, we evaluated menstrual cycle outcomes. The decisions for which outcomes to investigate were based on expert opinion of the likelihood of an effect on the outcomes, clinical importance, and estimates of the numbers of studies likely to be available. Studies that did not report numerical data on effect for these outcomes were not summarized; however, these studies were maintained in the database. For all endocrine outcomes, we required a minimum duration of 4 weeks (or one menstrual cycle).
Cancer and Tumor-Related Biomarkers: To evaluate whether soy may prevent cancer or reduce cancer risk factors, we included only studies that recruited subjects without a diagnosis of cancer. We limited our analyses to studies with tumor-related biomarkers or cancer risk factors as outcomes and to studies of clinical cancer outcomes (e.g., diagnosis of prostate cancer). We did not include studies that used soy products as “treatments” for cancer. The only outcome that fulfilled these criteria was testosterone. The studies that reported testosterone as an outcome in men without diagnoses of cancer were analyzed in the endocrine section. The decision to investigate only testosterone was based on expert opinion of the likelihood of an effect on the outcomes and of its clinical importance. For all tumor-related biomarkers, we broadened the eligibility criteria to include a minimum duration of 1 week.
Bone Endpoints: For bone resorption and/or formation biomarkers, the general inclusion criteria were used, including a minimum duration of 4 weeks. Because effects on bone mineral density occur slowly over time, we used minimum study duration of 1 year, although we did briefly review studies with a duration less than 1 year.
Miscellaneous Outcomes: For all other outcomes (neurocognitive, kidney, glucose metabolism), the general inclusion criteria were used in combination with the restriction to populations without the related specific diseases or conditions.
We conducted a comprehensive literature search to address the key questions.* Primary literature searches for English language publications on soy studies were conducted in EMBASE on March 25, 2004; in MEDLINE® on April 20, 2004; and in CAB Abstracts on June 24, 2004. Search terms included subject headings and textwords with filters to limit the publications to English language and primary studies of the adult and adolescent human populations. Subject headings and textwords were selected so that the same set could be applied to each of the different databases. A supplemental search was performed in MEDLINE on April 30, 2004, to retrieve articles using the textword “miso.” A search update was performed in MEDLINE In-Process & Other Non-Indexed Citations and MEDLINE on September 30, 2004, and in CAB Abstracts on October 4, 2004. A search of the TOXLINE® database was conducted in March 31, 2005, to identify additional reports of adverse events in humans. Additional sources of published and unpublished data were sought by contacting members of the TEP and from reference lists of selected review articles and meta-analyses.
We used a three-category grading system (A, B, C) to denote the methodological quality of each study. This system defines a generic grading system that is applicable to varying study designs, including randomized controlled trials, cohort, and case-control studies:
A: Least bias; results are valid; a study that mostly adheres to the commonly held concepts of high quality.
B: Susceptible to some bias but not sufficient to invalidate the results; a study that does not meet all the criteria in category A.
C: Significant bias that may invalidate the results; a study with serious errors in design, analysis, or reporting.
In this report, the focus is on the U.S. population and on specific subgroups within that population (i.e., post-menopausal women, peri-menopausal women, pre-menopausal women, men, and people with relevant medical histories such as breast cancer). Even though a study may focus on a specific target population, limited study size, eligibility criteria, and the patient recruitment process may result in a narrow population sample that is of limited applicability, even to the target population. To address this issue, we categorized studies within a target population into one of three levels of applicability, which are defined as follows: sample is representative of the target population; sample is representative of a relevant subgroup of the target population but not the entire population; sample is representative of a narrow subgroup of subjects only and is of limited applicability to other subgroups.
Meta-analysis was performed for several cardiovascular outcomes. We used the random effects model for continuous outcomes to combine studies. We also performed several random effects model meta-regression analyses to explore possible reasons for discrepancies across studies and to address Key Questions related to dose-response.
Soy supplements were used in about three-quarters of all the trials analyzed in this report; soy foods were used in the remaining trials. In this report, soy milk was categorized as a soy supplement. Among the soy supplement trials, 57 percent used soy protein with isoflavones, 36 percent used isoflavones alone, and 6 percent used soy protein without isoflavones. In about one-half of the soy foods trials, textured soy protein was used. Soy flour was used in about one-quarter of the soy foods trials. There are 146 separate treatment arms of soy supplementations and 68 separate treatment arms of soy foods or diets. Across studies, the total isoflavones ranged from 0 mg to 185 mg per day, and the total protein intake from soy ranged from 0 g to 154 g per day. It is notable that the median soy product dose across studies (36 g soy protein per day) was equivalent to over a pound of tofu daily or about 3 soy protein shakes daily.
No study evaluated clinical cardiovascular events. A total of 68 randomized studies reported data on total cholesterol, LDL, HDL, and/or triglycerides. The total isoflavones ranged from 0 mg to 185 mg per day, with a median of 80 mg. Among studies with soy protein, the total protein intake from soy ranged from 14 to 113 g per day, with a median of 36 g. There is a great deal of heterogeneity in the effects found on lipoprotein and triglyceride levels. Overall, the majority of studies reported small to moderate effects on the lipids, despite a wide range of net effects for total cholesterol, LDL, and triglycerides. Sixty-one studies reported data on the effect of consumption of soy products on total cholesterol levels. The median net change compared to control was approximately -5 (interquartile range -10, +1) mg/dL decrease (about -2.5 percent). A meta-analysis of 52 studies that reported data on the effect of soy consumption on LDL levels yielded a statistically significant net decrease of 5 (95-percent confidence interval [CI] -8 to -3) mg/dL (about -3 percent). A meta-analysis of 56 studies that reported data on the effect of soy consumption on HDL levels found a statistically nonsignificant net change of +0.6 (95-percent CI -0.5, +1.8) mg/dL. A meta-analysis combining 54 studies that reported data on the effect of soy consumption on triglyceride levels yielded a net change of -8 (95-percent CI -11, -5) mg/dL (about -6 percent). 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 (although not if studies with minimal soy protein doses are excluded) 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 two lipids; the effect of baseline HDL is unclear. For all lipids, in individual studies the effect of dose and baseline was generally inconsistent.
A total of 22 studies reported data on the effect of consumption of soy products on systolic and diastolic BP. Overall, across studies, there was no discernible effect.
Some of the well-known emerging risk factors for cardiovascular disease included for analysis in this report are: Lp(a), CRP, homocysteine, endothelial function, systemic arterial compliance, and oxidized LDL. The total numbers of studies that reported data on the effect of soy consumption are: 18 studies on Lp(a), 3 on CRP, 5 on homocysteine, 10 on endothelial function, 3 on systemic arterial compliance, and 13 on oxidized LDL. Across these studies, there is no discernible effect based on the type of soy products. The majority of studies were of poor quality with a narrow range of applicability. Given the limited evidence and poor quality of studies, no conclusions could be drawn on the beneficial or harmful effects of consumption of soy protein on these putative risk factors for cardiovascular disease.
A total of 21 trials examined the effects of soy and/or its isoflavones on hot flashes and night sweats in women. These trials generally measured frequency and severity of the symptoms. However, the investigators used a large number of vasomotor symptom scores or indexes that employed a variety of frequency intervals. These factors made meta-analyses unsuitable and limited the comparisons of results across studies. Furthermore, many of the studies had high withdrawal or dropout rates, which were frequently uneven between soy treatment and control arms, further limiting the validity of these trials. Overall, the effects of soy protein and/or its isoflavones are inconsistent across studies. Every trial found a decrease in hot flash frequencies or scores in both the treatment groups and the control groups. Thus, the results are difficult to interpret. A third of the studies found no or worsening effects compared to control; two-thirds showed that soy protein and/or its isoflavones either nonsignificantly or significantly decreased hot flash frequencies or scores compared to control in post-menopausal women. The evidence of a benefit was stronger among the randomized trials of isoflavone supplements, which mostly showed positive results—the net reduction in weekly hot flash frequency ranged from 7 percent to 40 percent. However, these trials are mostly rated as poor quality due to high dropout rates. Only four studies evaluated the effect of soy consumption on menopausal symptoms in peri-menopausal women or those receiving breast cancer therapy. Among these studies there is no evidence that soy consumption is better than control to reduce menopausal symptoms.
Measures of endocrine function from 50 trials were reported in 47 articles. Five studies with a total of 179 participants reported testosterone levels in healthy males before and after soy consumption. Four of these trials found a statistically nonsignificant decrease in testosterone levels. The small total number of subjects, as well as the low quality of these studies, precluded any meaningful conclusion. No statistically significant effect was found on FSH level, which is commonly measured in the initial evaluation of male and female infertility; results were conflicting.
Twelve studies reported estradiol levels at the follicular phase in 434 pre-menopausal women. The overall effect of soy on estradiol levels was not consistent. Most of the studies showed a trend for soy to reduce estradiol, although they failed to demonstrate a statistically significant effect. Six randomized trials reported the effect of soy on TSH. No overall effect of soy on TSH and thyroid function is clear.
An additional 11 trials (in 10 publications) evaluated the effect of soy on menstrual cycle length in pre-menopausal women. A wide range of soy interventions were used in these trials, making a conclusion on the effects from soy difficult. These trials did not show statistically significant changes in menstrual cycle length after treatments of soy and/or its isoflavones.
Twenty-four trials evaluated subjects without a history of cancer for effects of soy on tumor-related biomarkers. No study reported the development of cancer as an outcome. Most studies measured the effect of soy on estrogens and estrogen metabolites as well as on estrogenicity indicators. There were also trials that evaluated correlations between soy and possible cellular pathways of cancer prevention. No causal relationship could be established between these markers and cancer because they do not represent known risk factors for cancer disease. Only four studies reported on testosterone level, which is a risk factor for prostate cancer and is discussed above under “Endocrine Function.”
Overall, 31 studies evaluated various markers of bone health, including bone mineral density (BMD), bone formation biomarkers (bone specific alkaline phosphatase and osteocalcin) and bone resorption biomarkers (urinary hydroxyproline, urinary cross-linked N-telopeptide, urinary pyridinoline, and urinary deoxypyridinoline).
Because there are few long-term randomized trials and a wide variety of soy interventions used across studies, it is difficult to draw an overall conclusion about the effects of soy on bone outcomes. Overall, among the five studies of 1-year minimum duration, no consistent effect on BMD was seen with soy consumption. Studies of shorter duration likewise found no effect of soy. Similar to the results for BMD, studies of bone formation biomarkers generally found no effect of soy consumption when compared to control. While a number of studies reported reductions in two markers of bone resorption—urinary pyridinoline and deoxypyridinoline—no effects were found on the other markers of bone resorption, and the effects were not consistent across studies. For these markers, there is no clear evidence of a dose effect for either soy isoflavones or soy protein.
Only one study found a consistent effect on these markers. The study differed from other studies in that it evaluated a unique formulation of soy genistein and that it excluded subjects with denser femoral neck BMD.
Only one small study in patients with type 2 diabetes assessed the effect of soy on kidney function. No statistically significant change in glomerular filtration rate was seen after 8 weeks of soy protein diet. Four studies examined the effects of soy on cognitive function of post-menopausal women and college students of both sexes. Overall, no statistically significant or consistent effect was noted on neurocognitive functions such as verbal episodic memory. Six studies evaluated the effect of soy on fasting blood glucose. No statistically significant changes were reported.
In general, the rates of adverse events reported were greater in the soy treatment arms than in their respective control arms, but adverse events related to soy consumption were generally minor. Overall, soy products including isoflavones were well tolerated in the trials we examined.
The most frequently reported adverse events among a total of 3,518 subjects in 49 studies (including 5 nonrandomized and 3 pharmacokinetic studies) that reported adverse events were gastrointestinal in nature. These were reported in 33 of 41 comparison studies of soy diets, soy proteins, isoflavones, and phytoestrogen supplements. Most of the gastrointestinal adverse events were reported in soy diet and soy protein trials, especially the 12 studies that used purified isoflavone interventions in dosages ranging from 40 to 100 mg/day. The amount of soy protein in these trials ranged from 20 to 60 g/day, but there was no clear dose relationship between the amount consumed and subsequent adverse events. Menstrual complaints, reported in 15 studies, were also common. Six of these studies used purified isoflavone interventions in dosages ranging from 40 to 80 mg/day. However, most women in these studies were post-menopausal, and the controls frequently included hormone therapy regimens. Other adverse events included musculoskeletal complaints, headache, dizziness, and rashes. In addition, there were somewhat more withdrawals from the soy arms due to taste aversion.
Despite the large number of trials that have been performed, the health effects of soy for many conditions that have been studied remain uncertain. The methodological quality of over half the studies (about 55 percent) evaluated in this report was poor (Grade C). One-third of the poor-quality studies were either uncontrolled single-cohort studies, nonrandomized comparative studies, or comparative studies for which it was unclear whether they were randomized. Another third of the poor-quality studies had dropout rates that exceeded 20 percent or unequal dropout rates between the soy and control arms. Other reasons that studies were graded poor quality included lack of reporting of baseline data; inadequate accounting of important confounders; major discrepancies between text, tables, and/or figures or irreconcilable data that indicate likely improper statistical analysis; and substantial missing data.
There was also great heterogeneity among studies, particularly among the interventions analyzed. Comparisons across the myriad types of soy are intrinsically very difficult. This difficulty was compounded by the use of soy both as a supplement and as an integral part of the diet; furthermore, for numerous studies, it is difficult to distinguish between supplement and diet. It is likely that studies of supplements and diet are not easily comparable. Most studies involved a small number of study subjects and were of short duration. About one-half of studies were of less than 12 weeks duration and about one-third were shorter than 6 weeks. Few studies directly compared soy products, mostly comparing soy protein with varying amounts of soy isoflavones. Only one performed a factorial design study comparing both present and absent soy protein and present and absent soy isoflavones, thus allowing analysis of the effect of both soy protein and soy product. The universal issue of possible publication bias, where negative studies are less likely to be published and are more likely to be published later, is a potential concern. However, for most outcomes, the majority of studies reported negative outcomes, and there was no obvious evidence of publication bias among the lipid studies (where there is evidence of a positive effect).
Most of the studies evaluated the effects of soy on various biomarkers or measures, not clinical outcomes, although several of the endpoints, such as blood pressure, LDL, and bone mineral density, do have known meaningful correlations with clinical outcomes. Cardiovascular surrogate endpoints were assessed by the largest number of studies. Overall, soy was found to have a small effect on lipids. However, the duration of these studies was generally short, and it is uncertain whether the results would be sustained. No study evaluated clinical cardiovascular disease.
Reduction of hot flashes by soy was seen in trials involving post-menopausal and peri-menopausal women. Most of the trials lasted only 3 to 4 months; thus the long-term benefits remain unclear. In addition, different measurements were used to assess benefits across studies, making comparisons and synthesis difficult. Soy phytoestrogens are seen by some as an alternative to estrogen therapy to treat post-menopausal symptoms. However, the estrogenic effect of soy in potentially promoting tumor recurrence raises concern for its use by breast cancer survivors. The current literature provides no data to address this issue.
The evidence does not support an effect of soy products on endocrine function, menstrual cycle length, or bone health, although evidence was often limited and of poor quality. No study evaluated clinical endocrine or bone disease.
This report was limited to human studies, and thus was unable to fully respond to biological or biochemical hypotheses of benefits or harms of phytoestrogens suggested by various animal, in vitro, or assay detection studies: the correlations between specific nutrients and their effects remain unclear. While the evidence does suggest a greater likelihood of adverse events with soy consumption, these were mostly minor in nature. There were a limited number of studies with duration of 1 year or longer; thus the long-term adverse effect of soy in a large population is uncertain.
For all outcomes, including adverse events, there is no conclusive evidence of a dose-response effect for either soy protein or isoflavone. However, for LDL reduction, there is a suggestion of a possible dose-response effect for soy protein.
This report dealt with a broad range of health conditions and endpoints; thus it is difficult to focus research recommendations on a specific area. As is the case with most bodies of evidence regarding medical fields, better quality, well-reported, larger, and longer duration studies are needed to address the questions of interest. Future studies should fully report the components of soy products being tested; compare different doses, soy products, and populations; more closely evaluate the effects of different soy components, including non-protein, non-isoflavone components; fully consider the types of foods being replaced by soy products and the controls being used; and use the CONSORT statement as a guide to designing and reporting studies.3, 4
Conducting clinical trials in the area of health effects of food substances is fraught with difficulties. There is a complex interplay among the various components and potentially active substances within the foods and with other foods. Dietary variations, as well as other lifestyle and clinical variations among individuals, are also complex. Controlling for these factors is difficult within a trial. Interpreting discrepant results among trials is even more difficult. Isoflavones are believed to be the key active substance in soy, but this is by no means certain. Little data suggest that the amount of soy isoflavones is associated with an incremental effect, and studies of soy protein with little or no isoflavones frequently had similar effects as isoflavone studies. Difficulties with attempting to ascribe a food health benefit to a specific component of the food are highlighted by the recent spate of disappointing results from antioxidant trials, which suggest that the evaluation of potential nutrient benefits may need a paradigm different from the traditional clinical trial model.
The bioavailability of an ingested nutrient may also be an important factor in the determination of the beneficial effect. Several factors may affect the bioavailability of ingested nutrients: (1) absorption rate, which is affected by the interactions with competitive nutrients, the usual diet compositions, and types of foods or supplements; (2) incorporation rate into the blood stream, in which complex mechanisms might be involved, such as the functions of facilitated transporters, receptors on the membrane, or cellular binding proteins; (3) metabolism of the intestinal bacterial environment. Any one of these factors alone does not determine the bioavailability. In order to gain insights on the question of dose-response relationship, we need information not only on the soy isoflavone contents, including types and amount, but also on the bioavailability of the ingested soy isoflavones.
Unfortunately, studies that attempt to control for the myriad factors that interfere with clear interpretation of the effect of food products such as soy tend to be highly artificial, with little applicability to the average person. Clarity is needed to define what study questions are of interest. Metabolic laboratory studies or investigations of highly structured or restricted diets (such as those where soy protein constitutes the bulk of daily protein consumption) are of potential value only to possibly determine which components of soy are bioactive or to determine what extremes of diet may be necessary to achieve a benefit. Studies that substitute practical amounts of soy products into average people's diets would better address the question of whether people should make the effort to include more soy in their diets, but these studies will invariably be difficult to interpret. An exception to this may be studies of soy isoflavone supplements (e.g., nonfood capsules), which may be interpreted more like usual drug trials.
Carefully controlled efficacy studies (those conducted under the artificial conditions of a clinical trial) may still be useful to pin down the relative effects of various components of soy. Once this is better clarified, more practical effectiveness studies that aim to test the value of an intervention in more real-world scenarios with feasible interventions might be more important.
The full evidence report from which this summary was taken was prepared for the Agency for Healthcare Research and Quality (AHRQ) by the Tufts-New England Medical Center Evidence-based Practice Center under Contract No. 290-02-0022. It is expected to be available in August 2005. At that time, printed copies may be obtained free of charge from the AHRQ Publications Clearinghouse by calling 800-358-9295. Requesters should ask for Evidence Report/Technology Assessment No. 126, Effects of Soy on Health Outcomes. In addition, Internet users will be able to access the report and this summary online through AHRQ's Web site at www.ahrq.gov.
Balk E, Chung M, Chew P, Ip S, Raman G, Kupelnick B, Tatsioni A, Sun Y, Wolk B, DeVine D, Lau J. Effects of Soy on Health Outcomes. Summary, Evidence Report/Technology Assessment No. 126. (Prepared by the Tufts-New England Medical Center Evidence-based Practice Center under Contract No. 290-02-0022.) AHRQ Publication No. 05-E024-1. Rockville, MD: Agency for Healthcare Research and Quality. July 2005.
This evidence report has been prepared by the Tufts-New England Medical Center (Tufts-NEMC) Evidence-based Practice Center (EPC) concerning the effect of soy consumption on various diseases and conditions, including but not restricted to cardiovascular, kidney and gastrointestinal diseases, cancer, osteoporosis, menopausal symptoms, and reproductive health. This report was requested and funded by the National Center for Complementary and Alternative Medicine (NCCAM) and the Office of Dietary Supplements of the National Institutes of Health (NIH), through the EPC program at the Agency for Healthcare Research and Quality (AHRQ).
There is increasing interest in soy and health since the U.S. Food and Drug Administration (FDA) approved a health claim in October 1999 for use on food labels stating that a daily diet containing 25 grams of soy protein, also low in saturated fat and cholesterol, may reduce the risk of heart disease. This claim was based on the beneficial results in reducing plasma low-density lipoprotein (LDL) levels from dozens of human controlled clinical trials.1 The health claim, however, covers only soy protein, since research results surrounding soy isoflavones were controversial.2 The aims of this report are to summarize the formulations of soy products and/or soy food used in clinical trials, and to reflect the current evidence on the health effects of soy and its isoflavones on the following: cardiovascular disease (CVD), menopausal symptoms, endocrine function, cancer, bone health, reproductive health, kidney disease, cognitive function, and glucose metabolism. In addition, safety issues and drug interactions of using soy and its isoflavones as reported in the literature are summarized.
Soy is a legume commonly consumed by many Asians, although most Americans do not regularly eat it. Despite the increase in the consumption of soy products since the approval of the soy health claim in 1999, only a small number of Americans who believe that soy consumption is “healthy” consume soy products at least once a week, according to a survey by the United Soybean Board.2 Soybean and soy foods, such as tofu, tempeh, soy drinks, soy flours, and meat alternatives or analogs, are rich in soy protein and soy isoflavones. The isoflavones are among the phenolic compounds produced by soy, and a class of phytoestrogens, or plant-derived estrogens. Soy sauce and soybean oil are soy-derived foods, but lack substantial amounts of either soy protein or soy isoflavones.3
Soy isoflavones are one of the families of phytoestrogens that are similar in chemical structure to estrogen and thus may have similar effects (Figure 1
). The most prominent
isoflavones in soy are daidzein, genistein, glycitein, and their glucosides.
Most isoflavones occur in plants in the bound form of their glucosides, daidzin,
genistin and glycitin, and are biologically inactive. The glucoside forms of
isoflavones can be converted to free forms (aglycones) after consumption by
hydrolysis at the intestinal brush border membrane and by intestinal bacteria.
The efficiency of this conversion affects bioavailability, and subsequent
metabolism. Different food sources of soy isoflavones can also affect their
bioavailibilities. One aglycone, equol,
(7-hydroxy-3-(4'-hydroxyphenyl)-chroman), is an exclusive intestinal bacterial
metabolite of dietary isoflavones.4,
5 Equol has superior antioxidant potencies against destructive compounds,
known as free radicals1, than the parent isoflavones.6 However, many adults do not produce equol. This phenomenon has led to
the terminology of being an “equol-producer” or “non-equol producer” and may
explain the differential clinical effectiveness among individuals of isoflavones
from soy proteins.7 There are many other factors that might influence the health effects of
soy and its isoflavones, such as the relevant window of exposure (early in life,
peri-menopausal or post-menopausal) or the frequency of exposure.
The aglycone form of soy isoflavones and their metabolites can be absorbed by the gut and may exert several biological effects. There is a growing interest in the roles of these natural estrogenic or antioxidant substances in influencing disease progression or medical conditions in humans. Research suggests several mechanisms of action of soy isoflavones.
Estrogens produced in the ovaries and testes stimulate growth, blood flow, and water retention in the sexual organs and are also associated with the development of breast and endometrial cancers.8 Because of the structural similarity of soy isoflavones to estrogens, it has been suggested that these isoflavones might act as either an agonist or antagonist of estrogen. For example, studies suggest that genistein can induce responses similar to estradiol in breast, ovarian, endometrial, prostate, vascular and bone tissues, and in cell lines.9–12 Genistein can also act as an estrogen antagonist in some tissues. Animal studies have shown that genistein inhibits the development of chemically-induced mammary cancer.13, 14
Genistein was first identified in 1987 as an inhibitor of protein-tyrosine kinase, an enzyme that promotes cancer cell growth.15 Other cancer-related enzymes were also found to be inhibited by genistein, including DNA topoisomerases I and II16 and ribosomal S6 kinase.17 Possible mechanisms for the anti-proliferative properties of genistein include prevention of cell mutations by stabilization of cell DNA and reduction of cell oxidants, reduction in capacity of malignant cells to metastasize by inhibiting angiogenesis and subsequent tumor growth, as well as inducing cell differentiation.18
Isoflavones' antioxidant properties may enhance the resistance of LDL to oxidation and prevent free radical damage to DNA.19 Studies show that genistein has greater antioxidant activities than other isoflavones.6, 17 Genistein may also increase the production of antioxidant enzymes, such as superoxide dismutase.20
New concepts are still evolving about possible mechanisms of action of soy isoflavones. For example, some research suggests the potential health effects of soy isoflavones might occur via the transforming growth factor β signaling pathway.21
Although direct evidence is lacking, it has been suggested that soy and its isoflavones affect various diseases or conditions, such as CVD, cancers, osteoporosis, menopausal symptoms, and kidney disease, through different mechanisms of action described above.
Heart disease is, in part, a hormone-dependent disease, as illustrated by the lower age-related incidence in pre-menopausal women compared with males and a rise in incidence among both natural and surgical post-menopausal women.18 Although estrogen replacement therapy does not reduce long-term CVD risk in post-menopausal women,22 soy isoflavones may reduce CVD risk by several mechanisms. One possible pathway for the protective effects of soy isoflavones on CVD may be manifested through blood lipid changes.1 There has also been a suggestion that soy isoflavones may have regulatory effects on blood pressure (BP).23
The reported incidence of hot flashes varies among menopausal women in different countries. For example, it occurs in 70 to 80 percent of menopausal women in Europe, 57 percent in Malaysia, and 18 and 14 percent in China and Singapore, respectively. The estrogenic effects of soy isoflavones may be responsible for modifying the incidence rates, given that substantial dietary differences in soy consumption exist among these populations.18 Soy isoflavones may regulate menopausal symptoms by maintaining normal vascular function in both vasomotor tone and vessel wall compliance.24, 25 However, it is not clear which components of soy may be responsible for any putative effects.
Several observational studies have evaluated the effects of isoflavones on the risk of developing hormone-dependent cancers, such as breast, colon, prostate, endometrial, and ovarian. Much of the evidence is based on the differences in the consumption of soy products in different areas of the world.18 Case-control studies examining the association between soy-containing food intake and breast cancer risk suggest a marginally significant inverse relationship.26, 27 However, no significant association between soy-based foods and breast cancer risk was found in a large prospective cohort study in Japan or other prospective studies.26
A recent meta-analysis of stomach cancer studies conducted among Asians showed that non-fermented soy foods significantly reduced the risk of stomach cancer while fermented soy foods had no effect.28 But because various confounders such as dietary salt, fruit, and vegetable intake, were not adjusted for in the original studies, the role of soy could not be adequately assessed. A case-control study and the preliminary finding from an ongoing clinical trial suggest an inverse relationship between soy-based food and colon cancer risk.25, 29
The pathophysiological mechanism for involuntary bone loss in post-menopausal women can be explained by the deficiency of estrogen due to a rapid decline in ovarian function. It involves loss of both cancellous and cortical bone and continues throughout the remainder of life. It is caused by the loss of estrogen effects on extra-skeletal calcium homeostasis, leading to decreased intestinal calcium absorption, increased renal calcium wasting, and, perhaps also effects on vitamin D metabolism and loss of a direct effect on the parathyroid gland that decreases parathyroid hormone (PTH) secretion.30 These factors put post-menopausal women at an increased risk of developing osteoporosis and fracture.
Conventional therapies for treating osteoporosis in women, such as estrogen treatment alone or combination estrogen and progesterone, function as inhibitors of bone resorption. However, the effectiveness and safety of hormone replacement therapies in post-menopausal women remain controversial. The potential use of soy and its isoflavones to preserve bone tissue and delay or prevent the onset of osteoporosis has recently been addressed.9 Benefits from soy protein for bone health may also be related to changes in dietary protein intake, and calcium excretion. Since the animal protein-rich diet is associated with the highest excretion of undissociated uric acid and calcium, it increases the risk of kidney stones and osteoporosis.31 Substituting less hypercalciuric soy protein for animal protein may benefit calcium balance and bone health.
Low protein diets can halt or attenuate the progression of chronic kidney disease, and modifications in the quality of dietary protein can also affect the course of kidney disease. In animal models of kidney disease, studies have shown that soy protein diets limit or reduce proteinuria and kidney lesions associated with progressive kidney failure.32 However, it is not clear whether the kidney protective effects of soy protein diets are due to soy isoflavones or other soy components. The effect of soy protein intake on kidney function in humans has not been examined comprehensively. In a review of studies of people with different types of chronic kidney disease, soy protein moderated proteinuria and preserved kidney function, although most of the trials reviewed were of relatively short duration and involved small numbers of patients.32 The applicability of these findings to prevent the development of chronic kidney disease in people with normal kidney function is uncertain.
Reproductive health is defined by the World Health Organization (WHO) as a state of physical, mental, and social well-being in all matters relating to the reproductive system at all stages of life. This definition of reproductive health implies that people are able to have a satisfying and safe sex life and that they have the capability to reproduce and the freedom to decide if, when, and how often to do so (http://www.who.int/topics/reproductive_health/en). While reproductive health covers a wide range of issues including contraception safety and efficacy, fertility, infertility, and HIV/AIDS and other sexually transmitted diseases the focus of this report is on fertility. The putative estrogenic effect of soy may have a direct impact on endocrine function, which in turn may affect fertility.
Studies have also reported on the use of soy-based foods combined with hypocaloric diet in the treatment of obesity, but no significant effect of soy was found when compared to the control treatment.33, 34 Soy-rich foods have also been used to prevent malnutrition in patients with Crohn's disease, a chronic inflammatory disorder of the bowel. A study reported that patients with inactive Crohn's disease had a higher treatment compliance while on a soy-rich and lactose-free diet than on a conventional enteral diet, although body weight and lean muscle mass were increased equally in both groups of patients.35 A population-based cohort study in China showed an inverse relationship between intake of soy products and the risk of glycosuria in postmenopausal women, but not in pre-menopausal women.36 Estrogen loss associated with menopause may contribute to the development of Alzheimer's Disease.37 However, a cohort study of Japanese-American men reported that higher midlife consumption of soy in the form of tofu was associated with indicators of cognitive impairment and brain atrophy in late life.38
Although soy has been evaluated for treatment or prevention of a variety of diseases, this report - which evaluates the effect of soy on generally healthy people and not on disease treatment - evaluates only a limited number of other conditions, based primarily on the availability of data. These include endocrine function, cognitive function, and glucose metabolism.
This evidence report on the health effects of soy is based on a systematic review of the literature. The Tufts-New England Medical Center Evidence-based Practice Center (Tufts-NEMC EPC) held meetings and teleconferences with a technical expert panel (TEP) to identify specific issues central to this report. The TEP was comprised of technical experts in soy research and in relevant health areas of interest. A comprehensive search of the medical literature was conducted to identify studies addressing the key questions. Evidence tables of study characteristics and results were compiled, and the methodological quality and the applicability of studies were appraised. Study results were summarized with both qualitative and quantitative reviews of the evidence, summary tables, and meta-analyses, as appropriate.
A number of individuals and groups supported the Tufts-NEMC EPC in preparing this report. The TEP served as our science partner. It included technical experts, representatives from the Agency for Healthcare Research and Quality (AHRQ) and from the National Center for Complementary and Alternative Medicine (NCCAM) and the Office of Dietary Supplements at the National Institutes of Health (NIH). The TEP worked with the EPC staff to refine key questions, identify important issues, and define parameters for the report. Additional clinical domain expertise was obtained through local experts who joined the EPC. A draft version of this report was critically appraised by a panel of peer reviewers.* Revisions were made based on their comments; although all statements within the report are those of the authors only.
In the clinical trial literature, what formulations of soy were used? At what dose? For what purpose(s) (e.g., trial endpoints)? (Section 3.1)
Does current clinical trial evidence indicate that whole soy products and individual constituents of soy have an effect on (Sections 3.2–3.8):
cardiovascular events, risk factors, and measures;
menopausal symptoms;
endocrine function;
cancer and tumor-related biomarkers;
osteoporosis and osteoporosis risk factors;
reproductive health;
kidney function; and
other outcomes, based on results of Key Question 1 above?
What is the scientific evidence of a dose-response effect of different forms of soy and individual constituents of soy for the conditions specified in Key Question 1? (Section 3.9)
What are the frequency and type(s) of adverse events associated with consumption of soy that are reported in the scientific literature (both trials and epidemiology)? (Section 3.10)
What is the scientific evidence of a dose-response effect of whole soy products and individual soy constituents on their safety? (Section 3.10)
in
Chapter 1 illustrates the
diverse potential mechanisms of soy on various health conditions.
To guide the assessment and synthesis of the literature, we used an expanded version of the generally-referred-to “PICO” method (Participants, Intervention, Comparator, Outcomes) to define the parameters of interest. With input from the TEP, we asked the following questions to establish the literature review criteria:
What are the populations of interest?
What are the interventions of interest?
What are the comparators of interest?
What are the (marker/intermediate and clinical) outcomes of interest?
What are the health conditions of interest?
What are acceptable study designs?
This report encompasses several health conditions and many outcomes of interest. Therefore, specific inclusion criteria were needed for each of the health conditions and sometimes for different outcomes of the same health condition. In this section, we first describe the common inclusion criteria for any study included in this report, regardless of the health condition evaluated. This is followed by additional specific criteria for each of the health conditions.
The common inclusion criteria for studies analyzed in this report consist of: human subjects 13 years and older; prospective studies including randomized controlled trials, cohorts (including prospective epidemiological studies), cross-over and non-randomized comparison studies; at least 5 subjects in the soy arm; any health condition; quantification of the amount of soy; and reported outcomes of interest. In general, the minimum duration for all serum marker, urine marker, and vascular outcome studies was 4 weeks (exceptions are noted below, under Specific Inclusion Criteria for Health Conditions Examined).
For assessments of adverse events, we also included prospective observation studies and case-control studies, with no limitations on study size or duration, or quantification of soy product.
In addition to the health conditions of interest listed under Key Question 3, the TEP suggested the category of neurocognitive outcomes. NCCAM and the Office of Dietary Supplements were also interested in knowing about research that might have been done in other health conditions. Therefore, our literature search was conducted to broadly include soy studies for any health conditions. We screened all citations to identify health conditions not on the list agreed upon with the TEP.
During our review process, we created the category of endocrine outcomes, which incorporates many of the reproductive hormones and potential cancer risk factor studies.
We accepted studies that used soy supplements and foods that quantified the amount of soy ingredients or products. We categorized various soy products and soy foods into the following groups:
Refined soy products
Isolated soy protein with isoflavones
Isolated soy protein without isoflavones
Textured soy protein
Soy derived isoflavone
genistein/genistin
daidzein/daidzin
glycitein/glycitin
Soy/soya food products (ingested amount must be quantified)
Whole soy beans (edamame)
Soy flour
Soy drink (soy milk)
Tofu (bean curd)
Miso
Other processed soy bean products (tempeh, natto, okara, etc.)
We also categorized studies based on whether the soy products were consumed in the form of a supplement or as part of the overall diet. In general, we relied on the studies' descriptions of the products and their use to make this determination. The categorization of soy milk, however, was problematic as approximately equal numbers of studies described its consumption as either a dietary replacement of other beverages or as a supplement to be added to subjects' regular diet; many studies reported insufficient details to determine how the soy milk was being used. In consultation with the TEP, we (arbitrarily) categorized all soy milk studies as supplement studies, in order to standardize our evaluation. Where necessary, sensitivity analyses were done regarding this categorization.
For the purpose of this report, all study arms with a soy product of any type were considered to be soy interventions. Only study arms with a non-soy intervention were categorized as controls. This is in contrast to many studies that considered soy protein without isoflavones to be the control. Since we were interested in the effect of both soy protein and soy isoflavones, we categorized these study arms as soy interventions.
In addition to the above common inclusion criteria, with input from TEP members we established the following additional criteria and specific outcomes for each of the specific health conditions.
Cardiovascular outcomes
We included in our analyses cardiovascular outcomes listed in 5 in Chapter 3. We also sought studies of clinical cardiovascular outcomes (e.g., death, myocardial infarction, angina) but found none. The list of outcomes was determined in consultation with the TEP. The decisions for which outcomes to investigate were based on expert opinion of the likelihood of an effect on the outcomes, clinical importance, and estimates of the numbers of studies likely to be available.
Because of the relatively large number of available studies reporting on lipids, triglycerides, and blood pressure, it was decided with the TEP to limit inclusion of these studies to randomized controlled trials with a minimum of 10 subjects consuming a soy product. For all cardiovascular outcomes, we required a minimum duration of 4 weeks.
Menopausal Symptoms
We evaluated studies of peri-menopausal and post-menopausal women for menopausal symptoms. A minimum duration of 4 weeks was required for studies of menopausal symptoms.
Endocrine Function
We included in our analyses the following endocrine outcomes: testosterone, follicle stimulating hormone (FSH), total estradiol and thyroid stimulating hormone (TSH). Because menstrual cycle length is directly related to female reproductive hormones, this outcome is included in this section. The decisions for which outcomes to investigate were based on expert opinion of the likelihood of an effect on the outcomes, clinical importance, and estimates of the numbers of studies likely to be available. Studies that did not report numerical data on effect for these outcomes were not summarized; however, these studies were maintained in the database. For all endocrine outcomes, we required a minimum duration of 4 weeks.
Cancer and Tumor-Related Biomarkers
To evaluate whether soy may prevent cancer or reduce cancer risk factors, we included only studies that recruited subjects without a diagnosis of cancer. We did not include studies that used soy products as “treatments” for cancer. We limited our analyses to studies with tumor-related biomarkers or cancer risk factors as outcomes and to studies of clinical cancer outcomes (e.g., diagnosis of prostate cancer). The only outcome that fulfilled these criteria was testosterone. The studies that reported testosterone as an outcome in men without diagnoses of cancer were analyzed in the endocrine section. The decision to investigate only testosterone was based on expert opinion of the likelihood of an effect on the outcomes, and clinical importance. For all tumor-related biomarkers, we broadened the eligibility criteria to include a minimum duration of 1 week.
Bone outcomes
For bone resorption and/or formation biomarkers, general inclusion criteria were used, including a minimum duration of 4 weeks. Because effects on bone mineral density occur slowly over time, we used minimum study duration of 1 year; although we did briefly review studies of less than 12 months.
Miscellaneous outcomes
For all other outcomes (neurocognitive, kidney, glucose metabolism), the general inclusion criteria were used in combination with the restriction to populations without the related specific diseases or conditions. Thus, studies of cognitive function outcomes were restricted to populations without Alzheimer's disease, dementia, or mental retardation at baseline. Studies of kidney function outcomes were restricted to those populations without kidney disease at baseline. Studies of glucose metabolism were restricted to populations without diabetes.
Reproductive health
Based on the included studies in our systematic review, outcomes that could address reproductive health issues in men and women are hormones related to fertility status as well as menstrual cycle length in women, which is used in the initial assessment for female infertility. The goal of the initial infertility evaluation of the couple is to determine the likely cause of the infertility and to determine the most logical approach to treatment. The initial evaluation for male factor infertility should include a reproductive history and two properly performed semen analyses. An initial endocrine evaluation should include at least a serum testosterone and FSH. Endocrine evaluation should be performed if there is: (1) an abnormally low sperm concentration, especially if less than 10 million/mL; (2) impaired sexual function; or (3) other clinical findings suggestive of a specific endocrinopathy. The initial evaluation for female factor infertility should include a reproductive history and documentation of ovulation, which is usually done with over the counter ovulation kits. An initial endocrine evaluation includes day three FSH level and estradiol.
We excluded studies that investigated soy products that were mixed with other potentially active ingredients (e.g., soy and fish oil) where the effect of the soy could not be separated from the other ingredients. Studies that compared combinations to similar products without soy (e.g., soy + estrogen vs. estrogen) were included. We also excluded soy products that are used as an ingredient of enteral feedings. In addition, we excluded the following studies:
Review articles (no primary data)
Non-trial observational studies
Animal or in vitro studies
Age less than 13 years
Not English language
Fewer than 5 subjects in the soy arm of the trial (unless adverse event reported)
Ingested soy amount not quantified
Not soy, soy protein, soy isoflavone
Insignificant amount of total daily protein or isoflavones in the soy product (e.g., soy sauce, soy oil; case by case determination was made in collaboration with technical experts)
Mixed soy product or nutrition/diet drink (e.g., brands such as Boost, Ensure, GeniSoy, Met-Rx, Revival Soy, Slim Fast) where other active ingredients may be present
Studies that report serum or urine levels of isoflavones or amino acids achieved instead of amount ingested
Study of serum or urine isoflavone levels without data on clinical outcomes or risk factors
Studies that evaluated only soy allergies
No outcome of interest
We conducted a comprehensive literature search to address the key questions.* Primary literature searches for English language publications on soy studies were conducted in EMBASE on March 25, 2004, in MEDLINE on April 20, 2004, and in CAB Abstracts on June 24, 2004. Search terms included subject headings and textwords with filters to limit the publications to English language and primary studies of the adult and adolescent human populations. Subject headings and text words were selected so that the same set could be applied to each of the different databases. A supplemental search was performed in MEDLINE on April 30, 2004 to retrieve articles using the textword “miso”. A search update was performed in MEDLINE In-Process & Other Non-Indexed Citations and MEDLINE on September 30, 2004, and CAB Abstracts on October 4, 2004. A search of the TOXLINE database was conducted in March 31, 2005 to identify additional reports for adverse events in humans. Additional studies were sought by contacting members of the TEP, and from reference lists of selected review articles and meta-analyses.
All citations identified through the literature search were screened according to the inclusion criteria. A low threshold for acceptance was used at this stage to maximize the retrieval of potentially useful studies. Retrieved articles were evaluated against the complete inclusion criteria.
A single reviewer extracted each eligible study.† Data extraction problems were addressed during weekly meetings. Occasional sections were re-extracted to ensure that uniform definitions were applied across extracted studies. Problems and corrections were noted through spot checks of extracted data and during the creation of summary and evidence tables. A second reviewer independently verified the data in the summary tables using the original article. Items extracted included: factors related to study design (randomization method, allocation concealment method, blinding, study duration, and funding source), population characteristics (country, eligibility criteria, demographics, co-morbid conditions, concomitant medications, and baseline diet), interventions and comparison groups (description of soy product and control interventions or diets, including amount of specific protein), outcomes of interest (number enrolled and analyzed, intermediate and clinical outcomes, adverse events, reasons for withdrawals, results [including baseline value, final value, within-treatment change or between-treatment difference, and variance, as reported]), and whether each study addressed each of the key questions. In addition, each study was categorized based on applicability and study quality as described below.
Studies accepted in evidence reports have been designed, conducted, analyzed, and reported with varying degrees of methodological rigor and completeness. Deficiencies in any of these components can lead to biased reporting and interpretation of the results. While it is desirable to grade individual studies to highlight the degree of potential bias, the grading of study quality is a challenging process. Most factors commonly used in quality assessment of randomized controlled trials do not demonstrate a consistent relationship to estimates of treatment effects.39 Thus, there is still no uniform approach to grade studies. Our EPC has adopted the following approach in our previous evidence reports.
We used a 3-category grading system (A, B, C) to denote the methodological quality of each study. This grading system has been used in most of the previous evidence reports from the Tufts-NEMC EPC as well as in evidence-based clinical practice guidelines.40 This system defines a generic grading system that is applicable to varying study designs including randomized controlled trials, cohort, and case-control studies:
A. Category A studies have the least bias and results are considered valid. A study that adheres mostly to the commonly held concepts of high quality including the following: a formal randomized study; clear description of the population, setting, interventions and comparison groups; sufficient power (arbitrarily defined as minimum sample size of 30 subjects); clear description of the content of the intervention used (including both amount of soy protein and amount of soy isoflavones); appropriate comparator (with similar amount and distribution of fats); appropriate measurement of outcomes; appropriate statistical and analytic methods and reporting; double-blinding; no reporting errors; less than 20% dropout; clear reporting of dropouts; and no obvious bias.
B. Category B studies are susceptible to some bias, but not sufficient to invalidate the results. They do not meet all the criteria in category A because they have some deficiencies, but none likely to cause major bias. The study may be missing information, making it difficult to assess limitations and potential problems.
C. Category C studies have significant bias that may invalidate the results. These studies have serious errors in design, analysis or reporting, have large amounts of missing information, or discrepancies in reporting. Specific criteria included large (>20%) or unequal dropout rate, large discrepancy in baseline and final numbers of subjects, non-randomized or single-cohort studies, dissimilar baseline values among cohorts, unclear duration or numbers of subjects, missing baseline data, or irreconcilable apparent differences between data in figures, tables, and text.
Where different quality criteria applied to different outcomes within a study (e.g., missing baseline data for a specific outcome), quality grades may differ for different outcomes within the same study. In addition to applying these 3 grading systems, additional comments relating to potential sources of bias and other study limitations were recorded by each investigator during the data extraction process. Such comments are included in the evidence tables.
Methodological quality scoring was performed near the end of the review when we had the most experience and knowledge about the included studies. Each included study was graded by at least 2 people (with the exception of studies with major deficiencies, such as a non-comparative study design). When there were disagreements, 1 or 2 additional reviewers graded the studies and consensus was reached. Approximately half the studies had quality scoring by 3 or more reviewers.
Applicability addresses the relevance of a given study to a population of interest. Every study applies certain eligibility criteria when selecting study subjects. Most of these criteria are explicitly stated (e.g., disease status, age, sex). Some may be implicit or due to unintentional biases, such as those related to study country, location (e.g., community vs. specialty clinic), or factors resulting in study withdrawals. The question of whether a study is applicable to a population of interest (such as Americans) is distinct from the question of the study's methodological quality. For example, due to differences in the background diets, an excellent study of Japanese men may be very applicable to people in Japan, but less applicable to Japanese American men, and even less applicable to African American men. The applicability of a study is thus dictated by the questions and populations that are of interest to those analyzing the studies.
In this report, the focus is on the US population and on specific subgroups within that population (i.e., post-menopausal women, peri-menopausal women, pre-menopausal women, men, and people with relevant medical histories, such as breast cancer). Even though a study may focus on a specific target population, limited study size, eligibility criteria, and the patient recruitment process may result in a narrow population sample that is of limited applicability, even to the target population. To address this issue, we categorized studies within a target population into 1 of 3 levels of applicability that are defined as follows:
Sample is representative of
the target population. It should be sufficiently large to cover both sexes
(as appropriate for a given outcome), a wide age range, and other important
features of the target population (e.g., background diet).
Sample is representative of
a relevant sub-group of the target population, but not the entire
population. Limitations include such factors as narrow age range, single
ethnicity, narrow range of risk for relevant diseases (e.g.,
hyperlipidemia).
Sample is representative of
a narrow subgroup of subjects only, and is of limited applicability to other
subgroups. For example, a study of the oldest-old men or a study of a
population on a highly controlled diet.
Most of the outcomes of interest were continuous variables such as blood pressure and lipid levels. For these outcomes, the summary tables describe 3 sets of data: the mean baseline level in the soy arm, the net change of the outcome, and the reported P values of the difference between the soy and the control arms. The net change of the outcome is the difference between the change in the soy arm and the change in the control arm:
Net change = (SoyFinal - SoyInitial) - (ControlFinal - ControlInitial).
While some studies reported adjusted and unadjusted within-arm and between-arm (net) differences, to maintain consistency across studies, we calculated the unadjusted net change using the above formula for all studies when the data were available. All exceptions and caveats are described in footnotes to the summary tables.
We included only reported P values for the net differences. We did not calculate any P values, but, when necessary, used provided information on the 95% confidence interval or standard error (SE) of the net difference to determine whether it was less than 0.05. We included any reported P value less than 0.10. Those above 0.10 and those reported as “non-significant” were described as “NS” (non-significant) in the tables.
For measures expressed using standard or Systeme International (SI) units (e.g. lipid levels), the original units reported in the study were included in the evidence tables. However, all such measurements were converted to standard units in the summary and results tables to facilitate comparisons.
Meta-analysis was performed for several cardiovascular outcomes. We used the random effects model for continuous outcomes to combine studies. Studies were included only if they reported sufficient data to estimate both mean net change (or mean within-cohort change for specific sub-analyses) and SE of the net change (or of the within-cohort change).
The random effects model assigns a weight to each study that is based both on the individual study variance and the between-study heterogeneity. Compared with the fixed effect model, the random effects model is more conservative in that it results in broader confidence intervals when between-study heterogeneity is present.
For the meta-analyses, we required data on both the mean change in outcome level and the SE of the change. However, many studies provided only the SEs for the baseline and final outcome levels. In order to include these studies in analysis we had to make several assumptions to estimate the SE of the change. To do this we used the equation:
SE12 = √ (SE1 2 + SE2 2 - 2 x ρ x SE1 x SE2)
where SE1, SE2, and SE12 are the SEs for baseline, final and change, respectively, and ρ is the correlation between SE1 and SE2.41 We arbitrarily chose the correlation, ρ, to be 0.50, the mid-point value. In our experience, using different values for ρ generally results in similar estimates of SE.
For each soy cohort, the SE of the net change was then calculated using the standard calculation for determining the SE of 2 independent cohorts. Namely the above equation where the correlation factor ρ = 0 and thus the final term drops out. Where studies reported either within-cohort SEs or net change SEs, these numbers were used.
An important caveat in our analyses is that for studies with multiple soy cohorts but single non-soy cohorts, we assumed that the estimated net changes (and their SEs) are independent of each other despite their shared control groups. In addition, we made a number of “corrections” to the reported data where there were apparent errors (such as reporting standard deviation as SE or reporting mean values and SEs in different units [e.g., mg/dL and mmol/L]). Where within-cohort or net changes were reported graphically, we preferentially used tabulated data for baseline and final values; although we estimated values and SEs from graphs when necessary. Net changes reported as percentage changes were ignored when baseline values from the soy and non-soy cohorts were not identical.
Meta-analyses were performed for all eligible studies of low density lipoprotein (LDL), high density lipoprotein (HDL), triglycerides, and blood pressure. These outcomes were chosen for meta-analysis based on potential clinical relevance and the number of available studies.
To explore potential reasons for differences of results across studies and to evaluate possible dose-effects, we performed several meta-regressions analyses with the continuous variables soy protein dose, soy isoflavone dose, and mean baseline outcome value. We used a random-effects regression model as described by Berkey et al. 42 This model adjusts each study's weight in the regression by the degree of heterogeneity across all included covariates. Both multivariate and univariate analyses were performed. When appropriate, sub-analyses were performed to explain factors related to the meta-regressions.
In addition, we performed sub-group meta-analyses based on the following sub-groups, when appropriate:
Study treatment arms with normal versus abnormal mean baseline outcome values (e.g., LDL less than and greater than 130 mg/dL)
Study treatment arms with different types of soy products:
Protein with isoflavone
Protein with isoflavone (without soy milk)
Soy milk
Protein without isoflavone
Isoflavone alone
Study treatment arms where soy was consumed as dietary replacement and as a supplement
Study quality A or B versus quality C
Outlier studies omitted.
In order to complement a previously published meta-analysis,43 we also performed a meta-analysis of
High versus low dose soy isoflavones treatment arms among studies that investigated multiple isoflavone doses
All high-dose cohorts versus lowest dose
Only highest-dose cohort versus lowest dose
For all outcomes, when there were sufficient studies, we examined all these same factors, in addition to study population and sex (when relevant) to determine if there was evidence of a differential effect based on these factors. We also specifically evaluated studies that either performed direct comparisons between treatments or reported sub-group analyses. In consultation with the TEP, it was decided to not evaluate equol-production status of study subjects. This decision was made based on the current lack of applicability of any findings of differences between equol producers and non-producers, since no one outside soy studies knows their equol production status.
The evidence table offers a detailed description of the studies that addressed each of the key questions. The evidence table is available via the internet.* The table provides information about the study design, patient characteristics, inclusion and exclusion criteria, interventions evaluated, and comparison groups evaluated. Each study appears once regardless of how many interventions or outcomes were reported. Studies are ordered alphabetically by the first author, then by publication date, then by MEDLINE unique identifier number.
Summary tables are included in each Results section. They succinctly report summary measures of the main outcomes evaluated. They include information regarding study duration, study size (of subjects analyzed), intervention and control, outcome measures, study population, and methodological quality, and study applicability. These tables were developed by condensing information from the evidence tables. They are designed to facilitate comparisons and synthesis across studies. Studies reporting multiple outcomes may appear several times in summary tables.
Description of the soy interventions includes quantification of the amount of aglycone isoflavones and the amount of soy protein, when available. When studies did not specify whether measured isoflavones were aglycones or glucosides, or when studies reported only glucoside amounts, this is indicated in the footnotes. Outcome units and metrics are reported in standard units and as in common metrics, regardless of how these were reported in the articles. These tables include baseline values of relevant outcomes, within-cohort changes (Final - Baseline) in these values, reported P values of the within-cohort changes, net changes compared to control (as defined above, under Results of Comparative Studies), and reported P values of the net changes. In addition, reported P values of differences in effect between different soy treatment arms are included. Within-cohort and net changes were calculated from reported data when necessary (these values are italicized). Blank cells indicate that the relevant data were not reported in the articles.
Studies are categorized and ordered as follows. Diet studies are above supplement studies; randomized cross-over studies are above randomized parallel controlled trials, which are above non-controlled or other non-randomized studies; studies that used dairy controls are above animal protein or usual diet controls, which are above miscellaneous controls, which are above studies with no non-soy control. Within each of these categories, studies are ordered from largest to smallest number of subjects consuming soy products.
We used the term adverse event as defined by the World Health Organization (WHO) International Conference on Harmonization. An adverse event is “any untoward medical occurrence in a patient or clinical investigation subject administered a pharmaceutical product and which does not necessarily have to have a causal relationship with this treatment. An adverse event can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product.” An adverse drug reaction is any “noxious and unintended response to a medicinal product related to any dose…” (www.fda.gov/cder/guidance/iche2a). For the purpose of this report, soy or soy ingredient is substituted for pharmaceutical product. We reviewed all accepted and rejected human studies in the analyses of soy effects for data on adverse events and drug interactions. These reports included randomized trials, cohorts, case-control studies, and individual case reports and series. We excluded articles that reported “bad taste”, “aftertaste”, or “lack of palatability” as adverse events as well as those that reported negative conclusions such as “there were no immunological effects” or “there was no increase in pancreatic size”. While adverse events attributable to soy in animal studies and changes in biomarkers in in-vitro studies might be indicative of potential toxicities in humans and would be useful in the overall evaluation of safety issues, the review of this body of literature is beyond the scope of this report.
We summarized adverse events in several tables. Adverse events were grouped according to the study design and the type of soy product used. In these tables we report for each study the data on the number of subjects, trial duration, interventions and controls, and the adverse events. Based on the adverse event data we compiled, we grouped them into menstrual-related and gastrointestinal-related complaints, withdrawals due to any effects, and a miscellaneous category. A separate table was created for studies that stated that no adverse events occurred.
Searches in EMBASE and MEDLINE yielded 2,639 and 2,650 citations, respectively, with an additional 10 citations from the supplemental search. After removal of overlapping and duplicate publications, the final yield was 4,471 citations. An updated search conducted in early December yielded an additional 308 citations.
After screening of the titles and abstracts, 599 articles were retrieved for examination. A total of 178 clinical trials publications were included in sections of the report regarding soy effects. In addition to these studies, 5 prospective cohorts and 3 pharmacokinetics studies reported data on adverse events. The results are summarized in this chapter in the following order: soy products used in the trials (Section 3.1), effects of soy on cardiovascular endpoints (3.2), menopausal symptoms (3.3), endocrine endpoints (3.4), cancer and tumor related biomarkers (3.5), bone endpoints (3.6), reproductive health (3.7), and a miscellaneous category that includes kidney, neurocognitive, and glucose metabolism endpoints (3.8). Following is an overview of association of dose and product type with effect (3.9). Finally, adverse events from clinical trials and observational studies are summarized following the review of evidence on health outcomes (3.10).
All qualifying studies are presented in summary tables in the appropriate sections. Details regarding these studies are available in the evidence table.* Some additional studies that contained pertinent information, but did not qualify for inclusion are also discussed.
| Outcome Categories | Isoflavones Alone | Soy Protein with Isoflavones | Soy Protein without Isoflavones | Unclear Amount of Soy Protein and/or Isoflavones | Total # of Studies* |
|---|---|---|---|---|---|
| Cardiovascular | Total: 23 | Total: 60 | Total: 7 | Total: 4 | 94 |
| Advanced Care Products (1) | Abacor (2) | Essential Nutrition (1) | ADM (1) | ||
| Bonette (Novomed, Helsinki) (2) | Abalon (Nutri Pharma, Oslo) (1) | Protein Technologies International (6) b | ISP powder (not specific) (2) | ||
| Eugenbio (1) | Altima HP-20 (Protein Technologies International) (2) b | Scan Diet Shakes (1) | |||
| Genistein, Lab Plant (2) | Calcimel (1) | ||||
| NovaSoy (ADM) (3) | Eden (1) | ||||
| Novogen (1) | FXP HO 159 (1) | ||||
| PhytoLife (1) | ISP powder (not specific) (19) | ||||
| Protoveg (2) | Proderma (ALPRO, Belgium) (2) | ||||
| Soya hypocotyl Iso (Fuji Oil Co) (2) | Solae - powder (2) b | ||||
| Soycreme (1) | Supro - powder (12) b | ||||
| Total Life Co (1) | Supro - liquid/beverage (6) | ||||
| No brand name tablet (6) | Supro - tablet/cap (1) b | ||||
| Soymilk (8) | |||||
| Tofuline (1) | |||||
| Unilever Best Foods, Brazil (1) | |||||
| Menopausal Symptoms and Menstrual Cycle Length | Total: 12 | Total: 15 | 27 | ||
| Advanced Care (1) | Banyang Foods (2) | ||||
| Bonette (Novomed, Helsinki) (1) | ISP powder (ADM) (1) | ||||
| Genistein, Lab Plant (1) | ISP powder—Protein Technologies International (5) b | ||||
| PharmaVite (2) | ISP powder (not specified) (1) | ||||
| PHYTO SOYA (Glycine max. L. Merr.) (1) | Kibun Food Chimifa (1) | ||||
| Phytosoya (1) | Soya World (1) | ||||
| Protoveg (2) | Supro - powder (1) b | ||||
| Solgar Italia (1) | Supro 675 powder (1) b | ||||
| Soylife™ (Netherlands) (1) | TakeCare™ (PTI) (2) b | ||||
| SOYSELECT™ (1) | |||||
| Endocrine | Total: 21 | Total: 17 | Total: 1 | 39 | |
| Advanced Care Products (1) | ADM Euro-port (1) | Essential Nutrition (1) | |||
| Bonette (Novomed, Helsinki) (1) | Banyang Foods (1) | ||||
| Eugenbio (1) | ISP powder (not specific) (2) | ||||
| Genistein, Lab Plant (4) | Kibun Food Chemifa, Tokyo (3) | ||||
| Genistein-combined polysaccharide (GCP, Amino Up Chem Co) (1) | Soymilk (2) | ||||
| NovaSoy (ADM) (4) | Supro - powder (6) b | ||||
| PharmaVite (1) | TakeCare™ (PTI) (2) b | ||||
| Regen soy extract (Novogen Ltd) (1) | |||||
| SOYSELECT™ (1) | |||||
| Total Life Co (1) | |||||
| No brand name tablet (5) | |||||
| Tumor related | Total: 7 | Total: 9 | 16 | ||
| NovaSoy (ADM) (2) | Banyang Foods (2) | ||||
| PharmaVite (1) | Protein Technologies International- powder (5) b | ||||
| Protoveg (1) | Soymilk (1) | ||||
| SOYSELECT™ (1) | Supro - powder (1) b | ||||
| Total Life Co (1) | |||||
| No brand name tablet (1) | |||||
| Bone | Total: 7 | Total: 12 | Total: 4 | 23 | |
| Advanced Care Products (1) | ISP powder—Protein Technologies International (7) b | ISP powder—Protein Technologies International (4) b | |||
| Bonette (Novomed, Helsinki) (1) | ISP powder—Solae (1) b | ||||
| Genistein, Lab Plant (1) | Proderma, ALPRO Belgium (1) | ||||
| NovaSoy (ADM) (1) | SoGood Soymilk drinks (1) | ||||
| Total Life Co (1) | Supro 675 (2) b | ||||
| Isoflavone extracts (No brand) (2) | |||||
| All other outcomes | Total: 3 | Total: 5 | 8 | ||
| Healthy Woman soy menopause supplement (1) | DuPont Protein Technologies—powder (1) b | ||||
| PHYTO SOYA (Glycine max. L. merr.) (1) | Fortimel (1) | ||||
| Solgen (Solbar Plant Extracts) (1) | Soy beverage (1) | ||||
| Supro - powder (2) | |||||
| Total # of Studiesa | 74 | 118 | 12 | 4 | 207 |
The numbers do not add up to the total due to single study may examine multiple outcome categories
In October 1997 Ralson Purina completed its sale of Protein Technologies International (PTI) to DuPont. Solae is the trademark for PTI.
| Outcome Categories | Tofu Alone | Soybean Alone | Soy Flour or Foods Made from Soy Flour | Other Soy Food | Texture Soy protein or Soy Diet (Mixed Soy Foods) | Total N of Studies* |
|---|---|---|---|---|---|---|
| Cardiovascular | Total: 4 | Total: 2 | Total: 5 | Total: 3 | Total: 22 | 36 |
| Blue Lotus Foods (2) | Soybean diet (2) | Nutrisoy flour (ADM Europort) (2) | Isosoy Soy germ (1) | Texture Soy protein - Cholsoy L (3) | ||
| Mori-nu Silken Extra Firm tofu (1) | Soy crisp bread (1) | Nijiru (1) | Texture Soy protein (1) | |||
| Tofu (1) | Soy flour (1) | SoGood Soy nuts (1) | Soy protein and/or Isoflavone Unclear (18) | |||
| Soy flour (Soy Pro-ducts of Australia) (1) | ||||||
| Menopausal Symptoms and Menstrual Cycle Lengths | Total: 3 | Total: 2 | 5 | |||
| Nutlettes® -soy flour and corn cereal (1) | Soy protein and/or Isoflavone Unclear (2) | |||||
| Soy flour (1) | ||||||
| Soy-grits bread (George Weston Foods) (1) | ||||||
| Endocrine | Total: 3 | Total: 1 | Total: 3 | Total: 2 | Total: 4 | 13 |
| Blue Lotus Foods (2) | Soybean product (1) | ADM Euro-port (1) | Isosoy Soy germ (1) | Soy protein and/or Isoflavone Unclear (4) | ||
| Tofu (1) | Soy flour (2) | Soy nuts (1) | ||||
| Cancer | Total: 3 | Total: 2 | 5 | |||
| Nutrisoy flour (ADM Europort) (1) | Soy protein and/or Isoflavone Unclear (2) | |||||
| Soy Bread (1) | ||||||
| Soy flour (1) | ||||||
| Bone | Total: 1 | Total: 2 | Total: 4 | Total: 1 | 8 | |
| Soybean diet (1) | Soy flour (1) | Natto (1) | Soy protein and/or Isoflavone Unclear (1) | |||
| Soy-grits bread (George Weston Foods) (1) | Nijiru (1) | |||||
| Roasted hypocotyl germ of soybean & sesame(1) | ||||||
| Soy nuts (1) | ||||||
| All other outcomes | Total: 1 | Total: 1 | Total: 1 | Total: 4 | 7 | |
| Soya Bean (1) | Soy Flour, Rakosvolgye Co (1) | Almased (1) | Soy protein and/or Isoflavone Unclear (4) | |||
| Total # of Studiesa | 7 | 5 | 17 | 10 | 35 | 74 |
The numbers do not add up to the totals due to studies examining multiple outcome categories.
w/ = with; w/o= without; SP and/or Isoflavone Unclear = Unclear Amount of Soy Protein and/or Isoflavones
| Daily Dose of Isoflavones (mg/ day) per gram of Soy Protein | ||||
|---|---|---|---|---|
| Genistein | Daidzein | Glycitein | Total Isoflavones | |
| Abacor ISP (PTI) | -- | -- | -- | 3.70 |
| Abacor yogurt (PTI) | -- | -- | -- | 1.85 |
| Abalon (Nutri Pharma, Oslo, Norway) | -- | -- | -- | >3.3 |
| Calcimel | 4.44 | 3.50 | -- | 7.94 |
| Essential Nutrition ISP(Brough, UK) | 2.33 | 1.63 | 0.43 | 4.40 |
| FXP HO 159 (PTI) | -- | -- | -- | 3.70 |
| ISP96 powder (PTI) | 1.30 | 0.70 | -- | 2.40 |
| ISP90 powder (PTI) | 0.98 | 0.65 | 0.18 | 1.80 |
| ISP80 powder (PTI) | 1.20 | 0.60 | 0.20 | 2.00 |
| ISP56 powder (PTI) | 0.65 | 0.35 | 0.10 | 1.08 |
| ISP38 powder (PTI) | 1.74 | 1.16 | 0.26 | 3.16 |
| Protoveg, Direct Foods (Manchester UK) | 0.33 | 0.42 | -- | 0.75 |
| Solae ISP powder | 2.00 | 1.58 | 0.23 | ≥3.81 |
| Soymilk (Kibun Food Chimica,Tokyo, Japan) | -- | -- | -- | 6.42 |
| Soymilk (Proderma/ALPRO, Belgium) | 1.00 | 1.12 | 0.97 | 3.08 |
| Soymilk (Unilever Best Foods, Brazil) | 2.00 | 1.32 | 0.20 | 3.48 |
| Soymilk (Banyang Foods, Houston, TX) | 2.25 | 1.82 | -- | 4.06 |
| Supro powder high in isoflavones | 1.32 | 0.89 | 0.19 | 2.42 |
| Supro powder low in isoflavones | 0.66 | 0.45 | 0.09 | 1.21 |
| TakeCare™ (PTI) | 1.40 | 0.71 | 0.09 | 2.20 |
Individual isoflavone doses may not add up to total due to rounding errors and presence of other isoflavones.
ISP = isolated soy protein, PTI = Protein Technology Institute, Inc.
| mg/day | ||||
|---|---|---|---|---|
| Genistein | Daidzein | Glycitein | Total Isoflavones | |
| Acatris high-dose isoflavones (Netherlands) | 12 | 38 | 31 | 80 |
| Acatris low-dose isoflavones (Netherlands) | 6 | 19 | 16 | 40 |
| Bonette (Novomed, Helsinki, Finland) | 6 | 42 | 66 | 114 |
| Eugenbio (Seoul, South Korea) | 70 | 19 | 11 | 100 |
| Genistein, Lab Plant (Messina, Italy) | 54 | 0 | 0 | 54 |
| Healthy Woman soy menopause supplement | -- | -- | -- | 110 |
| Novasoy (ADM) | 44 | 44 | 2 | 90 |
| PharmaVite (San Fernando, CA) | -- | -- | -- | 76 |
| Regen soy extract (Novogen, N.Ryde, Australia) | -- | -- | -- | 40 |
| Solgar Italia (Padua, Italy) | 11 | 36 | 25 | 76 |
| Solgen (Solbar Plant Extracts, Ashdod, Israel) | -- | -- | -- | 60 |
| Soya hypocotyls isoflavones (Fuji Oil Co, Japan) | 2 | 17 | 10 | 40 |
| Soylife soy extract (Netherlands) | 12 | 40 | 28 | 80 |
| Total Life Co (Taipei, Taiwan) | -- | -- | -- | 150 |
Individual isoflavone doses may not add up to total due to rounding errors and presence of other isoflavones.
Key Question 1: In the clinical trial literature, what formulations of soy were used? At what dose? For what purpose(s) (e.g., trial endpoints)?
A total of 281 comparisons were made with soy supplements or with soy foods/diets in the experimental arms. About three-quarters of these were trials of soy supplements, as isoflavones alone, soy protein with or without isoflavones. About one-half of the trials of soy supplements or soy foods and diets focused on one or more of the many cardiovascular endpoints. About 20% of the trials studied endocrine function. About 10–15% each, of the trials, evaluated menopausal symptoms or menstrual cycle length, bone outcomes, or cancer markers/risk factors. Less than 5% of the trials evaluated neurocognitive function, kidney function, or glucose metabolism. Over 90% of the soy supplement trials used isoflavones alone or soy protein with isoflavones; there were more trials of soy protein with isoflavones than trials of isoflavones alone.
Across studies, the total isoflavones ranged from 0 mg to 185 mg per day and the total protein intake from soy ranged from 0 g to 154 g per day. Of note, the median soy product dose across studies (36 g soy protein per day) was equivalent to over a pound of tofu daily or about 3 soy protein shakes daily.
Key Question 2: Does current clinical trial evidence indicate that whole soy products and individual constituents of soy have an effect on:
a. cardiovascular events, risk factors, and measures (Section 3.2);
b. menopausal symptoms (Section 3.3);
c. endocrine function (Section 3.4);
d. cancer and tumor-related biomarkers (Section 3.5);
e. osteoporosis and osteoporosis risk factors (Section 3.6);
f. reproductive health (Section 3.7);
g. kidney function (Section 3.8.1); and
h. other outcomes, based on results of Key Question 1 above (Sections 3.8.2–4)?
| Analyzed Outcomes | Number of Studies a | Section | Other Outcomes | Number of Studies b |
|---|---|---|---|---|
| Total cholesterol c | 66 | 3.2.2 | Cardiovascular events or disease | 0 |
| Low density lipoprotein (LDL) c | 56 | 3.2.3 | ||
| High density lipoprotein (HDL) c | 62 | 3.2.4 | Apolipoprotein A-1 | 24 |
| Triglycerides c | 58 | 3.2.5 | Apolipoprotein B | 27 |
| Lipoprotein (a) [Lp(a)] | 21 | 3.2.7 | Apolipoprotein C-III | 0 |
| Blood pressure (BP) | 25 | 3.2.8 | Apolipoprotein E | 2 |
| C-reactive protein (CRP) | 3 | 3.2.9 | Endothelin | 4 |
| Homocysteine (Hcy) | 5 | 3.2.1 | E-selectin | 3 |
| Endothelial function | 10 | 3.2.11 | Factor VII | 3 |
| Systemic arterial compliance | 3 | 3.2.12 | Factor VIII | 0 |
| Oxidized LDL | 13 | 3.2.13 | Factor XII | 0 |
| Total | 83 | Fibrinogen | 4 | |
| Free or non-esterified fatty acids | 1 | |||
| Intercellular adhesion molecule (ICAM) | 2 | |||
| Interleukin 2R or 6 | 3 | |||
| Nitrous oxide (NO) | 0 | |||
| P-selectin | 1 | |||
| Remnant-like particles | 0 | |||
| Thrombomodulin | 0 | |||
| Vascular cell adhesion molecule (VCAM) | 2 | |||
| von Willebrand Factor (vWF) | 1 | |||
| Ankle-brachial index | 0 | |||
| Bleeding time | 0 | |||
| Carotid ultrasound/Doppler | 0 | |||
| Coronary angiography | 0 | |||
| Echocardiography | 0 | |||
| Electrocardiogram (ECG) parameters | 0 | |||
| Exercise tolerance testing (ETT) | 0 | |||
| Heart rate variability | 0 | |||
| Intima-media thickness (IMT) | 0 | |||
| Platelet aggregation | 1 | |||
Studies reported in multiple articles are not double-counted.
Estimates. Studies reported in multiple articles may be double-counted.
Only randomized trials with a minimum of 10 subjects who consumed soy products.
| Author Year (UI) | TC | LDL | HDL | Tg | Lp(a) | BP | CRP | Hcy | Endothelial Function | Systemic Arterial Compliance | Oxidized LDL |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Ashton 2000 10694766 | x | ||||||||||
| Ashton 2000 11194529 | x | x | x | x | x | ||||||
| Azadbakht 2003 | x | x | x | x | |||||||
| Bakhit 1994 | x | x | x | x | |||||||
| Baum 1998 | x | x | x | ||||||||
| Blum 2003 12659466 | x | x | x | x | x | ||||||
| Bricarello 2004 | x | x | x | x | x | ||||||
| Burke 2001 | x | ||||||||||
| Carroll 1978 | x | ||||||||||
| Chiechi 2002 11836040 | x | x | x | x | x | ||||||
| Crouse 1999 | x | x | x | x | t | ||||||
| Cuevas 2003 | x | x | x | x | x | x | |||||
| D'Amico 1992 | x | ||||||||||
| Dent 2001 | x | x | x | x | x | ||||||
| Dewell 2002 | x | x | x | ||||||||
| Gallagher 2004 | x | x | x | x | |||||||
| Gardner 2001 | x | x | x | x | |||||||
| Gardner-Thorpe 2003 | x | x | x | x | x | ||||||
| Gentile 1993 | x | ||||||||||
| Goldberg 1982 | x | x | x | x | |||||||
| Gooderham 1996 | t | t | |||||||||
| Han 2002 | x | x | x | x | x | ||||||
| Hermansen 2001 | x | x | x | x | x | x | x | ||||
| Hill 2004 | x | ||||||||||
| Jayagopal 2002 | x | x | x | x | x | ||||||
| Jenkins 1999 | x | x | x | x | x | x | x | ||||
| Jenkins 2000 10647066 | x | ||||||||||
| Jenkins 2000 10778882 | x | ||||||||||
| Jenkins 2002 12077742 | x | ||||||||||
| Jenkins 2002 12145008 | x | x | x | x | x | x | x | x | |||
| Kanazawa 1995 | x | ||||||||||
| Kreijkamp-Kaspers 2004 | x | x | x | x | x | ||||||
| Kreijkamp-Kaspers 2005 | x | x | |||||||||
| Kurowska 1997 | x | x | x | x | x | x | |||||
| Lichtenstein 2002 | x | x | x | x | |||||||
| Lissin 2004 | x | x | x | ||||||||
| Mackey 2000 | x | x | x | x | |||||||
| Meinertz 1988 | x | x | x | x | |||||||
| Meinertz 1989 | x | x | x | x | |||||||
| Meinertz 2002 | x | x | x | x | t | ||||||
| Merz-Demlow 2000 | x | x | x | x | x | ||||||
| Meyer 2004 | x | x | x | x | x | x | x | ||||
| Murkies 1995 | x | x | x | ||||||||
| Murray 2003 | x | x | x | x | |||||||
| Nestel 1997 | x | x | x | x | x | x | x | ||||
| Nikander 2003 14602747 | x | ||||||||||
| Nikander 2004 15240647 | x | x | x | x | x | t | |||||
| Nilausen 1999 | x | ||||||||||
| Onning 1998 | x | x | x | x | |||||||
| Petri 2004 | x | x | x | x | |||||||
| Potter 1993 | x | x | x | x | |||||||
| Puska 2002 | x | x | x | x | x | x | |||||
| Puska 2004 | x | x | x | x | x | x | |||||
| Rivas 2002 | x | ||||||||||
| Sagara 2004 | x | x | x | ||||||||
| Scambia 2000 | t | t | t | t | |||||||
| Scheiber 2001 | x | ||||||||||
| Shorey 1981 | x | x | x | ||||||||
| Simons 2000 | x | x | x | x | x | x | x | ||||
| Sirtori 1999 | x | x | t | t | t | ||||||
| Sirtori 2002 | x | x | |||||||||
| Squadrito 2002 | x | x | x | x | x | ||||||
| Squadrito 2003 | x | ||||||||||
| Steinberg 2003 | x | x | x | x | x | x | |||||
| Swain 2002 | x | x | x | x | x | ||||||
| Takatsuka 2000 | x | x | x | x | |||||||
| Teede 2001 | x | x | x | x | x | x | x | x | |||
| Teede 2004 | x | ||||||||||
| Teixeira 2000 | x | x | x | x | |||||||
| Tonstad 2002 | x | x | x | x | x | x | |||||
| Uesugi 2002 | x | x | x | x | |||||||
| Uesugi 2003 | x | x | x | ||||||||
| Upmalis 2000 | t | t | t | t | |||||||
| Van Horn 2001 | x | x | x | ||||||||
| Verrillo 1985 | x | x | x | x | |||||||
| Vigna 2000 | x | x | x | x | x | x | |||||
| Wangen 2001 | x | x | x | x | x | ||||||
| Washburn 1999 | x | x | x | x | x | ||||||
| Watanabe 2000 11216491 | t | t | t | ||||||||
| Wong 1995 | x | ||||||||||
| Wong 1998 Hyperlipidemia | x | x | x | x | |||||||
| Wong 1998 Normolipidemia | x | x | x | x | |||||||
| Yildirir 2001 | x | x | |||||||||
t = results reported in text only, no data reported
TC = total cholesterol; LDL = low density lipoprotein; HDL = high density lipoprotein; Tg = triglycerides; Lp(a) = lipoprotein (a); BP = blood pressure; CRP = C-reactive protein; Hcy = Homocysteine; UI = MEDLINE, EMBASE, or CAB Abstracts unique identifier.
We also searched for prospectively designed studies that investigated the effectiveness of soy consumption for reducing clinical cardiovascular disease or events. However, no such studies were found.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Jenkins 2002 12145008 | 4 wk | ISP w/Isoflavones | 73 | 50 | 41 | 261 | -17 | NS | -9 | <0.01 | post♀♂ |
| C | ||||
| ISP w/Isoflavones | 10 | 52 | 258 | -18 | -10 | <0.01 | |||||||||||
| Low fat dairy+egg protein | 264 | -8 | |||||||||||||||
| Potter 1993 | 4 wk | ISP + cellulose | 50 | 25 | 228 | -18 | <0.5 | nd | -26 | <0.01b | ♂ |
| C | ||||
| ISP + cotyledon | 50 | -17 | <0.05 | -25 | <0.01b | ||||||||||||
| Soy flour | 0 | -11 | NS | -19 | <0.05b | ||||||||||||
| Non-fat dry milk +cellulose | +8 | NS | |||||||||||||||
| Diet | RCT | Dairy | |||||||||||||||
| Crouse 1999 | 9 wk | ISP w/Isoflavones | 62 | 25 | 30 | 243c | -10 | NS | -11 | <0.05 | ♀♂ |
| B | ||||
| ISP w/Isoflavones | 37 | 25 | 30 | 235 | -4 | -5 | NS | (LDL 140–200) | |||||||||
| ISP w/Isoflavones | 27 | 25 | 27 | 248 | -8 | -9 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 28 | 239 | -3 | -4 | NS | ||||||||||
| Casein | 31 | 240 | +1 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 261c | -24 | <0.04 Trend | -24 | <0.03 | ♀♂ | ||||||||
| ISP w/Isoflavones | 37 | 25 | 12 | 260 | -20 | -20 | <0.03 | LDL 166–200 | |||||||||
| ISP w/Isoflavones | 27 | 25 | 15 | 264 | -14 | -14 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 12 | 261 | -9 | -9 | NS | ||||||||||
| Casein | 16 | 258 | 0 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 226 | +4 | NS | +3 | NS | ♀♂ | ||||||||
| ISP w/Isoflavones | 37 | 25 | 18 | 219 | +6 | +5 | NS | LDL 140–166 | |||||||||
| ISP w/Isoflavones | 27 | 25 | 12 | 229 | -2 | -3 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 16 | 223 | +1 | 0 | NS | ||||||||||
| Casein | 15 | 221 | +1 | ||||||||||||||
| Teixeira 2000 | 6 wk | ISP w/Isoflavones | 95 | 50 | 15 | 242 | -8 | nd | -16 | 0.03 | ♂ |
| A | ||||
| ISP w/Isoflavones | 76 | 40 | 18/17 d | 230 | -1 | -9 | NS | ||||||||||
| ISP w/Isoflavones | 57 | 30 | 18 | 232 | -4 | -12 | 0.04 | ||||||||||
| ISP w/Isoflavones | 38 | 20 | 15 | 231 | -4 | -12 | 0.04 | ||||||||||
| Calcium caseinate | 16 | 235 | +8 | NS | |||||||||||||
| Van Horn 2001 | 6 wk | ISP w/Isoflavones + oats | 39 | 19 | 19 | 32 | 238 | -8 | <0.02 | +1 | NS | post♀ |
| B | |||
| Milk+ oats | 32 | 244 | -9 | <0.02 | |||||||||||||
| ISP w/Isoflavones + wheat | 39 | 19 | 19 | 31 | 234 | 0 | NS | 0 | NS | ||||||||
| Milk+wheat | 32 | 240 | 0 | NS | |||||||||||||
| Baum 1998 | 24 wk | ISP w/Isoflavones | 35 | 23 | 6 | 90 | 40 | 21 | 250 | -13 | nd | -6 | NS | post♀ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 2 | 56 | 40 | 23 | 254 | -15 | -8 | NS | |||||||
| Casein +non-fat dry milk | 22 | 242 | -7 | ||||||||||||||
| Vigna 2000 | 12 wk | ISP w/Isoflavones | 76 | 40 | 40 | 246 | -16 | <0.01 | 0 | post♀ |
| C | |||||
| Caseinate | 37 | 253 | -16 | <0.01 | |||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002 e | 6 wk | ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 f | 42 | 238 | +3 | NS | -9 | Soy: 0.02 | post♀♂ |
| B | |
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 f | +8 | -4 | Iso: NSb | |||||||||
| Animal w/Isoflavones e | 27 | 21 | 4 | 52 | 0 | +10 | -2 | ||||||||||
| Animal w/o Isoflavones | +12 | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 f | 22 | nd | 258g | nd | -19b | Soy: 0.001 | post♀♂ | |||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 f | 268g | -9b | Iso: 0.02b | |||||||||
| Animal w/Isoflavones e | 27 | 21 | 4 | 52 | 0 | 272g | -5b | LDL>160 | |||||||||
| Animal w/o Isoflavones | 277g | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 f | 20 | nd | 222g | nd | +2b | Soy: NS | post♀ | |||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 f | 222g | +2b | Iso: NSb | ♂ | ||||||||
| Animal w/Isoflavones e | 27 | 21 | 4 | 52 | 0 | 221g | +1b | LDL<160 | |||||||||
| Animal w/o Isoflavones | 220g | ||||||||||||||||
| Ashton 2000 11194529 | 4 wk | ISP diet | 84 | 35 | 120 | 36 | 42 | 224 | -14 | -9 | 0.03b | ♂ |
| C | |||
| Lean meat diet | -5 | ||||||||||||||||
| Azadbakht 2003 | 7 wk | ISP diet | 19 | 14 | 201 | -13 | <0.05 | -16 | <0.01 | ♀♂ DM Proteinuria |
| B | |||||
| Usual diet | 197 | +4 | <0.05 | ||||||||||||||
| Wong 1998 h | 5 wk | ISP diet | >15% | 13 | 262 | -15 | +4 | NS | ♂ |
| B | ||||||
| Animal diet | 264 | -11 | |||||||||||||||
| Goldberg 1982 | 6 wk | ISP diet | 91 | 12 | 260 | -40 | <0.0001 | -8 | <0.05b | ♀♂ |
| B | |||||
| Animal diet | -32 | <0.0001 | |||||||||||||||
| Wong 1995 | 4 wk | ISP diet | >15% | 12 | 273 | -41 j | -33 | ♂ |
| C | |||||||
| Animal diet | -8 j | ||||||||||||||||
| Diet | RCT | Animal/Usual | |||||||||||||||
| Chiechi 2002 11836040 | 26 wk | ISP diet k | 47 | 58/24 d | 236 | -9 | NS | +6 | 0.07 | post♀ |
| C | |||||
| Usual diet k | 55/43 d | 220 | -3 | NS | |||||||||||||
| Shorey 1981 | 6 wk | ISP diet | 55 | 13 | 241 | -16 | 0.03 | +6 | ♂ |
| C | ||||||
| Animal diet | 11 | 221 | -22 | 0.002 | |||||||||||||
| Diet | Xover | Miscellaneous | |||||||||||||||
| Jenkins 1999 | 4 wk | ISP diet | 33 | 31 | 250 | -27 | -16 | <0.001 | post♀♂ |
| B | ||||||
| Vegetarian diet | 248 | -10 | |||||||||||||||
| Diet | RCT | Miscellaneous | |||||||||||||||
| Sagara 2004 | 5 wk | Soy powder baked goods | 80 | 20 | 25 | 240 | -15 | <0.05 | -10 | NS | ♂ |
| B | ||||
| Usual baked goods | 25 | 232 | -5 | NS | |||||||||||||
| Murkies 1995 | 12 wk | Soy flour | 23 | 235 | -10 | NS | -1 | NS | post♀ |
| B | ||||||
| Wheat flour | 24 | 229 | -9 | NS | |||||||||||||
| Diet | RCT | No Control | |||||||||||||||
| Verrillo 1985 L | 16 wk | ISP, supplement L | 31 | 38 | 336 | -100 | <0.01 | NS | -- | ♀♂ |
| B | |||||
| ISP, 60 g replacing dietary protein | 31 | 19 | 340 | -100 | <0.01 | -- | |||||||||||
| No control group | -- | ||||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Difference of final values (cross-over study)
Minor discrepancy between text and table. In text baseline value for all subjects reported as 241 mg/dL; within-cohort change the same. In text baseline value for “high-LDL” subjects reported as 260 mg/dL; within-cohort change the same.
N: baseline/final.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Final values. No data on baseline or change from baseline.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
Graph
No data on how fat content of 2 diets compare.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Bricarello 2004 | 6 wk | Soy milkb | 50 | 33 | 5 | 87 | 25 | 60 | 241 | -4 | NS | -4 | NS | ♀♂ |
| C | |
| Non-fat milkb | 0 | NS | |||||||||||||||
| Kurowska 1997 | 4 wk | Soy milk | 31 | 34 | 265 | +2 | -3 | NS | ♀♂ |
| B | ||||||
| Milk, 2% fat | +5 | ||||||||||||||||
| Blum 2003 12659466 | 6 wk | ISP w/Isoflavones | 85 | 25 | 24 | 270 | -30 | 0.0001 | +2 | NSc | post♀ |
| C | ||||
| Total milk protein | -32 | ||||||||||||||||
| Meyer 2004 | 5 wk | Soy milk/yogurt | 80 | 30 | 23 | 232 | -7 | 0 | NSc | post♀♂ |
| B | |||||
| Low fat milk/yogurt | -7 | ||||||||||||||||
| Bakhit 1994 | 4 wk | ISP + cellulose | 25 | 21 | 222 | -4 | NS | -8 | NSd | ♂ |
| C | |||||
| ISP + cotyledon | 25 | 0 | NS | +2 | |||||||||||||
| Casein + cellulose | +4 | NS | |||||||||||||||
| Casein + cotyledon | -2 | NS | |||||||||||||||
| ISP + cellulose | 25 | 11 | 239 | -16 | <0.05 | -9 | 0.04d | ♂ | |||||||||
| ISP + cotyledon | 25 | -19 | <0.05 | -5 | TC>220 | ||||||||||||
| Casein + cellulose | -7 | NS | |||||||||||||||
| Casein + cotyledon | -14 | NS | |||||||||||||||
| Sirtori 1999 | 4 wk | Soy milk | 20 | 11 | 1 | 32 | 35 | 21 | 337 e | -21 f | <0.05 | -12 | <0.05 | ♀♂ |
| C | |
| Milk, high protein | -9g | NS | |||||||||||||||
| Hermansen 2001 | 6 wk | ISP w/Isoflavones | >165 | 50 | 20 | 219 | -22 | -17 | 0.08h | ♀♂ DM |
| C | |||||
| Casein | 216 | -5 | |||||||||||||||
| Sirtori 2002 | 4 wk | Soy milk | 25 | 28 | 24 | 77 | 25 | 20 | 314 | -6 | NS | -9 | NS | ♀♂ |
| B | |
| Milk | 318 | +3 | NS | ||||||||||||||
| Cuevas 2003 | 8 wk | ISP w/Isoflavones | 48 | 24 | 8 | 80 | <40 | 18 | 286 | -46 | <0.05 | +4 | NSc | post♀ |
| B | |
| Caseinate | -42 | <0.05 | |||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Teede 2001 | 13 wk | ISP w/Isoflavones | 77 | 38 | 5 | 118 | 40 | 86 | 228 | -21 | <0.05 | -6 | NS | post♀♂ |
| B | |
| Casein | 93 | 228 | -15 | <0.05 | |||||||||||||
| Kreijkamp-Kaspers 2004 | 52 wk | ISP w/Isoflavones | 52 | 41 | 6 | 26 | 88j | 240 | -1 | +6 | NS | post♀ |
| A | |||
| Total milk protein | 87j | 236 | -3 | ||||||||||||||
| Puska 2004 | 8 wk | ISP w/Isoflavones | 153 | 41 | 69 | 291 | -16 | -15 | <0.001 | post♀♂ |
| B | |||||
| Yogurt | 74 | 293 | -1 | ||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Tonstad 2002 | 16 wk | ISP w/Isoflavones | 185 | 50 | 31 | 252 | -35 | nd | -9 | 0.01 | post♀♂ |
| B | ||||
| ISP w/Isoflavones | 111 | 30 | 34 | 265 | -33 | ||||||||||||
| Casein, 50 g | 36 | 264 | -21 | ||||||||||||||
| Casein 30 g | 29 | 266 | -30 | ||||||||||||||
| Gardner 2001 | 12 wk | ISP w/Isoflavones | 52 | 25 | 4 | 80 | 42 | 31 | 228 | -10 | 0.03 | -2 | NS | post♀ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 0 | 3 | 42 | 33 | 228 | -1 | 7 | NS | |||||||
| Milk protein | 30 | 236 | -8 | ||||||||||||||
| Dent 2001 | 24 wk | ISP w/Isoflavones | 80 | 40 | 24 | 209 e | 0 | NS | 0 | NS | peri♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 40 | 24 | 217 e | -6 | -6 | |||||||||||
| Whey protein | 21 | 212 e | 0 | ||||||||||||||
| Puska 2002 | 6 wk | ISP w/Isoflavones | 96 | 52 | 24 | 290 | -25 | -10 | 0.05 | post♀♂ |
| C | |||||
| Calcium caseinate | 28 | 297 | -15 | ||||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Jayagopal 2002 | 12 wk | ISP w/Isoflavones | 70 | 49 | 13 | 132 | 30 | 31 | 223 | -9 | <0.05 | -15 | 0.004 | post♀ DM |
| B | |
| Cellulose | 217 | +6 | NS | ||||||||||||||
| Gardner-Thorpe 2003 | 6 wk | Soy flour biscuits | 45 | 75 | 120 | 19 | 212 | -8 | 0 | NSc | ♂ |
| B | ||||
| Wheat flour biscuits | -8 | ||||||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Wangen 2001 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 18 | 215 | -22 | 0.0004 | NS c | -- | post♀ |
| C | |
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | 18/17 k | -24 | 0.0004 | -- | ||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | 18 | -18 | 0.0006 | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Supplement | RCT | No Control | |||||||||||||||
| Gallagher 2004 | 39 wk | ISP w/Isoflavones | 52 | 28 | 96 | 40 | 17 | 221 | 3 | NS | nd | -- | post♀ |
| C | ||
| ISP w/Isoflavones | 28 | 20 | 52 | 40 | 19 | 220 | -7 | NS | -- | ||||||||
| ISP w/o Isoflavones | 4 | 0 | 4 | 40 | 14 | 213 | 0 | NS | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Mackey 2000 | 12 wk | ISP w/Isoflavones | 65 | 28 | 25 | 281 | -14 | NS | nd | -- | post♀ |
| B | ||||
| Female study | ISP w/o Isoflavones | 4 | 28 | 24 | 288 | -12 | -- | ||||||||||
| No control group | -- | ||||||||||||||||
| Verrillo 1985 L | 16 wk | ISP, supplement | 31 | 38 | 336 | -100 | <0.01 | NS | -- | ♀♂ |
| B | |||||
| ISP, 60 g replacing dietary protein | 31 | 19 | 340 | -100 | <0.01 | -- | |||||||||||
| No control group | -- | ||||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors inestimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Unequal amounts of fat in soy milk (17.5 g/day) and cow milk (0 g/day).
Difference of final values (cross-over study)
Main effect of soy protein
Graph
Reported that 2 soy cross-over arms combined had decrease in total cholesterol of 6.2% (number used in this table), but that first arm had 6.5% reduction and second arm had 7.4% reduction.
Average reduction in 2 milk cross-over arms (-3.9% and –1.6%), assuming baselines were the same for both arms.
P=0.004 for difference between final values.
Intention-to-treat analysis (75 completed soy protocol, 78 completed control protocol)
N: baseline/final.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Change | Net Change a | Population | Applicability | Quality | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Base value | Value | P within | P btw Soy | Value | P vs Control | |||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002 b | 6 wk | ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 42 | 238 | +3 | NS | -9 | Soy: 0.02 Iso: NSd | post♀ ♂ |
| B | |
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | +8 | -4 | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | +10 | -2 | ||||||||||
| Animal w/o Isoflavones | +12 | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 22 | nd | 258e | nd | -19d | Soy: 0.001 Iso: 0.02 d | post♀ ♂ | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 268e | -9d | LDL>160 | |||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 272e | -5d | ||||||||||
| Animal w/o Isoflavones | 277e | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 20 | nd | 222e | nd | +2d | Soy: NS Iso: NS d | post♀ ♂ | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 222e | +2d | LDL<160 | |||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 221e | +1d | ||||||||||
| Animal w/o Isoflavones | 220e | ||||||||||||||||
| Supplement | Xover | Placebo | |||||||||||||||
| Nikander 2004 15240647 | 13 wk | Isoflavones | 6 | 42 | 66 | 0 | 56 | 227 | +5 | NS | +3 | post♀ Breast CA |
| A | |||
| Placebo | 225 | +2 | NS | ||||||||||||||
| Nestel 1997 | 5 wk | Isoflavones | 45 | 35 | 3 | 80 | 0 | 21 | 214 | -7 | NS | +4 | post♀ |
| C | ||
| Placebo | -11 | NS | |||||||||||||||
| Simons 2000 | 8 wk | Isoflavones | 80 | 0 | 20 | 226 | -16 | +3 | NSd | post♀ |
| B | |||||
| Placebo | -13 | ||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Han 2002 | 13 wk | Isoflavones | 70 | 19 | 11 | 100 | 0 | 40 | 226 | -27 | <0.001 | -27 | <0.001 | post♀ |
| A | |
| Placebo | 40 | 227 | 0 | NS | |||||||||||||
| Squadrito 2002 | 26 wk | Genistein | 54 | 0 | 30 | 205 | +8 | NS | +4 | NS | post♀ |
| A | ||||
| Placebo | 30 | 208 | +4 | NS | |||||||||||||
| Petri 2004 | 26 wk | Soy germ capsules | 60 | 0.8 | 25 | 230 f,g | -13 | NS | -13 | post♀ |
| C h | |||||
| Lactose capsules | 25 | 204 f,g | 0 | NS | |||||||||||||
| Lissin 2004 | 6 wk | Isoflavones | 44 | 44 | 2 | 90 | 0 | 20 | 243 | -5 | NS | +7 | NS | post♀ |
| B | |
| Placebo | 20 | 236 | -12 | <0.05 | |||||||||||||
| Dewell 2002 | 26 wk | Isoflavones | 40 | 50 | 90 | 0 | 20/18 j | 263 | -15 | -3 | NS | post♀ |
| C | |||
| Placebo | 16 | 243 | -12 | ||||||||||||||
| Uesugi 2002 | 4 wk | Isoflavones | 0 | 0 | 0 | 62 k | 0 | 12 | 226 | -11 | <0.05 | -14 | NS | peri♀ |
| B | |
| Placebo | 11 | 238 | +3 | NS | |||||||||||||
| Uesugi 2003 | 13 wk | Isoflavones | 0 | 0 | 0 | 62 k | 0 | 11 | 224 | -5 | -4 | NS | post♀ |
| B | ||
| Dextrin | 10 | 221 | -1 | ||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Difference of final values (cross-over study)
Final values. No data on baseline or change from baseline.
Significantly different (P<0.05) at baseline.
Graph
In contrast with other outcomes in this article, soy and control ams had significantly different LDL levels.
N: baseline/final.
31 mg daidzin, 7 mg genistin, 21 mg glycitin
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Meinertz 2002 | 4.5 wk | ISP w/Isoflavones, liquid | 2.39/g b | 20% | 12 | 160 | -29 | <0.001 | NS c | +5 | NSc | ♀♂ |
| C | |||
| ISP w/o Isoflavones, liquid | 0.11/g b | 20% | 161 | -35 | <0.001 | -1 | NSc | ||||||||||
| Casein diet, liquid | 155 | -34 | <0.001 | ||||||||||||||
| Meinertz 1989 | 4.5 wk | ISP diet, liquid cholesterol enriched | 112 | 11 | 171 | -39 | -7 | NSc | ♀♂ |
| C | ||||||
| Calcium caseinate | -33 | ||||||||||||||||
| Meinertz 1988 | 4 wk | ISP diet, liquid low cholesterol | 113 | 10 | 171 | -44 | +2 | NSc | ♀♂ |
| C | ||||||
| Casein diet, liquid low cholesterol | -46 | ||||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Carroll 1978 | ~5.5 wk | ISP diet | 75 | 10 | 163 d | -2 | -9 | <0.05c | pre♀ |
| C | ||||||
| Usual diet | +7 | ||||||||||||||||
| Diet | Xover | Miscellaneous | |||||||||||||||
| Wong 1998 e | 5 wk | ISP diet | >15% | 13 | 170 | -15 | +6 | NS | ♂ |
| B | ||||||
| Animal diet | 169 | -9 | |||||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Steinberg 2003 | 6 wk | ISP w/Isoflavones | 55 | 47 | 5 | 107 | 25 | 24 | 190 | -3 | NS | 0 | NS | post♀ |
| C | |
| ISP w/o Isoflavones | 1 | 0.5 | 0.5 | 2 | 24 | 0 | +3 | NS | |||||||||
| Total milk protein | -3 | ||||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Murray 2003 | 26 wk | ISP + Estradiol 0.5 mg | 66 | 44 | 10 | 120 | 38 | 8 | 212 | +1 | NS | +3 | NS | post♀ |
| C | |
| Total milk protein + Estradiol 0.5 mg | 7 | 211 | -2 | NS | |||||||||||||
| ISP + Estradiol 1.0 mg | 66 | 44 | 10 | 120 | 38 | 8 | 216 | -13 | NS | +4 | |||||||
| Total milk protein + Estradiol 1.0 mg | 7 | 261 f | -17 | NS | |||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Washburn 1999 | 6 wk | ISP w/Isoflavones once daily | 34 | 14 | 42 | 208 | -9 | nd | -9 | <0.01c | peri♀ |
| B | ||||
| ISP w/Isoflavones twice daily | 34 | 14 | -12 | -12 | <0.01c | ||||||||||||
| Carbohydrates | 0 | ||||||||||||||||
| Onning 1998 | 4 wk | Soy milk | 23/30 g | 12 | 170 | -4 | NS | +4 | ♀♂ |
| B | ||||||
| Oat milk | -8 | NS | |||||||||||||||
| Supplement | RCT | Miscellaneous | |||||||||||||||
| Takatsuka 2000 | 9 wk | Soy milk | 109 | 17 | 27 | 177 | -11 | 0.004 | -11 | 0.02 | pre♀ |
| B | ||||
| No supplement | 25 | 177 | 0 | NS | |||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Merz-Demlow 2000 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 13 | 144 | -5 | NS | -- | pre♀ |
| C | ||
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | +2 | -- | ||||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | +4 | -- | ||||||||||
| No control group | -- | ||||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors inestimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Per gram protein. No data on number of grams of protein.
Difference of final values (cross-over study)
Calculated mean of graphically displayed values for each cross-over arm. “Final” results are mean value of levels measured twice weekly.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
Significantly higher at baseline than other study arms because of single outlier with hypertriglyceridemia.
Women/Men
Measurements made at 4 menstrual cycle phases: early follicular, midfollicular, periovulatory, midluteal. Data extracted for periovulatory only. LDL change was greatest during this phase in high isoflavone group.
Abnormal levels of serum lipids, primarily low density lipoprotein (LDL), high density lipoprotein (HDL), and triglycerides (Tg) have long been recognized as an independent risk factor for cardiovascular disease (CVD). Of interest is whether soy protein and isoflavone consumption would be of value for improving lipid levels as part of a therapeutic lifestyle change, or at least that it would not be detrimental. Recent National Cholesterol Education Program (NCEP) guidelines recommend a goal for fasting total cholesterol of less than 200 mg/dL in all adults with lower levels recommended for people at elevated risk for CVD, including diabetics, smokers, people with hypertension or family history of premature CVD, or beyond middle age.44
We found 65 studies that reported on the effect of the consumption of soy products on total cholesterol. Of these, 4 reported only that there was no effect on total cholesterol.45– 48 The remaining 61 studies are described below.23, 24, 49–106
The range of daily soy isoflavone intake among studies of soy products with isoflavones was approximately 10 to 185 mg, with a median of 80 mg per day. The range of daily soy protein intake among relevant studies was approximately 14 to 113 g, with a median of 38 g per day. Most studies of hyperlipidemic subjects included post-menopausal women and/or men. The studies of normolipidemic subjects were more likely to include both men and women or be restricted to pre-menopausal women. The large majority of studies were of limited applicability, even within the categories of pre- or post-menopausal women, or men. Only 12 of the studies were graded as being broadly applicable. Among the 61 studies, 5 were rated good quality (A), 32 were rated fair quality (B), and 24 were rated poor quality (C)..
Across the 61 studies there was a wide range of effects of soy products on total cholesterol, although in most studies the net effect was negative, indicating that compared to control consumption of soy products resulted in reductions in total cholesterol. Approximately two-thirds of the cohorts of subjects consuming soy had a net reduction of total cholesterol compared to control. Across studies, net change ranged from -33 to +7 mg/dL, with the median net change equal to -6 mg/dL. In terms of percent net change (using the baseline level in the soy intervention cohort as the denominator), net change ranged from approximately -12% to +4%, with a median net change equal to -2.5%.
Analyses across studies
Figure 2
graphs the net change
compared to non-soy control of all cohorts evaluated (that included a
non-soy cohort). The left-most graph displays net change in total
cholesterol in relationship the baseline total cholesterol level; this
graph includes all studies. The middle graph compares net change to
daily soy isoflavone consumption (among those studies that report
isoflavone content). The right-most graph compares net change to daily
soy protein consumption (among those studies that report soy protein
content). Study cohorts who consumed soy with isoflavones are indicated
by squares; cohorts who consumed soy without isoflavones are indicated
by circles; and cohorts who consumed soy isoflavone without soy protein
are indicated by triangles. Black symbols represent cohorts where the
soy product was consumed as part of the regular diet; open symbols
represent cohorts where the soy product was consumed as a supplement to
the diet (including soy milk products). Larger symbols indicate cohorts
whose mean total cholesterol or LDL was elevated at baseline; smaller
symbols indicate cohorts with normal total cholesterol and/or LDL at
baseline.
Visual inspection of the graphs reveal no difference across studies in net effect on total cholesterol (as mg/dL or % change) based on baseline total cholesterol levels, amount of soy isoflavones consumed, or amount of soy protein consumed. Likewise, there are no clear differences across studies based on whether the soy products were consumed as part of the diet (i.e., specifically replacing other sources of protein) or as a supplement (consumed in addition to regular diet). Also comparing the net effects of soy protein with isoflavones to soy protein without isoflavones to isoflavones without soy protein reveals similar ranges of effects for all 3 types of products. Because of the relative weakness of total cholesterol as a cardiovascular risk factor as compared to LDL, HDL, and triglycerides, meta-analysis was not performed.
Effect of baseline total cholesterol on net change total cholesterol in individual studies
Effect of soy isoflavone dose on net change total cholesterol in individual studies
Effect of soy protein dose on net change total cholesterol in individual studies
Effect of soy as diet vs. supplement on net change total cholesterol in individual studies
Effect of sex and menopausal status on net change total cholesterol
See Section 3.2.6.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Jenkins 2002 12145008 | 4 wk | ISP w/Isoflavones | 73 | 50 | 41 | 176 | -16 | NS | -10 | <0.01 | post♀ ♂ |
| C | ||||
| ISP w/Isoflavones | 10 | 52 | 175 | -13 | -7 | <0.01 | |||||||||||
| Low fat dairy+egg protein | 178 | -6 | |||||||||||||||
| Potter 1993 | 4 wk | ISP + cellulose | 50 | 24 | 175 | -14 | NS | nd | -18 | <0.01b | ♂ |
| C | ||||
| ISP + cotyledon | 50 | -15 | <0.05 | -19 | <0.01b | ||||||||||||
| Soy flour | 0 | -9 | NS | -13 | NSb | ||||||||||||
| Non-fat dry milk+cellulose | +4 | NS | |||||||||||||||
| Diet | RCT | Dairy | |||||||||||||||
| Crouse 1999 | 9 wk | ISP w/Isoflavones | 62 | 25 | 30 | 166c | -10 | NS | -10 | <0.05 | ♀♂ (LDL 140-200) |
| B | ||||
| ISP w/Isoflavones | 37 | 25 | 30 | 161 | -5 | -5 | NS | ||||||||||
| ISP w/Isoflavones | 27 | 25 | 28 | 170 | -5 | -4 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 27 | 164 | -4 | -5 | NS | ||||||||||
| Casein | 31 | 165 | 0 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 185c | -22c | <0.04 Trend | -20c | <0.03 | ♀♂ LDL 166-200 | ||||||||
| ISP w/Isoflavones | 37 | 25 | 12 | 182 | -17 | -15 | <0.03 | ||||||||||
| ISP w/Isoflavones | 27 | 25 | 12 | 186 | -12 | -10 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 15 | 185 | -10 | -8 | NS | ||||||||||
| Casein | 16 | 182 | -2 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 147 | +1 | NS | -2 | NS | ♀♂ LDL 140–166 | ||||||||
| ISP w/Isoflavones | 37 | 25 | 16 | 146 | +4 | +1 | NS | ||||||||||
| ISP w/Isoflavones | 27 | 25 | 18 | 150 | +4 | +1 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 12 | 148 | +1 | -2 | NS | ||||||||||
| Casein | 15 | 146 | +3 | ||||||||||||||
| Van Horn 2001 | 6 wk | ISP w/Isoflavones + oats | 39 | 19 | 19 | 32 | 150 | -10 | <0.02 | -1 | post♀ |
| B | ||||
| Milk+ oats | 32 | 155 | -9 | <0.02 | |||||||||||||
| ISP w/Isoflavones + wheat | 39 | 19 | 19 | 31 | 147 | 0 | NS | +1 | |||||||||
| Milk + wheat | 32 | 151 | -1 | NS | |||||||||||||
| Vigna 2000 | 12 wk | ISP w/Isoflavones | 76 | 40 | 40 | 159 | -14 | <0.01 | -2 | post♀ |
| C | |||||
| Caseinate | 37 | 167 | -12 | <0.01 | |||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002 d | 6 wk | ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 42 | 160 | +6 | <0.05 | -5 | Soy: 0.04 Iso: NSb | post♀ ♂ | |||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | +8 | -3 | ||||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | +12 | +1 | ||||||||||
| Animal w/o Isoflavones | +11 | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 22 | >160 | 181 f | nd | -14b | Soy: 0.003 Iso: NSd | post♀ ♂ LDL>160 |
| B | |||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | 186 f | -9b | ||||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | 192 f | -3b | ||||||||||
| Animal w/o Isoflavones | 195 f | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 20 | <160 | 148 f | nd | +5b | Soy: NS Iso: NSd | post♀ ♂ LDL<160 | |||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | 147 f | +4b | ||||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | 149 f | +6b | ||||||||||
| Animal w/o Isoflavones | 143 f | ||||||||||||||||
| Ashton 2000 10694766 | 4 wk | ISP diet | 84 | 35 | 120 | 36 | 42 | 142 | -8 | <0.05 | +3 | <0.05b | ♂ |
| B | ||
| Lean meat diet | -5 | NS | |||||||||||||||
| Azadbakht 2003 | 7 wk | ISP diet | 19 | 14 | 145 | -6 | <0.05 | -8 | <0.04 | ♀♂ DM Proteinuria |
| B | |||||
| Usual diet | 144 | +2 | NS | ||||||||||||||
| Wong 1998 g | 5 wk | ISP diet | >15% | 13 | 181 | -23 | -9 | 0.03 | ♂ |
| B | ||||||
| Animal diet | 182 | -14 | |||||||||||||||
| Goldberg 1982 | 6 wk | ISP diet | 91 | 12 | 191 | -33 | <0.0001 | -10 | <0.05b | ♀♂ |
| B | |||||
| Animal diet | -23 | <0.001 | |||||||||||||||
| Diet | RCT | Animal/Usual | |||||||||||||||
| Chiechi 2002 11836040 | 26 wk | ISP dieth | 47 | 58/24 j | 159 | -6 | NS | -6 | post♀ |
| C | ||||||
| Usual dieth | 55/43 j | 145 | 0 | NS | |||||||||||||
| Diet | Xover | Miscellaneous | |||||||||||||||
| Jenkins 1999 | 4 wk | ISP diet | 33 | 31 | 170 | -21 | -11 | <0.001 | post♀ ♂ |
| B | ||||||
| Vegetarian diet | 169 | -9 | |||||||||||||||
| Diet | RCT | No Control | |||||||||||||||
| Verrillo 1985 k | 16 wk | ISP, supplement k | 31 | 38 | 243 | -89 | <0.01 | NS | -- | ♀♂ |
| B | |||||
| ISP, 60 g replacing dietary protein | 31 | 19 | 259 | -100 | <0.01 | -- | |||||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Difference of final values (cross-over study)
Minor discrepancy between text and table. In text baseline value for all subjects reported as 165 mg/dL; within-cohort change the same. In text baseline value for “high-LDL” subjects reported as 181 mg/dL; within-cohort change reported as -18 mg/dL, implying a net change of -16 mg/dL.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Final values. No data on baseline or change from baseline.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
No data on how fat content of 2 diets compare.
N: baseline/final.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P. btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Bricarello 2004 | 6 wk | Soy milkb | 50 | 33 | 5 | 87 | 25 | 60 | 157 | -9 | -10 | <0.05 | ♀♂ |
| C | ||
| Non-fat milkb | +1 | ||||||||||||||||
| Kurowska 1997 | 4 wk | Soy milk | 31 | 34 | 172 | -7 | -4 | NS | ♀♂ |
| B | ||||||
| Milk, 2% fat | -3 | ||||||||||||||||
| Blum 2003 12659466 | 6 wk | ISP w/Isoflavones | 85 | 25 | 24 | 178 | -36 | <0.0001 | +5 | NSc | post♀ |
| C | ||||
| Total milk protein | -41 | ||||||||||||||||
| Meyer 2004 | 5 wk | Soy milk/yogurt | 80 | 30 | 23 | 159 | -8 | -2 | NSc | post♀ ♂ |
| B | |||||
| Low fat milk/yogurt | -6 | ||||||||||||||||
| Bakhit 1994 | 4 wk | ISP + cellulose | 25 | 21 | 158 | -3 | NS | -8 | NSd | ♂ |
| C | |||||
| ISP + cotyledon | 25 | 0 | NS | +2 | |||||||||||||
| Casein + cellulose | 5 | NS | |||||||||||||||
| Casein + cotyledon | -2 | NS | |||||||||||||||
| ISP + cellulose | 25 | 11 | 169 | -9 | NS | -8 | 0.07d | ♂ TC>220 | |||||||||
| ISP + cotyledon | 25 | -11 | NS | -1 | |||||||||||||
| Casein + cellulose | -1 | NS | |||||||||||||||
| Casein + cotyledon | -10 | NS | |||||||||||||||
| Sirtori 1999 | 4 wk | Soy milk | 20 | 11 | 1 | 32 | 35 | 21 | 247 e | -19 | <0.05 | -11 | <0.05 | ♀♂ |
| C | |
| Milk, high protein | -8 | NS | |||||||||||||||
| Hermansen 2001 | 6 wk | ISP w/Isoflavones | >165 | 50 | 20 | 140 | -24 | -12 | 0.048f | ♀♂ DM |
| C | |||||
| Casein | 141 | -12 | |||||||||||||||
| Sirtori 2002 | 4 wk | Soy milk | 25 | 28 | 24 | 77 | 25 | 20 | 226 | +7 | NS | -3 | NS | ♀♂ |
| B | |
| Milk | 225 | +10 | NS | ||||||||||||||
| Cuevas 2003 | 8 wk | ISP w/Isoflavones | 48 | 24 | 8 | 80 | <40 | 18 | 195 | -35 | <0.05 | -1 | NSc | post♀ |
| B | |
| Caseinate | -34 | <0.05 | |||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Teede 2001 | 13 wk | ISP w/Isoflavones | 77 | 38 | 5 | 118 | 40 | 86 | 151 | -16 | <0.05 | -5 | NS | post♀ ♂ |
| B | |
| Casein | 93 | 147 | -11 | <0.05 | |||||||||||||
| Kreijkamp-Kaspers 2004 | 52 wk | ISP w/Isoflavones | 52 | 41 | 6 | 26 | 88g | 161 | -1 | +5 | post♀ |
| A | ||||
| Total milk protein | 87g | 159 | -6 | ||||||||||||||
| Puska 2004 | 8 wk | ISP w/Isoflavones | 153 | 41 | 69 | 197 | -14 | -15 | <0.001 | post♀ ♂ |
| B | |||||
| Yogurt | 74 | 198 | +2 | ||||||||||||||
| Tonstad 2002 | 16 wk | ISP w/Isoflavones | 185 | 50 | 31 | 177 | -37 | nd | -10 | 0.01 | post♀ ♂ |
| B | ||||
| ISP w/Isoflavones | 111 | 30 | 34 | 186 | -34 | ||||||||||||
| Casein, 50 g | 29 | 192 | -30 | ||||||||||||||
| Casein 30 g | 36 | 188 | -21 | ||||||||||||||
| Gardner 2001 | 12 wk | ISP w/Isoflavones | 52 | 25 | 4 | 80 | 42 | 31 | 151 | -15 | 0.01 | -3 | NS | post♀ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 0 | 3 | 42 | 33 | 151 | -3 | 9 | NS | |||||||
| Milk protein | 30 | 154 | -12 | ||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Dent 2001 | 24 wk | ISP w/Isoflavones | 80 | 40 | 24 | 127 e | +6 | NS | +4 | NS | peri♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 40 | 24 | 135 e | +1 | -1 | |||||||||||
| Whey protein | 21 | 134 e | +2 | ||||||||||||||
| Puska 2002 | 6 wk | ISP w/Isoflavones | 96 | 52 | 24 | 198 | -26 | -10 | <0.05 | post♀ ♂ |
| C | |||||
| Calcium caseinate | 28 | 199 | -16 | ||||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Jayagopal 2002 | 12 wk | ISP w/Isoflavones | 70 | 49 | 13 | 132 | 30 | 31 | 140 | -10 | <0.05 | -17 | 0.001 | post♀ DM |
| B | |
| Cellulose | 134 | +7 | 0.05 | ||||||||||||||
| Gardner-Thorpe 2003 | 6 wk | Soy flour biscuits | 45 | 75 | 120 | 19 | 139 | -4 | +6 | NSc | ♂ |
| B | ||||
| Wheat flour biscuits | -10 | ||||||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Wangen 2001 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 18 | 136 | -20 | 0.0003 | 0.02 c | -- | post♀ |
| C | |
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | 18/17 h | -18 | 0.0006 | 0.07 c | -- | |||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | 18 | -12 | 0.02 | -- | -- | |||||||
| No control group | -- | ||||||||||||||||
| Supplement | RCT | No Control | |||||||||||||||
| Gallagher 2004 | 39 wk | ISP w/Isoflavones | 52 | 28 | 96 | 40 | 17 | 141 | +3 | NS | nd | -- | post♀ |
| C | ||
| ISP w/Isoflavones | 28 | 20 | 52 | 40 | 19 | 138 | + 3 | NS | -- | ||||||||
| ISP w/o Isoflavones | 4 | 0 | 4 | 40 | 14 | 134 | -1 | NS | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Mackey 2000 Female study | 12 wk | ISP w/Isoflavones | 65 | 28 | 25 | 196 | -14 | 0.04 | nd | -- | post♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 28 | 24 | 197 | -13 | -- | |||||||||||
| No control group | -- | ||||||||||||||||
| Verrillo 1985 j | 16 wk | ISP, supplement | 31 | 38 | 243 | -89 | <0.01 | NS | -- | ♀♂ |
| B | |||||
| ISP, 60 g replacing dietary protein j | 31 | 19 | 259 | -100 | <0.01 | -- | |||||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Unequal amounts of fat in soy milk (17.5 g/day) and cow milk (0 g/day).
Difference of final values (cross-over study)
Main effect of soy protein
Graph
P=0.004 for difference between final values.
Intention-to-treat analysis (75 completed soy protocol, 78 completed control protocol)
N: baseline/final.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002 b | 6 wk | ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 42 | 160 | +6 | <0.05 | -5 | Soy: 0.04 Iso: NSd | post♀ ♂ |
| B | |
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | +8 | -3 | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | +12 | +1 | ||||||||||
| Animal w/o Isoflavones | +11 | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 22 | >160 | 181e | nd | -14d | Soy: 0.003 Iso: NSd | post♀ ♂ LDL>160 | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 186e | -9d | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 192e | -3d | ||||||||||
| Animal w/o Isoflavones | 195e | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 20 | <160 | 148e | nd | +5d | Soy: NS Iso: NSd | post♀ ♂ LDL<160 | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 147e | +4d | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 149e | +6d | ||||||||||
| Animal w/o Isoflavones | 143e | ||||||||||||||||
| Supplement | Xover | Placebo | |||||||||||||||
| Nikander 2004 15240647 | 13 wk | Isoflavones | 6 | 42 | 66 | 0 | 56 | 149 | +11 | NS | +13 | NS | post♀ Breast CA |
| A | ||
| Placebo | 147 | -2 | NS | ||||||||||||||
| Isoflavones | 6 | 42 | 66 | 0 | 28 | >162 | +25 | +42 | 0.009 | post♀ Breast CA LDL>162 | |||||||
| Placebo | -17 | ||||||||||||||||
| Nestel 1997 | 5 wk | Isoflavones | 45 | 35 | 3 | 80 | 0 | 21 | 138 | -5 | NS | +3 | post♀ |
| C | ||
| Placebo | -8 | NS | |||||||||||||||
| Simons 2000 | 8 wk | Isoflavones | 80 | 0 | 20 | 152 | -13 | -3 | NSd | post♀ |
| B | |||||
| Placebo | -10 | ||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Han 2002 | 13 wk | Isoflavones | 70 | 19 | 11 | 100 | 0 | 40 | 134 | -13 | <0.001 | -19 | <0.001 | post♀ |
| A | |
| Placebo | 0 | 40 | 134 | +6 | NS | ||||||||||||
| Squadrito 2002 | 26 wk | Genistein | 54 | 0 | 30 | 139 | 0 | NS | +4 | NS | post♀ |
| A | ||||
| Placebo | 30 | 147 | -4 | NS | |||||||||||||
| Petri 2004 | 26 wk | Soy germ capsules | 60 | 0.8 | 25 | 152 f,g | -20 | <0.05 | -32 | post♀ |
| C h | |||||
| Lactose capsules | 25 | 131 f,g | +12 | NS | |||||||||||||
| Lissin 2004 | 6 wk | Isoflavones | 44 | 44 | 2 | 90 | 0 | 20 | 163 | -1 | NS | +8 | NS | post♀ |
| B | |
| Placebo | 20 | 165 | -9 | <0.05 | |||||||||||||
| Uesugi 2002 | 4 wk | Isoflavones | 0 | 0 | 0 | 62 j | 0 | 12 | 148 | -10 | <0.05 | -12 | NS | peri♀ |
| B | |
| Placebo | 11 | 162 | +2 | NS | |||||||||||||
Or difference between final values, as noted.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Difference of final values (cross-over study)
Final values. No data on baseline or change from baseline.
Significantly different (P<0.05) at baseline.
Graph
In contrast with other outcomes in this article, soy and control ams had significantly different LDL levels.
31 mg daidzin, 7 mg genistin, 21 mg glycitin
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Meinertz 2002 | 4.5 wk | ISP w/Isoflavones, liquid | 2.39/g b | 20% | 12 | 83 | -21 | <0.001 | NS c | -6 | NSc | ♀♂ |
| C | |||
| ISP w/o Isoflavones, liquid | 0.11/g b | 20% | 84 | -18 | <0.01 | -3 | NSc | ||||||||||
| Casein diet, liquid | 79 | -15 | <0.01 | ||||||||||||||
| Meinertz 1989 | 4.5 wk | ISP diet, liquid cholesterol enriched | 112 | 11 | 102 | -32 | -16 | 0.02c | ♀♂ |
| |||||||
| Calcium caseinate | -19 | ||||||||||||||||
| Meinertz 1988 | 4 wk | ISP diet, liquid low cholesterol | 113 | 10 | 110 | -42 | +1 | NSc | ♀♂ |
| C | ||||||
| Casein diet, liquid low cholesterol | -43 | ||||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Wong 1998 d | 5 wk | ISP diet | >15% | 13 | 111 | -13 | -8 | 0.03 | ♂ |
| B | ||||||
| Animal diet | 109 | -5 | |||||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Steinberg 2003 | 6 wk | ISP w/Isoflavones | 55 | 47 | 5 | 107 | 25 | 24 | 112 | -1 | NS | -3 | NS | post♀ |
| C | |
| ISP w/o Isoflavones | 1 | 0.5 | 0.5 | 2 | 24 | -1 | -3 | NS | |||||||||
| Total milk protein | +2 | ||||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Murray 2003 | 26 wk | ISP + Estradiol 0.5 mg | 66 | 44 | 10 | 120 | 38 | 8 | 120 | 0 | NS | +7 | NS | post♀ |
| C | |
| Total milk protein+ Estradiol 0.5 mg | 7 | 125 | -7 | NS | |||||||||||||
| ISP + Estradiol 1.0 mg | 66 | 44 | 10 | 120 | 38 | 8 | 129 | -22 | NS | -26 | |||||||
| Total milk protein + Estradiol 1.0 mg | 7 | 164 e | +4 | NS | |||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Washburn 1999 | 6 wk | ISP w/Isoflavones once daily | 34 | 14 | 42 | 127 | -8 | nd | -9 | <0.05c | peri♀ |
| B | ||||
| ISP w/Isoflavones twice daily | 34 | 14 | -10 | -9 | <0.01c | ||||||||||||
| Carbohydrates | -2 | ||||||||||||||||
| Onning 1998 | 4 wk | Soy milk | 23/30 f | 12 | 112 | -8 | <0.05 | -4 | ♀♂ |
| B | ||||||
| Oat milk | -4 | NS | |||||||||||||||
| Supplement | RCT | Miscellaneous | |||||||||||||||
| Takatsuka 2000 | 9 wk | Soy milk | 109 | 17 | 27 | 91 | -5 | NS | -8 | NS | pre♀ |
| B | ||||
| No supplement | 25 | 94 | +3 | NS | |||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Merz-Demlow 2000 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 13 | 84 g | -5 | <0.05 | -- | pre♀ |
| C | ||
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | +2 | -- | -- | |||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | +4 | -- | -- | |||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Per gram protein. No data on number of grams of protein.
Difference of final values (cross-over study)
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
Non-significantly higher at baseline than other study arms because of single outlier with hypertriglyceridemia.
Women/Men
Measurements made at 4 menstrual cycle phases: early follicular, midfollicular, periovulatory, midluteal. Data extracted for periovulatory only. LDL change was greatest during this phase in high isoflavone group.
| Association Sub-group of studies | No. of Studies a | Association Beta (95% CI) | P-value b | ||||
|---|---|---|---|---|---|---|---|
| Baseline LDL v Net Change LDL | |||||||
| All studies | All studies | 59 | NS | ||||
| LDL >130 mg/dL | 45 | -0.14 (-0.28, +0.00) | 0.06 | ||||
| Isoflavone only | All studies | 11 | NS | ||||
| LDL >130 mg/dL | 11 | NS | |||||
| Protein w/Isoflavones | All studies | 42 c | NS c | ||||
| LDL >130 mg/dL | 30 | -0.10 (-0.23, +0.02) | 0.09 | ||||
| Soy Isoflavone Dose v Net Change LDL | |||||||
| All studies | All studies | 43 | NS | ||||
| LDL >130 mg/dL | 34 | NS | |||||
| Isoflavone only | All studies | 10 | NS | ||||
| LDL >130 mg/dL | 10 | NS | |||||
| Protein w/Isoflavones | All studies | 27 | NS | ||||
| LDL >130 mg/dL | 20 | NS | |||||
| Soy Protein Dose v Net Change LDLc | |||||||
| All studies | All studies | 52 | -0.10 (-0.20, +0.01) | 0.08 | |||
| LDL >130 mg/dL | 41 | -0.14 (-0.27, -0.00) | 0.04 | ||||
| Dose >10 g/day | 43 | NS | |||||
| Protein w/Isoflavones | All studies | 38 | NS | ||||
| LDL >130 mg/dL | 28 | -0.16 (-0.33, +0.02) | 0.08 | ||||
| Dose <80 g/day | 35 | -0.16 (-0.33, +0.02) | 0.07 | ||||
Studies with sufficient data for adjusted linear regression.
All associations were non-significant in multivariate analysis with baseline LDL, isoflavone dose, and soy protein dose.
With all 42 studies, beta = -0.07 (95% CI -0.13, -0.00), P=0.05; however, with the omission of a single study (Puska, 2004), which found a large net change (-15 mg/dL) with a high mean baseline (198 mg/dL), the association was non-significant (P=0.4)
–5)We found 55 studies that reported on the effect of the consumption of soy products on low density lipoprotein (LDL). Of these, 3 reported only that there was no effect on LDL.46–48 The remaining 52 studies are described below. 23, 24, 49– 51, 53, 55–60, 62, 64, 67–94, 96, 98–100, 102–106
The range of daily soy isoflavone intake among studies of soy products with isoflavones was approximately 10 to 185 mg, with a median of 80 mg per day. The range of daily soy protein intake among relevant studies was approximately 14 to 113 g, with a median of 34 g per day. Most studies of hyperlipidemic subjects included post-menopausal women and/or men. The studies of normolipidemic subjects were more likely to include both men and women or be restricted to pre-menopausal women. The large majority of studies were of limited applicability, even within the categories of pre- or post-menopausal women, or men. Only 11 of the studies were graded as being broadly applicable. Among the 52 studies, 4 were rated good quality (A), 28 were rated fair quality (B), and 20 were rated poor quality (C).
). Across
studies, net change ranged from -32 to +13 mg/dL, with the median net
change equal to -5 mg/dL. In terms of percent net change (using the
baseline level in the soy intervention cohort as the denominator), net
change ranged from approximately -21% to +9%, with a median net change
equal to -3%.
The summary estimate from a random-effects model meta-analysis of all soy cohorts with a non-soy control was statistically significant (Figure 4
); the net change in LDL
was -5 (95% confidence interval [CI] -7 to -3) mg/dL, where the adjusted
summary mean baseline LDL was approximately 150 mg/dL. This finding is
in agreement with a recent meta-analysis of the effect of soy products
on LDL and HDL by Weggemans et al. (2003)107, who reported a net change in LDL of -7 (95% CI -10, -3) mg/dL.
Our primary meta-analysis differed in that we used looser criteria for
inclusion - thus we included about 3 times as many studies. Both of
these meta-analyses differ markedly from an early meta-analysis by
Anderson et al. (1995)1 which reported a summary net change of -22 (95 % CI -32, -11)
mg/dL. However, that meta-analysis used much looser inclusion criteria,
including non-randomized trials, studies of children, very small sample
sizes, and short intervention durations. Their findings were highly
affected by several non-randomized trials and the inclusion of Verrillo
198567 (Tables 11 and 12), a study that lacked
randomization to a non-soy control.Analyses across studies
graphs the net change
compared to non-soy control of all cohorts evaluated (that included a
non-soy cohort). In a similar manner as the total cholesterol graph, the
left-most graph displays net change in LDL in relationship the baseline
LDL; this graph includes all studies. The middle graph compares net
change to daily soy isoflavone consumption (among those studies that
report isoflavone content). The right-most graph compares net change to
daily soy protein consumption (among those studies that report soy
protein content). Adjusted regression lines are included.
, left). Across studies,
for each additional 10 grams of soy protein consumed per day, the
average net reduction of LDL was 1.4 (95% CI 0.0, 2.7) mg/dL greater.
Inclusion of studies with lower baseline LDL reduced the degree and
significance of the association.
Furthermore, inclusion either of only studies with soy protein (dose >10 g/day) or only studies of protein with isoflavones yielded no association between protein dose and net change LDL. Excluding either soy protein with isoflavone studies of lower baseline LDL or the few outlier studies of higher protein dose (>80 g/day) yielded similar trends toward an association between higher dose soy protein and greater net reduction of LDL.
Evaluation of the effect of baseline LDL on net change LDL across studies revealed no association; however in sub-group of studies with elevated LDL (within the range of about 135 to 200 mg/dL), there was a trend suggesting that for studies with mean baseline LDL 10 mg/dL higher than other studies, the average net LDL reduction was 1.4 (95% CI 0, 2.8) mg/dL greater. This trend was similar for studies of protein with isoflavones and higher baseline LDL levels, but not for studies of isoflavones alone. No meta-regression analysis found an association between soy isoflavone dose and net change LDL.
These analyses, however, are subject to biases due to analyzing mean values across studies. In addition, these analyses are combining data from a very heterogeneous group of studies in regards to subject populations, soy products, and controls, among other factors. In all analyses, the studies were found to be statistically significantly heterogeneous. Associations may be either exaggerated or minimized because of the heterogeneity.
). All
sub-group meta-analyses were grossly similar. While the studies rated of
poor quality did, on average, find a larger effect of soy products than
better quality studies, the difference in effects was not statistically
significantly different (P=0.2). Of note, the studies
that evaluated soy isoflavones without soy protein had a high degree of
heterogeneity with mean net changes in LDL ranging from -32 mg/dL to +13
mg/dL (or +42 mg/dL in the subset of women with elevated baseline LDL).
Effect of baseline LDL on net change LDL in individual studies
Effect of soy isoflavone dose on net change LDL in individual studies
Meta-analysis of studies comparing 2 or more soy products with different levels of isoflavones
We performed random effects model meta-analyses of the 11 studies that reported sufficient data comparing 2 or more soy products with different levels of isoflavones. 49, 51, 56, 81, 82, 86–88, 98, 102, 106 In one analysis we included all cohorts with soy protein. This analysis is faulty in that it allows comparison of multiple high-isoflavone cohorts to the same low-isoflavone cohort, thus breaking the assumption of independence between all studies. Nevertheless, the summary net difference in effect on LDL was small and non-significant at -3 (95% CI -6, +1) mg/dL. In a second analysis, where we compared only the highest isoflavone dose cohort to the lowest, the net difference was similarly small and non-significant at -3 (95 % CI -8, +1) mg/dL. These results are in contrast to a recently published meta-analysis by Zhuo et al. (2004)43 largely because of the addition of the recently published Gallagher 200487 and Dent 2001,82 which was excluded because of a too long study duration, and Meinertz 2002,98 which might have been excluded because the amount of isoflavones is poorly reported. Importantly, all these studies found no difference in effect.
Effect of soy protein dose on net change LDL in individual studies
Effect of soy as diet vs. supplement on net change LDL in individual studies
) yielded a summary net
change in LDL of -7 (95% CI -9, -4) mg/dL, which was somewhat greater
than the net change among the soy supplements studies, -4 (95% CI -7,
-1) mg/dL.
Effect of sex and menopausal status on net change LDL
See Section 3.2.6.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Potter 1993 | 4 wk | ISP + cellulose | 50 | 24 | 37 | -2 | NS | nd | 0 | NSb | ♂ |
| C | ||||
| ISP + cotyledon | 50 | -2 | NS | 0 | NSb | ||||||||||||
| Soy flour | 0 | 0 | NS | +2 | NSb | ||||||||||||
| Non-fat dry milk+ cellulose | -2 | NS | |||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Wong 1998 | 5 wk | ISP diet | >15% | 13 | 39 | -2 | +3 | NS | ♂ |
| B | ||||||
| Animal diet | 41 | -5 | |||||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Gardner-Thorpe 2003 | 6 wk | Soy flour biscuits | 45 | 75 | 120 | 19 | 39 | +4 | 0 | NSb | ♂ |
| B | ||||
| Wheat flour biscuits | +4 | ||||||||||||||||
| Onning 1998 | 4 wk | Soy milk | 23/30 d | 12 | 42 | 0 | NS | 0 | ♀♂ |
| B | ||||||
| Oat milk | 0 | NS | |||||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Merz-Demlow 2000 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 13 | 48 e | +1 | NS | -- | pre♀ |
| C | ||
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | +1 | -- | ||||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | 0 | -- | ||||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Difference of final values (cross-over study)
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
Women/Men
Measurements made at 4 menstrual cycle phases: early follicular, midfollicular, periovulatory, midluteal. Data extracted for periovulatory only. LDL change was greatest during this phase in high isoflavone group.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | |||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | ||||||||||||||
| Diet | Xover | Dairy | ||||||||||||||||
| Jenkins 2002 12145008 | 4 wk | ISP w/Isoflavones | 73 | 50 | 41 | 51 | -3 | NS | +1 | <0.01 | post♀ ♂ |
| C | |||||
| ISP w/Isoflavones | 10 | 52 | 51 | -2 | +2 | <0.05 | ||||||||||||
| Low fat dairy+egg protein | 51 | -4 | ||||||||||||||||
| Meinertz 2002 | 4.5 wk | ISP w/Isoflavones, liquid | 2.39/g b | 20% | 12 | 61 | -7 | NS | NS c | +3 | <0.05c | ♀♂ |
| C | ||||
| ISP w/o Isoflavones, liquid | 0.11/g b | 20% | 60 | -8 | NS | +2 | NSc | |||||||||||
| Casein diet, liquid | 59 | -10 | <0.05 | |||||||||||||||
| Meinertz 1989 | 4.5 wk | ISP diet, liquid cholesterol enriched | 112 | 11 | 58 | -2 | +9 | <0.01c | ♀♂ |
| C | |||||||
| Calcium caseinate | -10 | |||||||||||||||||
| Meinertz 1988 | 4 wk | ISP diet, liquid low cholesterol | 113 | 10 | 47 | 0 | +1 | NSc | ♀♂ |
| C | |||||||
| Casein diet, liquid low cholesterol | -1 | |||||||||||||||||
| Diet | RCT | Dairy | ||||||||||||||||
| Crouse 1999 | 9 wk | ISP w/Isoflavones | 62 | 25 | 30 | 46 | +1 | NS | +2 | NS | ♀♂ (LDL 140-200) |
| B | |||||
| ISP w/Isoflavones | 37 | 25 | 30 | 45 | +1 | +2 | NS | |||||||||||
| ISP w/Isoflavones | 27 | 25 | 27 | 45 | 0 | +1 | NS | |||||||||||
| ISP w/o Isoflavones | 3 | 25 | 28 | 47 | -1 | 0 | NS | |||||||||||
| Casein | 31 | 45 | -1 | |||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 46 | -1 | NS | +1 | NS | ♀♂ LDL 166–200 | |||||||||
| ISP w/Isoflavones | 37 | 25 | 12 | 48 | -2 | 0 | NS | |||||||||||
| ISP w/Isoflavones | 27 | 25 | 15 | 45 | +2 | +4 | NS | |||||||||||
| ISP w/o Isoflavones | 3 | 25 | 12 | 45 | -2 | 0 | NS | |||||||||||
| Casein | 16 | 48 | -2 | |||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 47 | +2 | <0.04 Trend | +2 | NS | ♀♂ LDL 140–166 | |||||||||
| ISP w/Isoflavones | 37 | 25 | 18 | 42 | +4 | +4 | NS | |||||||||||
| ISP w/Isoflavones | 27 | 25 | 12 | 44 | 0 | 0 | NS | |||||||||||
| ISP w/o Isoflavones | 3 | 25 | 16 | 48 | 0 | 0 | NS | |||||||||||
| Casein | 15 | 41 | 0 | |||||||||||||||
| Teixeira 2000 | 6 wk | ISP w/Isoflavones | 95 | 50 | 15 | 44 | +1 | nd | +1 | NS | ♂ |
| A | |||||
| ISP w/Isoflavones | 76 | 40 | 17 | 43 | +2 | +2 | NS | |||||||||||
| ISP w/Isoflavones | 57 | 30 | 18 | 44 | +2 | +2 | NS | |||||||||||
| ISP w/Isoflavones | 38 | 20 | 15 | 41 | +1 | +1 | NS | |||||||||||
| Calcium caseinate | 16 | 42 | 0 | NS | ||||||||||||||
| Diet | RCT | Dairy | ||||||||||||||||
| Baum 1998 | 24 wk | ISP w/Isoflavones | 35 | 23 | 6 | 90 | 40 | 21 | 53 | +2 | nd | +4 | <0.03 | post♀ |
| B | ||
| ISP w/o Isoflavones | 2 | 1 | 2 | 56 | 40 | 23 | 52 | +3 | +5 | <0.01 | ||||||||
| Casein + non-fat dry milk | 22 | 53 | -2 | |||||||||||||||
| Vigna 2000 | 12 wk | ISP w/Isoflavones | 76 | 40 | 40 | 61 | 0 | NS | +1 | post♀ |
| C | ||||||
| Caseinate | 37 | 62 | -1 | NS | ||||||||||||||
| Van Horn 2001 | 6 wk | ISP w/Isoflavones + oats | 39 | 19 | 19 | 32 | 61 | 0 | NS | +1 | post♀ |
| B | |||||
| Milk + oats | 32 | 67 | -1 | NS | ||||||||||||||
| ISP w/Isoflavones + wheat | 39 | 19 | 19 | 31 | 64 | 0 | NS | +1 | ||||||||||
| Milk + wheat | 32 | 63 | -1 | NS | ||||||||||||||
| Diet | Xover | Animal/Usual | ||||||||||||||||
| Lichtenstein 2002 d | 6 wk | ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 42 | 51 | 0 | NS | -2 | Soy: 0.03 | post♀ |
| B | ||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | 0 | -2 | Iso: NSc | ♂ | |||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | +2 | 0 | |||||||||||
| Animal w/o Isoflavones | +2 | |||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 22 | nd | 55 f | nd | -1c | Soy: NS | post♀ | ||||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | 56 f | 0c | Iso: NSc | ♂ | |||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | 53 f | -3c | LDL>160 | ||||||||||
| Animal w/o Isoflavones | 56 f | |||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71 e | 20 | nd | 49 f | nd | +2c | Soy: 0.04 | post♀ | ||||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71 e | 49 f | +2c | Iso: NSc | ♂ | |||||||||
| Animal w/Isoflavones d | 27 | 21 | 4 | 52 | 0 | 48 f | +1c | LDL<160 | ||||||||||
| Animal w/o Isoflavones | 47 f | |||||||||||||||||
| Ashton 2000 11194529 | 4 wk | ISP diet | 84 | 35 | 120 | 36 | 42 | 48 | 0 | -3 | 0.01c | ♂ |
| C | ||||
| Lean meat diet | +3 | |||||||||||||||||
| Azadbakht 2003 | 7 wk | ISP diet | 19 | 14 | 47 | +3 | <0.05 | +2 | NS | ♀♂ |
| B | ||||||
| Usual diet | 46 | +1 | NS | DM Proteinuria | ||||||||||||||
| Wong 1998 g | 5 wk | ISP diet | >15% | 13 | 41 | -3 | 0 | NS | ♂ |
| B | |||||||
| Animal diet | 41 | -3 | ||||||||||||||||
| Goldberg 1982 | 6 wk | ISP diet | 91 | 12 | 45 | -1 | NS | -1 | NSc | ♀♂ |
| B | ||||||
| Animal diet | 0 | NS | ||||||||||||||||
| Diet | RCT | Animal/Usual | ||||||||||||||||
| Chiechi 2002 11836040 | 26 wk | ISP dieth | 47 | 58/24 j | 52 | -2 | NS | +2 | post♀ |
| C | |||||||
| Usual dieth | 55/43 j | 55 | -4 | <0.05 | ||||||||||||||
| Shorey 1981 | 6 wk | ISP diet | 55 | 13 | 52 | -8 | 0.001 | -4 | ♂ |
| C | |||||||
| Animal diet | 11 | 46 | -4 | NS | ||||||||||||||
| Diet | Xover | Miscellaneous | ||||||||||||||||
| Jenkins1999 | 4 wk | ISP diet | 33 | 31 | 49 | -3 | 0 | NS | post♀ |
| B | |||||||
| Vegetarian diet | 48 | -3 | ♂ | |||||||||||||||
| Diet | RCT | Miscellaneous | ||||||||||||||||
| Murkies 1995 | 12 wk | Soy flour | 40 g of flour per day | 23 | 67 | -2 | NS | 0 | NS | post♀ |
| B | ||||||
| Wheat flour | 24 | 63 | -2 | NS | ||||||||||||||
| Sagara 2004 | 5 wk | Soy powder baked goods | 80 | 20 | 25 | 54 | +4 | <0.01 | -1 | 0.06 | ♂ |
| B | |||||
| Usual baked goods | 25 | 55 | +5 | <0.01 | ||||||||||||||
| Diet | RCT | No Control | ||||||||||||||||
| Verrillo 1985 k | 16 wk | ISP, supplement k | 31 | 38 | 54 | -4 | nd | -- | ♀♂ |
| B | |||||||
| ISP, 60 g replacing dietary protein | 31 | 19 | 50 | 4 | -- | |||||||||||||
| No control group | -- | |||||||||||||||||
Or difference between final values, as noted.
Per gram protein. No data on number of grams of protein.
Difference of final values (cross-over study)
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Final values. No data on baseline or change from baseline.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
No data on how fat content of 2 diets compare.
N: baseline/final.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Bricarello 2004 | 6 wk | Soy milkb | 50 | 33 | 5 | 87 | 25 | 60 | 58 | +4 | <0.05 | +5 | <0.05 | ♀♂ |
| C | |
| Non-fat milkb | -1 | ||||||||||||||||
| Kurowska 1997 | 4 wk | Soy milk | 31 | 34 | 51 | +3 | +3 | 0.04 | ♀♂ |
| B | ||||||
| Milk, 2% fat | 0 | ||||||||||||||||
| Blum 2003 12659466 | 6 wk | ISP w/Isoflavones | 85 | 25 | 24 | 60 | -1 | NS | -3 | NSd | post♀ |
| C | ||||
| Total milk protein | +2 | ||||||||||||||||
| Steinberg 2003 | 6 wk | ISP w/Isoflavones | 55 | 47 | 5 | 107 | 25 | 24 | 60 | -2 | NS | -4 | NS | post♀ |
| C | |
| ISP w/o Isoflavones | 1 | 0.5 | 0.5 | 2 | 24 | 0 | -2 | NS | |||||||||
| Total milk protein | +2 | ||||||||||||||||
| Meyer 2004 | 5 wk | Soy milk/yogurt | 80 | 30 | 23 | 49 | +3 | +2 | NSd | post♀ |
| B | |||||
| Low fat milk/yogurt | 0 | ♂ | |||||||||||||||
| Bakhit 1994 | 4 wk | ISP + cellulose | 25 | 21 | 51 | -1 | NS | 0 | NSc | ♂ |
| C | |||||
| ISP + cotyledon | 25 | -1 | NS | +1 | |||||||||||||
| Casein + cellulose | -1 | NS | |||||||||||||||
| Casein + cotyledon | -2 | NS | |||||||||||||||
| ISP + cellulose | 25 | 11 | 53 | -3 | NS | -3 | NSc | ♂ | |||||||||
| ISP + cotyledon | 25 | -3 | NS | 0 | TC>220 | ||||||||||||
| Casein + cellulose | 0 | NS | |||||||||||||||
| Casein + cotyledon | -3 | NS | |||||||||||||||
| Hermansen 2001 | 6 wk | ISP w/Isoflavones | >165 | 50 | 20 | 51 | +3 | +1 | NS | ♀♂ |
| C | |||||
| Casein | 49 | +2 | DM | ||||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Cuevas 2003 | 8 wk | ISP w/Isoflavones | 48 | 24 | 8 | 80 | <40 | 18 | 54 | -1 | NS | +2 | NSd | post♀ |
| B | |
| Caseinate | -3 | <0.05 | |||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Teede 2001 | 13 wk | ISP w/Isoflavones | 77 | 38 | 5 | 118 | 40 | 86 | 56 | -2 | NS | +2 | NS | post♀ |
| B | |
| Casein | 93 | 58 | -4 | <0.05 | ♂ | ||||||||||||
| Kreijkamp-Kaspers 2004 | 52 wk | ISP w/Isoflavones | 52 | 41 | 6 | 26 | 88e | 59 | 0 | +2 | 0.09 | post♀ |
| A | |||
| Total milk protein | 87e | 59 | -2 | ||||||||||||||
| Puska 2004 | 8 wk | ISP w/Isoflavones | 153 | 41 | 69 | 64 | +2 | 0 | NS | post♀ |
| B | |||||
| Yogurt | 74 | 66 | +2 | ♂ | |||||||||||||
| Gardner 2001 | 12 wk | ISP w/Isoflavones | 52 | 25 | 4 | 80 | 42 | 31 | 58 | +4 | NS | +4 | NS | post♀ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 0 | 3 | 42 | 33 | 54 | +4 | +4 | NS | |||||||
| Milk protein | 30 | 58 | 0 | ||||||||||||||
| Tonstad 2002 | 16 wk | ISP w/Isoflavones | 185 | 50 | 31 | 52 | +5 | nd | +2 | NS | post♀ |
| B | ||||
| ISP w/Isoflavones | 111 | 30 | 34 | 57 | +3 | ♂ | |||||||||||
| Casein, 50 g | 29 | 51 | 0 | ||||||||||||||
| Casein 30 g | 36 | 49 | +3 | ||||||||||||||
| Dent 2001 | 24 wk | ISP w/Isoflavones | 80 | 40 | 24 | 58 f,g | -5 g | NS | -3g | NS | peri♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 40 | 24 | 56 f,g | -1 g | +1g | |||||||||||
| Whey protein | 21 | 57 f,g | -2 g | ||||||||||||||
| Puska 2002 | 6 wk | ISP w/Isoflavones | 96 | 52 | 24 | 61 | +3 | 0 | NS | post♀ |
| C | |||||
| Calcium caseinate | 28 | 60 | +3 | ♂ | |||||||||||||
| Murray 2003 | 26 wk | ISP + Estradiol 0.5 mg | 66 | 44 | 10 | 120 | 38 | 8 | 64 | +4 | NS | 0 | NS | post♀ |
| C | |
| Total millk protein + Estradiol 0.5 mg | 7 | 67 | +4 | NS | |||||||||||||
| ISP + Estradiol 1.0 mg | 66 | 44 | 10 | 120 | 38 | 8 | 69 | -5 | NS | +3 | |||||||
| Total millk protein + | 7 | 62 | -2 | NS | |||||||||||||
| Estradiol 1.0 mg | |||||||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Washburn 1999 | 6 wk | ISP w/Isoflavones once daily | 34 | 14 | 42 | 55 | -3 | nd | -1 | NS d | peri♀ |
| B | ||||
| ISP w/Isoflavones twice daily | 34 | 14 | -3 | -1 | NS d | ||||||||||||
| Carbohydrates | -2 | ||||||||||||||||
| Jayagopal 2002 | 12 wk | ISP w/Isoflavones | 70 | 49 | 13 | 132 | 30 | 31 | 51 | +0.4 | NS | 0 | NS | post♀ |
| B | |
| Cellulose | 50 | +0.6 | NS | DM | |||||||||||||
| Supplement | RCT | Miscellaneous | |||||||||||||||
| Takatsuka 2000 | 9 wk | Soy milk | 109 | 17 | 27 | 66 | -2 | NS | -4 | pre♀ |
| B | |||||
| No supplement | 25 | 63 | +2 | NS | |||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Wangen 2001 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 18 | 52 | +1 | NS | NS d | -- | post♀ |
| C | |
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | 18/17 h | +1 | NS | -- | ||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | 18 | 0 | NS | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Supplement | RCT | No Control | |||||||||||||||
| Gallagher 2004 | 39 wk | ISP w/Isoflavones | 52 | 28 | 96 | 40 | 17 | 53 | -2 | <0.05 | nd | -- | post♀ |
| C | ||
| ISP w/Isoflavones | 28 | 20 | 52 | 40 | 19 | 52 | -5 | <0.05 | -- | ||||||||
| ISP w/o Isoflavones | 4 | 0 | 4 | 40 | 14 | 52 | -4 | <0.05 | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Mackey 2000 | 12 wk | ISP w/Isoflavones | 65 | 28 | 25 | 59 | 0 | NS | NS | -- | post♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 28 | 24 | 64 | +2 | -- | |||||||||||
| Female study | No control group | -- | |||||||||||||||
| Verrillo 1985 j | 16 wk | ISP, supplement | 31 | 38 | 54 | -4 | nd | -- | ♀♂ |
| B | ||||||
| ISP, 60 g replacing dietary protein j | 31 | 19 | 50 | +4 | -- | ||||||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Unequal amounts of fat in soy milk (17.5 g/day) and cow milk (0 g/day).
Main effect of soy protein
Difference of final values (cross-over study)
Intention-to-treat analysis (75 completed soy protocol, 78 completed control protocol)
Graph
Median value, difference or net difference of median values.
N: baseline/final.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Baseline HDL > 50 (Women) / 40 (Men) | |||||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002 b | 6 wk | ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 42 | 51 | 0 | NS | -2 | Soy: 0.03 Iso: NS d | post♀♂ |
| B | |
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 0 | -2 | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | +2 | 0 | ||||||||||
| Animal w/o Isoflavones | +2 | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 22 | nd | 55e | nd | -1d | Soy: NS Iso: NS d | post♀ ♂ LDL>160 | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 56e | 0d | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 53e | -3d | ||||||||||
| Animal w/o Isoflavones | 56e | ||||||||||||||||
| ISP w/Isoflavones b | 27 | 14 | 5 | 46 | 55/71 c | 20 | nd | 49e | nd | +2d | Soy: 0.04 Iso: NS d | post♀ ♂ LDL<160 | |||||
| ISP w/o Isoflavones b | 0.4 | 0.8 | 0 | 1.3 | 55/71 c | 49e | +2d | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 48e | +1d | ||||||||||
| Animal w/o Isoflavones | 47e | ||||||||||||||||
| Supplement | Xover | Placebo | |||||||||||||||
| Nikander 2004 15240647 | 13 wk | Isoflavones | 6 | 42 | 66 | 0 | 56 | 69 | -1 | NS | -1 | post♀ Breast CA |
| A | |||
| Placebo | 68 | 0 | NS | ||||||||||||||
| Nestel 1997 | 5 wk | Isoflavones | 45 | 35 | 3 | 80 | 0 | 21 | 50 | -3 | NS | -3 | post♀ |
| C | ||
| Placebo | 0 | NS | |||||||||||||||
| Simons 2000 | 8 wk | Isoflavones | 80 | 0 | 20 | 55 | -3 | 0 | NS d | post♀ |
| B | |||||
| Placebo | -3 | ||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Uesugi 2002 | 4 wk | Isoflavones | 0 | 0 | 0 | 62 f | 0 | 12 | 66 | -1 | NS | -4 | NS | peri♀ |
| B | |
| Placebo | 11 | 65 | +3 | NS | |||||||||||||
| Uesugi 2003 | 13 wk | Isoflavones | 0 | 0 | 0 | 62 f | 0 | 11 | 65 | -1 | -1 | NS | post♀ |
| B | ||
| Dextrin | 10 | 71 | 0 | ||||||||||||||
| Baseline HDL < 50 (Women) / 40 (Men) | |||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Han 2002 | 13 wk | Isoflavones | 70 | 19 | 11 | 100 | 0 | 40 | 40 | +4 | <0.05 | 0 | NS | post♀ |
| A | |
| Placebo | 40 | 40 | +4 | <0.05 | |||||||||||||
| Squadrito 2002 | 26 wk | Genistein | 54 | 0 | 30 | 46 | 0 | NS | -4 | NS | post♀ |
| A | ||||
| Placebo | 30 | 46 | +4 | NS | |||||||||||||
| Petri 2004 | 26 wk | Soy germ capsules | 60 | 0.8 | 25 | 44 g | +13 | <0.05 | +13 | post♀ |
| B | |||||
| Lactose capsules | 25 | 48 g | 0 | NS | |||||||||||||
| Dewell 2002 | 9 wk | Isoflavones | 40 | 50 | 90 | 0 | 20 | 46 | -8 | 0 | NS | post♀ |
| C | |||
| Placebo | 16 | 46 | -8 | ||||||||||||||
Or difference between final values, as noted.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Difference of final values (cross-over study)
Final values. No data on baseline or change from baseline.
31 mg daidzin, 7 mg genistin, 21 mg glycitin
Graph
–7)We found 61 studies that reported on the effect of the consumption of soy products on high density lipoprotein (HDL). Of these, 5 reported only that there was no effect on HDL.45–48, 73 The remaining 56 studies are described below.23, 24, 49–60, 62–72, 74, 76–93, 95–100, 102–106
The range of daily soy isoflavone intake among studies of soy products with isoflavones was approximately 10 to 185 mg, with a median of 80 mg per day. The range of daily soy protein intake among relevant studies was approximately 14 to 113 g, with a median of 39 g per day. Three of the 5 studies of subjects with, on average, abnormal HDL were restricted to men; 1 included both men and women; 1 included only pre-menopausal women. The large majority of studies had, on average, subjects with normal HDL. These studies mostly included post-menopausal women or both men and post-menopausal women. The large majority of studies were of limited applicability, even within the categories of pre- or post-menopausal women, or men. Only 12 of the studies were graded as being broadly applicable. Among the 56 studies, 5 were rated good quality (A), 29 were rated fair quality (B), and 20 were rated poor quality (C).
). Random
effects model meta-analysis of non-soy-controlled studies with
sufficient data (Figure 7)
estimated a summary net change of +0.6 (95% CI -0.5, +1.8) mg/dL.Analyses across studies
graphs the net change
compared to non-soy control of all cohorts evaluated (that included a
non-soy cohort). The left-most graph displays net change in HDL in
relationship the baseline HDL level; this graph includes all studies.
The middle graph compares net change to daily soy isoflavone consumption
(among those studies that report isoflavone content). The right-most
graph compares net change to daily soy protein consumption (among those
studies that report soy protein content). Study cohorts who consumed soy
with isoflavones are indicated by squares; cohorts who consumed soy
without isoflavones are indicated by circles; and cohorts who consumed
soy isoflavone without soy protein are indicated by triangles. Black
symbols represent cohorts where the soy product was consumed as part of
the regular diet; open symbols represent cohorts where the soy product
was consumed as a supplement to the diet (including soy milk products).
Larger symbols indicate cohorts whose mean HDL was abnormally low at
baseline; smaller symbols indicate cohorts with normal HDL at baseline.
Visual inspection of the graphs reveal no difference across studies in net effect on HDL based on baseline HDL levels, amount of soy isoflavones consumed, or amount of soy protein consumed. Likewise, there are no clear differences across studies based on whether the soy products were consumed as part of the diet (i.e., specifically replacing other sources of protein) or as a supplement (consumed in addition to regular diet). Also comparing the net effects of soy protein with isoflavones to soy protein without isoflavones to isoflavones without soy protein reveals similar ranges of effects for all 3 types of products.
Random effects model meta-regression across all studies yielded a statistically significant association between mean baseline HDL and net change HDL, such that each increase in baseline HDL of 1 mg/dL was associated with an additional net increase 0.12 (95% CI 0.07, 0.18) mg/dL HDL (P=0.00002). However, exclusion of the 2 studies with atypically small standard deviations (Kreijkamp-Kaspers 2004 and Onning 1998) yielded a non-significant association (P=0.6). Meta-regression of both isoflavone and soy protein dose revealed no association across studies with net change HDL.
). However, again, the apparent differences in
effect are largely driven by the 2 studies with atypically small
standard deviations (Kreijkamp-Kaspers 2004 and Onning 1998). No
difference in effect size was found comparing low quality studies (rated
C) to studies of higher quality.
Effect of baseline level of abnormal lipids on net change HDL in individual studies
Effect of soy isoflavone dose on net change HDL in individual studies
Effect of soy protein dose on net change HDL in individual studies
Effect of soy as diet vs. supplement on net change HDL in individual studies
Effect of sex and menopausal status on net change HDL
See Section 3.2.6.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs | ||||||||
| Year | Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | Control | |||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Jenkins 2002 12145008 | 4 wk | ISP w/Isoflavones | 73 | 50 | 41 | 173 | +7 | NS | -3 | NS | post♀ |
| C | ||||
| ISP w/Isoflavones | 10 | 52 | 163 | -+12 | -22 | NS | ♂ | ||||||||||
| Low fat dairy+egg protein | 173 | +10 | |||||||||||||||
| Potter 1993 | 4 wk | ISP + cellulose | 50 | 23 | 173 | +4 | NS | nd | -26 | NS b | ♂ |
| C | ||||
| ISP + cotyledon | 50 | +4 | NS | -26 | NS b | ||||||||||||
| Soy flour | 0 | +15 | NS | -15 | NS b | ||||||||||||
| Non-fat dry milk + cellulose | +30 | NS | |||||||||||||||
| Diet | RCT | Dairy | |||||||||||||||
| Crouse 1999 | 9 wk | ISP w/Isoflavones | 62 | 25 | 30 | 153 | 0 | NS | -14 | NS | ♀♂ |
| B | ||||
| ISP w/Isoflavones | 37 | 25 | 30 | 150 | -5 | -19 | NS | (LDL 140-200) | |||||||||
| ISP w/Isoflavones | 27 | 25 | 27 | 166 | -12 | -26 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 28 | 144 | +8 | -6 | NS | ||||||||||
| Casein | 31 | 153 | +14 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 146 | -7 | NS | -37 | <0.03 | ♀♂ | ||||||||
| ISP w/Isoflavones | 37 | 25 | 12 | 147 | -2 | -32 | NS | LDL 166–200 | |||||||||
| ISP w/Isoflavones | 27 | 25 | 12 | 162 | -16 | -46 | <0.03 | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 15 | 156 | +15 | -15 | NS | ||||||||||
| Casein | 16 | 141 | +30 | ||||||||||||||
| ISP w/Isoflavones | 62 | 25 | 15 | 161 | +5 | NS | +8 | NS | ♀♂ | ||||||||
| ISP w/Isoflavones | 37 | 25 | 18 | 151 | -7 | -4 | NS | LDL 140–166 | |||||||||
| ISP w/Isoflavones | 27 | 25 | 12 | 171 | -13 | -10 | NS | ||||||||||
| ISP w/o Isoflavones | 3 | 25 | 16 | 136 | +2 | +5 | NS | ||||||||||
| Casein | 15 | 166 | -3 | ||||||||||||||
| Diet | RCT | Dairy | |||||||||||||||
| Teixeira 2000 | 6 wk | ISP w/Isoflavones | 95 | 50 | 15 | 193 | +32 | nd | +16 | NS | ♂ |
| A | ||||
| ISP w/Isoflavones | 76 | 40 | 17 | 181 | +7 | -9 | NS | ||||||||||
| ISP w/Isoflavones | 57 | 30 | 18 | 223 | -22 | -38 | NS | ||||||||||
| ISP w/Isoflavones | 38 | 20 | 15 | 157 | +14 | -2 | NS | ||||||||||
| Calcium caseinate | 16 | 189 | +16 | NS | |||||||||||||
| Baum 1998 | 24 wk | ISP w/Isoflavones | 35 | 23 | 6 | 90 | 40 | 21 | 154 | 0 | nd | -1 | NS | post♀ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 2 | 56 | 40 | 23 | 167 | -14 | -15 | NS | |||||||
| Casein+non-fat dry milk | 22 | 155 | +1 | ||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Ashton 2000 11194529 | 4 wk | ISP diet | 84 | 35 | 120 | 36 | 42 | 173 | -30 | -13 | 0.02 b,c | ♂ |
| C | |||
| Lean meat diet | -17 | ||||||||||||||||
| Azadbakht 2003 | 7 wk | ISP diet | 19 | 14 | 243 | -10 | <0.05 | -13 | <0.002 | ♀♂ DM Proteinuria |
| B | |||||
| Usual diet | 241 | +3 | <0.05 | ||||||||||||||
| Wong 1998 d | 5 wk | ISP diet | >15% | 13 | 210 | +56 | +66 | NS | ♂ | B | |||||||
| Animal diet | 260 | -10 | |||||||||||||||
| Diet | Xover | Miscellaneous | |||||||||||||||
| Jenkins 1999 | 4 wk | ISP diet | 33 | 31 | 158 | -11 | -25 | 0.07 | post♀ |
| B | ||||||
| Vegetarian diet | 158 | +14 | ♂ | ||||||||||||||
| Diet | RCT | No Control | |||||||||||||||
| Verrillo 1985 e | 16 wk | ISP, supplement e | 31 | 38 | 195 | -35 | NS | -- | ♀♂ |
| B | ||||||
| ISP, 60 g replacing dietary protein | 31 | 19 | 150 | -18 | -- | ||||||||||||
| No control group | -- | ||||||||||||||||
Or difference between final values, as noted.
Difference of final values (cross-over study)
However 95% confidence interval of net change crosses 0.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Kurowska 1997 | 4 wk | Soy milk | 31 | 34 | 155 | +17 | +8 | NS | ♀♂ |
| B | ||||||
| Milk, 2% fat | +9 | ||||||||||||||||
| Bakhit 1994 | 4 wk | ISP + cellulose | 25 | 21 | 152 | -21 | NS | -8 | NSb | ♂ |
| C | |||||
| ISP + cotyledon | 25 | -4 | NS | +9 | |||||||||||||
| Casein + cellulose | -13 | NS | |||||||||||||||
| Casein + cotyledon | -13 | NS | |||||||||||||||
| ISP + cellulose | 25 | 11 | 175 | -49 | <0.05 | -20 | NSb | ♂ TC>220 | |||||||||
| ISP + cotyledon | 25 | -16 | NS | +5 | |||||||||||||
| Casein + cellulose | -29 | NS | |||||||||||||||
| Casein + cotyledon | -21 | NS | |||||||||||||||
| Hermansen 2001 | 6 wk | ISP w/Isoflavones | >165 | 50 | 20 | 150 | -6 | +14 | 0.04c | ♀♂ DM |
| C | |||||
| Casein | 150 | +8 | |||||||||||||||
| Cuevas 2003 | 8 wk | ISP w/Isoflavones | 48 | 24 | 8 | 80 | <40 | 18 | 190 | -55 | <0.05 | -25 | NSd | post♀ |
| B | |
| Caseinate | -30 | NS | |||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Puska 2002 | 6 wk | ISP w/Isoflavones | 96 | 52 | 21 | 156 | -12 | +5 | NS | post♀ ♂ |
| C | |||||
| Calcium caseinate | 28 | 160 | -17 | ||||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Jayagopal 2002 | 12 wk | ISP w/Isoflavones | 70 | 49 | 13 | 132 | 30 | 32 | 195 | -3 | NS | -8 | NS | post♀ DM |
| B | |
| Cellulose | 193 | +5 | NS | ||||||||||||||
| Supplement | RCT | No Control | |||||||||||||||
| Verrillo 1985e | 16 wk | ISP, supplement | 31 | 38 | 195 | -35 | NS | -- | ♀♂ |
| B | ||||||
| ISP, 60 g replacing dietary proteine | 31 | 19 | 150 | -18 | -- | ||||||||||||
| No control group | -- | ||||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Main effect of soy protein
NS for difference between final values.
Difference of final values (cross-over study)
Verrillo: In both diet and supplement tables.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Tg <150 mg/dL | |||||||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002b | 6 wk | ISP w/Isoflavonesb | 27 | 14 | 5 | 46 | 55/71c | 42 | 136d | -24 | <0.05 | -15 | Soy:<0.0001 Iso: NSe | post♀ ♂ |
| B | |
| ISP w/o Isoflavonesb | 0.4 | 0.8 | 0 | 1.3 | 55/71c | -24 | -15 | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | -7 | +2 | ||||||||||
| Animal w/o Isoflavones | -9 | ||||||||||||||||
| ISP w/Isoflavonesb | 27 | 14 | 5 | 46 | 55/71c | 22 | nd | 119d,f | nd | -8e | Soy:<0.0001 Iso: NSe | post♀ ♂ LDL>160 | |||||
| ISP w/o Isoflavonesb | 0.4 | 0.8 | 0 | 1.3 | 55/71c | 110d,f | -17e | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 126d,f | -1e | ||||||||||
| Animal w/o Isoflavones | 127d,f | ||||||||||||||||
| ISP w/Isoflavonesb | 27 | 14 | 5 | 46 | 55/71c | 20 | nd | 107d,f | nd | -21e | Soy: 0.01 Iso: NSe | post♀ ♂ LDL<160 | |||||
| ISP w/o Isoflavonesb | 0.4 | 0.8 | 0 | 1.3 | 55/71c | 114d,f | -14e | ||||||||||
| Animal w/Isoflavones | 27 | 21 | 4 | 52 | 0 | 132d,f | +4e | ||||||||||
| Animal w/o Isoflavones | 128d,f | ||||||||||||||||
| Supplement | Xover | Placebo | |||||||||||||||
| Nikander 2004 15240647 | 13 wk | Isoflavones | 6 | 42 | 66 | 0 | 56 | 108 | +1 | NS | +1 | post♀ Breast CA |
| A | |||
| Placebo | 111 | 0 | NS | ||||||||||||||
| Nestel 1997 | 5 wk | Isoflavones | 45 | 35 | 3 | 80 | 0 | 21 | 128 | +7 | NS | +20 | post♀ |
| C | ||
| Placebo | -13 | NS | |||||||||||||||
| Simons 2000 | 8 wk | Isoflavones | 80 | 0 | 20 | 99 | 0 | +4 | NSe | post♀ |
| B | |||||
| Placebo | -4 | ||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Squadrito 2002 | 26 wk | Genistein | 54 | 0 | 30 | 133 | +27 | NS | +36 | NS | post♀ |
| A | ||||
| Placebo | 30 | 150 | -9 | NS | |||||||||||||
| Petri 2004 | 26 wk | Soy germ capsules | 60 | 0.8 | 25 | 150g | 0 | NS | -4 | post♀ |
| B | |||||
| Lactose capsules | 25 | 139g | +4 | NS | |||||||||||||
| Dewell 2002 | 26 wk | Isoflavones | 40 | 50 | 90 | 0 | 20/17h | 71 | +9 | 0 | NS | post♀ |
| C | |||
| Placebo | 16 | 115 | +9 | ||||||||||||||
| Uesugi 2002 | 4 wk | Isoflavones | 0 | 0 | 0 | 62j | 0 | 12 | 95 | +11 | NS | +19 | NS | peri♀ |
| B | |
| Placebo | 11 | 105 | -8 | NS | |||||||||||||
| Uesugi 2003 | 13 wk | Isoflavones | 0 | 0 | 0 | 62j | 0 | 11 | 127 | -14 | -32 | NS | post♀ |
| B | ||
| Dextrin | 10 | 118 | +18 | ||||||||||||||
| Tg >=150 mg/dL | |||||||||||||||||
| Supplement | RCT | Placebo | |||||||||||||||
| Han 2002 | 13 wk | Isoflavones | 70 | 19 | 11 | 100 | 0 | 40 | 204 | +7 | <0.05 | -3 | NS | post♀ |
| A | |
| Placebo | 40 | 176 | +10 | <0.05 | |||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Values log-transformed prior to statistical analysis.
Difference of final values (cross-over study)
Final values. No data on baseline or change from baseline.
Graph
N: baseline/final.
31 mg daidzin, 7 mg genistin, 21 mg glycitin
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Diet | Xover | Dairy | |||||||||||||||
| Meinertz 2002 | 4.5 wk | ISP w/Isoflavones, liquid | 2.39/gb | 20% | 12 | 88 | -15 | NS | <0.05c | -3 | NSc | ♀♂ |
| C | |||
| ISP w/o Isoflavones, liquid | 0.11/gb | 20% | 81 | -23 | <0.01 | -11 | NSc | ||||||||||
| Casein diet, liquid | 77 | -12 | NS | ||||||||||||||
| Meinertz 1989 | 4.5 wk | ISP diet, liquid cholesterol enriched | 112 | 11 | 58 | -4 | +1 | NSc | ♀♂ |
| C | ||||||
| Calcium caseinate | -5 | ||||||||||||||||
| Meinertz 1988 | 4 wk | ISP diet, liquid low cholesterol | 113 | 10 | 68 | -5 | +7 | NSc | ♀♂ |
| C | ||||||
| Casein diet, liquid low cholesterol | -12 | ||||||||||||||||
| Diet | RCT | Dairy | |||||||||||||||
| Vigna 2000 | 12 wk | ISP w/Isoflavones | 76 | 40 | 40 | 130 | -14 | NS | -2 | post♀ |
| C | |||||
| Caseinate | 37 | 117 | -12 | NS | |||||||||||||
| Diet | Xover | Animal/Usual | |||||||||||||||
| Lichtenstein 2002d | 6 wk | ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71e | 42 | 136f | -24 | <0.05 | -15 | Soy:<0.0001 Iso: NSc | post♀ ♂ |
| B | |
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71e | -24 | -15 | ||||||||||
| Animal w/Isoflavonesd | 27 | 21 | 4 | 52 | 0 | -7 | +2 | ||||||||||
| Animal w/o Isoflavones | -9 | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71e | 22 | nd | 119f,g | nd | -8c | Soy:<0.0001 Iso: NSe | post♀ ♂ LDL>160 | |||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71e | 110f,g | -17c | ||||||||||
| Animal w/Isoflavonesd | 27 | 21 | 4 | 52 | 0 | 126f,g | -1c | ||||||||||
| Animal w/o Isoflavones | 127f,g | ||||||||||||||||
| ISP w/Isoflavones | 27 | 14 | 5 | 46 | 55/71e | 20 | nd | 107f,g | nd | -21c | Soy: 0.01 Iso: NSe | post♀ ♂ LDL<160 | |||||
| ISP w/o Isoflavones | 0.4 | 0.8 | 0 | 1.3 | 55/71e | 114f,g | -14c | ||||||||||
| Animal w/Isoflavonesd | 27 | 21 | 4 | 52 | 0 | 132f,g | +4c | ||||||||||
| Animal w/o Isoflavones | 128f,g | ||||||||||||||||
| Wong 1998h | 5 wk | ISP diet | >15% | 13 | 89 | +9 | +16 | NS | ♂ |
| B | ||||||
| Animal diet | 92 | -7 | |||||||||||||||
| Goldberg 1982 | 6 wk | ISP diet | 91 | 12 | 116 | -12 | NS | +1 | NSc | ♀♂ |
| B | |||||
| Animal diet | -13 | <0.05 | |||||||||||||||
| Diet | RCT | Animal/Usual | |||||||||||||||
| Chiechi 2002 11836040 | 26 wk | ISP dietj | 47 | 58/24k | 120 | -3 | NS | -7 | post♀ |
| C | ||||||
| Usual dietj | 55/43k | 100 | +4 | NS | |||||||||||||
| Shorey 1981 | 6 wk | ISP diet | 55 | 13 | 97 | +40 | 0.046 | +28 | ♂ |
| C | ||||||
| Animal diet | 11 | 130 | +12 | NS | |||||||||||||
| Diet | RCT | Miscellaneous | |||||||||||||||
| Murkies 1995 | 12 wk | Soy flour | 23 | 95 | -2 | NS | -6 | NS | post♀ |
| B | ||||||
| Wheat flour | 24 | 96 | +4 | NS | |||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Per gram protein. No data on number of grams of protein.
Difference of final values (cross-over study)
Lichtenstein: In both Isoflavone and ISP tables.
Women/Men
Values log-transformed prior to statistical analysis.
Final values. No data on baseline or change from baseline.
Wong 1998: Sub-analyses of same study in LDL<130 and LDL>130 tables.
No data on how fat content of 2 diets compare.
N: baseline/final.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
| Genistein | Daidizein | Glycitein | T Isoflav | Soy Protein | |||||||||||||
| Supplement | Xover | Dairy | |||||||||||||||
| Bricarello 2004 | 6 wk | Soy milkb | 50 | 33 | 5 | 87 | 25 | 60 | 136 | -3 | NS | +2 | NS | ♀♂ |
| C | |
| Non-fat milkb | -1 | NS | |||||||||||||||
| Blum 2003 12659466 | 6 wk | ISP w/Isoflavones | 85 | 25 | 24 | 133 | +60 | 0.04 | -2 | NSc | post♀ |
| C | ||||
| Total milk protein | +62 | ||||||||||||||||
| Steinberg 2003 | 6 wk | ISP w/Isoflavones | 55 | 47 | 5 | 107 | 25 | 24 | 91 | +1 | NS | +5 | NS | post♀ |
| C | |
| ISP w/o Isoflavones | 1 | 0.5 | 0.5 | 2 | 24 | +4 | +8 | NS | |||||||||
| Total milk protein | -4 | ||||||||||||||||
| Meyer 2004 | 5 wk | Soy milk/yogurt | 80 | 30 | 23 | 115 | -7 | -2 | NSc | post♀ ♂ |
| B | |||||
| Low fat milk/yogurt | -4 | ||||||||||||||||
| Supplement | RCT | Dairy | |||||||||||||||
| Teede 2001 | 13 wk | ISP w/Isoflavones | 77 | 38 | 5 | 118 | 40 | 86 | 106 | -17 | <0.05 | -16 | NS | post♀ ♂ |
| B | |
| Casein | 93 | 106 | -1 | NS | |||||||||||||
| Kreijkamp-Kaspers 2004 | 52 wk | ISP w/Isoflavones | 52 | 41 | 6 | 26 | 88d | 120 | +2 | -8 | NS | post♀ |
| A | |||
| Total milk protein | 87d | 111 | +10 | ||||||||||||||
| Puska 2004 | 8 wk | ISP w/Isoflavones | 153 | 41 | 69 | 149 | +9 | 0 | NS | post♀ ♂ |
| B | |||||
| Yogurt | 74 | 148 | +9 | ||||||||||||||
| Tonstad 2002 | 16 wk | ISP w/Isoflavones | 185 | 50 | 31 | 118 | -18 | nd | -2 | NS | post♀ ♂ |
| B | ||||
| ISP w/Isoflavones | 111 | 50 | 34 | 113 | -14 | ||||||||||||
| Casein, 50 g | 29 | 129 | -9 | ||||||||||||||
| Casein 30 g | 36 | 139 | -15 | ||||||||||||||
| Gardner 2001 | 12 wk | ISP w/Isoflavones | 52 | 25 | 4 | 80 | 42 | 31 | 115 | 0 | NS | -9 | NS | post♀ ♂ |
| B | |
| ISP w/o Isoflavones | 2 | 1 | 0 | 3 | 42 | 33 | 115 | 0 | -9 | NS | |||||||
| Milk protein | 30 | 115 | +9 | ||||||||||||||
| Dent 2001 | 24 wk | ISP w/Isoflavones | 80 | 40 | 24 | 94e,f | +11f | NS | -33f | NS | peri♀ |
| B | ||||
| ISP w/o Isoflavones | 4 | 40 | 24 | 99e,f | -5f | -49f | |||||||||||
| Whey protein | 21 | 72e,f | +44f | ||||||||||||||
| Murray 2003 | 26 wk | ISP + Estradiol 0.5 mg | 66 | 44 | 10 | 120 | 38 | 8 | 116 | +3 | NS | +15 | post♀ |
| C | ||
| Total millk protein+ Estradiol 0.5 mg | 7 | 99 | -12 | NS | |||||||||||||
| ISP + Estradiol 1.0 mg | 66 | 44 | 10 | 120 | 38 | 8 | 97 | +53 | 0.02 | -14 | |||||||
| Total millk protein + Estradiol 1.0 mg | 7 | 191g | +67 | NS | |||||||||||||
| Supplement | Xover | Miscellaneous | |||||||||||||||
| Washburn 1999 | 6 wk | ISP w/Isoflavones once daily | 34 | 14 | 42 | 131 | -3 | nd | -28 | NSc | peri♀ |
| B | ||||
| ISP w/Isoflavones twice daily | 34 | 14 | +10 | -15 | NSc | ||||||||||||
| Carbohydrates | +25 | ||||||||||||||||
| Onning 1998 | 4 wk | Soy milk | 23/30h | 12 | 80 | +9 | NS | 0 | ♀♂ |
| B | ||||||
| Oat milk | +9 | NS | |||||||||||||||
| Gardner-Thorpe 2003 | 6 wk | Soy flour biscuits | 45 | 75 | 120 | 19 | 106 | +27 | 0 | NSc | ♂ |
| B | ||||
| Wheat flour biscuits | +27 | ||||||||||||||||
| Supplement | RCT | Miscellaneous | |||||||||||||||
| Takatsuka 2000 | 9 wk | Soy milk | 109 | 17 | 27 | 98 | -17 | NS | -1 | NS | pre♀ |
| B | ||||
| No supplement | 25 | 96 | -16 | NS | |||||||||||||
| Supplement | Xover | No Control | |||||||||||||||
| Wangen 2001 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 18 | 130 | -22 | NS | NSc | -- | post♀ |
| C | |
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | 18/17j | -22 | NS | -- | ||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | 18 | -27 | NS | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Merz-Demlow 2000 | 13 wk | ISP w/Isoflavones | 70 | 47 | 10 | 128 | 53 | 13 | 56k | -4 | NS | -- | pre♀ |
| C | ||
| ISP w/Isoflavones | 35 | 24 | 5 | 64 | 53 | -2 | -- | ||||||||||
| ISP w/o Isoflavones | 5 | 4 | 1 | 10 | 53 | +7 | -- | ||||||||||
| No control group | -- | ||||||||||||||||
| Supplement | RCT | No Control | |||||||||||||||
| Gallagher 2004 | 39 wk | ISP w/Isoflavones | 52 | 28 | 96 | 40 | 17 | 138 | +10 | NS | nd | -- | post♀ |
| C | ||
| ISP w/Isoflavones | 28 | 20 | 52 | 40 | 19 | 135 | +6 | NS | -- | ||||||||
| ISP w/o Isoflavones | 4 | 0 | 4 | 40 | 14 | 110 | +24 | NS | -- | ||||||||
| No control group | -- | ||||||||||||||||
| Mackey 2000 | 12 wk | ISP w/Isoflavones | 65 | 28 | 25 | 135 | +1 | NS | NS | -- | post♀ |
| B | ||||
| Female study | ISP w/o Isoflavones | 4 | 28 | 24 | 136 | -7 | -- | ||||||||||
| No control group | -- | ||||||||||||||||
Apparent discrepancies between Within-Cohort changes and Between-Cohort changes are due to rounding errors in estimated Within-Cohort changes compared to reported Between-Cohort changes.
Or difference between final values, as noted.
Unequal amounts of fat in soy milk (17.5 g/day) and cow milk (0 g/day).
Difference of final values (cross-over study)
Intention-to-treat analysis (75 completed soy protocol, 78 completed control protocol)
Graph
Median value, difference or net difference of median values.
Significantly higher at baseline than other study arms because of single outlier with hypertriglyceridemia.
Women/Men
N: baseline/final.
Measurements made at 4 menstrual cycle phases: early follicular, midfollicular, periovulatory, midluteal. Data extracted for periovulatory only. LDL change was greatest during this phase in high isoflavone group.
-9)We found 57 studies that reported on the effect of the consumption of soy products on triglycerides. Of these, 3 reported only that there was no effect on triglycerides.46, 47, 73 The remaining 54 studies are described below.23, 24, 49–52, 54–60, 62–64, 66–72, 74, 76–93, 95–100, 102–106
The range of daily soy isoflavone intake among studies of soy products with isoflavones was approximately 10 to 185 mg, with a median of 80 mg per day. The range of daily soy protein intake among relevant studies was approximately 14 to 113 g, with a median of 38 g per day. Sixteen studies included subjects with, on average, abnormal triglycerides; the remainder included, on average, subjects with normal triglycerides. The large majority of studies were of limited applicability, even within the categories of pre- or post-menopausal women, or men. Only 12 of the studies were graded as being broadly applicable. Among the 54 studies, 5 were rated good quality (A), 27 were rated fair quality (B), and 22 were rated poor quality (C).
). Approximately two-thirds of the cohorts of subjects
consuming soy had a net decrease in triglycerides compared to control,
implying a benefit of soy. Across studies, net change ranged from -49 to
+66 mg/dL, with the median net change equal to -3 mg/dL. In terms of
percent net change (using the baseline level in the soy intervention
cohort as the denominator), net change ranged from approximately -49% to
+31%, with a median net change equal to -2%. Random effects model
meta-analysis of non-soy-controlled studies with sufficient data (Figure 9) resulted in a
statistically significant net change estimate of -8 (95% CI -11, -5)
mg/dL.Analyses across studies
graphs the net change
compared to non-soy control of all cohorts evaluated (that included a
non-soy cohort). The left-most graph displays net change in
triglycerides in relationship the baseline triglycerides level; this
graph includes all studies. The middle graph compares net change to
daily soy isoflavone consumption (among those studies that report
isoflavone content). The right-most graph compares net change to daily
soy protein consumption (among those studies that report soy protein
content). Study cohorts who consumed soy with isoflavones are indicated
by squares; cohorts who consumed soy without isoflavones are indicated
by circles; and cohorts who consumed soy isoflavone without soy protein
are indicated by triangles. Black symbols represent cohorts where the
soy product was consumed as part of the regular diet; open symbols
represent cohorts where the soy product was consumed as a supplement to
the diet (including soy milk products). Larger symbols indicate cohorts
whose mean triglycerides was abnormally low at baseline; smaller symbols
indicate cohorts with normal triglycerides at baseline.
Visual inspection of the graphs reveal a possible negative association between baseline triglycerides levels and net change triglycerides across studies (net reductions are larger at higher baseline triglycerides. However, net reductions in triglyceride appear to be greater at both low doses of soy isoflavones and low doses of soy proteins. Similar patterns appear for all 3 types of products.
). Among studies with elevated baseline triglycerides, the net
reduction with soy consumption was greater (–11 [95% CI -16, -7]) than
among studies with normal baseline triglycerides (–5 [95% CI -9, 0]).
Sub-analysis of poor quality studies (rated C) revealed no significant
difference in effect compared to better quality studies.
Random effects model meta-regression across all studies found a similar effect, such that for each increase in mean baseline triglycerides of 10 mg/dL, the additional net change in triglycerides with soy consumption was -0.8 mg/dL (95% CI -0.13, -0.02). There was no clear threshold baseline level where there was a substantial, consistent change in the association between baseline triglycerides and net change triglycerides. By meta-regression, neither isoflavone or soy protein dose was associated with net effect on triglycerides.
Effect of baseline level of abnormal lipids on net change triglycerides in individual studies
Effect of soy isoflavone dose on net change triglycerides in individual studies
Effect of soy protein dose on net change triglycerides in individual studies
Effect of soy as diet vs. supplement on net change triglycerides in individual studies
Effect of sex and menopausal status on net change triglycerides
See Section 3.2.6.
| Randomized Design | Total Cholesterol | LDL | HDL | Triglycerides |
|---|---|---|---|---|
| Parallel | 29 | 21 | 27 | 25 |
| Cross-over | 32 | 31 | 29 | 29 |
| Total | 61 | 52 | 56 | 54 |
HDL = high density lipoprotein; LDL = low density lipoprotein
| Quality | Total Cholesterol | LDL | HDL | Triglycerides | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Population Category | A | B | C | A | B | C | A | B | C | A | B | C | ||||
| Total | 5 | 32 | 24 | Total | 4 | 28 | 20 | Total | 5 | 31 | 20 | Total | 5 | 27 | 22 | |
| Women & Men | 13 | 0 | 7 | 6 | 13 | 0 | 7 | 6 | 11 | 0 | 6 | 5 | 11 | 0 | 6 | 5 |
| Post-Menop Women & Men | 8 | 0 | 6 | 2 | 8 | 0 | 6 | 2 | 8 | 0 | 6 | 2 | 8 | 0 | 6 | 2 |
| Post-Menop Women | 24 | 4 | 11 | 9 | 20 | 4 | 7 | 9 | 23 | 4 | 11 | 8 | 22 | 4 | 8 | 10 |
| Peri-Menop Women | 3 | 0 | 3 | 0 | 3 | 0 | 3 | 0 | 3 | 0 | 3 | 0 | 3 | 0 | 3 | 0 |
| Pre-Menop Women | 3 | 0 | 1 | 2 | 2 | 0 | 1 | 1 | 2 | 0 | 1 | 1 | 2 | 0 | 1 | 1 |
| Men | 10 | 1 | 4 | 5 | 6 | 0 | 4 | 2 | 9 | 1 | 4 | 4 | 8 | 1 | 3 | 4 |
| Baseline Lipid Category | ||||||||||||||||
| Normal | 11 | 0 | 4 | 7 | 10 | 0 | 4 | 6 | 45 | 3 | 27 | 17 | 37 | 3 | 18 | 16 |
| Abnormal | 50 | 5 | 28 | 17 | 42 | 4 | 24 | 14 | 9 | 2 | 4 | 3 | 17 | 2 | 9 | 6 |
HDL = high density lipoprotein; LDL = low density lipoprotein; Post-Menop = post-menopausal; Peri-Menop = peri-menopausal; Pre-Menop = pre-menopausal
| Total Cholesterol | LDL | HDL | Triglycerides | |
|---|---|---|---|---|
| Soy Product (Study Arms) | ||||
| Protein with Isoflavones | 67 | 57 | 63 | 60 |
| Protein without Isoflavones | 11 | 10 | 11 | 11 |
| Isoflavones | 11 | 9 | 10 | 10 |
| Control Types (Study Arms) | ||||
| Dairy | 32 | 30 | 30 | 28 |
| Animal/Usual Diet | 10 | 8 | 8 | 8 |
| Placebo | 10 | 8 | 9 | 9 |
| Miscellaneous | 8 | 6 | 8 | 7 |
| No Non-Soy Control | 5 | 5 | 5 | 5 |
| Diet or Supplement (Studies)a | ||||
| Diet | 24 | 17 | 22 | 20 |
| Supplement | 38 | 35 | 35 | 35 |
HDL = high density lipoprotein; LDL = low density lipoprotein
One study is double counted because it compares diet to supplement.
| Comparison Characteristics | Total Cholesterol | LDL | HDL | Triglycerides |
|---|---|---|---|---|
| Protein with v without Isoflavones a | 11 | 10 | 11 | 11 |
| Different Soy Protein Dosages | 4 | 3 | 4 | 4 |
| Different Soy Isoflavone Dosages | 14 | 12 | 14 | 14 |
| Different Baseline LDL or TC Levels | 3 | 4 | 2 | 3 |
| Different Population Categories | 4 | 4 | 4 | 4 |
HDL = high density lipoprotein; LDL = low density lipoprotein; TC = total cholesterol
One study also included an isoflavone only treatment arm (for all lipids).
A total of 61 studies reported data on the effect of consumption of soy products on total cholesterol levels. The median net change compared to control found was approximately -6 mg/dL (or -2.5%) with a wide range of effects, from -33 to +7 mg/dL (–12% to +4%). Across studies, there were no discernable differences in effect based on baseline total cholesterol, soy protein consumption, soy isoflavone consumption, soy incorporated into diet or as supplement, or population (post-menopausal women, pre-menopausal women, men). However, 2 studies reported greater net effect of soy in subjects with more severely elevated lipids. Most studies that directly compared different doses of soy protein or soy isoflavones found no significant difference in effect, although results were mixed. Most studies that also directly compared effect in men and women found no difference.
A total of 52 studies reported data on the effect of consumption of soy products on LDL levels. A wide range of effects were reported, ranging from -32 to +13 mg/dL (or -21% to +9%). While few studies found a statistically significant benefit of soy consumption, meta-analysis across the diverse studies yielded a statistically significant net change of -5 (95% CI -8 to -3) mg/dL (roughly -3%). Across studies, there is possible evidence that the beneficial effect of soy products increases with increasing baseline LDL, particularly among studies where mean baseline LDL was greater than 130 mg/dL; although these associations were not statistically significant. Similarly, there is possible evidence of an association between higher soy protein dose and greater net reduction in LDL; however, only in the sub-analysis of studies with elevated baseline LDL was this association statistically significant. When studies with minimal doses of soy protein (<10 g/day) were omitted, the association was non-significant. No association was found between soy isoflavone dose and net effect. Qualitative analysis across all studies revealed no other associations between net change and other variables, including differences among soy products, , soy incorporated into diet or as supplement, or population (post-menopausal women, pre-menopausal women, men). The 3 studies that compared effect to baseline LDL level came to conflicting conclusions. Most studies that directly compared different doses of soy protein or soy isoflavones found no significant difference in effect, although results were mixed. Most studies that also directly compared effect in men and women found no difference.
A total of 56 studies reported data on the effect of consumption of soy products on HDL levels. The median net change compared to control found was +1 mg/dL. This estimate was in agreement with the meta-analysis estimate of +0.6 (95% CI -0.5, +1.8) mg/dL, which was not statistically significant. With only 2 exceptions, all studies reported a net effect on HDL of less than 10 percent, with an even distribution between net increases and net decreases or zero effect. Across studies, there were no consistent differences in effect based on baseline HDL, soy protein consumption, soy isoflavone consumption, soy incorporated into diet or as supplement, or population (post-menopausal women, pre-menopausal women, men). A possible associaton between baseline HDL and net change was found; although this association disappeared with the exclusion of 2 outlier studies. Studies that directly compared different baseline degree of abnormal lipids, doses of soy protein or soy isoflavones, or populations found no significant difference in effect.
A total of 54 studies reported data on the effect of consumption of soy products on triglyceride levels. The median net change compared to control found was approximately -3 mg/dL (or -2%), although a wide range of effects were reported, ranging from -49 to +66 mg/dL (–49% to +31%). Meta-analysis estimated a significant net effect of -8 (95% CI -11, -5) mg/dL. Meta-regression revealed a possible association between increased mean baseline triglyceride level and greater net reduction in triglycerides. Neither isoflavone or soy protein dose was associated with net effect on triglycerides. Within specific studies that investigated these possible associations, though, most studies found no associations. There was no evident association with whether soy was incorporated into diet or as supplement, or based on population (post-menopausal women, pre-menopausal women, men).
There is a great deal of heterogeneity of effects found on lipoprotein and triglyceride levels. None of the factors we evaluated, including population, quality, applicability, soy isoflavone dose, soy protein dose, or baseline lipidemia level satisfactorily explained the heterogeneity. Overall, the majority of studies reported small to moderate effects on the lipids, despite a wide range of net effects for total cholesterol, LDL, and triglycerides. With few exceptions, studies consistently reported a small benefit on HDL. While we cannot exclude the possibility of publication bias (negative studies being less likely to be published) as an explanation for the effect of soy on LDL, there was no clear evidence that negative trials were “missing.” However, the clinical heterogeneity of the trials makes this analysis difficult. Since most studies reported multiple outcomes, including lipids, it is possible that publication bias is less likely among these studies. It is also probably less likely that negative trials for HDL and triglycerides have not been published, unless the effect on LDL (and other outcomes) was also negative.
In order to guide future research, we have identified those studies that had a large effect on lipids. We arbitrarily defined “large” as an increase or decrease of at least 10 percent for total cholesterol, LDL, or HDL, and of at least 20 percent for triglycerides. We excluded studies with unequal baseline lipid levels between soy and control, studies with net improvements in subjects with normal baseline values, and studies with net worsening that resulted in still-normal lipid levels. These studies might provide insight into which formulations of soy product or which populations may benefit or worsen lipid levels most. If such factors are discernible, it may be most worthwhile to focus future research on these factors.
Among randomized trials that compared soy products to non-soy controls, we identified 7 studies that reported large effects on lipids in 8 treatment arms. Three (4 arms) had net reductions of LDL ranging from 10 to 14 percent,50, 84, 91 and 1 had a net reduction of total cholesterol of 12 percent.61 We also identified 2 studies with large increases in triglycerides of 27 and 31 percent.59, 92 However, overall, there were no characteristics of these studies that distinguished them from studies with smaller or no effects on lipids.
| Diet/Supplement | Design | Control | Dose | N | Base value | Change | Net Change a | Population | Applicability | Quality | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Author Year | Duration | Intervention | mg/day | g/day | Value | P within | P btw Soy | Value | P vs Control | ||||||||
